Thanks for your help understand this thing. I'm posting a clearer view of the schematic from a file. I'm thinking Ampeg made a couple changes to the amp and glued a schematic to it that was not updated, or vice versa, it's an older circuit board.
In that picture of the internals of the tank, the bottom is the input and the top is the output. The negative lead on the top (output) shows continuity to the tank chassis, and the amp chassis, when it's all bolted together. The bottom (input) is insulated from the tank with a fiberboard piece and the negative outside doesn't touch. AFAIK this is standard for spring reverb tanks. Yes the jacks are RCA.
Yes the schematic is confusing. The fuse is on there as if it's mounted in a holder next the xformer like a normal amp, but it's not. It's hiding under blown white "death" cap. I think this is the is the tank in the schematic, but I can't find the input output leads either. That would be interesting if C13 were backwards.
Does this help clarify anything?
Thanks!
The attachment below is my understanding of the reverb and suggestion for improving the reverb hum. So the driven transducer B+ is tied to the enclosure, and the output jack is isolated from the enclosure, allowing it to be grounded as is depicted in the Ampeg schematic. Is this correct?
So my proposed modification inserts a buffer transistor to deliver a filtered B+ to the reverb case and driven transducer. The newly added transistor will dissipate about 100mW, so I advise any TO-220 transistor for that spot. (A TO92 might be a bit wimpy re dissipation.) Will this be an easy mod?
The attachment below is my understanding of the reverb and suggestion for improving the reverb hum. So the driven transducer B+ is tied to the enclosure, and the output jack is isolated from the enclosure, allowing it to be grounded as is depicted in the Ampeg schematic. Is this correct?
So my proposed modification inserts a buffer transistor to deliver a filtered B+ to the reverb case and driven transducer. The newly added transistor will dissipate about 100mW, so I advise any TO-220 transistor for that spot. (A TO92 might be a bit wimpy re dissipation.) Will this be an easy mod?
Cool - so it's just the addition of one resistor and one transistor? I'm game to try it. It will help if I have one of those to-220 laying around, since I just made an order to Mouser.
Will check it out when I get a fuse in there and fire it up again. Hopefully the hot reverb short didn't blow anything.
That capacitor you showed in post #1 probably expired because of a dielectric failure due to age. Replacements should be X2 or Y2 which are rated for safety.
X2 capacitors are intended to be wired across the mains inputs (live-neutral) and Y2 from either to ground. Although the circuit shows the capacitor connecting to the amplifier ground for safety this should be wired to the common earth point rather than any actual connection in the audio stages. (Y2 type, but I suspect an X2 across the mains input would be better).
You may have a hard time replacing the transistors. I recall that RCA published this sort of circuit with a germanium PNP in TO-3 and a silicon NPN in either TO-66 or TO-220 which can be persuaded to fit in TO-66 footprints. If you did get to change the transistors you could try an MJE3055T for the NPN and an MJ2955 for the PNP, not an ideal match but would fit the locations (if that is indeed what you have). Being silicon, the bias arrangement would have to change to allow for the higher Vbe and two silicon diodes (such as 1N4002) might work although the driver current at 100mA or thereabouts may require higher current diodes for lower forward voltage (1N5402 perhaps) or a shunt resistor across one of the 1N4002's of around 100 ohms, but it would need to be able to handle 100mA and that would mean a relatively high power (at least 1W). Alternatively, a conventional Vbe multiplier transistor.
The driver transistors in TO-5 with their welded heatsinks are obsolete too, so you would need to use perhaps a 2N3019 (still available from Farnell last time I looked) with a heat clip (not a small one either, the dissipation is about 2.5W, and most of the traditional TO-5 heat clips are obsolete: some you might find are probably not low enough thermal resistance to be of use) or you might be able to use a BD139 with a good heatsink (20C/W or less). The input transistors could be BC557's.
X2 capacitors are intended to be wired across the mains inputs (live-neutral) and Y2 from either to ground. Although the circuit shows the capacitor connecting to the amplifier ground for safety this should be wired to the common earth point rather than any actual connection in the audio stages. (Y2 type, but I suspect an X2 across the mains input would be better).
