Restoring a 1964 Ampeg Reverberocket

[Schlumpfmeister]

Ok, back after a week of absence waiting for parts. I since replaced all electrolytics. For the two filter caps in paper left and right I took the inside out and stuffed 2x10u each in, so it even looks original.

So when I do an old amp it gets every electrolytic replaced. The cathode bypass caps really don't hurt anything if they blow but I've had so many issues go away with a fresh set.

Problem is the can, because now it buzzes. I was very careful to preserving the complete wiring. Turns out the can housing serves as star point for ground, as 3 of 4 outer legs are being utilized to connecting all kinds of black wire, then one of them connecting to the amp chassis with a very short wire. I kept this all in place which was a nightmare to work on, partially turning the amp upside down and soldering from the outside for the connections deep into the far end of the housing, then mounting it mechanically, then connecting the rest of the leads that are in reach from the top. And then it buzzed. So I suppose I either have a ground connection somewhere or a bad soldering joint, so back to square one and all over once more.

On a side note, the mechanical restauration is also pretty messy, I have it re-tolexed as someone had it covered with black leatherette. It took me 4 hours and all sorts of chemicals to eliminate the old (elastic, sticky) glue, just to learn from the amp builder doing the tolexing that he will spend another (paid for) 2 hours!

Oh if that amp uses a 5y3 rectifier tube,...

No, R12 R-B, it uses an integrated silicon diode network, the part is really tiny!

I know about the issue, rectifier tubes have a maximum filter capacity specified to limit startup current. I had that with a Deluxe Reverb that had gotten recapped using 100u filter caps. It displayed a lighting beam at startup!

 

[Schlumpfmeister]

In the States your amp wouldn't have had a grounded cord, other countries might have been ahead of the US in that regard. We didn't get them until around 1970 or so. Then of course a lot of them got cut off so you could still plug in to old outlets.

Yep, the 3 prong cable with internal wiring and ground connection looks original. Yet it still had a death cap…. Belt and braces as we say in Germany…

 

[Schlumpfmeister]

It has been 2 weeks now since my last posting and I am seeking some advice. I don't remember if the amp was buzzing / hissing before I replaced some of the tubes and all of the electrolytics, but now it does. I was very careful to maintain the whole wiring concept. The lead dress is not real tidy, but again, I haven't changed it from where it was. Here is what I have tried and observed:

If I pull V4 (PI) or if I ground pin 1 (PI input), the amp is dead silent.

If I Pull V3 the hiss/buzz reduces substancially.
If I pull V2, nothing changes.

If I gound pin 1 of V3, nothing changes.

Therefore I expected the problem to be somewhere at V3. I replaced the plate resistor, no change.

I doubled all of the filter caps individually with a spare 100u cap, one by one, nothing changes. I have been worried about the can cap as the CE mfct cans seem to have a bad repuation. Wasn't confirmed.

I physically tried all components by rattling, no effect.

What else could I try to find the source of the problem? The hiss/buzz increased considerably when I get closer to the input plugs. I always have one input short for all of the tests though.

By chance I gounded one of the inputs - most of the hiss is gone, what remains is a slew of 100Hz and overtones (200, 300, ..., peaks at 100 and 600Hz. We have 50Hz in Germany).

So this seems to be an input-ground problem (which I can easily fix, see above discussion), and filter buzz. But why and from where, given that all caps are new and I have tested by doubling the caps with no effect?

Thanks for suggestions!

Best Andreas


PS: I found the following errors in the schematics published above (some have been mentioned already)
- V1 1 and 2 reversed
- V1 1 plate voltage 110v cannot be correct, I read 155v and I have checked all my resistors involved
- death cap missing
- connection from V2 plate pin 2 to input V3 grid pin 1 not existent (it is also not shown in the original schematics)

 

[King Fan]

Re the filter hum, have we heard from @BobbyZ ? He knows a lot about RRs, and IIRC he maybe mentioned something about the original grounds being very sensitive to manipulation — or important to restore exactly?

