bigdrums925 9/23/2017 1:30 PM
Balder lightning guitar amp schematic please!
I am desperately searching for any documentation for a BALDER LIGHTNING GUITAR AMP.
I believe Balder was located in Sweden, then bought out by Nady Systems in Oakland, CA.
I have checked with Nady but they say they got rid of all Balder/Nady docs years ago.
If anyone can point me in a likely direction it would be appreciated greatly.
One of the problems I'm having with this amp is high frequency parasitic oscillation.
I'm told this was an ongoing problem with these, but a great SS amp otherwise

On the circuit board it reads "S.O.S. LIGHTNING LMk 82-2

Thank you
 
nickb 9/23/2017 3:05 PM
I see you asked once before Balder lightning schematic needed - it's not looking good.

I think you safely say you are on your own with this. Can you post some gut shots so we can see what you are dealing with? How do the controls affect the problem?
 
bigdrums925 9/23/2017 8:17 PM
Thanks Nick, I wasn't sure if I posted previously or not.

[ATTACH=CONFIG]45028[/ATTACH]
In this pic the 4ea TO3 output transistors have been removed.
When they were in I could not prevent it from going into oscillation.

[ATTACH=CONFIG]45029[/ATTACH]
This pic is of the circuit just before the output. I was able to determine one of the trim pots is for DC offset. Still don't know what the other is for

Recently I found upwards of 166 VAC nearest the missing output transistors. But this only occurs when a plug is inserted into the channel 1 input.
It doesn't have any signal on it, just inserting it causes the high AC voltage to appear, disappears when removed. It does not occur inserted into the
channel 2 input jack. Not sure but it may be related to the tube overdrive circuit in this amp.
The AC is strange but the oscillation is the one I need to tackle.
 
J M Fahey 9/23/2017 9:50 PM
Just thinking aloud and in no particular order:
1) expensive Darlington output transistors.
2) you can do nothing until you remount them, amp is running and showing nasty oscillation.
3) just a hunch but I would not be surprised if it had some family resemblance to old Hughes Ž Kattner amps.
They also once loved these Darlingtons, this designer might have "borrowed" that design (something way more common than most think), so that *might* help-
4)
How do the controls effect the problem?
5)
I found upwards of 166 VAC nearest the missing output transistors. But this only occurs when a plug is inserted into the channel 1 input.
"near" (physical proximity) means nothing, connections mean everything.
"166VAC" might very well be *60V DC* improperly understood by a cheap multimeter, measuring DC but set to the AC scale.

So in a nutshell:
6) rebuild the amp as original and weīll suggests some tests.
 
nickb 9/24/2017 1:18 AM
As Juan said + make sure no speaker is connected to avoid stressing the output transistors. Also, loose that badly wired jack plug before it shorts out and destroys the output transistors.
 
Jazz P Bass 9/24/2017 7:35 AM
What in the world are they using that UPC2002 for?

It's a 5W power amp.
 
nickb 9/24/2017 8:54 AM
Quote Originally Posted by Jazz P Bass View Post
What in the world are they using that UPC2002 for?

It's a 5W power amp.
Reverb driver (with attitude)?
 
J M Fahey 9/24/2017 10:38 AM
Most certainly.
Roland and Yamaha typically used a "car radio" type chip amp to drive low impedance (8 ohms ) reverb tanks.

They *might* also use it for an independent headphone out although that would be more unlikely.
 
bigdrums925 9/24/2017 2:09 PM
The UPC2002 probably does drive the reverb. I removed the "tank" to get a better look at the traces, and the "tank" mounts just above the 2002 chip on a set of four springs. Not that it matters but the reverb springs have no cover, never did.
I removed the Darlington's to see if the oscillation occurred prior to and without them in. Results were nonconclusive. It was at that time I noticed the high AC. I understand near means nothing, but the output drivers which connect directly to the Darlington's are there so.....and without the schematic was trying to get a general understanding of the signal flow paths.
 
J M Fahey 9/24/2017 5:54 PM
Ok.
The main point was that VERY probably you do NOT have 166VAC there, but your meter is lying to you.
 
nosaj 9/24/2017 6:19 PM
Quote Originally Posted by J M Fahey View Post
Ok.
The main point was that VERY probably you do NOT have 166VAC there, but your meter is lying to you.
Ground reference could be very important in this case, no?

nosaj
 
bigdrums925 9/24/2017 10:13 PM
Before I put the Darlington's back in can anyone suggest a way to test them with certainty?
I have a simple transistor tester but it gives me different results depending on which lead I put where.
The manual that came with it suggests it shouldn't matter which lead goes where
However it does say it might not test high gain Darlington's accurately
It is a radio shack cat No. 22-330 electronic components tester
 
J M Fahey 9/24/2017 10:44 PM
Quote Originally Posted by bigdrums925 View Post
it does say it might not test high gain Darlington's accurately
Test them using the diode function in your multimeter, they will test like any other NPN or PNP transistor with just two minor quirks because of their internal construction:
1) forward bias Vbe will be around 1.3V instead of around 0.65V for the very good reason that you are measuring *two* BE junctions in series
2) in many power Darlingtons , certainly in TIP142/147 , not sure on these, you will find an internal EC diode, connected "backwards" so it does not pass current under normal operation, so if you find it , donīt worry.
Itīs a "parasitic" diode which appears as a byproduct when the transistor is made, but since it does not cause any problem, they leave it there.

Example: TIP142 showing the internal parts, see the two BE junctions in series and the parasitic diode (pointing "the other way" so it does not pass current under normal operation) ; yours shouldnīt be very different.
[IMG]http://www.pinout.ca/wp-content/themes/directorypress/thumbs/pinout-diagram-TIP142-darlington-transistor1.jpg[/IMG]
 
Enzo 9/24/2017 11:54 PM
And to be thorough, the two internal resistors are parallel the two BE junctions and can confuse readings.
 
bigdrums925 9/26/2017 1:03 PM
I checked the Darlington's seemed ok but didn't find a double junction in every case, however the were all the same as its pair.
Put them in and brought up the variac with the scope on the output, no load. no input plug either 0-1/2 way very little current draw, then at 50%
the variac amp meter slams on full (1.5 amp meter) and about 20mV of oscillation appears on the scope. Same as before.
Turned it off. Any suggestions?
Btw the ground pin is missing from the power plug.
 
nickb 9/26/2017 2:55 PM
I don't think oscillation is your problem, at least not primarily. I'd say idle current is too high. This brings me to something I had wondered about from looking at the pictures. In order to balance the output transistors it is usual to put small value resistors in the emitters. The problem is I don't see any. There is a row of four resistors that seem to be the same value on the top left but I can't make out the value and they seem a bit wimpy for the purpose. You should be able to determine if they are going to the four emitters with a continuity check. The whole point of this is that you would normally measure the idle current by looking a the voltage drop across those resistors.

I see there is trimmer pot between what look like two driver transistors on black heatsinks. I suspect this is idle current adjustment. You can note the position and set then it to one end then the other with the power off and see what effect it has on the point at which the idle current shoots up. It might simply be a case of a bad trimmer. Failing that I don't think there are any great choices. My preference would be to lift the PCB and trace out the power amp section and draw a schematic of it.
 
J M Fahey 9/26/2017 2:59 PM
Were you using a speaker or t was unloded?
Try both ways.
20mV at the speaker out is *nothing*, it does not justify the 1.5A mains current jump which amounts to over 150W mains consumption.
Where would/could that current be going?
Place a scope probe *across* an emitter resistor (ground clip to one end, hot clip to the other, you might have to disable mains grounding on the scope) to check transistor current consumption.
 
The Dude 9/26/2017 3:38 PM
Quote Originally Posted by J M Fahey View Post
Were you using a speaker or t was unloded?......
That's a good question. If you are bringing the amp up slowly on a variac with a speaker attached, you could have problems. Often, power supplies do not come up in a linear fashion when you bring them up slowly. This causes DC on the output until the amp rails balance and, of course, lots of current. So, rule of thumb, don't use a load when bringing an amp up slowly on a variac.
 
nickb 9/26/2017 4:23 PM
I understood he did it with no load, see post #15. :

scope on the output, no load.
 
The Dude 9/26/2017 4:26 PM
Quote Originally Posted by nickb View Post
I understood he did it with no load, see post #15. :
Oops! Apologies, I missed that part. Carry on.
 
bigdrums925 9/26/2017 5:19 PM
Here's a look at the other side of the board. These are probably the resistors you spoke of. Except there are twice as many resistors as there should be (1.3 ohm)
[ATTACH=CONFIG]45066[/ATTACH]
Yes I thought it was idle current problem too at first. And assumed that trim pot would be for idle current.
At the point when the current shoots up the drop across one of the resistors is 1 volt.
again this is at half power on the variac, and more interesting is I get a 16 v p-p ac sine wave at the speaker output.
 
The Dude 9/26/2017 5:32 PM
On a side note, whatever that part is that's had it's solder pads lifted off (near the emitter resistors) does not look like it's very well attached. I'd unsolder it, clean the traces a bit and solder it in there better. Looks like it's barely (if at all) tacked in there.
 
nosaj 9/26/2017 5:46 PM
Quote Originally Posted by The Dude View Post
On a side note, whatever that part is that's had it's solder pads lifted off (near the emitter resistors) does not look like it's very well attached. I'd unsolder it, clean the traces a bit and solder it in there better. Looks like it's barely (if at all) tacked in there.
I understand at this point you all are trying to figure out whats wrong. I haven't seen it mentioned yet. The power transformer what kind of voltages is it giving out disconnected from the PCB? With the transformer disconnected from the circuit and a load put on the PT will it still give a high amp draw? just trying to start from the beginning of the circuit.

nosaj
 
bigdrums925 9/26/2017 5:59 PM
I understand. Does it help that the rail voltages are +- 46v and +- 12.6volts without disconnecting the trans?
If not I can do that.
 
J M Fahey 9/26/2017 8:54 PM
Yes I thought it was idle current problem too at first. And assumed that trim pot would be for idle current.
And it can still be an idle current problem: who says the trimmer or the circuit associated to it is working properly?
In fact I see the area where the Bias trimmer and bias sensing transistor should live, with signs of broken tracks and poor repairs.
That alone is an alarm siren going full blast.

