Sea Chief 6/5/2018 2:24 AM
Fender Twin UL > AB763 (re) build.
Howdy folks- hope all UK guys enjoying our fab weather! like hawaii here y'day.. but w'out lava in the garden.

I've rebuilt my UL TR (runaway bias issue/ conductive board diagnosed- redo needed) going the AB763 route inc a choke etc, as advised by Rift. All went well, everything checked 3x (as its a biggie) > comprehensive tests all going well..

..up to final checks: 1stly I wasn't getting anything from the spkr: no hum, hiss, zilch but dead silence. Then, once told to turn vol pot up (ch1) after treb & mid maxed too.. the (3A) fuse popped.

The fuse started crackling (for 1 sec b4 failing) -concurrently- with my nudging the vol pot up a mite, suggesting to the layman here of a clear cause/ result, ie the two linked (it -felt- like I caused it via the pot nudge). But told not so (by Rift, going thru tests on phone at time) only a coincidence the low amps @ vol pot just couldn't be the cause.

Anyway.. I'm back checking no4 time as advised. Thoughts if you would please. Thanks, SC
 
pdf64 6/5/2018 2:38 AM
I suggest that it's a bad idea to subject those expensive components to fault current, and to avoid that, you build and use a light bulb limiter before you do anything else.
 
Sea Chief 6/5/2018 2:48 AM
Hi pdf64,

I used a CL- I wouldn't have dreamt of doing without one tbh, esp such a comprehensive & fairly complicated re-build.
 
pdf64 6/5/2018 4:25 AM
Quote Originally Posted by Sea Chief View Post
...I used a CL...
What exactly is meant by 'CL' (current limiter?)?
Whatever, the fuse blowing indicates that a light bulb limiter would be better.
 
J M Fahey 6/5/2018 4:52 AM
1stly I wasn't getting anything from the spkr: no hum, hiss, zilch but dead silence. Then, once told to turn vol pot up (ch1) after treb & mid maxed too.. the (3A) fuse popped.
This hints at shorted speaker out or terrible OT miswiring.
 
Sea Chief 6/5/2018 5:28 AM
pdf.. I used a light bulb Current Limiter.
 
pdf64 6/5/2018 6:18 AM
Yikes, what wattage bulb is in it?
 
Sea Chief 6/5/2018 8:05 AM
Hi Pdf (site went down for a while there..)

60w. No the fuse didn't pop with the CL.. by this time I'd done the checks enough to plug straight into mains.

So godamn frustrating.. I bought the amp nearing a year ago now, been at fixers for 9 months, ccame back with chassis ruined in transit so I had to strip then hammer/ bend for a wknd/ then me rebuilding for last month.. all finially rebuilt, tests all going good, then @ last leg of testing. pop. urghhh.
 
Sea Chief 6/12/2018 10:07 AM
Hi out there- can anyone help me out here? Fuse issue seems under control (scrubbed 9-pin sockets, odd board resolder spot.. may have been it). Not one blown since.

I tested v1-6, & a pair of 6L6's in v7 / v10 (not yet 4 together). They started to redplate, not cherries/ more subtly up a side of part of the tube plate. The bias pot is in the midway position. On for 5 mins max when I realised- hope tubes ok.

Chris has suggested I might 'just run them colder, maybe'.. presumably that means just turn the bias pot down? But are there any drawbacks to turning it to min- doesn't that in turn lead to stress somewhere else? may have got that wrong.

Got gtr sound for 1st time, during 6L6 test! woohoo ~(2 tubes/ one 8ohm spkr). Bad reverb buzz thing- but cross that bridge later.

Thanks. SC
 
pdf64 6/12/2018 11:44 AM
If your overall bias is adjustable then adjust it to suit the tubes that you've got in there, ie to idle at about 30mA plate or cathode current each.
There's no point in having it adjustable if it's just left in the middle
Care needs to be taken regarding terminology, eg does 'turning the bias down' mean that the control grid voltage gets more or less negative?
I suggest to think of things in terms of results, ie adjust bias to achieve hotter or colder idle point.
If the plates are even mildly red then they are way too hot.
 