You may have a hard time replacing the transistors. I recall that RCA published this sort of circuit with a germanium PNP in TO-3 and a silicon NPN in either TO-66 or TO-220 which can be persuaded to fit in TO-66 footprints. If you did get to change the transistors you could try an MJE3055T for the NPN and an MJ2955 for the PNP, not an ideal match but would fit the locations (if that is indeed what you have). Being silicon, the bias arrangement would have to change to allow for the higher Vbe and two silicon diodes (such as 1N4002) might work although the driver current at 100mA or thereabouts may require higher current diodes for lower forward voltage (1N5402 perhaps) or a shunt resistor across one of the 1N4002's of around 100 ohms, but it would need to be able to handle 100mA and that would mean a relatively high power (at least 1W). Alternatively, a conventional Vbe multiplier transistor.
The driver transistors in TO-5 with their welded heatsinks are obsolete too, so you would need to use perhaps a 2N3019 (still available from Farnell last time I looked) with a heat clip (not a small one either, the dissipation is about 2.5W, and most of the traditional TO-5 heat clips are obsolete: some you might find are probably not low enough thermal resistance to be of use) or you might be able to use a BD139 with a good heatsink (20C/W or less). The input transistors could be BC557's.
For this do you think a TIP31C or TIP32C would work? I have these in my recycled junk drawer.
Thanks very much for your suggestion on this. Unfortunately, the added transistor did not have any effect on the hum. The problem must be elsewhere. I think I am going to just live with it.
Well, that's disappointing to hear--- no pun intended. I can suggest further diagnostic experiments if you're still so inclined. BTW, do you have a 'scope?
A number of items to explore:
Re reverb hum, if I understand correctly, with reverb control at 0, the hum is acceptable? And I assume the reverb function still performs correctly? That is, the added TIP31 must be powering the reverb tank properly? The tank's revised B+ is roughly +38V?
Is the hum due to inadequate filtering of tank B+ or does it originate from the pickup path? The VR1 volume pot may establish if the hum source comes from the input. If the hum remains internal, a definitive experiment would ground C10 instead connection to VR3 wiper. If hum's still present, tank B+ remains under suspicion. In this event, I suggest lifting negative end of C13 and adding 10uF/50V cap with in series with C13 negative to negative 10uF. The positive end of the added 10uF cap can be used as a probe to search for source of the hum; interesting inspection points would include circuit common, C12+, tank B+, Q4 collector, Q4 emitter.
Another issue is the safety ground. My thinking is that C32 should be deleted and safety ground should be tied to analog common.
A number of items to explore:
Re reverb hum, if I understand correctly, with reverb control at 0, the hum is acceptable? And I assume the reverb function still performs correctly? That is, the added TIP31 must be powering the reverb tank properly? The tank's revised B+ is roughly +38V?
Is the hum due to inadequate filtering of tank B+ or does it originate from the pickup path? The VR1 volume pot may establish if the hum source comes from the input. If the hum remains internal, a definitive experiment would ground C10 instead connection to VR3 wiper. If hum's still present, tank B+ remains under suspicion. In this event, I suggest lifting negative end of C13 and adding 10uF/50V cap with in series with C13 negative to negative 10uF. The positive end of the added 10uF cap can be used as a probe to search for source of the hum; interesting inspection points would include circuit common, C12+, tank B+, Q4 collector, Q4 emitter.
Another issue is the safety ground. My thinking is that C32 should be deleted and safety ground should be tied to analog common.
Last edited:
Okay, I tried the 'definitive' test lifting c10 and grounding instead of lead to VR3. The hum is still present and varies with the VR4 Reverb pot. What does this tell us? Does that rule out the signal chain before the reverb circuit?
I haven't dealt with c32 yet. When I got it, it hummed loudly with the blown capacitor in there, but maybe because it was blown but still attached. Could you clarify what should go to ground there? You mean the ground off the three prong plug which grounds to chassis, to the ground on the circuit?
Your C10 experiment shows hum originates in reverb section. My chief suspect is failed C11, assuming that the added buffer is performing as intended. DC bias checks should add confidence re buffer. If new C11 doesn't yield improvement, the added cap experiment I described in post 29 will allow you to probe around for the source of the hum.
