|shzmm||2/8/2018 11:06 AM|
|Ampeg SVT-II no pro no go|
after my last adventure into SVT land, I am back with an older re-issue model-- the SVT II (2) non pro model. This one appears to be from the 90's and is in bad condition. There have been a number of PCB repairs and the like. The main problem after having to replace the fan, replace one of the 12AU7s, and so on is that the signal is severely limited and distorted-- it appears to be almost square and only achieves a few volts output.
One issue is that this amp is wired for 240V and we have 220VAC here. That leaves the heater voltage around 5V, however I can boost the mains voltage up to 240 and it comes up to a respectable 6V so I don't think there's a problem there per se. In fact, all voltages check out ok except it isn't possible to draw enough current through the output tubes. Although I can adjust the bias voltage quite low, the recommended .072V reading at the output tubes cathodes can't be reached. I get at best around .035 and there is still no change in the output waveform which stays square and very weak. I have also tried another set of tubes with no difference.
The two 12AU7s pass a clean sine to the output tubes so I tested the output transformer for a short but that also tested out ok. My best guess is still that there is a trouble with the output transformer but normally a test of a short will demonstrate this.
Any other ideas of something to check before this tired old beast becomes a door stopper? As well, if anyone has a wiring diagram for the power transformer leads please pass that on!
|Tom Phillips||2/8/2018 12:40 PM|
|g1||2/8/2018 12:56 PM|
|shzmm||2/8/2018 2:12 PM|
|Thank you for the reply. |
All 6 plates are sitting at around 610 and screen voltage is around 320-- the voltages are a bit down from the norm due to the mains voltage difference.
Otherwise, I haven't noticed anything unusual in the amp. All the other voltages are in line with the schematic.
|g1||2/8/2018 3:32 PM|
|I think for your testing you should use the 240V you said you can, I assume you have a variac.|
At 5V heaters you could get weird issues.
So at 240V you still can't get beyond .035 at the bias test points? What is the DC on your power tube grids?
|shzmm||2/9/2018 4:07 AM|
|The best I get is .045 at the higher mains voltage... Grid voltage is then around -35V |
Thanks for the help
|Chuck H||2/9/2018 7:24 AM|
|There's something hinky with the filament supply. I've never seen an amp that had low filament voltage from the factory (not saying it doesn't happen, but I haven't seen it). Usually the filament voltage is a little high. So if the amp has 6.3V at 240V from the mains that SHOULD math out to 5.77V at 220V. That's still in spec. Further, if you're getting 5V at 220V you should be getting 5.44V at 240V. Both below spec and not up to 6V which you report. The filament voltages don't math out, why? |
Regarding the bias voltage, have you measured it at the supply with 240V mains voltage? It should be 180V as spec'd on the schematic. If it is at least close then the low bias voltage condition at the grids is occurring on the grid side of the 47k grid load resistors.
If I'm interpreting the schematic correctly there's a ribbon cable for many of those board connections. Including the OT connection to the output jack. I'd be especially critical of those connections and the condition of the board where the pins contact pads.
|shzmm||2/9/2018 10:37 AM|
Bias voltage is at spec of -180 so I don't think that is a problem. I am not sure what you mean however with a "low bias voltage" The bias voltage can be adjusted from around -58 to -35V. I have also checked all the 47k grid resistors and they appear o.k.
There are no ribbon connections on the amp but it is a good point to make as these amps have a number of connections that need to be pulled and reconnected each time when removing the main PCB. I have cleaned and reseated all of these.
Thanks for your help, hopefully we can figure something out...
|Chuck H||2/9/2018 11:54 AM|
|With other voltages being on spec and the filament voltage being low even with no tubes in the amp I have to suspect the filament circuit is compromised. Math says 6.54V at 240V into the primary with no tubes. |
Did you pull the preamp tubes as well as the power tubes when checking the no load voltage on the filament winding?
Have you checked the hum balance circuit for shorts?
You mention that you have clean sine out of the PI. At what voltage? Was it sufficient to drive the power tubes? Is it balanced?
Have you tested for continuity between the actual output transformer secondary lead and the output jack to be sure nothing is open or shorted between those two points?
|g1||2/9/2018 11:57 AM|
|I'm very suspicious of your meter. What kind is it?|
Schematic shows around -47V bias at the grids. You have -35V there which should make the tubes run quite hot. Yet you can't get more than .035V at the test points which would mean the bias is too cold.
