1. Please take 30 seconds to register your free account to remove most ads, post topics, make friends, earn reward points at our store, and more!  
    TalkBass.com has been uniting the low end since 1998.  Join us! :)

Ampeg SVT 3-Pro Problem

Discussion in 'Amps and Cabs [BG]' started by Gowron, Dec 3, 2012.

  1. Gowron


    Jun 26, 2011
    Qu 'Nos
    My 3-Pro has been serving me faithfully for over a year, and all of a sudden I'm getting only about half the volume it should be putting out. This amp was always a beast until last Friday at a gig, I couldn't get any stage volume. Enough made it through the mic apparently as the recording sounded fine from off stage. At this particular dive bar I usually had the problem of being too loud, that master volume knob is normally pretty sensitive when turning up. Tone sounded normal though.
    So anyways, I'm not super tech savvy, but my friend/band-mate is and we(he mostly) went through to try and diagnose the problem. First thing we made sure it's not the cab (410HLF). It worked just dandy with his Gallien Krueger 700RBII head. We sent the 3-Pro preamp into the GK and it sounded normal. But when running the GK preamp into the 3-Pro, the volume problem was there again, so we figure the problem is in the driver section. We tested the first test point in the 3-Pro poweramp schematic (R1), and it checked out. Second test point did not however, voltage was a bit high I believe. (Neither us of has money for a fancy function generator so we used a laptop with a program I can't remember the name. Think it did the job though.)
    So we thought maybe the problem is somewhere around the 12AX7 circuit. It was a pain in the arse but we pulled the whole poweramp section out and he checked for cold solder joints. Nothing stuck out. We put everything back together as it was getting late, but my friend's next guess was to check the BJT transistors. He can fix anything electronic, but isn't as familiar with this particular amp as an amp tech or bass junkie who sees them all the time. I told him TalkBass is always my goto.
    Anyone have any similar issues? Anyone have experience with those transistors in the driver section going bad, or know of what could have happened there? We didn't think it could be the mosfets. Also changed the tubes as I already had a new set of JJ's. They didn't make a difference. I'm sending my band-mate a link so he can register and chime in with anything I missed. If anyone has any suggestions I would be forever grateful. Hopefully my friend and I will be able to reconvene and solve this riddle. Thanks for reading.
  2. RickenBoogie


    Jul 22, 2007
    Dallas, TX
    Try a short patch cable in the fx loop, (send/return). Barring that, you'll need an amp tech.
  3. Jerrold Tiers and Hodgy hang here and can be of the most help. I have no schematics so I don't know where TP2 is in the circuit.
  4. Jerrold Tiers

    Jerrold Tiers

    Nov 14, 2003
    St Louis
    OK, so TP1 is the input to the power amp, basically the grid of the first tube. Signal there means you are putting in a signal OK, although I am not clear if you were measuring DC or AC voltage.

    TP2 is the plate of the first tube. If the DC voltage is high, or AC voltage is low, the tube may have low gain, or something in the cathode circuit may be bad. If the AC voltage is high, that won't make less sound, but might make for more distortion. Again there might be a problem in the cathode circuit.

    The more interesting test points are 3, 4, and 5. They should all be "similar" for AC voltage, 4 and 5 a little lower than 3. DC voltages will be similar on 3 and 4 at around 160V, but zero for TP5.

    Without looking at 3, 4, and 5, don't worry about the bipolar parts. They have no gain function, they are just "current sources" for the bias and DC offset correction circuits.

    If you find that 3, 4, and 5, or some combination, are significantly lower than the schematic says, then the bipolar parts could be an issue. But so could an open cap, C4, C6 or C7, or a bad solder joint associated with one of them.

    C4 is part of the drive, it causes the lower tube half to work with the upper one. If a bad C4, or other problem, causes the lower half of the AU7 NOT to work with the upper half, it can cause low output.

    Also, turning down the tube gain, which reduces the plate voltage, can cause low output, although it usually increases distortion, that's what it is for...... since I assume you know about that, I will point out that any failure in that circuit which acts like turning it down will do the same thing.
  5. izytang


    Dec 4, 2012
    Hi there, this is the friend that Gowron was talking about. First of all, thank you very much for sharing your experience and expertise - it's much appreciated.