At the point when the current shoots up the drop across one of the resistors is 1 volt.
again this is at half power on the variac,
Another red flag: 1V means >1.5A per transistor, or over 3A , at idle and with only half voltage applied or less.

and more interesting is I get a 16 v p-p ac sine wave at the speaker output.
What frequency sinewave?
Are you injecting a tone while you do this test? You shouldnīt.
And no load connected.

To be able to suggest tests, we need that output stage schematic, specifically so you can follow instructions and we all speak the same, so grab pencil and paper and draw it.

Not all of the amp (hopefully) but at least the power rails, ground, main filter caps, speaker out, Zobel network, grounding, power transistors and mainly the bias circuit.
 
bigdrums925 9/27/2017 11:25 PM
Curiously just before the amp meter swings full on, a 600Hz signal @ 14Vp-p appears at the output with no input signal.
The sine wave output remains after the meter swings full on and I see no oscillation. However the speaker plug was only half way in.
So it seems possible that the 600Hz is on the ground plane perhaps. When the speaker plug is in I see the oscillation and no 600 Hz.
Now can't recreate the 600Hz at all or the oscillation but the meter still swings hard 1/2 way on. I ramped up the variac all the way momentarily and measured 3.4 volts drop across the emitter resistor.

The schematic for the output will probably take a little time because I've never drawn one before and following the traces on this amp is a nightmare. But I will try in the next day or so. I don't know what a Zobel network is, can you explain?
 
nickb 9/28/2017 12:29 AM
The 1.3 ohm resistors may be in four parallel pairs giving 0.65 ohms. That makes your idle current huge so there is no question there is a problem there. It might help to look at other solid state power amp schematics to get a feel of what's there as there are often great similarities between designs.

A Zobel network is a capacitor and resistor connected across the speaker. It's there to improved the amplifier's stability especially at higher frequencies. Zobel networks

 
MarkusBass 9/28/2017 2:54 AM
It's clear that the amp failed in the past and the question is whether it was fixed correctly, or not. You can easily see that several components were replaced. The bias can be altered with the upper trimmer (2k2). It works together with the little transistor on the top of the board. So I would check the trimmer and the transistor. Also, what are the voltages on collector and emitter of the transistor. It seems to me that there is no other way to fix the amp then drawing its schematic and verifying whether it is correct. This shouldn't be to difficult since I think that this is typical solid state amp.

Mark
 
J M Fahey 9/28/2017 7:49 AM
Quote Originally Posted by bigdrums925 View Post
The schematic for the output will probably take a little time because I've never drawn one before and following the traces on this amp is a nightmare. But I will try in the next day or so. I don't know what a Zobel network is, can you explain?
We donīt need the full schematic, just the power transistors, whatīs directly connected to them, and mainly
the bias circuit , which basically is a transistor, a trimmer and 4 or 5 resistors.
To be able to suggest what to measure and how to adjust it, we need at least that, specially so we all talk the same.
Draw what you can and if we find something missing, weīll ask you to add it ... but we have to start with *something*

This is the area we are interested in:
 
nickb 9/28/2017 12:21 PM
Attempt:

[ATTACH=CONFIG]45099[/ATTACH]

Did you acquire the amp in this condition?
 
MarkusBass 9/29/2017 2:06 AM
What worries me is the resistors that are in series with the BIAS trimmer are both 33k and they both were replaced. 33k seems to much for me. But it look like the OP lost interest of fixing the amp
 
J M Fahey 9/29/2017 4:22 AM
Hey, maybe we have good news in a week or two
 
dmeek 9/29/2017 11:49 AM
Obviously I have too much time on my hands, but I do love a puzzle. The 220 ohm feedback resistor
probably should be 2.2k -looks like it has been replaced with the wrong value, might even solve the
oscillation problem. It's the 1W resistor on the left

Doug



http://music-electronics-forum.com/a...1&d=1506712143



(I've edited this post to make a few corrections)
 
nickb 9/29/2017 12:54 PM
Way to go Doug! I think there are some errors a so I'll try a redraw in due course...


Agree with other suggestions plus I'd check the small TO92 transistor next to the heatsink. I suspect it's dead.


... this is my interpretation, with suggested value corrections based on Doug's hard work
[ATTACH]45115[/ATTACH]
 
MarkusBass 9/29/2017 2:35 PM
Nice . I wonder whether some resistors that we think are 33k, are actually 2k2 (the stripes are red instead of orange). But if someone replaced the resistors in series with the bias trimmer with 33k, such high values make it impossible for the bias transistor to work correctly. The OP could measure some of the 33k resistors.

Mark
 
J M Fahey 9/29/2017 2:37 PM
Hey!!! thanks to you both, dmeek and nickB

following the latest fad, Iīll also suggest a small correction and a couple tests,

1) I like this input pair connection better

[ATTACH=CONFIG]45116[/ATTACH]

2) after checking that schematic and actual PCB match, Iīd suggest measuring voltage difference between points 1 and 2 ; 3 and 4 ; 5 and 6.
I know that amp does not currently stand getting full voltage , so please rise Variac slowly, until idle or mains current starts to dramatically rise, then back down a little into a safe zone, we donīt want to overheat or damage anything, and *then* measure suggested pairs.

I bet we will find overbiasing signs there ( and that at way less than mains voltage)

Once we turned a suspect into a culprit, weīll try to solve that, but letīs go one step at a time

EDIT: suggest using a plain graphic format, such as gif or jpg, to make editing and reposting easier
 
dmeek 9/29/2017 3:02 PM
[QUOTE=nickb;466678]Way to go Doug! I think theee are some errors a so I'll try a redraw in due course...


Thanks for sorting out the input stage, it had me confused.
btw the zener at lower left is backwards
 
bigdrums925 9/29/2017 3:11 PM
No, it was working some years ago, friend gave it to me. At that time it needed some minor repairs that were more or less obvious. I don't remember what that was though. It was working but not used much until about 6 months ago I found out that some gtr players really like it and had some value. I had a buyer coming to see it and I checked it the night before. The input jack was making noise so I assumed it was like many others just cracked solder joints on the jack at the board. I resoldered them but became curious about the 2 trim pots. I don't remember exactly but after marking their positions and adjusting them slightly, the resistor between the drivers smoked in the center. I could see the 2 red stripes but not the multiplier so measured it at 220 ohms. I replaced that and then started blowing the output transistors. Got it to settle down but then started getting the oscillation.
 
bigdrums925 9/29/2017 3:29 PM
No not at all losing interest. Just had a few obligations to deal with this week. I started to draw a schematic of the PS and amp section but I have never drawn one before so it will take some time before I have anything I can post.. I know one of the trimmers is for DC offset, the other seems to have some effect on the bias based on the emitter resistor voltage drop changing between 3.x volts and 6.x volts last time I measured it.
How can you tell from just the pics the 33k resistors raised off the board are in series with the trim pot? You are correct of course, I checked, one to each fixed leg.
From much input here it seems the idle current is not being controlled properly, from what little I know it should be down at about 10-20 mA
I have been trying to get that down by changing a few of the resistors that were suspect, out of spec etc.
So how can I reduce the idle current to the point where the trim pot is in range?
 
nickb 9/29/2017 3:42 PM
Mark - I had changed the Vbe multiplier resistor values to 330 and 5.6K based on a requirement of 6 Vbe's and to give the trimmer a decent range. With 33K's that Vbe multiplier transistor would likely be current starved and could cause high idle current as you say. But with two equal resistors the Vbe multiplier would only be x2 which is not enough.

Hmm..just read post 38. Kinda blows that idea out of the water. Time to measure volts.

Juan: Yes, I forgot the check the phase

Doug: Yes. Thx I see on the schematic I got one of the zeners the wrong way round on the input offset trim.

Updated (jpeg):

[ATTACH=CONFIG]45117[/ATTACH]
 
bigdrums925 9/29/2017 3:43 PM
OMG Doug! You drew this from the 2 pictures alone? You all are incredible.
If the 1watt 220 ohm resistor you are referring to is the carbon, that value was because its pair is 220 ohm.
Much of that section seems to be laid out symmetrically. That's all I had to go on.
Please forgive any dumb remarks or observations on my part because I know very little. That's why I'm here
 
bigdrums925 9/29/2017 4:01 PM
Bear with me please. I may sound quite dumb here. Disconnected from the board with no load I measure 2 pair that are 34.x VAC.
Not sure what amount of load to use though or how to determine the current when loaded.
 
bigdrums925 9/29/2017 4:31 PM
The TO92 next to the heat sink checks good. Its a BC182 BTW. I believe its the one near the "idle" on your PDF? Originally it was pushed down to touch the heat sink with some heat sink compound.
Once again I'm blown away at the help here, now a proper PDF with labels. Wow Thank you!
I noticed the transistors are mostly labeled Q?. Would it help if I provide the transistor numbers?
Or to provide any other part values?
 
J M Fahey 9/29/2017 5:26 PM
Quote Originally Posted by bigdrums925 View Post
Would it help if I provide the transistor numbers?
Or to provide any other part values?
Oh, please DO.

I suggest you leave schematic as is so we donīt lose track, but add actual PCB numbers, labels, or whatīs written (or painted) on components by them, either in another colour (say, violet) or between brackets or similar, such as (R4 100k) or whatever.

Amplifier and sound strongly remind me of Pearce amplifiers (evolution of Gibson/Moog LAB L5 and family).
 
bigdrums925 9/29/2017 5:41 PM
I have changed the 33k trim pot resistors to 330 Ohms and 5k6 as per the new drawings, I changed resistor that was burning up back to 220 Ohms.
And changed the resistor that was in the speaker output path parallel with the inductor near the big filter caps.
I still see a fairly sharp rise in current ramping up the variac.
Will check this later tonight.
Thanks everyone
 
J M Fahey 9/29/2017 10:28 PM
Please measure the voltage pairs I suggested, there lies the path to solve the problem.
We are not guessing here but troubleshooting, which does have a standard method to follow.
 
