Hoffman AB763 Twin Reverb HUM when intensity is turned all the way down

[proaudioguy]

Last update for now.


Everything is back together.  The idle rizz noise needs to go but overall its all much quieter.


The speakers move when the tremolo is on and I really would like as many people as possible that have built these 6G16 Tremolos to tell me if that’s normal.



I adjusted the BIAS DOWN (hotter plate disappation) and the ticking from the JBL came back and then I adjusted it further and it went away again, but by that point its too weak.  I plan to make tweaks to the oscillator circuit to get stronger tremolo when biased hotter, later on.  The point is, there is a spot where the trem is moving the speakers in and out I guess more approaching a square wave perhaps, and the JBL pops.  Lower bias and its OK, higher bias and its OK but the trem intensity is too low. 


I intend to get another JBL for this amp cause the JBL just sounds better than the EVs and the amp is insanely heavy with 2 EVs in it.  These JBLs are lighter.  The EVs tend to be a bit harsh and the JBL mellows that without taking out the sparkle.  In fact I can run with the bright switch OFF.  With the EVs I had to have the bright switch on to get the sparkle but then turn down the highs to get away from harsh and it was a moving target based on how loud it was.  The JBL just seems to work.  I guess that’s why so many use the D120F.  This is a K120.  It works well with the EV but I need to find another one to try it with both.  End of the day I will use whichever combination sounds the best.  SO, I think I need to figure out why the speakers are moving.


Its as if the oscillator frequency itself is hitting the speakers as AUDIO.  Perhaps the frequency is just to low to reproduce audible sound from these speakers. I need to get an O-scope I guess.  I actually have one that has never been plugged in since I’ve had it.  Its so old, I assume it doesn’t work.  Buddy gave it to me.  I would like to have a 1MHz compact dual trace unit and I would love it to be analog.  I think that might be a unicorn though.

[ac427v]

I'm too slow--you already closed up. But here are ideas for next time...I avoid having signal and high voltage wires run parallel to each other. The four power tubes in your Twin Reverb might benefit from rewiring to reduce possible interaction. You could place each 470 ohm screen resistor on the tube socket and feed them with a single wire. This would daisy chain from one tube to the next and be located between the socket and the back chassis wall. Daisy chain the signal wires to each tube pair but locate them in front of the sockets. The wires are still parallel but far fewer of them and spaced much farther apart. Sluckey shows this very nicely on his layout diagram for a 6V6 Plexi, attached.My other idea is Sluckey-related also. No originality here! I have had great success using his Trem-O-Nator Circuit to wiggle the signal in the preamp stage. The tremolo volume swells are very smooth, like a good bias-vary trem but without affecting the amps bias or fading out at high volume. The circuit only uses one triode so the downside is finding a use for the extra triode

[proaudioguy]

I'm too slow--you already closed up. But here are ideas for next time...I avoid having signal and high voltage wires run parallel to each other. The four power tubes in your Twin Reverb might benefit from rewiring to reduce possible interaction. You could place each 470 ohm screen resistor on the tube socket and feed them with a single wire. This would daisy chain from one tube to the next and be located between the socket and the back chassis wall. Daisy chain the signal wires to each tube pair but locate them in front of the sockets. The wires are still parallel but far fewer of them and spaced much farther apart. Sluckey shows this very nicely on his layout diagram for a 6V6 Plexi, attached.My other idea is Sluckey-related also. No originality here! I have had great success using his Trem-O-Nator Circuit to wiggle the signal in the preamp stage. The tremolo volume swells are very smooth, like a good bias-vary trem but without affecting the amps bias or fading out at high volume. The circuit only uses one triode so the downside is finding a use for the extra triode

RIGHT!  I always thought of those high voltage lines as DC but I realize more recently there isn’t enough filtering on them and that could certainly cause some buzz, only thing is when all the little tubes are out, its dead silent.  The power tubes aren’t amplifying their own noisy power, but I wonder if some of that noise is getting into prior stages.  I still suspect grounding.


Heck, I’m still baffled as to why my TRS foot pedal setup was so problematic.  It was exactly the same as the twin reverb reissue, and that works fine.  I got fed up trying to figure it out and re-wired it to vintage, which filled the other RCA hole on the chassis, which is good.  Magically it fixed a terrible footswitch pop when turning off the tremolo.


My number one question right now is if the 6G16 tremolo is supposed to be MOVING the speakers when nothing is plugged into the amp.