An excerpt from post 4:
And another excerpt from post 8:
A lot of speculation here:
I continue to be baffled by the original explosion of C32. I believe at one time, there must have been a direct connection of C32 across the power line since a 0.047uF cap could not have enough stored energy to self-explode.
Convention normally places the power switch and fuse on the hot side of the of the power line; assuming this convention was followed at design time, the schematic shows C32 tied to the power line, probably to the hot side! I speculate the former owner experienced loud hum with his original 2-wire power cord and contrived an external ground to his guitar as a hum reduction bandaid--- and hence connection of C32 hot-side-power to ground. This probably worked for for him until C32 exploded and/or he got shocked one time too many. Then the 2-wire cord got replaced with a 3-wire with ground.
Since you report that C32 has to be present to avoid loud hum, I'd wager that inspection will show that C32 is now tied to power Neutral (white wire if the power cord has color coding) since if it were tied to hot side, hum would be extreme.
If my conjecture is correct, everything points to probable transformer leakage. An interesting experiment would connect C32 to safety ground. Hum might improve slightly. Connecting circuit common to safety ground might cure hum, but maybe not if leakage current is too large.
And another excerpt from post 8:
A lot of speculation here:
I continue to be baffled by the original explosion of C32. I believe at one time, there must have been a direct connection of C32 across the power line since a 0.047uF cap could not have enough stored energy to self-explode.
Convention normally places the power switch and fuse on the hot side of the of the power line; assuming this convention was followed at design time, the schematic shows C32 tied to the power line, probably to the hot side! I speculate the former owner experienced loud hum with his original 2-wire power cord and contrived an external ground to his guitar as a hum reduction bandaid--- and hence connection of C32 hot-side-power to ground. This probably worked for for him until C32 exploded and/or he got shocked one time too many. Then the 2-wire cord got replaced with a 3-wire with ground.
Since you report that C32 has to be present to avoid loud hum, I'd wager that inspection will show that C32 is now tied to power Neutral (white wire if the power cord has color coding) since if it were tied to hot side, hum would be extreme.
If my conjecture is correct, everything points to probable transformer leakage. An interesting experiment would connect C32 to safety ground. Hum might improve slightly. Connecting circuit common to safety ground might cure hum, but maybe not if leakage current is too large.
Hmm, okay let's see if we can sort out C32. I don't know what happened before I bought the amp a couple weeks ago, but the previous owner seemed none too sorry to be rid of it. I don't think he ever opened the chassis.
It looks like C32 (the one that was in there blown, and now the one I installed) is in the original position labeled on the circuit board itself and is not a modification. The leads from the power cord to the circuit board suggest that the 3-prong is not original.
C32 is wired as in the schematic to the lead from the on/off switch and the fuse on one side, and leading to the rectifier diodes and so on to the filter caps on the other end. So yes, neutral is as it should be, right? That white lead on the switch is from power cord neutral to C32. So I can't just delete it can I? It's not unsafe in this condition, right?
On the reverb, I tried to replace C11 with a 100uf 100v cap (says 64uf on schemo, 47uf in the amp) but then reverb didn't work at all. Maybe bad cap. Jumpering a cap in parallel has no effect. I have limited close spares to work with.
Using the DIY signal tracer probe, I did find the hum at C11 at the 'earliest' in the circuit, if I'm not mistaken, as you suggest. But then much louder on on R19, R20... Am I just hearing the unfiltered 43v AC there? I didn't keep the buffer TIP31 in there, just alligator clipped it in for the experiment but it didn't have an effect. It might be easier if I just post of video of using the signal tracer, to show where the hum 'begins.'
Hi szegedin,
Thanks! I've copied your text and will embed some replies:
Hmm, okay let's see if we can sort out C32. I don't know what happened before I bought the amp a couple weeks ago, but the previous owner seemed none too sorry to be rid of it. I don't think he ever opened the chassis.
It looks like C32 (the one that was in there blown, and now the one I installed) is in the original position labeled on the circuit board itself and is not a modification. The leads from the power cord to the circuit board suggest that the 3-prong is not original.