    We troubleshooted the SVT-3 Pro again tonight and discovered there was a misreading on the schematic from the previous session. Thus, we restarted the whole process and determined that the problem seems to lay in the pre-amp board. How do we know this? We routed a GK 700RB-II's Send output to the Power Amp input of the SVT-3 Pro and the volume was fine. Alternating that by using the SVT-3 Pro's pre-amp output to the GK 700RB-II's Return yielded the same volume issue. See attached JPG for reference.

    Thus, by narrowing the problem to the pre-amp board, we sought to evaluate test points from the input to wherever the point of failure is. We injected a 1000 KHz sine wave of 200 mVACp-p from the audio output of my laptop (using GoldWave), to the input of the amp. The first four test points checked out just fine (see following post), but the fifth did not. The expected value of V1's pin 6 was 150 volts AC, but was measured at 22 volts AC. All subsequent readings were off. We initially thought it might the 12AX7 pre-amp, but replacing it yielded no change. Then I measured the voltage between P1 and R12, getting 800 mVAC, with the gain control turned fully clockwise. Interesting, 800 mV / 4.7 V and 22 V / 150 V result in very similar ratios, so I wonder if they are related.

    I suspect there is something wrong with the gain potentiometer, but the peak LED still operates and the voltages in later test points are larger, not smaller, so I'm not sure here. Feel free to review the attached PDF to see the measurements made to the circuits. I hope this helps to understand where the problem may be. Thank you very much!

    Attached Files:

  6. izytang


    Dec 4, 2012
    These are the two other diagrams from the pre-amp schematic (TalkBass only allows two per post).

    Attached Files:

  7. First I believe you should be measuring Volts DC at those points. If a new tube made no difference I would be suspecting plate resistor R13 first. I don't have the schematics so I can't offer a DC voltage drop across R13 that you should expect to see.
    Jerrold Tiers will be along to correct any of this or concur with additional info. Look closely at the notes on the schematic to see if the voltage test points are with signal or without a signal present.
  8. Jerrold Tiers

    Jerrold Tiers

    Nov 14, 2003
    St Louis
    The schematic gives both DC and AC readings for those points, so there should be no problems.

    The 800 mV seems correct, the divider ratio of 50K vs the 220K gives that.

    I would be inclined to check the DC at TP5. It might be lower than the expected 140V, and if so, that will indicate that the capacitor C3 may be leaky (bad). That would do it by drawing down the DC voltage.

    Something wrong in the cathode circuit could also be an issue, C4 and associated circuitry, if open, could lower the gain.

    And a DC problem would make the TP5 voltage way too high, with associated low AC signal voltage.
  9. izytang


    Dec 4, 2012
    Thanks for the input, it's much appreciated. I attached a PDF with all test point measurements made in my first post in this thread, but I probably should have listed them here too:

    Pre-Amp Section
    Measured Expected
    1 0.2 V 0 V 0.2 V 0 V
    2 0.2 V 0 V 0.2 V 0 V
    3 0.2 V 0 V 0.2 V 0 V
    4 4.7 V 0 V 4.5 V 0 V
    5 22 V 153 V 150 V 140 V
    6 30 V 223 V 22 V 200 V
    7 38 V 123 V 0.7 V 120 V

    As you can see from above, the measured DC voltage at TP5 was 153 VDC. Expected was 140 VDC - so it's actually higher. Could this be related to C4? If so, this would be simple enough to replace. Thanks for the continued input!
  10. Had a similar issue. Mosfets needed to be biased. Fixed the problem instantly.
  11. izytang


    Dec 4, 2012
    We initially troubleshooted the powerboard by itself and all AC and DC test points checked out; all values within tolerance. Also, we isolated each board by testing different preamp and poweramp combinations. Connecting a GK preamp to the SVT3 poweramp yielded no problems in volume or tone. Doing it the other way around however was a different story. This conclusively determined that the problem is in the preamp section, specifically in the signal chain near R12 and C4.