In regards to Slucky’s tremonator.  Once all the noisy bits are gone, then I start with mods.  The first will be to do all the STRONG tremolo mods to facilitate a stronger bias setting.  I will put an LED on the cathode, etc.  We’ll see how that works.  If all is said and done and I pick up a real set of 6L6 GCs and I can’t get them biased AND have the tremolo I like, I will change to Slucky’s mod.  Next will be channel 1 tone stack mod to Marshall or tweed or bassman.  I’m leaning toward the marshall stack.
Next after that, I might try something else like using the polarity switch on the back panel to switch NFB or replacing it with a knob for the same purpose.  I really don’t want to make the amp TOO versatile though.  I am OK with a 1 trick pony here.  I think I will ultimately end up with a dumble clone, or I might start from scratch with another fender and impliment all the early dumble mods to tighten up the bass and the changes to the tone controls, and cascading the channels, which is something I’ve wanted to try for 25 years on this amp actually, which was why I originally bought it.  I’m not sure I’ll be adding any pedal PAB or GAIN mods to this one though cause it sounds SO SO good as is and just needs this noise gone and to be able to be biased hot with the correct power tubes.  It will be in my rental stock.  I don’t really rent backline but I might do it.  I have a dream of having a rehearsal studio too.  No one really does that right where I want to be living.

[proaudioguy]

Hey guys, I made a video so you could hear the noise in reference to the actual sound of the amp.  I suck at playing but I know you guys are tone fanatics.  I actually changed the setup of the guitar the other day which killed a lot of my tone but its a LOT easier to play now.  I’ll deal with it.  I still have another one with fat strings and high setup.

&feature=youtu.be


Let me know what you think of the amp.  I just gotta finish getting rid of that rizz/buzz and maybe I can get the hiss down a tiny bit more with some more metal film resisters.
The knob settings in the photo are the way it was set at after I turned the speed down on the tremolo.

[proaudioguy]

Guys, I think I found a small error.  Some years ago I had one of the 470Ohm resisters smoke when I first got it going.  I think it shorted on a chassis screw.  I replaced with a carbon comp in the parts kit I bought (which had a lot of errors, one of which I missed apparently).  Looks like I used a 47k.  I found it looking at the gut shot I posted earlier in this thread.  I’ll measure it when I open the amp.


What is the effect of the second grid one one of the power tubes being 47k instead of 470?  Would that cause enough of an imbalance to make the rizz I’m chasing audible, if just barely?


I’m taking your advice and re-wiring the octals in the more vintage way like the sluckey attachment.


So does that look like an Orange ring or a brown ring on that carbon comp?  I’ll measure it.  That’s my SOP before installing so its hard to believe I missed it.


EDIT: its brown, just looks orange int eh drawing.  Ordering all the parts to swap now.

[proaudioguy]

Better?

[proaudioguy]

I tried wiring the bias like the 6G16 and it brought a little bit of the hum back so I left it as I left it with the 2 caps cause its nice and quiet with intensity up or down, other than that rizz I am chasing.




I have a different problem.


This was there before I just did these changes today. 


If I crank up the amp LOUD it can jump to a BUZZY distortion sound.  I can play a note and keep turning it up while I keep playing the note over and over and if I play the note hard enough with the volume up enough the tone, which is awesome converts to this nasty buzz.  I raised the bias up a bunch to see if that was related but its not.  I didn’t mention it because I thought it might be a speaker issue, but its not. 


The buzzy distortion is the worst thing I’ve ever heard.  If I let the note ring when it drops enough it turns back into normal tone.


 This distortion is far worse than the idle noise issue.  I’ll try swapping some tubes with another amp just to make sure its not V4B or V6.

[proaudioguy]

Removing V3 (reverb driver) eliminated this buzzing distortion issue.  I’m thinking blocking distortion but I’ve never heard it before so I can’t say for sure that’s what it is and I don’t have a working scope to track it down.  I did move the lead dress around a bit a couple weeks go after some comments so maybe I created the problem.  I swapped in another 12AT7 and that is a known excellent tube out of another amp and the problem persists.

&feature=youtu.be
I am turning it up gradually while hitting the A string.  The difference in volume from the first not to the later notes is substantial but the phone compresses it.