 

[Schlumpfmeister]

By now I suspect leakage from the new CE can cap, so I will replace it with the original again. If it then works, I will do a cap job using individual parts. I see no other reason why it should hum with an even multiple of 50Hz!? I will report back as soon as I have tried it out.

 

[Schlumpfmeister]

Ok, update after an (spoiler: unsucessful) soldering night:

By now I suspect leakage from the new CE can cap, so I will replace it with the original again. If it then works, I will do a cap job using individual parts. I see no other reason why it should hum with an even multiple of 50Hz!? I will report back as soon as I have tried it out.

I bonded out all 4 connections to the can cap so that it was no longer used and connected individual caps outside of the amp. I used 100u + 100u instead of 60u + 40u and thus it did buzz a little less, but not much. So the cap can seems ok and there seems to be no grounding issue at the cap can.

...I was very careful to preserving the complete wiring. Turns out the can housing serves as star point for ground, as 3 of 4 outer legs are being utilized to connecting all kinds of black wire, then one of them connecting to the amp chassis with a very short wire....So I suppose I either have a ground connection somewhere or a bad soldering joint, so back to square one and all over once more.

Here is an interesting observation: Pretty much all grounds on the board are daisy-chained - with the exception of one section, related to tremolo, reverb and PI. There, a the 2 tremolo pots have lugs connected to their housing, a tone cap is connected to the related pot housing, and a ground connection going from the the reverb pot housing to the related section on the board. It is not a ground loop, without this connection, the related section of the board wouldn't be grounded at all. I changed this such that all ground connections follow the star rule, see pic. Unfortunately, no change. So I wonder if I should revert back, the Ampeg guys must have had a plan when they did the ground concept the way they did it?

Here are all the steps I have taken, see pic:
- removed ground link from reverb pot housing to board (replaced by red crocodile cable for now daisy-chaining ground on the board)
- moved chassis ground of tone cap to board ground on the outer lug of the volume pot
- took the free ground connection and board-grounded the ground lugs of both tremolo pots and reverb pot

Any opinions why it was done the way it was done and if I should revert back?

I also disconnected the input plugs from the chassis, now it should be true star grounding, but the amp is then unstable, the input plugs seem to require the chassis connection!?

...By chance I gounded one of the inputs - most of the hiss is gone, what remains is a slew of 100Hz and overtones (200, 300, ..., peaks at 100 and 600Hz. We have 50Hz in Germany). So this seems to be an input-ground problem (which I can easily fix, see above discussion), and filter buzz. But why and from where, given that all caps are new and I have tested by doubling the caps with no effect?

Unfortunately, I was unable to figure out a input section configuration that allows both for grounding unused inputs and not substancially loading down the grid of the input tube.

So 5 hours tinkering and soldering and no success. At this point, I have no clue how to reduce hum and hiss any more. Filtering seems ok, grounding is almost ideal now.

Any suggestions?

 

[timfred]

Long shot, but I had one of these with a partially failed rectifier. Bad hum, replaced with discrete diodes, fixed.

 

[Schlumpfmeister]

Long shot, but I had one of these with a partially failed rectifier. Bad hum, replaced with discrete diodes, fixed.

Thanks for the hint, but I don't think this is the problem: As stated above, if I ground the input to the PI, the whole amp is dead silent. If I ground the input 1 to V3 which is one stage prior to the PI, all the buzz and hiss is there. I have replaced the V3-1 plate resistor with no improvement.

 

[Schlumpfmeister]

Pretty much all grounds on the board are daisy-chained - with the exception of one section, related to tremolo, reverb and PI. There, a the 2 tremolo pots have lugs connected to their housing, a tone cap is connected to the related pot housing, and a ground connection going from the the reverb pot housing to the related section on the board. It is not a ground loop, without this connection, the related section of the board wouldn't be grounded at all. I changed this such that all ground connections follow the star rule, see pic. Unfortunately, no change. So I wonder if I should revert back, the Ampeg guys must have had a plan when they did the ground concept the way they did it?