It doesn't make sense, something is not right with the measurements.
Check that you have not shorts at R35/D20 or R36/D21. They should all measure about 1 ohm.
|Chuck H||2/9/2018 12:05 PM|
|nevetslab||2/9/2018 12:44 PM|
|I have additional schematics for this SVT II. I looked at the Universal Primary, and it only has 100V/120V taps on the two primary coils. But, with them in series, yhou should be able to wire it for 220VAC (100V Primary in series with the 120V Primary> I've attached the schematics below.|
The Output stage is much like the SVT-VR, in it's use of a common cathode resistor for all three tubes in each half. Have you tried using one pair of tubes at a time? I'd do that, so you can measure what each tube's plate/screen current is. It could be these tubes are tired, and can't get full current, even with the bias voltage applied set hot. I've seen that happen over the years with all the SVT's I've serviced.
Now, in the filament end of things, my schematic only shows the High Voltage transformer having the universal primary. It's possible the Filament transformer also has one.. I don't have the color code for that. It might just be a 2-coil/4-wire primary (120V/240V) I believe we have just one SVT II here in our inventory. It's been a few years since I've been inside it, but could pull it out to have a look.
"m not surprised at the cage in heater voltage betwen On Power Tubes and all six power tubes installed. They pull a good amount of current
|shzmm||2/9/2018 1:40 PM|
were defective, but everything remained the same.
|shzmm||2/9/2018 1:47 PM|
I will also look into the wave form and voltage going to the power tubes tomorrow but I seem to remember this being ok. If it helps, I can post photos of the output waveforms.
Thanks again for all the replies!
|nevetslab||2/9/2018 3:27 PM|
|While it still sounds like an output transformer issue, one of the changes made some time back on the SVT output circuit is the change to 220 ohm 1-2W Screen resistors, and the elimination of the diodes across them. I'll see if I have that service not handy.|
You certainly have sufficient plate voltage, screen voltage as well as sufficient bias adj range to turn the power tubes all on hard. The notes on the screen resistor change don't suggest it would have anything to do with what you're dealing with. I also assume the 5 ohm/5W plate resistors are all good.
I've never built one of the Transformer shorted turns testers....that might be handy in this case. If you have a working output transformer handy, you could try discoonnecting this one in the SVT-II, then drive the primaries of it with the primaries of the spare O/T, and drive that spare thru the 4 or 8 ohm tap, so you'd get high AC voltage to feed the primary of this suspect transformer, then see what you get both loaded and unloaded on the secondary of the SVT O/T. This would at least get you into more meaningful signal current. I've been hoodwinked before by driving a suspect OT with an oscillator, or for that matter, AC mains via Variac on the primary and see what I get on the secondary. Now, you do have 220VAC available. That would actually be the easiest test.....feed it AC mains on the isolated Primary leads (disconnect the CT!!) and see what you get on the O/T secondary. If it pulls a lot of current, that could be it.
|Chuck H||2/9/2018 3:42 PM|
|shzmm||2/10/2018 4:43 AM|
|at 240VAC no tubes filament voltage is 6.8|
hum balance is dead center
setting the bias to .038 allows for a clean output of 4VAC into 4 Ohms, or a bit off the normal 220 Watts or so!
at this point, the PI is outputting cleanly and delivering 19VAC-- without the output tubes installed they deliver up 63VAC clean and reasonably well balanced (within 5%)
continuity check on output jacks made early on and now repeated
the 22 Ohm resistors and diodes were replaced with 220 Ohm early on
no 5Watt plate resistors on this model, it uses ferrite beads.
putting 220VAC across the primary of the OT gives 7.7VAC on the 2Ohm tap which works out to around a 1625r load. That seems about right too and with no strange current load.