MarkusBass 9/29/2017 11:38 PM
Quote Originally Posted by bigdrums925 View Post
I don't remember exactly but after marking their positions and adjusting them slightly, the resistor between the drivers smoked in the center. I could see the 2 red stripes but not the multiplier so measured it at 220 ohms. I replaced that and then started blowing the output transistors. Got it to settle down but then started getting the oscillation.
Can you mark on the photo which exactly resistor have you replaced with 220 Ohms resistor? The resistor between the drivers looks like 3k6 and not 220 Ohms. Also, I wouldn't replace any resistors that were not replaced previously. You are trying to fix the amp amp not to design and manufacture a new one. The amp worked previously and you just need to fix it.
Can you verify whether the resistors that we think are 33k, are 33k, or 2k2? Just measure them.

Mark
 
MarkusBass 9/29/2017 11:53 PM
Quote Originally Posted by J M Fahey View Post
I suggest you leave schematic as is so we donīt lose track, but add actual PCB numbers, labels, or whatīs written (or painted) on components by them, either in another colour (say, violet) or between brackets or similar, such as (R4 100k) or whatever.
It's the easiest to add component numbers in KiCad - Nick can do it. But we also need capacitor values, which should be reported by Bigdrums.

Mark
 
nickb 9/30/2017 8:00 AM
Quote Originally Posted by MarkusBass View Post
It's the easiest to add component numbers in KiCad - Nick can do it. But we also need capacitor values, which should be reported by Bigdrums.

Mark
As requested:

[ATTACH=CONFIG]45133[/ATTACH]

Now we need those voltages. Bring the variac to a little below the point where the current rises out of control then measure the voltages requested.
 
bigdrums925 9/30/2017 3:29 PM
#5 is -2.14 v
#6 is -1.65 v
#4 is - .28 v voltage here seems to change note: the resistor near #4 is 470 ohms not 47 ohms
#3 is - .15 v voltage here seems to change note: the resistor near #3 is 470 ohms not 47 ohms
#2 is .44 v
#1 is .46 v

for reference these were all taken with the variac on only 15% with a current draw of 1.4 amps
 
nickb 9/30/2017 3:48 PM
Quote Originally Posted by bigdrums925 View Post
#5 is -2.14 v
#6 is -1.65 v
#4 is - .28 v voltage here seems to change note: the resistor near #4 is 470 ohms not 47 ohms
#3 is - .15 v voltage here seems to change note: the resistor near #3 is 470 ohms not 47 ohms
#2 is .44 v
#1 is .46 v

for reference these were all taken with the variac on only 15% with a current draw of 1.4 amps
I think #5 & #6 are transposed. So taking #6 as -2.14V I'd expect #4 to be around -3.4V yet it's -0.28V. This means the lower transistors should pass no current but we can see that the current is high at 350mA per transistor. Pull those lower transistors - one or both is likely bad.

Also, I was wrong about the 5.6k /330. Please restore to the original 33k values. The explanation is complex so I'll leave it for later.
 
bigdrums925 9/30/2017 8:13 PM
Yes I will post updated pics to reflect the recent changes, see below. The 3k6 resistor is now 220 ohms as original. I am only changing the resistors I changed previously.
Also I began writing in the transistor numbers and cap & resistor values in bright orange on this drawing. Please comment as needed. Never done this before.

[ATTACH=CONFIG]45142[/ATTACH]

[ATTACH=CONFIG]45143[/ATTACH]

If I understood correctly the 33k pair should be 330 and 5k6 ohms which I replaced last night
Take a look at the new pic because I'm am not sure about this.
However if you are referring to the resistors mounted perpendicular to all the rest, one of them being carbon, they are 220 ohms
My head is spinning but I am trying!
 
MarkusBass 10/1/2017 1:03 AM
Quote Originally Posted by bigdrums925 View Post
If I understood correctly the 33k pair should be 330 and 5k6 ohms which I replaced last night
And just in the previous post Nick asked you not to do it . Are you reading the posts in this thread?
You are also using old schematic - the new one with component symbols, posted by Nick, is just 3 posts below. Please use the new one.

I asked you also to verify the value of some resistors (like e.g. 33k). Can you do it? Now, on the photo with different light, it seems to me (Nick can you verify this please) that 330R resistors are actually 220R resistors and 33k resistors are actually 2k2 resistors (the stripes are red not orange). We can only guess it based on the photo but you can verify it in less than 1 minute. Can you do it please?

Quote Originally Posted by bigdrums925 View Post
for reference these were all taken with the variac on only 15% with a current draw of 1.4 amps
I never use variac with solid state amps. Instead, I use bulb limiter. The reason is simple: solid state amps are DC coupled and one voltage missing will cause all other voltages be incorrect. In your case you have two Zener diodes in the first stage of the amp. They are most probably 15V (again, could you read the symbol on the diodes?) and if you provided only 15% of the expected voltage, they will not work. It means that the input stage of the amp does not work, has incorrect voltages and all other voltages are incorrect as well. And you don't know whether this is because of a failed component, or due to missing 15V rail voltage. Variac can be only used with tube amps.

Can you also provide symbols of the transistors used in the amp (are the all BC182?) and values of capacitors. This would allow for having complete schematic of the power amp. Wouldn't you be glad having such a schematic?

Mark
 
nickb 10/1/2017 1:50 AM
Bigdrums - don't rely on email notifications as they often don't work. Also, if you are not using linear thread display mode, it's easy to miss new responses as they might be anywhere.

Towards the top of the page is a horizontal grey bar. At the right of it is a drop down marked 'display'. Go and select linear. Then look over the replies to see if there are any you missed.

My plan for attacking the DC problem that Mark mentioned was to leave the lower pair of output transistors off the board until we get the DC right.

Yes, it's hard to be sure of the colors from the picture so do please check them. I am updating the schematic as we go and I'll not re-post it for now to avoid confusion.
 
MarkusBass 10/1/2017 2:11 AM
Great work Nick. To complete the schematic you could add a Zobel network on the output (a capacitor and I think 4 parallel resistors to the ground) and a Thiele network on the output (a resistor parallel to inductor).
We are waiting for transistor types, capacitor values and 33k and 330 Ohms resistors verification.

Mark
 
bigdrums925 10/1/2017 2:57 AM
[QUOTE=MarkusBass;466824]And just in the previous post Nick asked you not to do it . Are you reading the posts in this thread?

Apparently I am not getting the correct sequence of the posts on my screen, didn't see his comment to put the 33k's back in until tonight, they are back in the board now

You are also using old schematic - the new one with component symbols, posted by Nick, is just 3 posts below. Please use the new one.

I will try to put transistor numbers/resistor & caps values as I am able to the newest drawing I'm able to edit.

I asked you also to verify the value of some resistors (like e.g. 33k). Can you do it? Now, on the photo with different light, it seems to me (Nick can you verify this please) that 330R resistors are actually 220R resistors and 33k resistors are actually 2k2 resistors (the stripes are red not orange). We can only guess it based on the photo but you can verify it in less than 1 minute. Can you do it please?

The 2ea 33k trim pot resistors are 33k. Even looking at the stripe colors in person it's hard to tell weather red or orange. So I am measuring them
I am having a little confusion as to which transistors/resistors/caps on the board are which on the drawing. I will verify any values I can on the latest post of the drawing I can. Also I think it may help if I assign T/R/C numbers on the drawing?
But there are none printed on the board to reference. Hopefully then the communication between us will be more clear, particularly when suggesting verification of values.


In your case you have two Zener diodes in the first stage of the amp. They are most probably 15V (again, could you read the symbol on the diodes?)

If you are referring to the ones on each side of the first trim pot they read 82x. That's all I can see without removing them

and if you provided only 15% of the expected voltage, they will not work. It means that the input stage of the amp does not work, has incorrect voltages and all other voltages are incorrect as well. And you don't know whether this is because of a failed component, or due to missing 15V rail voltage. Variac can be only used with tube amps.

I will attempt to verify the rail voltages in this circuit area and report back.

Can you also provide symbols of the transistors used in the amp (are the all BC182?) and values of capacitors. This would allow for having complete schematic of the power amp. Wouldn't you be glad having such a schematic?

I believe the BC182 is the only transistor of that type/number used in this amp. It is one of the drivers, correct? Not sure what you mean by symbols but am trying to determine which transistors on the drawing are which on the board. This should be simple but I'm not sure the drawing is correct yet. I put the transistor numbers I knew on the drawing hoping someone can tell me if I'm doing this correctly. I will continue this effort and update as I am able. I mentioned previously I have no experience making or editing drawings and only basic experience reading them. So I will be a bit slow getting the drawings/values/numbers posted and want to make sure when I do they are correct. Yes I would be very glad to have the schematic. I probably won't be able to do so tomorrow as I have a wake to go to.
 
bigdrums925 10/1/2017 3:08 AM
Just saw the trace drawing. Shall I attempt to verify? i.e. I only see two traces going to the temp sensor/bias transistor?
It will take me some time to do this though as I have to do them one by one. Please bear with me as I very much appreciate the help!
 
MarkusBass 10/1/2017 3:18 AM
Quote Originally Posted by bigdrums925 View Post
I believe the BC182 is the only transistor of that type/number used in this amp.
That's not true. Half of the transistors are of PNP type and they must be of different type (BC182 in NPN).

EDIT: complementary transistors to BC182 are BC212. Please check whether some of the transistors on the board are BC212.

If you are not sure where the transistors are on the schematic, you can simply put its symbol on the photo you posted previously. You can do it even with Paint. The same applies to capacitors. Just put the values on the photo and Nick will put them on the schematic.
I read the forum in linear mode. In this case, the latest post is always on the top and it is not possible to miss posts. You can select it in the top-right corner of the screen. Like this:

[ATTACH=CONFIG]45144[/ATTACH]


Mark
 
bigdrums925 10/2/2017 1:59 AM
I marked up the picture with the transistor #/resistor values/cap values or number/diode value or number

[ATTACH=CONFIG]45148[/ATTACH]

Hopefully now when I'm asked to check this part or value we can do it with callouts from the schematic and I will know exactly where.
 