Since I have reverted back to the original grounding setup.

As noted above, if I ground the input to the PI, the amp is dead silent. If I ground the input to V3-1 which is the stage right before PI, I get all of the hiss and buzz. So I subsequently replaced
- 47k plate resistor to V3-1
- 68k power supply resistor prior to the plate resistor
- 5k power supply resistor prior to the 68k resistor
- initial 1k power supply resistor prior to the 5k resistor
Now there is no input from anything prior to V3-1 as its input is grounded right at the socket, and all of the filter caps and plate resistor as well as voltage reduction resistors into V3-1 have been replaced - and it still buzzes and hums. As stated above I have tested the function of the filter can cap and it seems ok.

At this junction I am clueless, I think I have tried everything now, there are only 5 other components involved with V3-1,
- the cathode resistor,
- a resistor connecting the cathodes of V2-1 and V3-1 (no clue what it is for?!)
- the coupling cap
- a 1nF cap from the plate to ground,
- and a 470k resistor from the coupling cap output to ground

I will replace those also, but not with great hope!?

 

[NTC]

Have you tried testing the amp with a temporary extra filter cap in parallel with the one that filters the voltage to V3?

 

[Schlumpfmeister]

Have you tried testing the amp with a temporary extra filter cap in parallel with the one that filters the voltage to V3?

Thanks for the hint. Yes, of course. Also different tube. No effect.
I am getting to the point of accepting the fact that this (or these?) amp is a bit noisy.

 

[BobbyZ]

Finding the source of hum is something I'm absolutely no good at. Seriously I'm completely an idiot at it!
Here's how dumb I am.
Get the electrolytic caps in my second early Reverberocket and it's humming like crazy. So I cleaned pot and input jack grounds and a bunch of things like that, probably clipped a rest lead or ten to ground and tried that, I know I pulled a lot of hair out!
Finally gave up for the day just before becoming completely suicidal. Go back to the basement a day or two later and had a light bulb moment. From the buzzing of the florescent light above the amp.
Shut that $%&\£€ light off and all was well!
Put the back of the amp on with its tin foil shield and it was fine.
That's when I figured out how important that shield is on a chassis. Aluminum furnace tape is a good replacement for the old Reynolds Wrap Ampeg used, normally it's ripped up.

Other than that I'm just guessing on hum.

 

[timfred]

- a resistor connecting the cathodes of V2-1 and V3-1 (no clue what it is for?!)

* V2-1 splits the signal into wet (plate to the reverb stage) and dry (cathode follower to V3-1).
* V3-1 is a summing stage with 2 inputs: pin 1 (grid) and pin 3 (cathode). On the grid is the wet return from the reverb stage. On the cathode is the dry signal from V2-1.

If you only grounded V3-1 pin 1, then you only eliminated the reverb stage. Turning Dimension down to 0 would have done the same thing. If you want to eliminate the entire preamp, you should lift a leg of the cathode resistor connecting these stages (R15 - 22K in original schematic). EDIT: Or yank V2 tube completely.

 

[timfred]

I still think this could be a noisy power supply. Just because the power section (PI + Power tubes) is silent when you ground the PI input does not mean there is not residual PS ripple. One feature of a push-pull amp is that PS ripple will be cancelled out in the power section if it is well-balanced.

What are your AC voltages at the power nodes A-F?

 

[Schlumpfmeister]

... back to the basement a day or two later and had a light bulb moment. From the buzzing of the florescent light above the amp.
Shut that $%&\£€ light off and all was well!
Put the back of the amp on with its tin foil shield and it was fine.
That's when I figured out how important that shield is on a chassis. Aluminum furnace tape is a good replacement for the old Reynolds Wrap Ampeg used, normally it's ripped up.