Thanks again for all the replies, I am very curious how this one will end up.
|g1||2/10/2018 11:38 AM|
It won't give more than 19VAC with power tubes in? Then what happens, it clips square?
|Chuck H||2/10/2018 6:23 PM|
|So we know the 4 ohm tap fails to make power and the 2 ohm tap tests fine... Why didn't you check one tap (or all taps actually) for both tests? But I don't think it matters. A short in the secondary that still works fine on the 2 ohm tap shouldn't be causing the problem you have.|
And g1 is asking a good question here. I would think the PI should be able to deliver a minimum of inverse the bias voltage with tubes in in order to drive the amp to full power. But when you plug in tubes you only get 19VAC. That's way too low unless you're not putting enough test signal into the amp. Which I don't suppose is the case because with the power tubes out you get 64VAC. Since you replaced the screen resistors I guess those are off the plate too. So... The OT tests ok and something about installing the power tubes makes the amp misbehave. Please do like g1 asks and identify the PI test points you're using and what happens to the wave form there when power tubes are installed. I think you may have more than one bad power tube.
|shzmm||2/11/2018 4:15 AM|
Thanks for the replies!
Sorry for the misunderstanding, I was trying to answer a lot in one post. The voltage readings were off pin 8 V2 & 3 or the 12AU7 drivers' cathodes. The 19VAC reading is just the value at which the amplifier output starts to clip-- I believe the driver will stay cleaner longer than that actually but I will check that tomorrow and try to explain things a little better in the future!
I think the suggestion of trying out a single pair of the 6550s is a good one. I will run that test as well tomorrow and see if it turns up anything. However, since I have tried two different sets of tubes now, I don't find it likely to be an issue with the tubes themselves.
Out of curiosity, I wonder whether there is a difference in testing a transformer with current draw or not. When the leads are free such as while testing for shorts with a neon bulb, or passing a large voltage through the primary perhaps there is a difference versus when there is a load? Would it be a disaster to connect the secondary to a power resistor load while slowly bringing up a voltage on the primary? I have had power transformers that tested fine without a load but once under a bit of duress the voltage drops significantly making it unusable.
|g1||2/11/2018 12:04 PM|
|Yes you can test transformers with a resistive load, if you use an appropriate value.|
Have you tried other AU7's in the driver spots?
Have you had a good look at the impedance switch?
|nevetslab||2/11/2018 4:11 PM|
|Regarding testing the Output Transformer, having been first disconnected from the plates & C/T connection to the P/S (having the power tubes removed would be sufficient, I think), and driving the primary with the output from your Variac, sourced from 220VAC.....under no load, there should be just excitation current, probably less than 100mA. Not knowing the turns ratio, what to load the secondary with is a bit of a guessing game. Maybe 8 or 16 ohms, just to see that it will deliver voltage looking like the same waveform as it does without a load. We're just trying a different test than the shorted turn test, though if there is a shorted turn issue, I think it would show up in this procedure. It would, of course, draw more current loaded than unloaded. Granted, I haven't tried this on the O/T of an Ampeg SVT amp, so I don't know first hand what the open current load would be (excitation current).|
We assume you're driving the amp thru the Power Amp input jack on the rear, and not thru the preamp. Though, if your signal source doesn't have enough output level, then you would need to go thru the preamp to get enough level
The Screen resistors are sourced from a lower voltage supply, and not off the Plate supply.
|shzmm||2/12/2018 8:53 AM|
|ok I tested the OT with an 8 Ohm load and really didn't see much of a difference there... at 220VAC on the primary, I got 6.5VAC on the 2 Ohm tap which is about a volt less than unloaded. |
I have already replaced both 12AU7s and they read similar values and certainly deliver enough voltage to drive the 6550s so I don't see an issue there.
I double checked the impedance switch again and that appears to be fine.
Then, I checked the amp with only a pair of output tubes and got exactly the same results adjusted for fewer output tubes.
All tests have been performed at the power amp input, although I also just tried the regular instrument input for kicks.
I appreciate all the input so far, but it seems we are all grasping at straws with no clear direction any longer. The problem lies within some interaction between the OT and the output tubes I would say.
Unfortunately, I don't have a spare OT handy and probably won't pony up the money to purchase one any time soon. I guess the SVT won this battle!
|nevetslab||2/12/2018 11:51 AM|
|Later this week I'll look for the one Ampeg SVT II we have (assuming it's still there), and have a look at it for doing some side-by-side comparasons.|
Meanwhile, looking at the older SVT II Power Amp schematic, I see the expected AC drive signal at the 12AU7 cathodes is called out as 30VAC for getting 35VAC on the 4 ohm output. On your amp, with no power tubes installed, you stated getting 64VAC and only 19VAC with the power tubes in (regardless of one pair of all six, based on your stating you got the same results). So, we know SOMETHING is loading down the signal drive to teh power tubes.