MarkusBass 10/2/2017 11:25 AM
OK. Now it looks much better but there is still some work to be done. It's because I see that several components were replaced in a random way and I'm not sure whether the values are correct.
What I see is:
- power transistors are from different production series (in pairs), which is not allowed. Did you replaced just two of them instead of four?
- low power transistors are of different types (previously you said that all of them are BC182). Did you replace any of the low power transistors, or it was like this from the very beginning? I hope that you are aware that BC182 has different pinout than MPSA06. You have to be very careful when replacing one with the other.
- some capacitors are still not labeled (for example on the top of the board there are three capacitors but only two of them are labeled). Can you fix it?
- can you check what is the Zener diode on the top of the board?
At least two resistors are labeled incorrectly: the resistors on the left of the board are either 68 Ohms, or maybe 6R8. Can you measure them directly in the board?
- the resistor in the second column from the left (third from the top) is marked as 33k but I can see that this is either 330R, or 220R. Can you measure it in the board? BTW, it looks like it was replaced. Did you do it?
- the resistors that we thought that we are 33k, are actually 2k2 (as I thought). The 330R resistors are actually 220R.
- Zener diodes on the input are 15V (I thought so).
- it looks like some of the resistors were replaced (by you or someone else). For example, the famous resistor 220R, which you replaced with 3k6 resistor. What else did you replace?
- are there any additional resistors below BD419/420 heatsinks?
- the bias trimmer is very sensitive in this amp. You marked its initial position and started turning it. You shouldn't do it if you don't know how to verify the bias settings.
- what are the power supply rails? +/-40V, or 45V?

BTW, 33k resistors in series with the trimmer (or even 2.2k) in the bias circuit will not work correctly. Did you replace any of them?

I know how the amp works (even without the information you haven't provided). I could fix it but it's to far. Please provide the information listed above because we need to have a correct schematic first.

Mark
 
bigdrums925 10/2/2017 1:09 PM
I replaced both mj11013's and mj11014's randomly based on what was blown and what was available. At one point both the 13's were bad and I ordered 7 of them, replacing the 2 bad one's. Then one of the 14's was bad and I only had 1 so replaced that. They're expensive!

None of the low transistors (TO92's) have checked bad throughout this whole time so I have not replaced any of them, I got it this way in that regard. Never even unsoldered any except the BC182 next to the big heatsink (part of the bias I believe)

The 3 disk caps: left (0.2uF) & center (33nF) are original. The green on right needs replacement with another 0.2uF. I don't have one to replace it with so I had paralleled 2ea 0.1uF together and since then removed 1 leaving the green at 0.1uF. (Hoping to find one so I don't have to order just 1 small disk cap.)

The Zener is labeled "C3 V6" Motorola. Or perhaps "V6 C3". Not sure but could be 3.6v Zener? Checks good

4ea resistors on top left of board are 68 ohms ea. I mistakenly labeled them (67 ohms 4ea)


The resistor 2nd column, third from top measures 33k. (Pic is of previous configuration.) I think this one and it's pair are bias trim pot resistors I was asked to replace with 330 & 5.6k, then told I should change them back to 33k. That is the current configuration in spite of the resistor in the picture.

Yes I replaced various resistors, all within the picture. The replacements were all left above the board. The famous 220 ohm, once replaced with 3k6, is now back to 220 ohm, like when I got the amp.

No hidden resistors below BD419/BD420, They are the driver transistors I believe.

Last check on power rails were +/- 46.6 Do you need me to check again as this would require removing the outputs?

Yes the 33k in the bias were changed and changed back. (see above)

Hope this helps
 
nickb 10/2/2017 1:49 PM
OK guys here is where we're at. We still need part numbers for the rest of the TO92's.


Schematic:

[ATTACH=CONFIG]45164[/ATTACH]



DC Simulation showing expected voltages. Ignore the component references as they do not match the schematic above. Note I tweaked R26 to give a sensible idle current with the pot about midway and not to have the trimmer be too sensitive.

[ATTACH]45156[/ATTACH]
 
bigdrums925 10/2/2017 2:44 PM
Nick: I placed the numbers for the remaining TO92's far left and right of the devices in the latest picture because of space in the middle by the trim pots.
So verify please but I think Q5 then is a MPSA56, Q1 & Q3 are BC182's, Q7 MPSA06, Q2 & Q4, BC560's
 
nickb 10/2/2017 4:11 PM
Quote Originally Posted by bigdrums925 View Post
Nick: I placed the numbers for the remaining TO92's far left and right of the devices in the latest picture because of space in the middle by the trim pots.
So verify please but I think Q5 then is a MPSA56, Q1 & Q3 are BC182's, Q7 MPSA06, Q2 & Q4, BC560's
So you did! I missed it I've edited the post above with the TO92 part numbers .


I see a problem with the BC560's and MPSA56/06. The max voltage for a BC560 is 45V. There parts have 46.6+0.6= 47.2V on them so they will be operating out of spec. Even the BC182 at 50v rating is marginal. MPSA06/56 are 60V and they could have nearly the full supply of 90V on them. Could the power supply voltages reported at +46.6V and -46.6V be wrong? If they are right then this amp has been brutalized at some point in the past, beyond what we already knew.
 
bigdrums925 10/2/2017 4:37 PM
For what it's worth the transformer says it's rated at 2 x 31.8 V at the secondary's. Not sure what that is once rectified. I know I measured the rails several times previously and got +/- 46.6, and the 15v rails were +/- 12.3. Someone here in a previous post said I should check the output of the transformer disconnected and check for current capacity with a load but I wasn't sure how to do that or what the load should be. Nonetheless if I need to do that I will with some directions.

I should note here that while getting the part values and numbers last night and replacing wrong value and incorrect replaced parts, I got what appeared to be a short measuring across 2 of the output transistors MJ11013's. I removed 1 of them but can't verify a short in the transistor or the board, but my cheap tester says the transistor is bad. That may not mean much since it has given false readings before with these Darlington's
At the end of the night I decided to see if the sudden current draw at 15% variac was still there. It was not, but started to smoke the famous 220 ohm resistor located between the drivers.
So I'm not sure how to check the rails at this point without some problems. Please let me know how to proceed.
 
nickb 10/3/2017 12:51 AM
Well 31.8VAC x 1.4 = 45.5VDC so that confirms the 46.6V measurement. We'll need to find substitutes for the driver trannies before too long. I had mentioned one or both of the the MJ11013's were bad way back in post 51.

Test Q8 and Q9, if they are OK, then test Q5, Q5 & Q7. If that is inconclusive remove Q9, power up and measure the voltages around Q5, Q6 and Q7
 
MarkusBass 10/3/2017 1:19 AM
Quote Originally Posted by nickb View Post
I see a problem with the BC560's and MPSA56/06. The max voltage for a BC560 is 45V. There parts have 46.6+0.6= 47.2V on them so they will be operating out of spec. Even the BC182 at 50v rating is marginal. MPSA06/56 are 60V and they could have nearly the full supply of 90V on them. Could the power supply voltages reported at +46.6V and -46.6V be wrong? If they are right then this amp has been brutalized at some point in the past, beyond what we already knew.
It looks like the transistors were there originally. Also BC182 is rated for 50V only. Somehow the amp worked. The capacitors in the power supply are rated for 50V so power supply rails between 40 and 45V are expected. Bigdrums, can you make a photo of the capacitors (showing how they are connected to the board)? Are they original, or they were replaced?
Nick, I see few minor problems with the schematic and the simulation. The Q9 transistor on the schematic is connected incorrectly (it should be collector to V- rail). There are two 100k resistors on the board, which on the schematic are connected to the output. In the simulation they are missing. Are they connected to the output, or to the ground (I don't see it on the photos)?
There should be also 33n capacitor between collector and emitter of the bias transistor. Input capacitor is 220nF.
I'm not sure whether the amp could ever worked with two 33k resistors in the bias circuit. I have to check it but it seems to me that such values made setting the bias impossible. Even if they worked, the bias circuit is very sensitive and you shouldn't even touch it without undrstanding how it works.
Also, the 220R resistor in the feedback circuit (in series with 220uF) looks also suspected to me. It sets very high gain of the amp. I would expect there something like 820R, or 1k resistor. Bigdrums, did you replace any of these resistors previously?

Further steps depends on Bigdrums soldering skills. Can you desolder BD419/420 without destroying the board? I'm asking because the 220R resistor (replaced with 3k6) was desoldered in the way that PC board tracks were destroyed. We don't want this.

Mark
 
nickb 10/3/2017 2:46 AM
It's nice to have more than one pair of eyes

I've fixed Q9 and updated (post 62 above).

I left out the 100K's in the sim on purpose - they seem to have no discernable effect, especially on the DC conditions. I changed the Vbe multiplier lower res (R26 in the schematic) in the sim to 8.2K get a sensible idle current at mid piont and make the trimmer is less sensitive (see post 62).
.
PS: I set the gain to x100 using 22k /220 ohm in the feedback. Was stable in sim and 0.9Vpp in for full out seems reasonable
 
MarkusBass 10/3/2017 6:37 AM
Quote Originally Posted by nickb View Post
PS: I set the gain to x100 using 22k /220 ohm in the feedback. Was stable in sim and 0.9Vpp in for full out seems reasonable
0.9Vpp is 0.32RMS. A little bit high for me but maybe the amp is designed like this.
I think that we should wait for a photo of the power supply (capacitors) and information about soldering skills of the OP.
I was looking for replacement transistors and ZTX753 and ZTX653 look good. But this is for the future.
I've just noticed two additional resistors on the schematic R31 and R32. Are they used in the amp?

Mark
 
bigdrums925 10/3/2017 9:46 AM
I can solder and repair well. I can't always prevent damage from occurring though.
This is an overview of the PS:
[ATTACH=CONFIG]45166[/ATTACH]
And a closeup:
[ATTACH=CONFIG]45167[/ATTACH]
A sideview:
[ATTACH=CONFIG]45168[/ATTACH]
And the bottom:
[ATTACH=CONFIG]45169[/ATTACH]
 
nickb 10/3/2017 10:12 AM
Quote Originally Posted by MarkusBass View Post
I've just noticed two additional resistors on the schematic R31 and R32. Are they used in the amp?

