Adding a 1-tube reverb to a single channel AB763

[guitylerham]

Hey everyone,

I posted a while ago when I was starting my project: a PR with Power Scaling. Well, after some back and forth, I ended up rebuilding the PR into a DR but in an attempt to simplify, I used a single triode+mosfet trem on the last preamp and a single tube reverb. The amp is up and running but appears to have a LOT of gain and it's not clean above 2 on the volume and very loud above that. I believe I know where the problem is, I just haven't wrapped my head around it yet.

A standard AB763 taps the reverb signal before a 3m3/10pf and inserts it after. The schematic I followed of the 1-tube reverb shows a 100k+ resistor instead. I have a 470k there now. I'm thinking I need to adjust the related voltage dividers and combine both circuits because I believe the next stage is getting slammed with a lot of signal that the stock ab763 used to dump to ground. Below is what I have built and I would really be grateful if someone could explain what's going on with the voltage dividers and what might be a place to start with tweaking because I don't think I can just up one value without adjusting another.

Updated schematic:


Here's a comparison of the AB763:

[tubenit]

Do you have a 500p -.001 range cap between R and the dwell pot?


I'd try 5uf cathode caps instead of 22uf on the three preamp gain stages and a 5751 in V1


With respect, Tubenit

[ac427v]

I think you are on the right track with your thinking on the 3M3 resistor being the culprit. Your schematic is really more of a 1 1/2 tube reverb because you have the post reverb mixing gain-stage from the original DR schematic. The one-tube reverb needs a much smaller resistor because it omits that post reverb mixing triode.-Craig

[guitylerham]

Do you have a 500p -.001 range cap between R and the dwell pot?


I'd try 5uf cathode caps instead of 22uf on the three preamp gain stages and a 5751 in V1


With respect, Tubenit

I’ll have to check on that input cap before the dwell.

I need to update that schematic because I did takevyour advice and used 4.7uf bypass caps on the first preamp stages to cutvsome lowend. I think I’ll enjoy what those do. Thanks for the heads up!

[guitylerham]

I think you are on the right track with your thinking on the 3M3 resistor being the culprit. Your schematic is really more of a 1 1/2 tube reverb because you have the post reverb mixing gain-stage from the original DR schematic. The one-tube reverb needs a much smaller resistor because it omits that post reverb mixing triode.-Craig

That makes sense. There’s an extra gain stage after the Reverb that used to be part of the cathodyne when it was a Princeton Reverb. The 1-tube verb doesn’t require the gain it provides. So perhaps I should use the 3m3 resistor as the mixer and see how it works. Just not sure the effect of the 1m vs 100k Reverb mix pot. Thanks!

[tubenit]

I am NOT suggesting that you "should" do it this way.  Having said that,  I typically have the insertion point for the "one tube reverb" just before the LTPI as in this illustration. 

I am wondering that IF you do not have a 500p to .001 range cap going into your dwell pot, that you may be "reverbing" a lot of mids from the .02 cap prior to the mixing resistor for the reverb?  Maybe that could be contributing to a "gain-y" tone?

Regarding the schematic, I have NOT built this particular amp but it stems from an idea I had a couple of years ago.  Not sure about the vibrato?

with respect, Tubenit

[PRR]

> an attempt to simplify ..... have a LOT of gain

I don't think you want V1b.

After removal you will have to patch-up some other values. But I think this will lead to a happier amplifier.

[guitylerham]

> an attempt to simplify ..... have a LOT of gain

I don't think you want V1b.

After removal you will have to patch-up some other values. But I think this will lead to a happier amplifier.

Whoa, that's blowing my mind. Do you think it would still retain the essence of the AB763? I wonder how much character is coming from that stage.  I agree that I think I have too much gain. Guess it comes down to either dumping it a la fender style or just not amplifying it as in removing a stage.

Tubenit and PRR, what are you thoughts regarding removing V1b in regards to your schematic that you posted? V1b is still intact there (though you mentioned you haven't built that circuit yet...). That'll leave an extra triode!

P.S. I do need to add a 500pf between the tap and dwell pot!

Edit: Does this schematic look like a decent starting place?

[guitylerham]

Regarding the schematic, I have NOT built this particular amp but it stems from an idea I had a couple of years ago.  Not sure about the vibrato?

with respect, Tubenit

I decided to bias-vary the last preamp stage before the LTPI so that I didn't have to figure out how to scale the vibrato with the power scaling (also the cathode bias mode renders a less effective vibrato). Seems to work ok, but it thumps majorly and I'll have to address that somehow.

[tubenit]

what are you thoughts regarding removing V1b in regards to your schematic that you posted?

You can do that IF you want to.  My Tweed BluezMeister is foot switchable between 2 (clean) and 4 gain stages (overdrive) going into the reverb into the LTPI and power amp.  However, I am typically using a 12AY7 or 5751 in V1 and typically a 5751 or 12AX7 in V2.  As a reference point, I typically have the clean volume pot at 12 noon.

