Ampeg SVP-Pro Schematic Tutorial

Discussion in 'Amps and Cabs [BG]' started by bgavin, Jul 23, 2007.

  1. This is an open invitation to the tube gurus to do an in-depth tutorial on the SVP-Pro schematics.

    The intent is for others to learn about the various stages of the SVP-Pro, and how they work. Collaboration by the bright minds here will most likely lead to a series of improvements that can be made to this preamp.

    I would like to see every section explained, and the pros/cons pointed out in understandable terms.

    Link to Main Board Schematic PDF

    Link to Tube Board Schematic PDF
  2. I'm no expert on this stuff - but I guess the question most want answered would be how to get rid of the noise/hiss common in these preamps?
  3. I'd like to see an illuminating discussion lead by tube gurus such as Psycho Bass Guy, and the others.

    As with most commercial products, I'm certain the SVP-Pro has compromises made to cheapen the product. The smart tube guys will understand these, and can probably suggest corrections. This might be as simple as replacing caps and resistors with premium parts, or could be a change to the circuit itself. I dunno, I'm not a tube guy.

    My area of expertise is speakers and cabinets, which I share freely here. I'm hoping one of the tube guys will do the same.
  4. Jerrold Tiers

    Jerrold Tiers

    Nov 14, 2003
    St Louis
    One source of hiss is the FET switch elements used for mute, etc.

    It the FET is damaged, particularly ones at input, where they can be abused from outside, it can leak and make a very nasty hissing. It MAY "almost work" also, so that is not necessarily a clue.
  5. I bought an svp-cl last year new from a local shop, which had been there awhile. I wanted to try the pro, but they were hard to find and the notorious hiss problem seemed to be very prevalent. My question is why does the pro hiss, yet the cl doesn't???????

    Is it as simple as tube design arrangements-combinations/too much gain in the last stage x7's v. au7's? I've also read here that the circuitry board needed re-worked in the way of solder joints, etc., but once done, everything was fine!!

    ...these preamps are so powerful and tone enhancing...I wish I knew the answer to help out. :crying:


    Ampeg Club Member
  6. JGR

    JGR The "G" is for Gustav Supporting Member Commercial User

    Jun 29, 2006
    President, CEO, CFO, CIO, Chief Engineer, Technician, Janitor - Reiner Amplification
    let's not forget the physical layout and packaging of everything - there is an aweful lot crammed into that single rack space with 5 tubes, graphic eq, etc.

    things such as component spacing and orientation, location of the transformers, location and dress of high voltage and heater wires/traces, grounding scheme, etc. etc. all play a huge role in tube amps.

  7. James Demeter has it figured out. So did Steve Rabe.

    Granted, the SVP-Pro probably has more tubes involved.

    That's why I'm hoping somebody will step up and tutor the rest of us.
  8. fdeck

    fdeck Supporting Member Commercial User

    Mar 20, 2004
    Madison WI
    HPF Technology LLC
    Granted, I am not a tube guru. However, I have designed Class-A preamps, and low-noise circuits. Tubes and JFETs have the same empirical noise model, so I am not completely out of my element here.

    First, I looked at the schematic. It's hard to read because the tubes are all at the top of the page. It might help to re-draw it in left-to-right format using LTSpice (, leaving out anything that isn't in the main signal path.

    Second, PBG will correct me if I am wrong here, but you can disable successive stages of a tube preamp by effectively grounding each grid with a 100 uF capacitor to ground (in case a grid is biased). This allows you to find out which portion of the circuit is the worst offender. During this process, it would help a lot to actually measure the noise using a freeware spectrum analyzer such as Virtual Analyzer. This lets you narrow down your analysis of the circuit, since there is usually one dominant source.

    Don't turn on your speakers during measurement -- why risk blowing them or damaging your hearing. Also, beware lethal high voltages inside the amp, very close to where you are working. General knowledge of safe amp repair practices is essential.

    The two main causes of noise in a tube preamp are going to be thermal noise in the resistors and statistical (shot) noise in the plate current of tubes. Assuming modern resistors, the composition of the resistor is a relatively small effect.

    An idealistic design process is to have lots of gain in the first stage, so that subsequent stages contribute minimally to the noise picture. This is difficult in a mainstream tube preamp because of insertion losses of passive attenuators and tone networks. For this reason, the worst offender could be anywhere in the circuit.

    Good luck. I know that I have not pinpointed the cause, but have only given some ideas of how to search. You can get triode models for LTSpice (Duncan's Amp Pages has them I think), and LTSpice can do a noise analysis. But I have found the results difficult to interpret, and have resorted to a combination of rule-of-thumb theory and measurement.

    Here are some quantitative comparisons, using units of nV/rtHz (nanoVolts per square root of Hz): The quietest tube I know of is the 12AX7, which has around 10 nV/rtHz of equivalent input noise (the shot noise in the plate current, divided by the transconductance). A common IC op amp in bass amps is the TL071, at 18 nV/rtHz. Premium audio amps are around 4 to 8 nV/rtHz. Specialized op amps and discrete design can get you below 1 nV/rtHz. Thus there should be no inherent reason why a tube preamp is noisy, but low-noise design may unacceptably increase the complexity of a circuit.
  9. BbbyBld


    Oct 13, 2005
    Meridian, MS
    Wow! That's a confusing schematic! Drawing a legible schematic would be a huge improvement! Some people make them confusing on purpose, though.

    I looked at it for about 5 minutes, and one thing I would do is change the graphic EQ design to a constant-Q topology instead of using gyrators. The EQ design is dated and cheap.