Sea Chief 6/12/2018 11:57 AM
Understood pdf. When I say turn bias down I mean less negative voltage/ a lower numerical figure/ cooler tubes.

Ok I just plucked up courage to check current (at my 1.5r resistor, V7), in an attempt to turn down this pot/ cooler tubes as Chris alluded to. I switched on & saw 85mA! so quickley dialed down the pot to get me to 35mA. Great > played a bit (gtr in for test, trem ON) & heard good tones.. & trem worked ~ok. 1st time I'ver got to hear amp, properly, since I bought almost a year ago!

But hang on.. noticed the current creeping up, very slowly, to 45mA.. then I switched off.
 
Justin Thomas 6/12/2018 12:10 PM
Sea Chief,

A lower NUMERICAL number is what RAISES the idle current. You want to INCREASE the numerical value to lower the idle current.

I like to think of it as a number line, with positive on the right and negative on the left. Say you go from -100V to +100V with Zero in the middle. Even though you are making a "larger" numerical value in the negative direction, -50V is a "greater" value than -100V.

Bias is screwy because we almost never think in terms of "negative" voltage. But you want to turn your bias pot the opposite direction - you want your pot at a LOWER resistance value ("down") because you want a "larger" (numerical) value of voltage, which in actuality is a SMALLER amount than what you started with.

Remember, to turn your bias pot (resistance) up means to lower the numerical value of your voltage, which makes your idle current go up, which makes you tubes redplate. And since it's thoroughly possible to wire a bias pot to work in either direction because they are linear taper pots, then don't worry about which physical direction the pot is going. Just turn it the way that gives you a larger NUMERICAL value of voltage (which in actuality is a lower quantity of voltage, because you're moving in the left-hand direction on a number line.)

Confused yet? :P

Justin
 
pdf64 6/12/2018 12:49 PM
Quote Originally Posted by Sea Chief View Post
Understood pdf. When I say turn bias down I mean less negative voltage/ a lower numerical figure/ cooler tubes...
Less negative voltage (ie a lower magnitude) at the control grids would cause the power tubes to run hotter (ie more cathode current).

Quote Originally Posted by Sea Chief View Post
...Ok I just plucked up courage to check current (at my 1.5r resistor, V7), in an attempt to turn down this pot/ cooler tubes as Chris alluded to. I switched on & saw 85mA! so quickley dialed down the pot to get me to 35mA. Great > played a bit (gtr in for test, trem ON) & heard good tones.. & trem worked ~ok. 1st time I'ver got to hear amp, properly, since I bought almost a year ago!

But hang on.. noticed the current creeping up, very slowly, to 45mA.. then I switched off.
1.5 ohms is a weird value? 1 ohm is more usual. A 1 ohm resistor will tend to measure ~1.5 ohms unless a special low ohm meter is used (they use 4 test leads).

Regarding the upwards creeping current, it would be useful to use 2 meters, one for the cathode current, the other for the bias.
Are the 1k5 power tube control grid stoppers new?
 
Sea Chief 6/12/2018 1:04 PM
Ok sure its a 1r then.. whatever Chris put in there. V7-V10 completely new sockets & components, again by Rift, including the bias 0.1 caps on the board.

Ive been reading that the current isnt best read asap after amp switch on/ cold tubes.. is that right? I read the 85mA fig only a few secs after amp on > then dialed down the bias pot a few sec later: so I read 35mA prolly only 20 secs after tubes stone cold.

Alas only 1 meter pdf. thx SC
 
Sea Chief 6/13/2018 2:37 AM
Can anyone help me out here?

I tested current at the next 6L6, v8.. & get the same thing happen/ steady @ 35mA for a few mins, then creeps up. More slowly though than v7.. say 10 mins to get to 45mA, rather than 5 mins for v7.

Does this suggest my innitial problem, ie it had "runaway bias" (which I rebuilt the whole amp to fix).. hasn't been fixed/ remains??
 
Sea Chief 6/13/2018 4:20 AM
I just tested vdc at pin 3 of v7 to ground.. & it went up to, & flew around at between 680v- 750v.