    It's my understanding that the MOSFET driver circuit does not need biased/offset, withstanding any dramatic modifications to the amp. But I'm always open to any possible points of failure. Thanks for the input.
  12. Jerrold Tiers

    Jerrold Tiers

    Nov 14, 2003
    St Louis
    150V, 140V, no big problem there.....

    if the tube is of the right type and good, voltages are good, and input signal is good, and plate load is good, there pretty much has to be the correct gain.

    Looking at the circuit, I tend to think that looks like a lot of gain. Did you have the bright switch on, or off when you measured? it isn't specified.
  13. izytang


    Dec 4, 2012
    Yeah, those were my thoughts too. I tried my best to measure and measure again, and to document every DC or AC reading at each test point, but if it's possible there was an error I can do it again. I was using a digital multimeter with the common/ground probed attached to the case chassis. So all voltage readings are with respect to neutral/ground.

    Conditions and Settings of the SVT3-Pro front Panel:

    Input: 1000 Hz 200mVAC p-p sine wave, 0% THD
    Bright: Out
    -16 dB: Out
    Gain: 12' O Clock (but other variations were tried, to see dynamics)
    Hi: Out
    Low: Out
    Bass: 12'o Clock
    Midrange: 12'o Clock
    Frequency: 1
    Treble: 12'o Clock
    Master: 7'o Clock (0)
    Tube Gain: 5'o Clock (10)
    Mute: Out
    EQ: Out
    9-band EQ: All flat
    EQ level: 0 dB
    Power Switch: On

    Those are the initial conditions, to the best of my documentation. There wouldn't be anything weird about this circuit that could give incorrect readings? I could try and use a different DMM to make sure they're all the same. Unfortunately, I don't have access to an oscilloscope, else I would send plots of the waveforms at each test point.

    One other thing I did, just to see if the gain potentiometer was bad, was to bypass the signal by placing a cabled wire that connected R75 directly to R12. The readings of those test points changed as follows:

    6 57 VAC
    7 83 VAC

    Which is about twice the amount of the measured AC voltages at the same test points. I apparently didn't measure TP5 with the gain pot bypass.

    As always, thank you for your help!
  14. Jerrold Tiers

    Jerrold Tiers

    Nov 14, 2003
    St Louis
    Generally the tests are assumed to have gains at full, but tones in middle or as specified. With gain at full I suspect you would get more than with it bypassed but at 12:00. The loading of teh pot on R75 is less when turned up than at 12:00 and a bypass from the top of the pot to R12.

    That 12:00 setting may be enough to account for some of the low gain.

    I don't know if the gain is measured with bright on or off, but with it "on" the gain is higher, more likely to be the sort of gain shown in the chart. try it and see.

    Not all gain charts are completely described, and not all are perfectly correct either. Not Ampeg, and not some others I could name (but won't).

    Your original complaint was low volume.

    The only thing needed for good volume is to have the thing working and presenting a suitable signal to the power amp. It actually looks as if you can get a good signal to the power amp..... now that things in the amp have been all disturbed, possibly cleaned etc in the process, are you still getting the low volume?
  15. Gowron


    Jun 26, 2011
    Qu 'Nos
    I want to say that the volume level is still lower, but maybe my ears are deceiving me. I'd like to try it at band practice tomorrow and then I'll know for sure if the problem was fixed. Before my friend and I were comparing it to his GK 700RBII head, but maybe that's not a fair comparison, and maybe changing the tubes or cleaning up solder joints did in fact fix the problem. I know the GK head is 30 watts higher, but I always thought it didn't seem that much more powerful. Maybe I was wrong though.
    One thing I just noticed though is that the peak LED isn't working anymore. I maxed the gain, master, tried hitting the bright switch and boosting all EQ knobs, would not light up. Amp was distorting as usual with the gain maxed. I tried with tube gain at 0 and 10, still nothing. When I hit the mute button it DOES light up though. Don't know if this is related or something new that happened when we put the preamp board back in.
  16. izytang


    Dec 4, 2012
    Sounds like Gowron's amp is still having volume problems (plus a random peak LED anomaly). I see it as one of those things that you're sure when it's working, and not sure when it's not working.