Have yhou looked at the DC voltages under drive conditions? You indicated with the power tubes removed, you had a bias adjustment range of -58V to -35V. What is the bias adjust range with the power tubes installed (1 pair or all )? And, on each bias adjust, since there are two. Do the power tubes restrict the bias adjust range? Are the plate and cathode voltages on the voltage gain stage for each 12AU7 similar to what' shown on the schematic? Plate voltage of 160V, cathode voltage of 7V is shown. And, does the grid voltage of the 12AU7 driver stage sit around -77V. It will change as per setting of each bias adjust pot.
And, under signal drive, does the plate supply voltage and screen supply voltage change. I assume both do drop some between NO power tubes installed and power tubes installed. We're all pondering why you can't get sufficient bias to obtain the 24mA plate current per tube on each half. What does the raw bias supply change to with the power tubes installed? Schematic calls it out at -180V. Full wave rectifier via diodes D12 & D13, followed by C8 100uF/250V.
Since the Primary of the HT transformer has a universal primary, each coil having 100V taps, you can strap the primary for 220VAC, and in the process, since one of the primaries is 100VAC with the other at 120V, having the heater transformer across the 120V coil, that should raise the heater voltage output..
|shzmm||2/12/2018 1:40 PM|
I will check again, but I don't think the range of bias volage was at all affected by the fact of the power tubes being present or not.
I also remember the 12AU7s being within a reasonable range from the schematic but I haven't checked the voltages under drive.
The -180 raw bias voltage is also present.
|g1||2/12/2018 2:53 PM|
|nevetslab||2/12/2018 5:13 PM|
|With 220VAC on the isolated primary, and you stated at the 2 ohm tap, unload is was about 1V higher than it loaded with 8 ohms, which would be about 7.5V. So, the turns ratio found here is 220/7.5 = 29.33. Using the AC voltage shown on the schematic of 372V to gnd, or 744VAC plate to plate, we'd see about 25.4VAC unloaded at the 2 ohm tap. Disregarding the loading for the moment, computing the power, with this turns ratio, 25.4V sq'd / 2 ohms = 322.6W. That sounds about right for a 300W amp.|
With the primary wires disconnected, what do you get at the primary when you feed 1V RMS into the secondary. You might try this at a few frequencies, like 50Hz, 400HZ, 1kHz. Try it thru both the 2 ohm and 4 ohm output taps (where you're connecting your signal generator). It will add further confidence to the O/T being ok, which it's beginning to look like it is.
|Chuck H||2/12/2018 6:12 PM|
|shzmm||2/13/2018 10:57 AM|
|nevetslab||2/15/2018 3:02 PM|
|I brought over the Ampeg SVT-II amp from our Guitar Dept so I can open it up and have a look and get some measurements to help (I hope) track down what's happening with your amp. The Fender Hot Rod Deluxe popped up, so I've been occupied with details in a mod I had posted some time back, so it will probably be tomorrow before I can did into the SVT II.|
|shzmm||2/16/2018 3:12 AM|
Thanks a lot for your help!
|nevetslab||2/17/2018 5:39 PM|
|Svt-ii photos-orientation time|
Saturday, and I finally got the covers off the working SVT-II from our rental Dept @ CenterStaging, LLC Burbankc,CA. I first thought I'd capture and present a series of photos of the amp, just to put things into perspective when what is easy to see on the schematic, and a nightmare to try and get to in the real world. My apologies for not having a current-higher-resolution camera than this old Sony Mavica that I'm still beating up...low res, stores on 3.5" floppy disc!
[ATTACH=CONFIG]47192[/ATTACH] [ATTACH=CONFIG]47193[/ATTACH] [ATTACH=CONFIG]47194[/ATTACH] [ATTACH=CONFIG]47195[/ATTACH] [ATTACH=CONFIG]47196[/ATTACH]
[ATTACH=CONFIG]47197[/ATTACH] [ATTACH=CONFIG]47198[/ATTACH] [ATTACH=CONFIG]47199[/ATTACH] [ATTACH=CONFIG]47200[/ATTACH]
There's plenty more images but this will get us started. First, I'm curious to see what the heater voltage is. Here in California, I'm running on 120VAC. 5.84VAC. At 110VAC Mains, its' 5.37VAC, and at 127VAC (hi line), it's 6.24VAC. Interesting. This has the same Heater voltage problem (is it?) as you've found. No matter...that's where they set the Heater xfmr windings.