    Another thing, the op amps are running at +/- 8V and that supply is half-wave rectified. I would have used a full-wave bridge rectifier and I would have run the op amps at +/- 15V for more dynamic range and lower noise.
  10. fdeck

    fdeck Supporting Member Commercial User

    Mar 20, 2004
    Madison WI
    HPF Technology LLC
    All good ideas. Also, use a quieter op amp for the EQ -- 13 nV/rtHz is not stunning.
  11. A9X


    Dec 27, 2003
    Sinny, Oztraya
    You can do a lot better than a 12AX7. Req of a tube is about 2.5/gm. The first stage in the SVT (an any amp really) will determine the best possible SNR and in this case with a 12AX7 at about 400uA Ik gm will be circa 1.1mA/V or Req = 2270 ohms. Plug that into Boltzman (20kHz / 40*C) and I get 880nV. Add in all the other highish value resistors' noise contribution in the circuit and it's hardly quiet.

    Alternatives would be a 6DJ8 in common cathode (around 10mA/V) or better yet an optimised cascode (look at the front end of an old HP or Tek scope - also Valley/Waldman). If you can deal with the crippling Miller C a 6C45 at about 45mA/V is great, but so are the cascodes and hybrid cascodes I prefer in low level circuits.

    Best option would be to tweak the front end (V1) design using an ECC88/6DJ8 running at decent currnt to get the transconductance up, as well as modifying the pad and putting the mute FET shunted across the wiper of the 250k pot. Not sure if the existing PS would deal with the extra 20mA or so required to do this well, but if it could, I can see an easy 8dB (ottomh numbers) improvement in SNR by doing this alone. Gain is lower though.

    I don't see there's a lot to do in this amp without a major redesign as it has too many (noisy) stages, and an afterthought of a PSU. PSRR in a common cathode stage is quite low (usually under 10dB) and so the PS also needs to be quiet.

    Agree with earlier comments that it's a dog of a drawing, and that the opamps used could be better, and the opamp PS could be better designed and run at a higher voltage, but it works off the heater taps to save money.

    Been tossing around the idea of doing a thorough modern tube DIY bass pre design/build as a group discission/idea like ax84 does with guitar amps. Would anyone else be interested?
    elkkid2 likes this.
  12. deaf pea

    deaf pea

    Mar 24, 2005
    Cuernavaca 1 hr S Mexico City
    Seymour Duncan/Basslines SMB-5A Endorsing Artist
  13. This may explain why the SVP-Pro seems to have limited headroom when the EQ is engaged. Thankfully you can remove the graphic EQ from the circuit and get an all-tube signal path. Also, IIRC, mine does not use OP275s or NJM2060s. I think it has some TL074s in the EQ circuit.

    FWIW, mine (the earlier blackface version with variable D.I. from 1998 or so) is not noticeably noisy.
  14. I'd be absolutely no help - but I'd love to see it. Great Idea.
  15. I would be interested. I'd be happy for a well designed preamp, with a Drive control, and without a fancy graphic EQ. It would have line level balanced out. Something like the Demeter HBP-1, but with tube overdrive
    elkkid2 likes this.
  16. Jerrold Tiers

    Jerrold Tiers

    Nov 14, 2003
    St Louis
    The SVP-PRO is also somewhat dated, so that may not be totally surprising.

    As for the cheap, evidently none of the preamps were cheap enough to get acceptable sales, even the SVP-CL.

    The net result of the higher voltage would be a maximum of about 6 dB added headroom. Since the EQ is followed by the master and a gain stage, the slightly lower headroom is not unduly limiting.

    The real limit on actual output is the output buffer stage, which is a cathode follower having a 150K resistor to ground. A lower resistor value there could allow a higher output into the typical low input impedance input of an SS power amp. The 150k will really limit output into a 10K impedance, which is possible for some power amps.

    However, it looks as if the 12DW7 cannot be used in the location due to the wrong tube half used as the gain vs follower stage. So there is some limit on how low the resistor can be.

    That V5 tube is after the master, so a hum or noise in it will not be controlled by the master. Hence the suggestion of a good tube in that location.
  17. deaf pea

    deaf pea

    Mar 24, 2005
    Cuernavaca 1 hr S Mexico City
    Seymour Duncan/Basslines SMB-5A Endorsing Artist
    I haven't noticed any lack of output level . . . I've got lots more headroom with the SVP-PRO (compared to the SVT-3PRO). So I don't see a NEED for MORE output level . . .
  18. kringle77

    kringle77 Supporting Member

    Jul 30, 2004
    Massena NY
    Yeah, something to make the tube eq section more powerful would be nice. Especially the mid controls. I find them murky.
  19. With mine, I find that the graphic EQ section will clip (LED lights up constantly) if I run a significant amount of Drive (for dirty sounds). This means that the EQ's gain slider has to be taken down to prevent clipping. Seems to me that another 6-10dB of headroom would be welcome in this section.

    BTW, dUg Pinnick from King's X used to use a couple of the SVP-Pros in his rig. Massive bass tone!
  20. Passinwind

    Passinwind I Know Nothing Supporting Member Commercial User

    Dec 3, 2003
    Columbia River Gorge, WA.
    Owner/Designer &Toaster Tech Passinwind Electronics
    I think there's definitely critical mass for that. I'd rather do a solid state one, but I'm still in if you do it with tubes.;)

    I've neither seen nor heard the preamp in question in this thread, so I'm watching rather than talking for now.