WTF?
 
pdf64 6/13/2018 4:31 AM
Hanging additional circuits off a power tube plate can facilitate oscillation.
Why test the plate's Vdc? Just measure the HT, the difference in Vdc between them at idle is negligible, and as you may be seeing, probing the plate can disrupt idle.
How about setting the bias for a much cooler idle and monitoring cathode currents for 20 mins, to see if they stabilise at a reasonable level?
Perhaps remove the phase splitter tube in V6 to eliminate the possibility of oscillation.
If the cathode currents rise too high, shut it down.
 
Sea Chief 6/13/2018 4:45 AM
Additional circuits- sorry I don't understand.

Why test vdc? I thought this was inherrantly done in any amp build.. & also re. the dissipation fig needed, so B+ divided by cathode current.. I thought all that was entirely relevant to what I'm doing/ current/ bias issues? I don't understand the Q.

Why wasn't I seeing a steady 495v though? was my fluctuating 680-750v to be expected, considering what I was doing (& not symptomatic neccessarily.. of an imminent series of filter cap explosions?).. if so.. what was I doing wrong??
 
pdf64 6/13/2018 7:54 AM
Quote Originally Posted by Sea Chief View Post
Additional circuits- sorry I don't understand.
To be specific, the probes, leads and meter are the additional circuit which you are connecting to the power tube plate when taking a Vdc reading there.
That node is a powerful transmitter; the meter etc that you've connected there can now act as its antennae. The signal from it gets picked up by the sensitive 'receiver' grids of previous stages, creating both negative and positive feedback loops.
The plate Vdc may be reading high because there's a massive amount of Vac, due to oscillation, also there, which may cause your meter to give an erroneous Vdc reading.
Hence I suggest to use the HT voltage, rather than the plate voltage, for the dissipation calculation. Or break the feedback loop by removing V6.
But I suggest that you just bias them each to ~30mA cathode current, dissipation will be fine, the particular value is immaterial for guitar amps as long as it isn't excessive, and 30mA won't be excessive, plus you'll avoid digging another hole of confusion, misunderstanding and general misery here.
 
Sea Chief 6/13/2018 10:50 AM
Ah I see.. so I got the place to read B+ wrong then?! my oh my.. well its been over a year since I did this in any other amp/ nr a year this one's been at the doctors- Im very rusty then.

Right so please remind me where to read the B+, with all the tubes in place. I understand I can take a reading across the 2 red wire HT where they go into the bias board.. but this isn't reading the B+ at the business place/ the tubes tho is it?

IE don't I need to read each individual tubes' B+ at a socket pin?
 
Sea Chief 6/13/2018 12:51 PM
Quote Originally Posted by pdf64 View Post
To be specific, the probes, leads and meter are the additional circuit which you are connecting to the power tube plate when taking a Vdc reading there.
That node is a powerful transmitter; the meter etc that you've connected there can now act as its antennae. The signal from it gets picked up by the sensitive 'receiver' grids of previous stages, creating both negative and positive feedback loops.
The plate Vdc may be reading high because there's a massive amount of Vac, due to oscillation, also there, which may cause your meter to give an erroneous Vdc reading.
Hence I suggest to use the HT voltage, rather than the plate voltage, for the dissipation calculation. Or break the feedback loop by removing V6.
But I suggest that you just bias them each to ~30mA cathode current, dissipation will be fine, the particular value is immaterial for guitar amps as long as it isn't excessive, and 30mA won't be excessive, plus you'll avoid digging another hole of confusion, misunderstanding and general misery here.
pdf I'm lost. Ive just gone to youtube to refresh myself where to probe for b+, to get my approx 495v figure (I measured this before.. but cannot remember where). The amp tech did -exactly- what I did, black probe to chassis > red probe to the plate/ pin 3 of the power tube (el34 in his case). 495v steady.

Please remind me.. why was mine 680v-750v & fluctuating wildly? (is it bc I had a guitar plugged in??)
 