Just looking to see what the bias voltage is set at...it's probably been 3 years since I've had this in the shop. TP 1 is 62mV DC, TP2 is 59mV, so it's a little low from the nominal spec of 72mV. Voltage at the cathodes of the two 12AU7 drivers are -47.9VDC, right in the ball park.
I'll post this for now, and resume either tomorrow after church or Monday morning. My eyes are in terrible shape, and driving at nighttime is now very scary...scheduling for Cataract surgery on both eyeballs.
|Jazz P Bass||2/17/2018 7:55 PM|
|"scheduling for Cataract surgery on both eyeballs"|
They 'do' one eye at a time.
So you have to wait a week or 2 until you get the second one done.
I am here to tell you that you will not want to wait!
I took the patch off of the first one on the way home.
I could not believe how vibrant colors where.
Bluer than Blue.
Greener than Green.
Redder than Red.
|nevetslab||2/19/2018 5:52 PM|
|SVT II Output Waveforms, Data|
I had hoped to have the whole ay with this amp, but didnt' get started until nearly 2PM. But, got some waveform meassurements & scope photos. First, the amp won't put out 300W into 4 ohms It will do 200W with slipping, and 225W into 4 ohms with severe clipping, as you'll see. Here's some saveforms, with details below going left to right [ATTACH=CONFIG]47224[/ATTACH] [ATTACH=CONFIG]47225[/ATTACH] [ATTACH=CONFIG]47226[/ATTACH] [ATTACH=CONFIG]47227[/ATTACH] [ATTACH=CONFIG]47228[/ATTACH] All of the output waveforms are from the monitor output of the Amber 3501a, and aren't to any specific scale. The other waveforms ARE to scale, and will be noted. The first scope photo is @ 20VAC Out into 4 ohm resistive load (100W). The positive-going half-wave is the cathode voltage across the 1 ohm cathode resistor with it's diode across it The image next to it is the Amber 3501a Audio Analyzer, monitoring the output voltage, showing 20VAC on the meter, and is also sourcing the 400Hz sine wave input to teh power amp stage. The next waverforms are @ 28VAC into 4 ohms, with the output now showing some clipping, along with the voltage across the 1 hm cathode voltage (now @ 500mV/Div), definitely showing saveform clipping, while we don't see much of it on the output. The next two waveforms are showing teh same 28VAC output, along with the positive and negative driver tube outputs to the grids of the power tubes. 20V/Div via X10 probe. We're starting so see some assymetrical clipping on them at this drive level [ATTACH=CONFIG]47229[/ATTACH] [ATTACH=CONFIG]47230[/ATTACH] These two waveforms are at 30VAC output ito 4 ohms (225W), and definitely clipping. The driver stage waveform is starting to show a bit more distortion, but still only moderate non-linearity The next two waveforms are driving higher resistance loads on the 4 ohm output tap, with the[ATTACH=CONFIG]47231[/ATTACH] cathode voltage showns with the output waveform. [ATTACH=CONFIG]47235[/ATTACH] [ATTACH=CONFIG]47236[/ATTACH] These last two waveforms are at 40VAC output, 1st one into 16 ohms, and 2nd one into 24 ohms.. The half-wave is the waveform across the cathode resistor on the positive half of trhe power tubes. I'll have to add the voltage measurements on the amp @ idle when I get home At 120VAC Mains, I find the following: Power Tube Plate voltage: 647VDC Screen supply voltage 383VDC Driver tube cathode (bias) -47.8VDC Driver tube grid voltage -70VDC @ V3B, -72VDC @ V2B V3A Plate Voltage 181VDC V3A Cathode Voltage 7.4VDC V2A Plate Voltage 182VDC V2A Cathode Voltage 7.4VDC V1A Plate Voltage 217VDC V1A Cathode Voltage 70.5VDC V1B Plate Voltage 186VDC V1B Cathode Voltage 1.2VDC I forgot to get grid voltages on V1A, V2A & V3A. The Cathode Voltages @ TP1 & TP2 were recorded in the prvious post, as was the heater voltage. I haven't yet made measurements under 2 ohm load, and should get that tomorrow.