Thanks BobbyZ, yes, the amp is super sensitive to ambient conditions, just waving with my hands above the chassis already induces extra hum. I have however tried getting rid of all outside disturbances (just tried again a minute ago), no computers, monitors, light sources on, also covered with aluminium foil, no change to the noise pattern though.

* V2-1 splits the signal into wet (plate to the reverb stage) and dry (cathode follower to V3-1).
* V3-1 is a summing stage with 2 inputs: pin 1 (grid) and pin 3 (cathode). On the grid is the wet return from the reverb stage. On the cathode is the dry signal from V2-1.

If you only grounded V3-1 pin 1, then you only eliminated the reverb stage. Turning Dimension down to 0 would have done the same thing. If you want to eliminate the entire preamp, you should lift a leg of the cathode resistor connecting these stages (R15 - 22K in original schematic). EDIT: Or yank V2 tube completely.

Thanks timfred! I was wondering all along how the wet/dry signals got mixed! I have tried pulling the connecting resistor yesterday with no improvement. I just tried dimension = 0 and V2 pulled out and the amp buzzes a bit more w/o V2 then with V2 plugged in, but no big difference.

Turning dimension all the way up increases buzz to some extent, maybe plus 50% or so.

I still think this could be a noisy power supply. Just because the power section (PI + Power tubes) is silent when you ground the PI input does not mean there is not residual PS ripple. One feature of a push-pull amp is that PS ripple will be cancelled out in the power section if it is well-balanced.

What are your AC voltages at the power nodes A-F?

Here the measures, but my multimeter is limited to 500v AC, so the only trustworthy (if anything) AC measure would be for E which I have highlighted:
A: 342v DC / 757v AC
B: 341v DC / 751v AC
C: 256v DC / 553v AC
D: 249v DC / 544v AC
E: 131v AC / 278v AC
F: 322v AC / 709v AC

I had firmly expected something close to 0v AC, why are the readings so high?

As stated above I have tried testing the (new) can cap and it seemed ok:

I bonded out all 4 connections to the can cap so that it was no longer used and connected individual caps outside of the amp. I used 100u + 100u instead of 60u + 40u and thus it did buzz a little less, but not much. So the cap can seems ok and there seems to be no grounding issue at the cap can.

 

[Schlumpfmeister]

I since also replaced the rectifier with 4 discrete diodes - no improvement.

 

[timfred]

I had firmly expected something close to 0v AC, why are the readings so high?

Yes, AC voltages at those nodes should be very small - 0-15 VAC maybe. Those readings are what I would expect if the filter caps were not grounded at all.

 

[timfred]

Also (after you fix the power supply):
- Every single red "Tiny Chief" capacitor I have ever seen has been bad.
- Every 470K+510K resistor pair in RR PIs I have seen have drifted in value far enough to cause poor PI balance.

 

[Schlumpfmeister]

Yes, AC voltages at those nodes should be very small - 0-15 VAC maybe. Those readings are what I would expect if the filter caps were not grounded at all.

I took the new can cap out - this is the one part I really didn't want to touch again, everything is so tight! I bonded all connections out and swapped the old and the new can cap in, no difference, it always hums, also, the E voltage reading stays the same as shown above.

I am officially giving up now. I cannot see what else I could try. All caps have been replaced, all resistors have been measured, there are no obvious broken cables or connections, all mass points are connected, everything is grounded. Maybe I am just too picky on the noise level and this amp experiences a certain hum and hiss level?

Also (after you fix the power supply):
- Every single red "Tiny Chief" capacitor I have ever seen has been bad.
- Every 470K+510K resistor pair in RR PIs I have seen have drifted in value far enough to cause poor PI balance.

Thanks for the hints!

I have replaced all of the red "Tiny Chief" caps now.

I unsoldered the two resistors and their readings are 493kOhms and 536kOhms. Do you think that this is ok or should I replace them, I could select some that fit rather tight with the spec?

 

[slider313]

Is there hum or is it hiss and buzz? Is it audible with the volume control turned all the way down?