C32 is wired as in the schematic to the lead from the on/off switch and the fuse on one side, and leading to the rectifier diodes and so on to the filter caps on the other end. So yes, neutral is as it should be, right? That white lead on the switch is from power cord neutral to C32. So I can't just delete it can I? It's not unsafe in this condition, right? To prove the power wiring is what we believe, would you measure AC volts between safety ground and the white lead tied to the power switch? Assuming it's neutral, there should be only a small difference voltage between safety ground and neutral. The presence or absence of the 0.047uF won't have any effect on safety. Said differently, having the cap present wouldn't protect you. Using the guitar amp with safety ground connected to circuit common would be analogous to using a metal-cased power tool having a safety ground--- you should be protected. A GFI breaker still seems a good precaution.
On the reverb, I tried to replace C11 with a 100uf 100v cap (says 64uf on schemo, 47uf in the amp) but then reverb didn't work at all. Maybe bad cap. Jumpering a cap in parallel has no effect. I have limited close spares to work with. This I don't understand. If the new cap was defectively shorted, there would be no function. But I would think R19 would become noticeably hot if the cap were a short. You should be able to parallel a new cap with the original--- the more capacitance the better. Be certain the supply to R20 is taken from the buffer transistor. If it's left connected to +43V, there will be very little improvement in hum. Whenever an experiment doesn't make sense, be sure to probe setup with DVM to ensure circuit is powered as you intend.
Using the DIY signal tracer probe, I did find the hum at C11 at the 'earliest' in the circuit, if I'm not mistaken, as you suggest. But then much louder on on R19, R20... Am I just hearing the unfiltered 43v AC there? I didn't keep the buffer TIP31 in there, just alligator clipped it in for the experiment but it didn't have an effect. It might be easier if I just post of video of using the signal tracer, to show where the hum 'begins.' That you found lower hum at C11 than at +43V is encouraging. I'm going to hope that you forgot to move R20 power to the buffer. You have a DIY tracer? Would you provide details?
Thanks! I've copied your text and will embed some replies:
Hmm, okay let's see if we can sort out C32. I don't know what happened before I bought the amp a couple weeks ago, but the previous owner seemed none too sorry to be rid of it. I don't think he ever opened the chassis.
It looks like C32 (the one that was in there blown, and now the one I installed) is in the original position labeled on the circuit board itself and is not a modification. The leads from the power cord to the circuit board suggest that the 3-prong is not original.
C32 is wired as in the schematic to the lead from the on/off switch and the fuse on one side, and leading to the rectifier diodes and so on to the filter caps on the other end. So yes, neutral is as it should be, right? That white lead on the switch is from power cord neutral to C32. So I can't just delete it can I? It's not unsafe in this condition, right? To prove the power wiring is what we believe, would you measure AC volts between safety ground and the white lead tied to the power switch? Assuming it's neutral, there should be only a small difference voltage between safety ground and neutral. The presence or absence of the 0.047uF won't have any effect on safety. Said differently, having the cap present wouldn't protect you. Using the guitar amp with safety ground connected to circuit common would be analogous to using a metal-cased power tool having a safety ground--- you should be protected. A GFI breaker still seems a good precaution.
On the reverb, I tried to replace C11 with a 100uf 100v cap (says 64uf on schemo, 47uf in the amp) but then reverb didn't work at all. Maybe bad cap. Jumpering a cap in parallel has no effect. I have limited close spares to work with. This I don't understand. If the new cap was defectively shorted, there would be no function. But I would think R19 would become noticeably hot if the cap were a short. You should be able to parallel a new cap with the original--- the more capacitance the better. Be certain the supply to R20 is taken from the buffer transistor. If it's left connected to +43V, there will be very little improvement in hum. Whenever an experiment doesn't make sense, be sure to probe setup with DVM to ensure circuit is powered as you intend.
Using the DIY signal tracer probe, I did find the hum at C11 at the 'earliest' in the circuit, if I'm not mistaken, as you suggest. But then much louder on on R19, R20... Am I just hearing the unfiltered 43v AC there? I didn't keep the buffer TIP31 in there, just alligator clipped it in for the experiment but it didn't have an effect. It might be easier if I just post of video of using the signal tracer, to show where the hum 'begins.' That you found lower hum at C11 than at +43V is encouraging. I'm going to hope that you forgot to move R20 power to the buffer. You have a DIY tracer? Would you provide details?