|shzmm||2/20/2018 10:39 AM|
|[QUOTE=nevetslab;480659]I had hoped to have the whole ay with this amp, but didnt' get started until nearly 2PM. But, got some waveform meassurements & scope photos. First, the amp won't put out 300W into 4 ohms It will do 200W with slipping, and 225W into 4 ohms with severe clipping, as you'll see. Here's some saveforms, with details below going left to right [ATTACH=CONFIG]47224[/ATTACH] [ATTACH=CONFIG]47225[/ATTACH] [ATTACH=CONFIG]47226[/ATTACH] [ATTACH=CONFIG]47227[/ATTACH] [ATTACH=CONFIG]47228[/ATTACH] All of the output waveforms are from the monitor output of the Amber 3501a, and aren't to any specific scale. The other waveforms ARE to scale, and will be noted. The first scope photo is @ 20VAC Out into 4 ohm resistive load (100W). The positive-going half-wave is the cathode voltage across the 1 ohm cathode resistor with it's diode across it The image next to it is the Amber 3501a Audio Analyzer, monitoring the output voltage, showing 20VAC on the meter, and is also sourcing the 400Hz sine wave input to teh power amp stage. The next waverforms are @ 28VAC into 4 ohms, with the output now showing some clipping, along with the voltage across the 1 hm cathode voltage (now @ 500mV/Div), definitely showing saveform clipping, while we don't see much of it on the output. The next two waveforms are showing teh same 28VAC output, along with the positive and negative driver tube outputs to the grids of the power tubes. 20V/Div via X10 probe. We're starting so see some assymetrical clipping on them at this drive level [ATTACH=CONFIG]47229[/ATTACH] [ATTACH=CONFIG]47230[/ATTACH] These two waveforms are at 30VAC output ito 4 ohms (225W), and definitely clipping. The driver stage waveform is starting to show a bit more distortion, but still only moderate non-linearity The next two waveforms are driving higher resistance loads on the 4 ohm output tap, with the[ATTACH=CONFIG]47231[/ATTACH] cathode voltage showns with the output waveform. [ATTACH=CONFIG]47235[/ATTACH] [ATTACH=CONFIG]47236[/ATTACH] These last two waveforms are at 40VAC output, 1st one into 16 ohms, and 2nd one into 24 ohms.. The half-wave is the waveform across the cathode resistor on the positive half of trhe power tubes. I'll have to add the voltage measurements on the amp @ idle when I get home At 120VAC Mains, I find the following: Power Tube Plate voltage: 647VDC Screen supply voltage 383VDC Driver tube cathode (bias) -47.8VDC Driver tube grid voltage -70VDC @ V3B, -72VDC @ V2B V3A Plate Voltage 181VDC V3A Cathode Voltage 7.4VDC V2A Plate Voltage 182VDC V2A Cathode Voltage 7.4VDC V1A Plate Voltage 217VDC V1A Cathode Voltage 70.5VDC V1B Plate Voltage 186VDC V1B Cathode Voltage 1.2VDC I forgot to get grid voltages on V1A, V2A & V3A. The Cathode Voltages @ TP1 & TP2 were recorded in the prvious post, as was the heater voltage. I haven't yet made measurements under 2 ohm load, and should get that tomorrow.[/QUOTE] nice work! I will post my shots tomorrow, but your amp is much cleaner than what I have here! Looks like the voltages are pretty much in line with what I had already posted however...|
|nevetslab||2/20/2018 1:10 PM|
|Some added notes to what i posted last night. V1A Grid voltage 25.2VDC Test Point 1 68mV Test Point 2 63mV AC Mains @ idle 120VAC @ 2.01A/204W Power @ 20VAC Output/4 ohms (100W) 4.36A/442W @ 120VAC mains Power @ 28VAC out/4 ohms (196W) 5.8A/563W @ 120VAC Power @ 30VAC out/4 phms (225W) 6.4A/647W @ 120CAC I haven't opened up the insides thus far. And, I don't have component layout drawing for the PCB's. Thus far, I haven't identified the bias supply. Nor have I checked the min/max range of the bias control and the corresponding voltages at the Test points TP1 & TP2. Some client work came in so I need to address that for now.|
|shzmm||2/21/2018 8:31 AM|
|ok here goes,|
Power Tube Plate voltage: 657VDC
Screen supply voltage 308VDC
Driver tube cathode (bias) -45VDC
Driver tube grid voltage -66VDC @ V3B, -66VDC @ V2B
V3A Plate Voltage 191VDC
V3A Cathode Voltage 7.1VDC
V2A Plate Voltage 186VDC
V2A Cathode Voltage 7.4VDC
V1A Plate Voltage 237VDC
V1A Cathode Voltage 86VDC
V1B Plate Voltage 188VDC
V1B Cathode Voltage 1.3VDC
V1A Grid voltage 25.5VDC
Test Point 1 19mV
Test Point 2 20mV
The waveform doesn't get any prettier after around 4 VAC at 4 OHM. 2OHM doesn't look any better either.