I am happy with it and even with the overdrive channel switched in, I feel like it's more of a slight overdriven boost to the clean channel then a distorted overdrive.  The overdrive channel has more sustain but still sounds somewhat like a boosted clean channel.  The clean channel using two gain stages stays reasonably clean, IMO.  However, it should be noted what some of the pot values are in the OD section and what the post LTPI coupling cap values are.

With your amp, I would be inclined to leave the first 3 gain stages in.  However, I might use a 12AY7 in V1 or perhaps use 2.7k or 3.3k cathode resistor in the 3rd gain stage with a 2.2uf cathode cap to lower gain.  Just a thought.  I do like the 12AY7 and 5751 tubes and think the ones I have seem to have sweet musical tone to them.

With respect, Tubenit

[guitylerham]

With your amp, I would be inclined to leave the first 3 gain stages in.  However, I might use a 12AY7 in V1 or perhaps use 2.7k or 3.3k cathode resistor in the 3rd gain stage with a 2.2uf cathode cap to lower gain.  Just a thought.  I do like the 12AY7 and 5751 tubes and think the ones I have seem to have sweet musical tone to them.

I used an audio probe and had clean audio into the grids of the preamp stages but the plates are distorted and loud. Is this normal behavior? If so, then how does clean quiet audio into a tube come out grossly distorted and yet present an undistorted tone at the following stage's grid? If not, then something's way wrong.

I'll keep tracing the circuit and making sure everything is correct.

Update: all readings look normal as far as resistance to ground on cathodes and plate to D node. I’m thinking the input to a tube is small, clean on the audio probe. Then it’s amplified by the tube which leaves a loud signal on the plate that could be overdriving the little solid state audio probe amp. Then after the EQ loss for example, the next stshe’s grid now has a small clean signal again. Correct me if I’m wrong. I tried a 3m3 resistor as the mixer but I’m still feeling too much gain is the problem.

[tubenit]

I'd double, triple and quadruple check the resistor values in the signal path & plate/cathode resistors.  Everything may be fine, but it won't hurt to go over those again to be sure.

I'd also check voltages for plates and cathodes in the preamp. 

And I'd look for weak solder joints.

Tubenit

[guitylerham]

Well I spent the day going over the schematic and amp, checking values for plate and cathode resistors, nothing out of the ordinary. Here are some voltages but I can't see anything wrong either.

A 433vdc
B 415vdc
C 344vdc
D 276vdc

V1a
plate(1): 189vdc
grid(2): 0
cathode(3): 1.3vdc

V1b
plate(1): 175vdc
grid(2): 0
cathode(3): 1.2vdc

V2a reverb driver
plate(1): 410vdc
grid(2): 0
cathode(3): 3.25vdc

V2b reverb recovery
plate(1): 176vdc
grid(2): 0
cathode(3): 1.5vdc

V3a trem oscillator
plate(1): varies 177-211vdc
grid(2): 0
cathode(3): 1.5vdc

V3b 3rd gain stage
plate(1): 183vdc
grid(2): 0
cathode(3): 1.3vdc

V4a PI
plate(1): 192vdc
grid(2): 46vdc
cathode(3): 72vdc

V4b PI
plate(1): 193vdc
grid(2): 49vdc
cathode(3): 72vdc

There is some ugly blatty clipping when I have the volume up past two. I'm gonna keep at at it. stay tuned!

I've traced the amp and noted each component on this schematic so this is what's in the amp for anyone following along:

[tubenit]

If you have it in cathode biased with the reverb & vibrato tubes removed does it still have this problem?

What happens when you disconnect the negative feedback?  Sometimes the oscillation if the OT wires are backwards isn't a loud squeal.

[guitylerham]

If you have it in cathode biased with the reverb & vibrato tubes removed does it still have this problem?

What happens when you disconnect the negative feedback?  Sometimes the oscillation if the OT wires are backwards isn't a loud squeal.

Cathode or fixed bias presented the same problem buuuuut...

I disconnected the NFB and didn't notice any tonal difference at low volume but when I cranked it, I could turn it up past noon and it had none of the ugly clippling that was present past even 2 on the dial! I would have never thought of NFB being the culprit because it didn't squeal at all. The reverb still oscillates but it doesn't even start to squeal till I turn the mix up. I'll swap the OT leads and replace the NFB and see where we are.

I still need to figure out how to smooth the reverb levels (i.e. find the correct value mixing resistor or deal with gain levels, which may be causing the reverb squeal on the recovery side) but it's sounding so much better!

Thanks Tubenit and everyone else who chimed in. I've read countless exchanges between the fine people here and this is one of greatest resources for people like us who are compelled to torture ourselves over these fascinating endeavors. I truly appreciate the effort everyone spends helping each other.

[tubenit]

I disconnected the NFB and didn't notice any tonal difference at low volume but when I cranked it, I could turn it up past noon and it had none of the ugly clippling that was present past even 2 on the dial! I would have never thought of NFB being the culprit because it didn't squeal at all. The reverb still oscillates but it doesn't even start to squeal till I turn the mix up. I'll swap the OT leads and replace the NFB and see where we are.