Mark
They are under the heatsinks of the drivers. Can't really see to be sure.
 
dmeek 10/3/2017 4:40 PM
They are visible in the photo in post #59 - 470 ohms. Also the 100k resistors are soldered to the ground plane.
 
bigdrums925 10/3/2017 11:34 PM
Removed and tested Q8, Q9, Q1, Q3, Q5, Q2, Q4, Q7. All checked good. I checked Q6 recently.
With the 4 output transistors in I powered up slowly with the variac and the current draw rises sharply at about 15%
I removed Q9 but still have the excessive current draw so until I remove at least some of the outputs I won't be able to measure around Q5, Q6, Q7.
Any suggestions beyond removing the outputs?
Still not showing bad outputs though....
 
bigdrums925 10/3/2017 11:37 PM
Is there a updated schematic available?
 
bigdrums925 10/4/2017 12:15 AM
OK I just put everything back in after finding and replacing Q13, which was shorted.
I fired it up with low (normal) current draw, no oscillation, no smoke.
However the output is very low and channel 2 doesn't work.
I'll check the outputs later or tomorrow and post the results.
 
MarkusBass 10/4/2017 2:07 AM
The latest schematic is here: http://music-electronics-forum.com/t45081/#post466975 . i thought that you fired up the amp without the output transistors (and then got smoke). You didn't tell us this .
What is the voltage on any of R36-46 resistors? Can you change it with bias trimmer?

Mark
 
nickb 10/4/2017 2:53 AM
Quote Originally Posted by bigdrums925 View Post
Is there a updated schematic available?
The edit post period has expired so I'll have to post fresh. Here is the latest sch V0.5


PS: Thanks again to Doug and his eagle eyes else we would have not got here for a long time.
 
MarkusBass 10/4/2017 3:18 AM
I think that this is version 1.0. The only thing that seems to me that is incorrect, is that 4x68 Ohms resistors are in series with capacitor in the Zobel network and the output coil is in parallel to a single resistor (1R6) - at least it seems to me.
Great job.

Mark
 
J M Fahey 10/4/2017 9:18 AM
Quote Originally Posted by bigdrums925 View Post
Removed and tested Q8, Q9, Q1, Q3, Q5, Q2, Q4, Q7. All checked good. I checked Q6 recently.
With the 4 output transistors in I powered up slowly with the variac and the current draw rises sharply at about 15%
I removed Q9 but still have the excessive current draw so until I remove at least some of the outputs I won't be able to measure around Q5, Q6, Q7.
Any suggestions beyond removing the outputs?
Still not showing bad outputs though....
Trying to get back in the action but again dismayed when seeing that instead of **essential** voltage measurements we get endless useless shotgunning and parts removal and reinsertion with no clue or purpose.
Doubly so in an amp which is already suffering from track destruction, even worse because it uses outdated taped PCB design, guaranteed to produce thin flimsy tracks, which must not be messed with if at all possible.

Not surprisingly parts check good but amp does not work.

Hint: I couldnīt care less about how a transistor measures outside the amp, all that matters is Vce and Vbe *in circuit*.
Outside measurement is only a confirmation of what I *already* found in the circuit, not the other way around.
 
nickb 10/4/2017 11:14 AM
Bigdrums - there seems to be a real problem here. Either your are missing posts in the forum or are simply ignoring instructions. Don't rely on email updates at you wont get all the posts. Did you switch to linear mode?

We did get voltage readings which told us one or both of the PNP output transistors were bad. After that it went off the rails as the next step response was either not followed or missed. Example, you had been told to remove both the lower transistors and the do some more testing. This was get things to a good state to avoid blowing up more transistors but you went ahead and replaced them.

Never mind. They are in now so go and redo the voltage measurements. Did you look at the pdf of expected voltages? Use it as a guide on what you should see.
 
bigdrums925 10/4/2017 4:08 PM
My apologies for any misunderstandings. I don't think I am missing any posts since switching to linear mode. And I don't rely on email updates. I have this thread in my favorites and it takes me here every time. Also I am not ignoring instructions, rather occasionally am confused by what exactly I should do and in what order. In one post for example I was told it doesn't matter what measurements say with parts changed or removed, and to get it back the way it was with all the parts re-installed. Then when I read I should remove the lower transistors. What does "lower" mean? That particular instruction was a difficult one to interpret.
We recently have a mostly complete schematic and I would hope for clarity we could used the callouts i.e. Q10, Q13. Are those the 2 transistors lower means?
Again, I apologize for any misunderstanding, and I am aware some if not all of you guys are becoming frustrated with me. That is the last thing I want to happen.

If I remove the 2 lower transistors (Q10, Q13 I believe) I should be able to power up fully. Q10 was shorted and has been removed. I will remove Q13 as well and take the measurements. I was asked to take them at R36-R46 I think. For the record I started doing that last night and got 0 volts. Found the 2 PS fuses were blown. I replaced them this morning only to find the 1 shorted output Q10 which I removed. BTW in post #75 above I said that Q13 was shorted. I believe it was in fact Q10.

I was also asked to test Q8 and Q9. If ok test Q5, Q5, & Q7. If that is inconclusive, remove Q9, power up and measure voltages around Q5, Q6, Q7.

volts to ground with Q10 & Q13 removed:
Q5: e 45.6, b 45, c 0.5
Q6: c 0.5, b -1.1, e -1.8
Q7: c -1.8, b -44.8. ,e -45.3

r36: .53
r37: .53
r38: .53
r39: .53
r40: .53
r41: .53
r42: .53
r43: .53
r44: .53
r45: .53
r46: .53

I hope this is complete and what you were asking for. I will wait for further instructions before changing anything, and thank you all
 
bigdrums925 10/4/2017 11:56 PM
Quote Originally Posted by MarkusBass View Post
The latest schematic is here: http://music-electronics-forum.com/t45081/#post466975 . i thought that you fired up the amp without the output transistors (and then got smoke). You didn't tell us this .
What is the voltage on any of R36-46 resistors? Can you change it with bias trimmer?

Mark
Checked voltages on R41 to ground. Still 0.53v and the trim pot does not change that number.
 
nickb 10/5/2017 12:55 AM
Success! Those voltages are good. Q10 and Q13 are not fitted and that part of why you cannot bias it. The other reason is R11 is too high but we'll get to that.

Do not connect a load until we say to do so.

The next step is to fit Q10. Please note Mark's point about transistors being from the same batch. If you don't follow his advice there is a very good chance of blowing them up again on high output volume. With Q10 fitted you should be able to bring up to full voltage, but do so carefully. Measure across R38 and check you get no more than 30mV. If excessive back off the variac and check the voltages around Q6 again. More likely you won't be able to get to 30mV and will need to change R11 to 8.2k. Also check the output voltage is <1V. Now adjust RV1 to get close to zero on the output. Next adj RV2 to to get 25mV across any of the 1.3 ohms resistors.
 
bigdrums925 10/5/2017 2:19 AM
Quote Originally Posted by nickb View Post
Success! Those voltages are good. Q10 and Q13 are not fitted and that part of why you cannot bias it. The other reason is R11 is too high but we'll get to that.

Do not connect a load until we say to do so.

The next step is to fit Q10. Please note Mark's point about transistors being from the same batch. If you don't follow his advice there is a very good chance of blowing them up again on high output volume. With Q10 fitted you should be able to bring up to full voltage, but do so carefully. Measure across R38 and check you get no more than 30mV. If excessive back off the variac and check the voltages around Q6 again. More likely you won't be able to get to 30mV and will need to change R11 to 8.2k. Also check the output voltage is <1V. Now adjust RV1 to get close to zero on the output. Next adj RV2 to to get 25mV across any of the 1.3 ohms resistors.
Installed Q10 and ramped up variac. I know you know your stuff but is it really R11 that needs to change to 8.2k?
I ask because R26 has a side note of recommend 8.2k instead of 33k.
Problem: while checking for 30mV the second time both PS fuses blew and it shorted Q10 between emitter and collector. No significant current draw noted.
I very carefully ramped up the variac each time watching the amp meter, very little meter needle movement less than 0.2 amps.
VOM Meter read 0.2mV max
Looks like Q10 socket has a problem, each time one blows its in that position.
Compared Q10 & Q13 terminals (without transistors) with meter and they are the same
When should I put Q13 back in?

Awaiting your instructions....
 
nickb 10/5/2017 2:31 AM
Sorry it wa a typo, did mean R26 but we're not ready for that yet.

Leave Q10 and Q13 out, bring up variac and measure around Q8 qnd Q9. I suspect Q9 is bad. Also check R33 is not open as you did once mention it smoked.
 
bigdrums925 10/5/2017 3:18 AM
Q8 & Q9 both check ok and the voltages seem good, 46v other side -46v, 1.1v, -1.1v, 0.5v, -0.5v etc.
Every place I checked it's opposite was pretty much right there in negative.
R33 still checks 220 ohms.
 
nickb 10/5/2017 5:27 AM
Measure again but this time on the actual pads of Q10-Q13.

Also look again at R26. From your measurements you have 0.7v across the base and emitter of Q6 and 1.6 from base to collector. If both resistors are 33K and the trimmer is midway then you should 0.7v across both, ignoring any small base current. To get what you see, I'd expect R26 would to be around 10K. So something is wrong there.
 
bigdrums925 10/5/2017 10:42 AM
Q10 & Q13 collector - 46.1, emitter 0.5, base - 0.5

for what its worth voltage drop across R25 is 1.68 and R26 is 0.65

Yes 0.68 across base and emitter and 1.76 across base and collector on R26

both 33k resistors measure the same and I adjusted slightly trim pot to (equal resistance) center before taking above measurements
 
MarkusBass 10/5/2017 11:01 AM
Nick,

I also have simulation of this power amp. But I got different result regarding the 33k resistor changed to 8.2k. I wonder what is the reason of the discrepancy. One of the possibilities is that I use different transistors than you do. I use MPSA06 , MPSA56 and MJL21193/194. I simulated the quiescent current through one of the ballast resistors (so the current through output transistor will be twice higher) and here is what I got:
[ATTACH=CONFIG]45185[/ATTACH]
This is the resistor set as a parameter to the following list: 8.2k, 22k , 33k. With 8.2k I get the current from 140mA to 240mA. This is definitely to much. With 22k the current is from 2mA to 18mA. This one looks very good. And with 33k the current is exactly 0mA. It means that the amp works in class B and who knows - maybe it was designed in this way. Don't you think? I check also that in this case the output signal is correct.
To see it better I remover the 8.2k from the list (so now the graph is only for 22k and 33k) to see more details. This is what I got:
[ATTACH=CONFIG]45186[/ATTACH]
Looks very good. If Bigdrums claims that he didn't change the 33k, maybe this was in the original design of the amp? I would change only one resistor in the amp: 220 Ohms in the feedback loop (to e.g. 820 Ohms).