Can anyone confirm this is blocking distortion?  Is this a thing that happens on the reverb driver?  The amp has to be way loud before it happens and I know even if the reverb return is down all the way, the driver is getting nailed by V2.  I’m not sure how it ends up in the signal path though.  Does it go back up from the grid and then get amplified by V4B along with the guitar signal?  Should I add a resister between the 500pF and the 1Meg at the grid?  If so, what size?  I think I’ll go dig up a chop stick and move the wire around while hitting the offensive note and see if that makes any improvement.

[proaudioguy]

I moved the leads around and separated them on V3, but of course the 2 loops that cross the top of the tube can’t really be moved.  It made no difference.  I played for a few seconds without the tank plugged in and that also got rid of this distortion.  That confuses me even more.  If it is oscillating, I would think it would be more likely to do it without the tank plugged in.  I didn’t want to burn up the transformer so I kept that to about 5-10 seconds.  FWIW, the reverb cables are braided shield type.  I can’t see them being an issue.  I checked at lower volume that the reverb works good.  I then used the foot switch to turn it off which shorts the return.  Cranked it up and the issue is still there.

[pdf64]

That’s not blocking distortion.
It’s either a bad connection (may be internal to a component) or a parasitic oscillation.
Is it vibration related, eg does moving the amp chassis away from the speaker cab, or plugging into a different, physically separated cab, affect it?

[proaudioguy]

https://el34world.com/Forum/index.php?topic=23989.msg281848#msg281848


I’ll plug into another cabinet first but someone else on FB said to put in a 10K grid resister between the 1meg and the grid on V3 (which is V2 on that Princetion). Of course the sound clip is gone so I can’t hear if its the same.

[proaudioguy]

So this is strange.  I clipped in a 10K, then 5K, then 2.8K, then for the heck of it cause someone said 200Ohms might work, I tried 470.  They all worked.  So I soldered in 470, doesn’t work.  So I soldered in 5600, doesn’t work.  So I soldered in 10K, doesn’t work.  Clipping in the reverb works, and there is no distortion.  Soldering in, I get the distortion.  I have to completely remove the amp each time and put it back in the cabinet because if the reverb isn’t plugged in, there is no issue so testing outside of the box is of no use.

[PRR]

> if the reverb isn’t plugged in, there is no issue so testing outside of the box is of no use.

Why? Is the tank glued to the box with too-short cables?

[proaudioguy]

Yea, I’m not pulling the reverb tank or the cables out.  Its the original 1968/ 69 reverb tank from this cabinet and everything is nice and tidy.  Every time I pull a screw out it goes in looser than before.


I pulled the cathode bypass cap on V3 (as Sluckey suggested in that other thread) and the problem is gone (the 10K grid stopper is still there) but the reverb does NOT sound as good as it did before.


Perhaps Sluckey will explain what is going on here?



I cannot understand why I could clip the resister in with alligator clips and the reverb would work and sound great, then when I soldered it in, it would go back to this mess.

[pdf64]

I’ve encountered that same issue previously, and ascribed it to ‘parasitic motorboating’ (my conjoined terminology to describe an unusual merging of instability  subsets), via the 0V return rather than the HT supply.
Due to the V3 reverb driver being supplied via the screen grid HT node, which makes the 0V return path for its HT current rather convoluted and intermingled with preamp 0V returns.
Note that the fine engineers of the SF era got rid of V3 cathode bypass.
My solution was to add a small resistor (470ohm) in series with the cathode bypass cap, so as to limit peak currents.
I used to have a sketch of that arrangement on photobucket, but they’re charging even to access the old stuff now. Let me know if you’d like it uploading somewhere else, and I’ll get around to it

[proaudioguy]

I’ve encountered that same issue previously, and ascribed it to ‘parasitic motorboating’ (my conjoined terminology to describe an unusual merging of instability  subsets), via the 0V return rather than the HT supply.
Due to the V3 reverb driver being supplied via the screen grid HT node, which makes the 0V return path for its HT current rather convoluted and intermingled with preamp 0V returns.
Note that the fine engineers of the SF era got rid of V3 cathode bypass.
My solution was to add a small resistor (470ohm) in series with the cathode bypass cap, so as to limit peak currents.
I used to have a sketch of that arrangement on photobucket, but they’re charging even to access the old stuff now. Let me know if you’d like it uploading somewhere else, and I’ll get around to it

Right so its as if the V3 is actually interacting with the V2 and that is what is making it audible?