Dave H 6/13/2018 1:17 PM
Sometimes (but not always) the amp will oscillate when a meter probe is connected to the plate of a power tube giving wild meter readings. If your amp is as the layout below and you really want to know the B+ voltage you can measure it at the top right of the diode/bias board where the two strings of diodes connect together but I'd do as Pete said just set the bias current to 30mA. Check and reset it back to 30mA after 10-15 min and call it done. It will wander around a little with the mains voltage after that but it shouldn't go up to 45mA unless there's a fault.

[ATTACH=CONFIG]49296[/ATTACH]
 
Sea Chief 6/13/2018 1:22 PM
pdf help me out here would you. I dont understand a word of your post #18, now that Ive seen eg after eg on youtube & in discussion forums answered by amp techs.. of the -exact- thing I have done to measure plate voltage @ pin 3 of the power tube. Here for eg:

[a Q on a forum like this, he is talking about the power tube]. Q: "I definitely want to avoid getting electrocuted. I am only planning on using the multimeter touching off at pin 3 to ground, with speaker plugged in for sure. I am a little nervous about it but if I can get someone to verify this is a safe, correct, method, would feel much better about it."


A: "I can tell you that I used to run my own amp repair business in Austin, and this is how I did it. Just make sure the chassis is stable and you should be good to go. On Fenders, if I was just doing a quick bias adjustment, I used to put the chassis upside down on the top of the amp. But depending on the size of the transformers, you may have to put something underneath one side to balance it."

And a Youtube clip, another amp tech @ 1.40 https://www.youtube.com/watch?v=B46w_B_Aoq8
plenty of eg's online concur with what I did.. but you allude to it being wrong for xyz reasons.

I am totally bewildered.
 
Sea Chief 6/13/2018 1:34 PM
Hi Daveh. I thought it was inherrantly page 1 stuff to need to know the B+. Its the one thing Ive known from any amp built, the fundamental Q, is what this important figure is. So I dont understand why you say 'if you really want to know it..'. You HAVE to know it in order to find out x, y, z (dissipation/ aiming for that 60-70% area is inherrantly important> you need to know the B+ to then divide * etc.. ie whether the amp's running at the correct voltage, for the tubes etc etc/ primary stuff- surely?).

What the heck am I missing here? its like the carpet's been pulled from under me.
 
Sea Chief 6/13/2018 1:38 PM
I don't have a string of diodes, I have a bridge rectifier so Ive no idea where I masure here.

Let me go back a step. Is it correct that in order to guage if the amp is working as it should.. you should find out what the plate voltage is, at each of the 4 power tubes-? is that correct?
 
Dave H 6/13/2018 2:12 PM
Quote Originally Posted by Sea Chief View Post
I don't have a string of diodes, I have a bridge rectifier so Ive no idea where I masure here.

Let me go back a step. Is it correct that in order to guage if the amp is working as it should.. you should find out what the plate voltage is, at each of the 4 power tubes-? is that correct?
You can measure the B+ at the + terminal of the bridge then. The plate voltage of the power tubes will only be a few volts less than that (because of the voltage drop across the OT primary).

We are recommending that the bias current is set at 30mA because at that value the tube dissipation is sure to be OK even if the B+ is over 500V.
 
Sea Chief 6/13/2018 3:20 PM
Ok Dave & pdf.. Ive just taken your word for it, & I'm forgetting all that voltage faff for now (Ive got alot of + terminals at the bridge.. so bllx to the b+ I say). I turned on & it went up to 45mA, so I set it to 30mA/ it went up to 40mA so I backed the pot back again (prolly at close to min now) to 30 again.. & monitoring it (& playing- proper 1st go woo bloody hoo) for an hour: & it stays there~therabouts! actually 27.9 when I turned off.

So, is that creep 'issue' resolved? & is it deterimental to have it so low at 27.5 mA at all? if it stays there that is I guess.
 
Sea Chief 6/13/2018 3:25 PM
And bar it being very groundy-noisy when Im not touching strings.. it sounds bloody great too.

Better now than a UL SF twin- no doubt about it.