@Bigdrums: for me it is very difficult to advise you what you should do because you are not following what you are asked for. You showed desoldered power transistors so I thought that they are desoldered all the time. But then, after few failures of the amp, you say that one of the output transistors was shorted. So you soldered and desoldered the transistors several times and couldn't check that some of the transistors are shorted? . For me it's clear that you still have some transistors shorted and if I were you, I would try to power up the amp without the output transistors and verify whether it works (of course without any load).
You've got a great present - amp schematic, which is not available anywhere else in the world, you got simulations proving that the amp works correctly (if not the shorted transistors). Please use it wisely.

Mark
 
bigdrums925 10/5/2017 11:16 AM
I think R44,R45,R46, & R47 are tied to the power transistors output ahead of the 1.6 ohm resistor R48 and inductor L1 instead of after.
and that there is another (disk) cap labeled "683M" before the 0.22uF cap C9 going to "ground"
This "ground" is going directly to the collectors of Q11 & Q12.
Does this make any sense? That is to the voltage supply 46.6v
[ATTACH=CONFIG]45187[/ATTACH]

[ATTACH=CONFIG]45188[/ATTACH]
 
nickb 10/5/2017 12:11 PM
Quote Originally Posted by bigdrums925 View Post
Q10 & Q13 collector - 46.1, emitter 0.5, base - 0.5

for what its worth voltage drop across R25 is 1.68 and R26 is 0.65

Yes 0.68 across base and emitter and 1.76 across base and collector on R26

both 33k resistors measure the same and I adjusted slightly trim pot to (equal resistance) center before taking above measurements
Thanks - all the measurements are very helpful.

That could only happen if either the current through Q6 was high or the gain of Q6 was lower than I expect. From post #81 (I assume the supply is 46.6) the drop across R27 is 1v so the current is 1/220 = 4.5mA. Q6 should have an HFE of at least 120 so making the base current 4.5/120 = 37.5uA

Now the current through R25 is 1.7/34 = 50uA, much too close to the base current. This explains why Mark gets a different result from me in the simulation and thanks for the cross check.

In my sim the base current is just 4uA, I bet his is much higher. The basic intention of Q6 is to regulate the idle current with temperature and that is the reason Q6 is thermally coupled to the heatsink. With the values we have here the idle current is going to be affected by HFE variations with temperature (and ageing). We don't know what the designer had in mind. I believe that HFE rises with temperature so the overall effect of this current starved design is that as temperature rises the Vbe lowers and so the collector to emitter voltage follows, also the rising HFE will reduce that voltage even further. This is a good thing so long as there not too much of it and we get crossover distortion. This seems a really cheapskate way of achieving thermal regulation as it's so dependant on transistor characteristics. A better and commonly seen design adds a thermistor this part of the circuit.

I did mention much earlier that the reason for leaving the resistors at 33K was complex....

So, the conclusion is to leave things as they are with 33K resistors for now. Turn the trimmer all the way towards the R25 end and then measure the voltage from Q11 base to Q10 base with no load and no Q10 or Q13 fitted. Repeat with the trimmer all the other way.

Whether we need to change R26 to something smaller depends on the result of this test.
 
nickb 10/5/2017 12:20 PM
Quote Originally Posted by bigdrums925 View Post
I think R44,R45,R46, & R47 are tied to the power transistors output ahead of the 1.6 ohm resistor R48 and inductor L1 instead of after.
and that there is another (disk) cap labeled "683M" before the 0.22uF cap C9 going to "ground"
This "ground" is going directly to the collectors of Q11 & Q12.
Does this make any sense? That is to the voltage supply 46.6v
[ATTACH=CONFIG]45187[/ATTACH]

[ATTACH=CONFIG]45188[/ATTACH]
Noted
 
bigdrums925 10/5/2017 12:56 PM
Trimmer turned to R25, voltage between pins 1 (base) on Q11 & Q10 is 1.11v

Trimmer turned to R26, voltage between pins 1 (base) on Q10 & Q11 is 1.17v
 
nickb 10/5/2017 1:24 PM
Quote Originally Posted by bigdrums925 View Post
Trimmer turned to R25, voltage between pins 1 (base) on Q11 & Q10 is 1.11v

Trimmer turned to R26, voltage between pins 1 (base) on Q10 & Q11 is 1.17v
.

Not a great range and a bit high.
It could be that Q6 has gone low gain since this worked once so try another or use a bigger value for R26 say 39K.
 
bigdrums925 10/5/2017 2:10 PM
Don't have a BC182 replacement on hand so changed R26 to 39k

Trimmer turned to R25, voltage between pins 1 (base) on Q11 & Q10 is 1.01v

Trimmer turned to R26, voltage between pins 1 (base) on Q10 & Q11 is 1.07v

both these voltages climb slowly with time
 
nickb 10/5/2017 3:04 PM
Quote Originally Posted by bigdrums925 View Post
Don't have a BC182 replacement on hand so changed R26 to 39k

Trimmer turned to R25, voltage between pins 1 (base) on Q11 & Q10 is 1.01v

Trimmer turned to R26, voltage between pins 1 (base) on Q10 & Q11 is 1.07v

both these voltages climb slowly with time
Time to try a good Q10 in again. Set RV2 to the R25 end, bring the variac up carefully while monitoring the voltage across R36 and the variac current meter. If the R36 voltage gets to 35mV or you see the current surge, stop raising the voltage. Be careful things will happen very suddenly.

PS: I was wondering why Q10 blew earlier when the VOM reading never got over 0.2mV. I realised that we have a schematic but no component layout diagram. Because of that there could be confusion. So, it's best to be sure: It's very important that you monitor the current across the resistor that is connected to Q10 emitter.
 
bigdrums925 10/5/2017 10:38 PM
I have data sheet for the Motorola MJ11013 series and have verified that the resistors R36 & R38 are in fact tied to the emitter on Q10.
However with the VOM across R36 and slowly raising the variac, I see nothing on the VOM 00.0 on mV scale.
I too am concerned about blowing another Darlington, and more perplexing is it takes out the PS fuses as well.
This in spite of never seeing the current rise on the variac. I have 3 MJ11013's left and need 2 for this amp to work.
I raised the variac to 50% with no measureable voltage drop across R36, and less than .1 amps on the variac.
I suppose I need to raise the variac to full very slowly, but this is where I lost the Darlington's in the past.
Any suggestions before I proceed?
 
nickb 10/6/2017 1:00 AM
I share your worry BD. I really don't know why it blew the first time as the base voltage was not that that high. The current will rise fairly quickly but not that quick. You could replace R36 with something bigger, say 100 to 330 ohms while you bring it up an remove R38. The wattage doesn't really matter as it will just burn out if the current does suddenly rise and will give a measure of protection.
 
bigdrums925 10/6/2017 12:02 PM
Awesome, thanks. I'll try that when I get home tonight
 
bigdrums925 10/6/2017 8:17 PM
Alright then. I put a 100 ohm 2 watt resistor in R36, removed R38. Meter across the 100 ohm set on mV and ramped up the variac very slowly.
The voltage drop across the resistor slowly climbed to 15mV. Things look better as I was seeing no voltage across R36 last time.
Looks like the fuses held and so did Q10.

I'm guessing I now replace the 100 ohm with the 1.3 ohm resistors (one at a time) that were there before and try again
and if that works replace R13 and test for operation?

Awaiting your directive.....
 
bigdrums925 10/6/2017 11:03 PM
BTW I also checked the voltage drop across R37 & R39 but nothing.
Is that because of the lesser resistance?
Don't we need some voltage drop to indicate current in the transistor?
Anyway, seems like the amp is close to working now but I don't dare proceed on my own,
so please tell me what the next step should be without taking unnecessary risks.
Thanks again
 
nickb 10/7/2017 12:49 AM
OK. Change the R36 100 ohm to 10 ohm and try again. This resistor has two effects. It limits the current and also slows the rate at which the current rises. By now you should have a good feel of where on the variac you start to see the R36 voltage rise so you know to take is very slowly from that point. Leave R38 out.

PS: Did you change R23 currently 220 ohms to 22K? Do that now. If you don't the amp will probably oscillate.
 
MarkusBass 10/7/2017 2:30 AM
Bigdrums, I hope that you know that you can check the amp without assembling the output transistors (and without the risk of burning them again) just by using signal generator and oscilloscope? This would increase your confidence in the circuit. Do you have generator and oscilloscope?
It would be also nice if you tell us what resistors are now used. For example, what resistors are assembled in the bias circuit, 2x33k or something else. Also what resistors are assembled in the feedback loop, 22k + 220 Ohms, or something else? When you do the test are the output transistors attached to the heatsink, isolated from the heatsing and assembled using two screws (for each transistor)?
I found out the reason for the discrepancy between my simulation and Nick's simulation. It was my fault - I used output transistors that are not Darlingtons. Sorry for that. But it proves that the amp could work with standard transistors (just with one or two resistors changed).
So here are my findings again. Please ignore the previous one.
First I wanted to find out what it the reasonable value of the R26 resistor (in the bias circuit).
The following simulation assumes that the bias trimmer is in the middle and the R26 resistor is changed from 6k8 to 12k. The current through one of the output transistor is plotted.
[ATTACH=CONFIG]45210[/ATTACH]
You can see that useful range is from 8k2 to 12k. For 6k8 the current is very high and it would destroy the output transistors very quickly.
Now lets check the current range with 8k2 resistor (the bias trimmer is changed from 0% to 100%):
[ATTACH=CONFIG]45211[/ATTACH]
The maximum current is 130mA which is also to much (unless you'll be very careful when turning the bias trimmer).
The next simulation is for 9k1.
[ATTACH=CONFIG]45212[/ATTACH]
The current can be altered from 6mA to 30mA. This one looks quite good.
Here is another test for 10k.
[ATTACH=CONFIG]45213[/ATTACH]
This one is again a little bit better. The current can be changed for 6mA to 12mA. Please note that the current change is almost linear.
One more simulation - with 12k resistor.
[ATTACH=CONFIG]45214[/ATTACH]
The current can be changed from 3.6mA to 7.6mA. The current change is linear.
The last simulation shows what happens when I change the R26 resistor back to 33k (as it was originally):
[ATTACH=CONFIG]45215[/ATTACH]
The current is in a range of 50uA - very difficult to be measured. This is most probably incorrect value of the resistor, unless the amp was designed to work in class B.