As you can see the boards have had a lot of work done on them not by me however. As well there is quite a lot of heat damage as well as the caps all look tired.
I'm still pretty convinced this amp is a lost cause but let's see...
|Chuck H||2/21/2018 9:02 AM|
|Is it just me, or do all those electrolytics look crowned?|
|shzmm||2/21/2018 9:25 AM|
Either way, you can see the heat buildup in these amps is just silly. Even with a fan, having 6 6550s in close proximity in a rack mounted amp was probably not the best idea of the engineers of Ampeg.
|Chuck H||2/21/2018 9:56 AM|
|Well I can't imagine those covers are bulging on their own.?. Get those caps outta there! They may be shorting while the amp is conducting and ruining performance. It's not uncommon for a bad cap to seem ok when the amp is just idling.|
|Justin Thomas||2/21/2018 10:00 AM|
|Perhaps they're bulging on top because the "explosion end" is blocked with goop. Sucks that the goop used to secure the caps is the exact same color as electrolytic cap spooge...|
And, not all hum is bad caps, but not all bad caps hum, either...
|Chuck H||2/21/2018 10:09 AM|
|That was my thinking too. They're bulging = bad. (<period)|
Might be the whole problem. Don't forget the bias caps, etc. I'd strip every single aluminum cap out of that amp and replace it.
|shzmm||2/21/2018 10:51 AM|
|I just checked the caps. The covers are there to protect against shorts I suppose. There is no bulging on the top or bottom. The goop is ancient silicon or whatever they used back in the 90's...|
I have seen lots of amps with bad caps but never had an issue like this one. I can replace them all of course but it will take some time-- my usual shop doesn't have these sizes.
|Enzo||2/21/2018 11:46 AM|
|Just heat around them often causes end covers on caps to bulge, it is not a sign of a bad cap. Slit the plastic sleeve to free the end cap and look underneath. I bet the caps are perfectly fine under there.|
|g1||2/21/2018 2:45 PM|
|nevetslab||2/21/2018 9:59 PM|
|I'm wondering about the screen supply. The Power Tubes plate voltage is in the same ballpark as I see on CenterStaging's amp, but the screen supply being only 308VDC makes me wonder about the supply caps. It's almost as though one of the rectifier diodes is open, or, the caps are just tired. It might be a good idea to replace at least the C9 100uF/450V, C8 100uF/250 & C10 47u/F/450V. Certainly no harm in replacing the rest of the HV supply caps, but if you're tight on funds, I'd see about getting the screen supply back up where it needs to be. You did say all the filter caps do look a bit tired. |
I haven't independently varied the screen supply on one of these amps to see what the effect is if it's too low. The screens on the amp in my shop is a bit higher than the 340VDC called out on the schematic (earlier schematic). And, you haven't been able to get the bias voltage up in yours. It is suggesting power supply issues. All the other voltage you're reading jive with what I'm seeing, as you had pointed out.
In looking at the power supply schematic, I could lift out D11 and D14, which are the full wave rectifier diodes for the Screen supply (as well as the plate supply for V2B/V3B), and patch in my HP 712C HV Power Supply to see what the varying Screen supply voltage does on the performance of the amp. It may Saturday before I get a chance to do that, but I am curious.
I still haven't looked at the 2 ohm load conditions....obviously lower output voltage swing.....and C/S's SVT II amp doesn't reach the specified output voltage spec either.
|shzmm||2/22/2018 2:55 AM|
|I'm wondering if changing the screen resistors to 220r --from a Ampeg service note-- is what is bringing down the screen voltage. I am not reluctant to change the caps just have to source them first!|
Also, on the schematic I am using, I don't see D14. D11 appears to allow a draw for c9 after the doubler of d12,13 and c8,11.