What can I do re. the groundy noise.. is that the gtr or amp? cant say like so with my 5f1 so gotta be amp?

cheers chaps SC
 
g1 6/13/2018 9:00 PM
If the noise is only when you are not touching the strings then I think that is fairly normal. At least I get that, others may have different experience but I would think it would depend on shielding in guitar, room environment (emf) etc.

For the B+/plate voltage issue. These guys were not saying you don't need it, they were saying you don't have to check it directly at the plate (pin3). Sometimes this will create squealing/oscillation and weird readings like you experienced. When you have a problem measuring it directly at pin3, you can instead measure the B+ going in to the OT. This will be very close to the same as pin3 voltage, but will get around any problems (oscillation) induced by the test equipment.
In most old style fender amps, you can measure the B+ at the standby switch.
 
Sea Chief 6/14/2018 2:23 AM
Ok thanks guys. Yes standby switch I seem to recall I measured xyz before.. must have been that.

Well all back together for 1st time, 2 playing sessions now: bar the normal buzz/ bit of motorboating/ noise etc minor stuff I'll sort w'another thread maybe.. amp sounds really great. Like a new amp, less hard than a UL & far more complexity- huge result! the only thing I can check with amp together & in daylight- is the heat from tubes: nothing hot or OTT both sessions/ amp only just warm.

Ok so without being able to measure the bias.. having got success with keeping it steady at 28.5mA say (ie between 27- 30) in my last test.. is this now safe to say it'll not do the creeping up 35>45>.. thing again? or should I be monitoring it on a bench for days to make sure?
 
J M Fahey 6/14/2018 2:58 AM
Quote Originally Posted by Sea Chief View Post
Ok Dave & pdf.. Ive just taken your word for it, & I'm forgetting all that voltage faff for now (Ive got alot of + terminals at the bridge.. so bllx to the b+ I say). I turned on & it went up to 45mA, so I set it to 30mA/ it went up to 40mA so I backed the pot back again (prolly at close to min now) to 30 again.. & monitoring it (& playing- proper 1st go woo bloody hoo) for an hour: & it stays there~therabouts! actually 27.9 when I turned off.

So, is that creep 'issue' resolved? & is it deterimental to have it so low at 27.5 mA at all? if it stays there that is I guess.
Congratulations, you are doing fine.

The main point is to be able to set power tube bias :

a) at some reasonable current which does not cause buzzy crossover distortion.
You are there

b) that current must be **stable** ... if it uncontrollably runs up it will soon reach damaging values.
You finally reached that too, so ...

c) Enzoism follows: once you have fixed it, stop "fixing"

3) As of:
Ok so without being able to measure the bias.. having got success with keeping it steady at 28.5mA say (ie between 27- 30) in my last test.. is this now safe to say it'll not do the creeping up 35>45>.. thing again? or should I be monitoring it on a bench for days to make sure?
You *are* measuring it .
You have *already* checked its stable.
Again: once you have fixed it, stop "fixing" meaning: close it and go play some Blues (or whatever).
 
pdf64 6/14/2018 3:11 AM
It's worth bearing in mind that the whole chassis will tend to run somewhat hotter when buttoned up inside the cab, than when it's out on the test bench, and that heat can cause problems to manifest.
And that the AB763 mod will put more stress on the power tubes than the original 135W circuit.
Good tough power tubes are needed in there; if they're gassy (eg due to previous redplating), grid current can rise, and end up in them being prone to redplating.
The proof of the pudding is in the eating.
 
olddawg 6/14/2018 1:20 PM
Quote Originally Posted by pdf64 View Post
It's worth bearing in mind that the whole chassis will tend to run somewhat hotter when buttoned up inside the cab, than when it's out on the test bench, and that heat can cause problems to manifest.
And that the AB763 mod will put more stress on the power tubes than the original 135W circuit.
Good tough power tubes are needed in there; if they're gassy (eg due to previous redplating), grid current can rise, and end up in them being prone to redplating.
The proof of the pudding is in the eating.
Also... unless you are a purist... there is nothing wrong with including a whisper fan in the cab. Many times Ive had to put a house fan behind the occasional amp on stage...