    Anyway, to answer some questions: Yes, the bright switch was out (off) during all tests and measurements. The gain was set to 12:00 for most tests, since that's around where it usually would be for a full signal (from when it used to work). However, the gain was increased for a couple of side tests, and no real difference was made.

    I'm sure it's buried somewhere in one of the original posts in the thread, but the fundamental complaint is that the volume output is considerably less now than it was for over a year. Therefore, it's not a matter of the SVT-3 Pro putting out enough volume, but rather why it got cut all of a sudden.

    So in recapping, the SVT-3 Pro's power amp board is working perfectly, but in hearing the bass volume, but also with checking all test points. The pre-amp, however, is still not outputing a suitable signal level. The test points break down at TP5, specifically AC measurements.

    Do you think we should change out some of those surrounding capacitors, or perhaps re-measure the circuit?
  17. Gowron


    Jun 26, 2011
    Qu 'Nos
    If I hear the amp in a full band setting, then I will definitely know if the problem still exists. But I can't be certain hearing the amp by itself. At the gig last weekend it was very noticeable to me. But that was before we changed the tubes out. So I think Jerrold's right, we need to be certain the problem is still there, and if it is, then we should think about re-testing with Jerrold's suggestions. Maybe we goofed on the AC measurements somehow, or something wasn't working correctly, or maybe the AC voltage is off but unrelated to the prior issue. Anyways, I'll post again when I have more information. Thank you guys so much for all the info and suggestions, it has been a tremendous help.
  18. Jerrold Tiers

    Jerrold Tiers

    Nov 14, 2003
    St Louis
    The thing about turning gain UP all the way is important..... most tests are intended to be made that way.... too many variable if you do not. One person's "turned to 12:00" is a little off from the next, control pots have a tolerance, and it usually is considerable compared to any other part, so for consistency the gains are usually measured with any gain control "UP" full, unless specifically stated otherwise.

    That control is a 70-503-22, an audio taper part. Audio taper pots tend not to be exceptionally accurate in terms of resistance vs position. That's the reason for testing with it full up.

    There would be a considerable difference in volume with it all the way up. Only if the level doesn't come up significantly would I assume a serious problem.

    Another point...... unfortunately, those sorts of tests mean a whole lot more if one uses an oscilloscope as well.... That will show what the signal LOOKS like, meaning you can see any gross distortion, clipping, check the basic waveform, etc, and make much better sense of the readings. it's an essential basic piece of test equipment, without which you can't be sure of anything much past power supply voltages, and even there you will miss some things. You probably don't want to hear that.......

    At the moment, best to check that the problem still seems to be there.

    The peak light not lighting is indeed odd, because if it lights with mute, it should light from at least ONE of the several other points that are monitored. Either signal levels are rather low, or something else strange is going on.
  19. izytang


    Dec 4, 2012
    Hi Jerrold,

    After your previous post, we decided to do a complete re-measurement of the pre-amp circuit by invading a lab with professional instrument measurement equipment, including oscilloscopes, function generators, variable power supplies, station multimeters, etc.

    After taking detailed documentation, I prepared a report for your review, which gives great information on the setup and the results; complete with data tables and oscilloscope plots at various test points.

    The link for the PDF report is here: https://docs.google.com/open?id=0B6r6RJHNYI0QMkEtZVhzMGprT2M

    In summary, we tested all the DC points. Every single one checked out. At TP5, the signal was zero, and at V1's pin 7 (input), the signal was just fine. Again, the report details everything you would be interested in, but below summaries the AC measurements:


    Since the plate voltages are fine, I'm inclined to think that C3 or C4 (or both) are bad and have leaked; causing AC signals to be blocked in the circuit. Would you agree with this conclusion? If so, I'll go ahead and order their equivalents from Mouser.

    As always, thank you for the input!
  20. wilfredes


    Oct 11, 2012
    The schematic gives both DC and AC readings for those points, so there should be no problems.