Yet another rethink here. Tell me to go away if I'm becoming irksome.
Maybe I'm on the wrong track in suspecting the 43V B+. Q4 provides current source drive to the reverb driven transducer. In principle, it provides some rejection to B+ hum, but I just assumed it was insufficient and proposed buffer as remedy.
But this may be another simple experiment: if I understand correctly, the reverb box is connected via RCA plugs. Leave the receive pickup connected, but unplug the source cable. (Make sure the shell of the plug can't touch any circuitry--- it's voltage hot.) What happens re hum? Confirm with your meter that the reverb case is floating, and try grounding the reverb case. Any insights? Another experiment is to unplug pickup RCA and ground its center. Hum should be absent at all reverb gain settings. What about the reverb pot's case? is it grounded?
Maybe I'm on the wrong track in suspecting the 43V B+. Q4 provides current source drive to the reverb driven transducer. In principle, it provides some rejection to B+ hum, but I just assumed it was insufficient and proposed buffer as remedy.
But this may be another simple experiment: if I understand correctly, the reverb box is connected via RCA plugs. Leave the receive pickup connected, but unplug the source cable. (Make sure the shell of the plug can't touch any circuitry--- it's voltage hot.) What happens re hum? Confirm with your meter that the reverb case is floating, and try grounding the reverb case. Any insights? Another experiment is to unplug pickup RCA and ground its center. Hum should be absent at all reverb gain settings. What about the reverb pot's case? is it grounded?
Thanks for the consultations - not irksome at all, on the contrary!
Okay I put a 470uf cap in C11; I don't think it had an effect. No luck with the buffer idea.
Yes the reverb tank chassis is grounded to chassis through the return ground lead. With tank input unplugged, it still hums with VR4 turned up. When I unplug return cable and ground to chassis, hum is still present with VR4 turned. It does seem like just a grounding issue somewhere more than anything else. Reverb pot case is grounded.
This amp chassis is not easy to get in and out, and has to be come fully in and out every time - I broke the panel-mount reverb pot now from repeated removal. I had another amp that I could not eliminate reverb hum - these are just inherently noisy things. At a certain point one does have to give up.
I still have to address some DC voltage on the input because of the JFET, I think. The guitar volume pot causes scratchiness. I didn't put a diode in circuit with that JFET, just dropped it in, in line with suggestion on other thread. I read about 10-15mv DC on the guitar input.
It is a quiet amp now, with the JFET, which is satisfying.
You may have a different opinion, but I think you're making good headway! I think I've led you astray and I apologize for not being more systematic. I should have thought of the testing Q5 by grounding C13(-) much earlier. That's a good experiment for assessing remain hum.
Roughly estimated, the Q5 stage has gain of about 150, or about 43dB. The power amp stage presents another 21dB, so there's about 64dB of gain from Q5 to the speaker--- enough to be challenging. If hum is acceptable with reverb pot at 0 but excessive when at max, then hum is in the Q5 stage. Suspect inadequate filtering at C20. Radiated hum from the transformer is another possibility. Once you're satisfied that hum is acceptable or as good as it will get, return to reverb issues.
Check for DC access pot elements as scratchy pots can be aggravated by DC leakage from blocking caps.
If hiss performance is acceptable, especially if the 120K remain in place, don't worry about the protection Zeners.
Good luck.
Roughly estimated, the Q5 stage has gain of about 150, or about 43dB. The power amp stage presents another 21dB, so there's about 64dB of gain from Q5 to the speaker--- enough to be challenging. If hum is acceptable with reverb pot at 0 but excessive when at max, then hum is in the Q5 stage. Suspect inadequate filtering at C20. Radiated hum from the transformer is another possibility. Once you're satisfied that hum is acceptable or as good as it will get, return to reverb issues.
Check for DC access pot elements as scratchy pots can be aggravated by DC leakage from blocking caps.
If hiss performance is acceptable, especially if the 120K remain in place, don't worry about the protection Zeners.
Good luck.