Otherwise I can also jig up an external screen voltage and see what happens; I am also curious. However, first I will do some measurements again at the screen voltage at drive conditions. I would think if there is a huge loss there, then indeed that could be the culprit. If not, it is back to the drawing board!
|nevetslab||2/22/2018 10:52 AM|
|The SVT II I have here has 220 ohm screen resistors, and no diodes that used to be there with the 22 ohm value. Usually we see screen resistors in the range of 470 ohm to 2.2k on power tubes. Orange uses 1K on their 6550's in the AD-200 bass amp, for instance. And the maintenance note I included only stated it was to extend the life of the power tubes (by reducing the screen current). |
We could compare the open circuit secondary voltage on that 5-leaded winding that sources both the bias supply and the screen supply. Your plate voltage is in agreement with the stated schematic voltage for the plate supply (660VDC), so all that fits.
Did you measure the screen supply voltage at the power tugbes or at the filter cap (which also feeds the driver tubes plate voltage?
Where do you source parts over there? Farnell? Nichicon, United Chemicon, Cornell-Dublier are typical sources, all being 10mm L/S snap-in low-proile caps. I probably have the data sheets in my database for the Nichicon & United Chemicon parts, having replaced some in the past.
I just looked at the schematic you're using. Does that match your amp? I see a voltage doubler circuit in use (C8 & D13), followed by D11 & C9 to form a half-wave rectifier circuit for the screen supply. I haven't lifted out the motherboard in my chassis, so I don't know if this one here is the same as yours. So, you have one additional cap C8 that could be suspect.
|Enzo||2/22/2018 11:03 AM|
|We see screen resistors of 470-2200 ohms, yes, but they are on screens that sit at essentially the plate voltage. These 220 ohm resistors are for screens that sit at half the plate voltage or less.|
|nevetslab||2/22/2018 11:23 AM|
|Enzo||2/22/2018 3:20 PM|
|I am not the theory guy, but I don't recall the SVTs burning up 220 ohm resistors. What are the odds someone putting the Orange together chose those values based on tradition?|
Though I didn't phrase it as such, my post #47 was more of a question than a statement.
|nevetslab||2/22/2018 8:32 PM|
|Nor am I the theory guy. And of course you're absolutely correct, where that range of screen resistors are from the plate potential and not from a lower supply as in the SVT & the Orange. I've seen a number of screen resistors in SVT's burn up, both when they were 22 ohms as well as 220 ohms, even though the protection circuit acts pretty quickly to turn off the HT pwr xfmr. Sorry about the previous reply...no offense intended. |
Any thoughts on if the screen supply being much lower in shzmm's SVT II than the working amp I have in the shop? 380VDC vs 308VDC. I've never been down this road, so mostly, like the rest of us, looking for clues.
|Enzo||2/23/2018 4:53 AM|
|No offense perceived, no problem, I was just thinking out loud so to speak.|
Yes, any shorted tube can burn up whatever resistor is there.
I guess if my screens node was way low, I'd track it down through the power supply. My prints have voltages everywhere, so we can determine if we started out with too low to start with, or if a later node was being loaded. And certainly look for ripple.
|christarak||2/26/2018 12:48 AM|
|[QUOTE=Chuck H;479968I hate telling a customer that something is above my skill set. I'm more likely to call them and tell them I need more time if they're ok with that. Then I charge for the repair (or finishing task) based on what time it likely would have taken someone with specific experience to handle. Seems fair and I'm never sorry for the experience.[/QUOTE]|
I am having a similar hair-pulling frustration with a JVM410 at the moment. I have spent way more time on it than I should, at the cost of other repairs. When do you throw in the towel?
|Chuck H||2/26/2018 8:23 AM|
|Enzo||2/26/2018 9:41 AM|
|I don't tell them it is above my skill set. Frankly it probably isn't. But the answer eludes me this time. I tell them I am unable to solve the puzzle. I also usually continue by telling them that doesn't mean it is unrepairable, just that I can't seem to do it. I tell them another tech might spot the problem quickly.|
We have all managed to miss simple solutions to problems, just as we all have/will find answers to very difficult problems in a short time. No one bats 100%, do not be ashamed when you miss.
In certain circles, I am a guy techs turn to when they get stuck. And it often is the case that what I find is simple, an oversight, a mistake, a faulty assumption. I am sure someone who followed up on ones I couldn't solve many times found simple solutions I somehow could not see.
|christarak||2/26/2018 12:05 PM|