So, as a conclusion, I would start with 10k as R26 resistor as see what it gives. Depending on the results, I would increase, or decrease it.

I would also considered the 220 Ohms resistor in the feedback loop. It decreases the feedback and increases the gain of the amp (to 39dB). This seems to much for me and could cause the amp to oscillate. I hope that the other resistor is 22k. Was this value originally in the amp? I would change to to at least 560 Ohms. But not now. First you have to be able to switch the amp on.

Mark
 
bigdrums925 10/7/2017 1:54 PM
Nick: I changed R23 to 22k.
Then changed R36 from 100 to 10 ohms. Left R38 out.
With the trim pot at the R25 end and ramped up the variac slowly.
With the meter across R36 got a reading of 2.2mV rising very slowly.
Since it was that low I tried turning the trim pot slightly and the mV went crazy so spray cleaned it and left it at R25 end.
Now what?
 
bigdrums925 10/7/2017 1:55 PM
Nick: I changed R23 to 22k.
Then changed R36 from 100 to 10 ohms. Left R38 out.
With the trim pot at the R25 end and ramped up the variac slowly.
With the meter across R36 got a reading of 2.2mV rising very slowly.
Since it was that low I tried turning the trim pot slightly and the mV went crazy so spray cleaned it and left it at R25 end.
Now what?
 
nickb 10/7/2017 2:16 PM
I did wonder if the trimmer might be a problem. It would certainly explain a thing or two. I should have tested it earlier - sorry.

Let's back up and re-install the 100 ohm then twiddle the trimmer slowly and see how that R36 voltage behaves. If it's not a steady ramp as you turn the trimmer now that you've cleaned it would be best to replace it. I've never had much luck with cleaning pots. Any improvement seems to last a day or too.

PS: Mark has been very busy looking at the feedback resistors incl R23. We'll circle back to that later. 22k should be conservative so leave that in for now.
 
bigdrums925 10/7/2017 5:08 PM
Mark:
Yes I have a signal generator and o-scope(s)
In the bias circuit R26 measures 38.5k but is marked orange blue orange gold. Perhaps I need to lift one end to measure properly.
It may be time for me to buy a Fluke VOM. Can you recommend which model Fluke I should be looking at?
R25 is 33k.
I'm not certain which is the feedback circuit, consisting of R23, R24? And these 2 are mounted on the board vertically on each side of trimmer RV2 in the picture?
R24 is 220 Ohms, R23 is 22k.

Currently Q11 & Q12 are installed and have been for a while. There are insulators under each of them. There is only 1 screw holding each down at this time.
Q10 is also installed, insulated from the heat sink, and has 1 screw per post 96.
Q13 is not installed at this point.


So should I change R26 from 39k to 10k now? if so what measurements on what terminals should I monitor? Across R36 in mV?
Note R38 is not installed, Nick & I have been adjusting R36 downward, currently 10 ohms.

Please be as simple and straightforward as possible as I just am able to follow what we are doing.

Awaiting your reply, and Thank you!
 
bigdrums925 10/7/2017 9:21 PM
Installed 100 ohm in R36. Trimmer is a little shaky, 15mV to about 30mV across R36 But it isn't smooth and has a dead spot I think where the voltages jump to 100mV briefly.
What can we do while I order and install the part?
 
MarkusBass 10/8/2017 1:37 AM
Bigdrums, it's clear that you have to replace the trimmer. You cannot do much if it's not working correctly. You can test it without the output transistors. Just by measuring the voltage between the slider and one of the other pins (in mV range). You shouldn't get any dead spots while doing this.
Quote Originally Posted by bigdrums925 View Post
Mark: Yes I have a signal generator and o-scope(s)
So why haven't you use them? You can test the amp without output transistors just by providing signal to the input and watching the output on oscilloscope. Is it clean and undistorted?
Quote Originally Posted by bigdrums925 View Post
In the bias circuit R26 measures 38.5k but is marked orange blue orange gold. Perhaps I need to lift one end to measure properly.
It looks like it is 36k. Don't change it now. I'm surprised with the high current you get. It is with 100 Ohms resistor, or 1.3 Ohms?
Quote Originally Posted by bigdrums925 View Post
It may be time for me to buy a Fluke VOM. Can you recommend which model Fluke I should be looking at?
DMMs are very cheap. You can buy them even for $10. What you need is to measure voltages in the range of 0-200mV. What are you using now?
Quote Originally Posted by bigdrums925 View Post
R25 is 33k.
It's OK - leave it as it is.
Quote Originally Posted by bigdrums925 View Post
I'm not certain which is the feedback circuit, consisting of R23, R24? And these 2 are mounted on the board vertically on each side of trimmer RV2 in the picture? R24 is 220 Ohms, R23 is 22k.
Yes, I'm taking about R23 and R24. They may be changed in the future but for the moment leave them as they are.
Quote Originally Posted by bigdrums925 View Post
Currently Q11 & Q12 are installed and have been for a while. There are insulators under each of them. There is only 1 screw holding each down at this time.
Q10 is also installed, insulated from the heat sink, and has 1 screw per post 96.
Q13 is not installed at this point.
One screw per transistor is a very bad idea. When the current starts ramping, the transistors will not be cooled properly. Also mounting two transistors from one side and one transistor for the other is a very bad idea. You should use 1+1 or 2+2.
Quote Originally Posted by bigdrums925 View Post
So should I change R26 from 39k to 10k now? if so what measurements on what terminals should I monitor? Across R36 in mV?
No, don't change it now since you get voltages on R36 in mV. Yes, you should monitor voltage on R36 and it should be in mV range (I think that less than 10mV).
Quote Originally Posted by bigdrums925 View Post
Note R38 is not installed, Nick & I have been adjusting R36 downward, currently 10 ohms.
I'm not sure what Nick advised you to do but I think that the output circuit has to be symmetric, meaning: just two output transistors Q10 and Q11, R38 and R39 removed, 10 Ohms resistors installed in place of R36 and R37.

Mark
 
bigdrums925 10/8/2017 3:31 AM
I removed the trim pot and cleaned it, seems to be working now.
Put it back in with 100 ohm in R36, climbs to about 17mV across R36, variac full on.

Now this is in the configuration Nick and I had it to solve the bias problem: Q11 & Q12 in place while only Q10 in place this side. No Q13

I was using the O-scope when I posted originally, that's how I knew I had oscillation.

My VOM is a Sun DMM-1230

Yes I was a little concerned about testing asymmetrically but hoping we would be able to put Q13 back in soon. But I can't afford to take chances at this point
so I'll wait to hear back with baby steps till we have this in the bag.

So what next?
 
MarkusBass 10/8/2017 6:10 AM
Quote Originally Posted by bigdrums925 View Post
I removed the trim pot and cleaned it, seems to be working now.
Hmm, "it seems" is not enough, you have to be 100% (or even more) sure that it will work now and in the nearest future. Otherwise, better order some new output transistors .
Quote Originally Posted by bigdrums925 View Post
Put it back in with 100 ohm in R36, climbs to about 17mV across R36, variac full on.
Now this is in the configuration Nick and I had it to solve the bias problem: Q11 & Q12 in place while only Q10 in place this side. No Q13
Bigdrums, I think that you are ignoring some little details and as a result you are not testing what Nick asked you to do. And these little details are the most important in your case. I don't think that Nick told you to leave Q10, Q11 and Q12 in the circuit and change only R36. What about R37, R39, R40 and R42? Are they still in circuit with 1.3 Ohms value?
Quote Originally Posted by bigdrums925 View Post
I was using the O-scope when I posted originally, that's how I knew I had oscillation.
My question is still valid: why don't you use it now when it seems that the amp has a big chance to be working correctly?
Quote Originally Posted by bigdrums925 View Post
Yes I was a little concerned about testing asymmetrically but hoping we would be able to put Q13 back in soon. But I can't afford to take chances at this point
so I'll wait to hear back with baby steps till we have this in the bag.
It's fine that you are concerned about it but why you are still doing it (testing the amp in incorrect way)?

Next steps depend on your answer how Q11 and Q12 are connected to the circuit. Also remember (I remind you for the second time) that assembling the power transistors with 1 screw only it extremely risky. I hope you understand it.

Mark
 
bigdrums925 10/8/2017 1:04 PM
I noticed while trying to clean the circuit board numerous times that it does not clean like others I done. I can dissolve the coating (flux and the like) but it does not clean up. I have a working microscope, the kind with some distance under the first optics for soldering and the like. I could see there was also this difficult coating on the trim pots electrical contact surface. Because this trim is an "open" type, I was able to clean it effectively. I then tested it with the meter and was satisfied it would perform reliably. Reinstalled, and was able to adjust smoothly from about 4mV to 17mV.

I'm quite sure said nothing about Q11, Q12. At the time there were only the MJ11014's installed and the 8ea 1.3 ohm resistors in place. He told me to put in Q10 and remove R38, and change R36 to either 100 or 330 ohms. Since the others were not mentioned, I did nothing with them. I am trying to follow the directions verbatim. I understand that we should be balancing both sides but there has been no sign of trouble on the MJ11014 side. However I mentioned I could measure no voltage drop across the resistors on that side either. Of course I am concerned that removing and reinstalling these parts is taking it's toll on the traces. I gather from you that I should balance both sides, either 2 or 4 MJ1101x's in place and remove the other resistors as well for further testing. I never know who will be responding next, and sometimes the directives from different tutors here don't necessarily agree, making it difficult to know exactly what I should be doing and not doing. R37, R39, R40 and R42 are still in the circuit as the 1.3 ohm values.

I have yet to have the amp working with a signal applied and a load since I first posted. I have already removed and reinstalled the MJ's a few times so I was only using the 1 screw to hold it in place before soldering. Haven't had a heat issue since posting.

So I haven't heard from Nick for a bit so what you want me to do as I understand it is to remove Q12 and the resistors for Q11 replacing them with 1ea 100 ohm resistor to make both sides even. Then put the 2ea screws in for Q10, Q11. Is that correct? If so, what shall I test for at that point?
 