OK so assuming a stock AB763 Reverb circuit, are you lifting the cap at the ground connection and inserting the resister, or at the cathode end?  I considered doing this but I didn’t have a value for the resister.  Is a 1W resister OK?  I happen to have several 470Ohm 1 Watt carbon comp so I could try that easily.

[pdf64]

The positive feedback loop somehow involves the power amp, hence it only occurring at higher power outputs. And I’ve only encountered it on TRs, so higher power amp currents than other reverb amps. But not (yet) the 135W version, which has a different power amp configuration / 0V plan.
The 470ohm resistor should only dissipate low power, even 1/2W is probably far more than necessary.
It’s perhaps better have the cap’s negative terminal (ie its outer foil) connected to 0V.

As it seems unlikely that these amps behaved like this when new, it may be worth considering what might have changed over time. One possibility is that one or more  0V to chassis connections has developed excessive resistance, due to eg loosened fasteners, oxide build up, failed solder joints.
Problem being that a Fender relies on the numerous such connections all being good.
Every jack and pot and chassis tab fastener has a star washer that needs to bite into clean shiny metal for its connection to be properly established. And then there’s the brass plate and the corrosion that can build up between it and the chassis, due to dissimilar metallurgy.
Carbon type resistors should be avoided for use as screen grid resistors, now that the far more appropriate MO are available. As the carbon is a fuel load and catches fire in the case of a severe screen grid overload / short. Whereas MO are extremely unlikely to, and have a flame retardant coating just in case (pretty much anything can burn if it gets hot enough).
Plenty of old amps show the signs of the above happening.

[proaudioguy]

I just pulled the input jacks out a few days ago and re-wired them so they are very solidly to ground.  Several small changes have been made to attempt to clean up the noise I was getting in this amp and in the process, I may have created this problem.


The CAP is an electrolytic.  I’ll connect the - to the ground bus and the + to the 470Ohm and the other side of the 470Ohm to the cathode with the 2.2K.


BTW, this is the layout....

[proaudioguy]

I used Rob’s amp picture and dropped in my turret board and did my best to show how its wired.  I have no idea what might have changed, as you put it.  I can tell you other than doing a scratch build, if I do any more Princeton kits, I’ll be buying an eyelet board rather than using the cardboard crap that comes in the kits.  I picked up the PR I built in 2014 cause it has a scratchy pot and the thing has been in air conditioning and shade the whole time and the board is actually crinkling up a bit like the old ones did.  Its not been to a gig and since his arthritis has taken over big time, its barely played.  I digress. 


Here is the layout of my amp.  The actual lead dress is different but you have seen that in the photos.

[proaudioguy]

I currently have everything put back to normal.  The buzzy distortion persists. 

I have the amp out of the cabinet now, but still attached to everything.  Makes no difference to the distortion.  I still need to try adding the resister in series with the cathode bypass cap but I still want to know why I need to reduce the gain of the tube to fix this when there are thousands of these amps out there working just find without needing that.  I would really like to find the source of the problem rather than put a bandaid on it.


Here are my voltages. EDITED with fluke!
All controls DOWN

All controls DOWN


V1
1   248
2   0
3   1.72
6   250.9
7   0
8   1.81


V2
1   250.1
2   0
3   1.75
6   246.9
7   0
8   1.787


V3


1/6   410
2/7   0.027
3/8   7.97


V4
1   244
2   0
3   179.4
6   237
7   0 hum
8   1.89


V5 Trem footswitch off controls down
1/7   169
2   0
3   1
6   415 small pop
8   1


V6
1   203.7
2   54.7
3/8   82.9
6   207.9
7   54


V7
3   414
4   414
5   -43.6
8   0


V8
3   414
4   414
5   -43.6
8   0


V9
3   414
4   414
5   -43.6
8   0


V10
3   414
4   414
5   -43.6
8   0


I moved the ground of the middle cap on the board from the ground of the series caps to the ground of the other 2 preamp filters.  That made no difference to the operation.  It may or may not have affected the noise level, but I’ll determine that after the distortion thing is fixed.  Its easy to move back.  Open to comments on that.

[pdf64]

I'm confused by the P1, P2.. on robrob's referencing for connections off to the doghouse cap board; sorry if I've overlooked it but are those defined anywhere?

...I moved the ground of the middle cap on the board from the ground of the series caps to the ground of the other 2 preamp filters...