MarkusBass 10/8/2017 1:13 PM
I thought that initially all four power transistors were removed (because you had a short circuit in the output section of the amp). Wasn't there a photo with removed transistors? Now, it seems to me that it was not the case. Let's wait for Nick. I don't want to interfere with his ideas how to proceed. I'm sure he will be here tomorrow in the morning.

Mark
 
nickb 10/8/2017 1:34 PM
I'm back! Sunday is my day off so we went out to enjoy a rare nice October day.

Our immediate aim here is to get the DC conditions right. I can see the merit in using the scope to check its not oscillating but only after we are sure that we are safe DC wise. Running asymmetrically is beneficial for now. First if it all goes horribly wrong you will only loose the one PNP transistor Q10, the second is as least twice the voltage is developed across the R36 emitter resistor making it a more sensitive measurement. It will do no harm. On the other hand you would not want to drive a load with too much power like this.

If you are 100% sure that the trimmer is now good and go back to where we were, i.e Q10, Q11, Q12, R37, R39, R40 and R42 fitted. Q13 out and R36 = 100 ohms. Set the trimmer to the R25 end again, monitor the R36 voltage and do the careful variac thing again. We should be able to get to 100% voltage without the current zooming off. Now check the operation of the trimmer. Report what range of voltages you get across R36 with the trimmer at extremes. If it gets over 1 volt something is not right.

BTW the 39K resistor will have a tolerance, probably 5% meaning about 37k minimum. Now add in you meter accuracy, often about 3% on the resistance range and you can see a measurement of 36K is easily possible.
 
MarkusBass 10/8/2017 1:52 PM
Great. This explains the procedure. If you manage to fire up the amp without the fire , I would also measure voltages on R37 and R40 (and report here what values you got).

Mark
 
bigdrums925 10/8/2017 5:08 PM
25-45 mV across R36 sweeping the trim pot, pot is linear as I sweep.

0.0 mV across both R37 & R40

What should be the next step?
 
nickb 10/9/2017 3:20 AM
Quote Originally Posted by bigdrums925 View Post
25-45 mV across R36 sweeping the trim pot, pot is linear as I sweep.

0.0 mV across both R37 & R40

What should be the next step?
The range is OK, perhaps a tad on the high side. I expect it will come down a little once we have a idle current flowing. Q6 needs to be in good thermal contact with the heatsink and all transistor screws fitted. Install Q13 and all the 1.3 ohm resistors again, set the trimmer to the R25 end and run up the variac. Set the trimmer for 25mV across R36 (it should be about the same on any of the 1.3 ohms). Let it warm for a few minutes and recheck.

Next check the DC voltage on the output and adj R1 to get close to zero. With no load, hook the scope to the output and look for any sign of oscillation. Now apply a 1Kz 200mVpp signal to the input and check the output you should be able to get about 82Vpp without clipping.

With no input, Recheck the 25mV idle and the output offset voltage. If all is OK you can connect the load and give it a try. Play for a bit to warm it up and check the idle once more.
 
bigdrums925 10/9/2017 2:40 PM
Well it is working hooray. The idle current is only about 0.6mV however
Tried it with a speaker load and a guitar plugged in and that works.
0.0 offset, nothing is getting hot and no oscillation.
1kZ 200mVP-P renders a max of 53 volt output
What do you think about the idle current being so low?
 
nickb 10/9/2017 3:54 PM
I have a strong suspicion that the trimmer was the problem all along. With only 0.6mV I think you will be able see crossover distortion on the output. Now, please....don't hit me... change the R26 39k back to 33k and try once more.

200mVpp gives 53V(pp?) under what settings of the controls?
 
bigdrums925 10/9/2017 6:18 PM
Changed the R26 to 33k but now only get a 0.0 reading across R36-R43
Haven't tried all the controls yet but both gain and master volumes all the way up.
 
nickb 10/10/2017 6:22 AM
R26 went from 39k to 33k and the volatge went from 0.6mV to 0? That's crazy. Measure the voltages around Q6 and check the value of R26.
 
MarkusBass 10/10/2017 6:47 AM
I think it's (almost) correct. The fact that with 100 Ohms as R26 you get 25mV doesn't prove that with 1.3 Ohm you should be able to get 25mV as well. We discussed this quite a long time. Could it be that the drop from 0.6mV to 0.0mV is caused by meter accuracy?
I would solder another 33k resistor on top of R26 (to get 16k as R26). The simulation showed that the value of R26 should be as low as 12k. Let's try 16k for now. Unless, 33k is not soldered correctly now .

Mark
 
nickb 10/10/2017 7:13 AM
Quote Originally Posted by MarkusBass View Post
I think it's (almost) correct. The fact that with 100 Ohms as R26 you get 25mV doesn't prove that with 1.3 Ohm you should be able to get 25mV as well. We discussed this quite a long time. Could it be that the drop from 0.6mV to 0.0mV is caused by meter accuracy?
I would solder another 33k resistor on top of R26 (to get 16k as R26). The simulation showed that the value of R26 should be as low as 12k. Let's try 16k for now. Unless, 33k is not soldered correctly now .
Mark
Note of caution: If you do this please raise the variac very slowy in case the iddle curent is too high. I couldn't bear to see a grown man cry
 
MarkusBass 10/10/2017 7:33 AM
Another note of caution: you wanted to have DC conditions of the amp correct. But now I see that Bigdrums provides input signal and get full power out of the amp. Shouldn't he wait for all DC voltages be correct (including voltage on R26)? Unless, we agree that the amp works in class B and 0mV on R26 is OK.

Mark
 
dmeek 10/10/2017 7:46 AM
I suggest a slight change to the bias circuit. This way if the trimmer fails the idle current goes to minimum.


http://music-electronics-forum.com/a...1&d=1507643168
 
bigdrums925 10/10/2017 12:28 PM
OK I put another 33K on top of R26 and touched up 1 solder joint that looked bad
Now the range of the trimmer is 3.3mV to 5.6 mV
The trimmer seems solid per watching the VOM

dmmek: your suggestion is noted and I think I should wait for nick and markus to tell me when. Thank you

I just plugged in a guitar and speaker and the amp sounds as good as I might hope for. Both gain and master are at 1/2.

Awaiting instructions
 
MarkusBass 10/10/2017 12:50 PM
This is the question to Nick. How much current would you like to have? For the start I wouldn't go higher that 10-15mA. What do you think?
Do you remember that my simulation showed that value 12k seems to be a good value for R26? Now we have 16.5k. Should we change it to 15k, or 12k? (But remember that with 12k you shouldn't check the maximum current because it will be 30mA, or more).
Dmeek's modification is excellent but maybe we could wait and check what we get with the 15k resistor?
Bigdrums, I assume that now we have all power transistors soldered, all 1.3 Ohm resistors soldered and you are trying to set the bias of the amp (its quiescent current), right? The power transistors are attached firmly to the heatsink and variac is fully on. yes?

Mark
 
nickb 10/10/2017 1:07 PM
Doug's idea is a good one. Go for it!

We could probably go lower for R26 yet. Before we do that feed a 1Khz sine wave in and set the controls for say 10Vpp with a load connected ( got a dummy load?) and look for any signs of cross over distortion. If it looks clean and it sounds good to you I'd be tempted to leave things as they are. Otherwise, try a small value for R26, say 15k and try again. Remember to always set the trimmer to the R25 end to start with.

Mark: The magic number is around 26mV as at this voltage at which the output impedance of the transistor is equal to the 0.65 ohm emitter resistor. But it is just a guide and without any other information I'd choose it as a do not exceed value. Anything less than that where (a) there is no crossover distortion and (2) the idle is current stable when the amp is hot works for me.
 
MarkusBass 10/10/2017 1:22 PM
Currently, Bigdrums has 33k resistor soldered on top of another 33k (it gives 16.5k). I'm sure this does not look nice and I would replace R26 with a single 15k resistor.
Thanks for explanations regarding the current.
 
bigdrums925 10/10/2017 3:08 PM
I put a 15K in R26 and dissipation is now 7.1 mV max until I put a signal through it.

Hard to check for crossover distortion as my scope gets fuzzy as I turn up the volume but from what I can see it's fine

It works well with the guitar and speaker hooked up

I think we should call it good.

Should I add Doug's jumper?
 
J M Fahey 10/10/2017 6:11 PM
I think if you can play normally, plus you can play LOUD for an hour and amp does not self destroy, then better stop repairing it and put the cover back
 
bigdrums925 10/10/2017 11:05 PM
Thanks to all that contributed to the successful outcome of this restoration, I can't thank you all enough.
There is a bit of buzz affected by the master volume only and I need to get the reverb working again
But without this website and the tireless efforts by all, this amp would have probably never worked again.
You guys are awesome!

I will graciously attempt to make the last few changes to the schematic for anyone that may need it.
 
MarkusBass 10/11/2017 12:32 AM
There were still some doubts regarding the feedback resistor (220 Ohms) but if the amp works OK, I would leave it as it is. Happy playing.
What about a final photo of the assembled amp? Just for goodbye .

Mark
 
nickb 10/11/2017 12:49 AM
Here is the final power amp schematic.
 
bigdrums925 10/11/2017 2:27 PM
Mark: Absolutely, be happy to post pictures, give me a day or 2 as I put the unit back in the combo box but will have it out soon.

Nick: Thanks so much for the updated schematic. I think I unintentionally made things difficult for you guys, but we got through it! I am so happy about it.

How can I ever repay?
 
nickb 10/11/2017 3:16 PM
Your show of gratitude is appreciated. It's more than many do.

To my mind, this is the major part of what MEF is about, a few guys working together to help another out. Sometimes it's repairs like this, sometime it's theory...
 
bigdrums925 10/11/2017 6:10 PM
Here's pics of the working amp
[ATTACH=CONFIG]45275[/ATTACH]
[ATTACH=CONFIG]45276[/ATTACH]
[ATTACH=CONFIG]45277[/ATTACH]
[ATTACH=CONFIG]45278[/ATTACH]
[ATTACH=CONFIG]45279[/ATTACH]
[ATTACH=CONFIG]45281[/ATTACH]
[ATTACH=CONFIG]45280[/ATTACH]
 
dmeek 10/12/2017 10:57 AM
I could be wrong but R33 220 ohms looks a little burnt. Would be better to put a 1W resistor there.