And then there's your referencing system, presumably for the doghouse ecaps; some specific, common terminology is required for effective communication. eg the AB763 schematic refers to the HT cap nodes as B, C, D, and we can assume 'A' as the reservoir cap.

[proaudioguy]

I’m sorry about the confusion.  Also all those voltages i posted earlier are WRONG.  My fluke JUST arrived and I already confirmed the plates are 414 on the 5881s and the bias is -43.


I had posted a layout of MY amp earlier and I marked P1 P2 P3.


Basically P1 is the MIDDLE cap, P3 is the next one of the same size, P3 is the same sized one on the end.  The other 2 caps on the other side are in series to double the voltage handling.  I think they are 70µF each so 35µF total.


I am going to revise my voltage numbers that I already posted so once I edit that, those should be right.

[proaudioguy]

Here is where I am with it......


I should probably start a new thread since the tremolo hum was solved several days ago.....


I have a great sounding Twin Reverb with the 6G16 BIAS VARI TREMOLO. Only problem is when I dime it, something goes wrong with the reverb driver tube and it causes a horrible distortion you can hear in this video. The volume was being turned up gradually in the video. The distortion gets a lot worse with more signal.  https://youtu.be/fUi-ziwMOkII built the amp several years ago and lately I have been chasing noise and I’m fairly sure this was not a problem before that. I cleaned up the ground bus that was soldered to the back of the pots just to make sure I didn’t have multiple paths to the chassis through the pots. This did not affect the noise. I have swapped out carbon comp grid and plate resisters with metal film. This made a huge difference in the noise. The Tremolo intensity pot and bias are intermingled in this circuit. When the intensity was turned all the way down, I had a hum. Putting the bias on the center of the bias pot as in the 6G16 didn’t help but putting it on the outside lug, adding a cap to the middle lug and also attaching the middle lug to the intensity made that hum go away. I had the 470ohm resisters on the board with 1ohm resisters for the cathode to check bias. Someone thought that mess of wires might contribute to noise so I changed that whole area to the more traditional style with the resisters on the tube sockets. It cleaned up the look but had no affect on the noise. During all of this the amp sounds fantastic, but I have a small rizz left. A buddy that used to have a Twin, asked me to dime it and record some noodling for him and thats when I discovered this nasty distortion.  So far I have done the following.Pulled out V3 reverb driver tube - problem solved, new problem created, no reverbPulled out the reverb send cable - problem solved, new problem created, no reverbDisconnected the cathode bypass capacitor - problem solved, new problem created, reverb sounds ringy and is weaker at normal decibel levelsPlaced a 500pF cap between the plate and ground. This did not work but it moved the distortion to a higher gain setting, it also killed the reverb.Placed a 500pF cap between the plate and the cathode - problem almost solved but new problem created, VERY weak reverb, mostly lower mids, required the reverb return to be up around 8 or 9 where 3-4 is usually plenty, consider there is a 500pF cap between the plate of V2 and the grid of V3 already so the other 500pF is sending everything above the cutoff to ground and the 500 in series with the grid is filtering out everything below the cut off so there isn’t much signal left. The resistance in the circuit will change the cut off frequencies, and I don’t have the math to predict that but its not a solution anyway.I placed a .1µF orange drop between the B+ going to the reverb transformer and ground, made no difference, but may have affected the reverb sound, but I didn’t dig into that.Tried 5 values of grid resisters from 470-10k.I have moved the leads around but that doesn’t make any difference.I have swapped the tube with a known good tube. The tube is a JAN Phillips 12AT7 and was new going in. I roll tubes periodically when chasing noise on this amp so that tube may well be brand new. I have a stack of these NOS.What I have not yet tried is adding a 470Ohm resister between the cathode and the cathode bypass cap. That was suggested on a forum. It reduces the gain of the tube and may be a compromise, but I would like to find the root of the problem rather than bandaid it.A few people have indicated that a bad filter cap could be the cause. I will order some filter caps. If I don’t need them for this, I can use them for something else later. The filter caps were all new F&T when I installed them.The file size limit is too small so I can’t add any pictures. Here are links but they ill eventually expire unfortunately.


https://www.dropbox.com/s/6rey6utl1k1dn ... s.pdf?dl=0
https://www.dropbox.com/s/bf7sih2yjeeb5 ... 3.png?dl=0
https://www.dropbox.com/s/fo26h6ts7vo9v ... .heic?dl=0