I still need to figure out how to smooth the reverb levels (i.e. find the correct value mixing resistor or deal with gain levels, which may be causing the reverb squeal on the recovery side) but it's sounding so much better!

Glad you got rid of the ugly clipping noise.  I had a similar thing happen on a build and it proved to be the NFB & reversed OT wires thing also.

I am wondering IF you can resolve the reverb issue by doing the reverb insertion just prior to the .001 cap going into the LTPI?   Have you tried that yet?

With respect, Tubenit

[guitylerham]

I disconnected the NFB and didn't notice any tonal difference at low volume but when I cranked it, I could turn it up past noon and it had none of the ugly clippling that was present past even 2 on the dial! I would have never thought of NFB being the culprit because it didn't squeal at all. The reverb still oscillates but it doesn't even start to squeal till I turn the mix up. I'll swap the OT leads and replace the NFB and see where we are.

I still need to figure out how to smooth the reverb levels (i.e. find the correct value mixing resistor or deal with gain levels, which may be causing the reverb squeal on the recovery side) but it's sounding so much better!

Glad you got rid of the ugly clipping noise.  I had a similar thing happen on a build and it proved to be the NFB & reversed OT wires thing also.

I am wondering IF you can resolve the reverb issue by doing the reverb insertion just prior to the .001 cap going into the LTPI?   Have you tried that yet?

With respect, Tubenit

I will try that reverb injection point but my only hesitation is reverberating the preamp tremolo (and all it's noises) through later stage reverb vs. trem-modulating the signal with verb already. In recording, I'd much prefer reverb to be the last effect but on amps I always thought best to put trem last a la output tube bias trem. I'm very curious to try!

PSA: I'd read the advice before but it didn't hit home till now. Build your amp without NFB until it's working!

[tubenit]

I will try that reverb injection point but my only hesitation is reverberating the preamp tremolo (and all it's noises) through later stage reverb vs. trem-modulating the signal with verb already

.

The signal chain of reverb and tremolo is a valid point.  Having said that, I am not sure it will make much of a difference with the reverb immediately following the tremolo?  I think this is a "try it and see" moment. 

Your other option would be to connect the tremolo like Tremolux  6G9-B     https://el34world.com/charts/Schematics/files/Fender/Fender_tremolux_6g9b_schem.pdf

or  https://el34world.com/charts/Schematics/files/Fender/Fender_tremolux_6g9.pdf

I realize you'd lose your cathode/fixed biased switching which I personally have never tried.  I have read from a few who have done the cathode/fixed biased switching that they don't actually switch things much and have a preferred favorite and leave it there?  However, since I haven't tried it, I don't know if that would be the case?

With respect, Tubenit

[guitylerham]

Just swapped the OT primaries and hooked the NFB up and the blatty clipping is back. So the problem seems to be the NFB. I could live with it permanently gone but I'd like to understand why it's behaving that way.

[tubenit]

Just swapped the OT primaries and hooked the NFB up and the blatty clipping is back.

That doesn't make sense???     

I would anticipate a problem with one of those wiring approaches.  Either "original wiring" or "swapped around wiring"  should've cause a squeal or some type of oscillation issue when the NFB was hooked up (I think)?   Any chance that you accidently wired it back exactly the same? 

Something seems quite odd about this from my limited experience?

With respect, Tubenit

[guitylerham]

Just swapped the OT primaries and hooked the NFB up and the blatty clipping is back.

That doesn't make sense???     

I would anticipate a problem with one of those wiring approaches.  Either "original wiring" or "swapped around wiring"  should've cause a squeal or some type of oscillation issue when the NFB was hooked up (I think)?   Any chance that you accidently wired it back exactly the same? 

Something seems quite odd about this from my limited experience?

With respect, Tubenit

Your “limited” experiences far outweighs mine and this is an annoying, if not interesting problem to solve!

I definitely swapped the blue and brown leads to the opposite tubes.

My nfb wire is tapped from the green 8r OT secondary.

Isolated speaker jack, sleeve grounded to PA filter caps.

Correct 820r/47r-to-ground network.

No squealing before or post swap.

The amp sounds great, save for the squealing Reverb and thumpy trem! I am tempted to just rebuild to Slucket’s Deluxe Lite style (as suggested by PRR) with your Reverb injected before the PI but that’ll come later if needbe.

Beats me too but I’ll keep at it. Thanks!!

Thought: One thing I do notice is the volume doesn't turn on until 1-2 and then it's fairly loud. When I turn the volume off, I can hear some quiet guitar signal coming through the speaker that disappears with my guitar volume off. This is like what someone else on the forum has mentioned in the past where the signal is being impressed on the power supply and making its way to the speaker. This could certainly be related. I'm going to use another 22uf cap grounded, and probe the positive side to various points in the power supply to see if the amp needs more filtering.

Reworking my grounding scheme as well...

[guitylerham]

I cleaned up the grounding and in the process discovered I used an isolated speaker jack... without grounding it to the chassis. Ugh. So, hooked up NFB and got squeal, replaced OT primaries how they were and with NFB, no squeal, reverb doesn't oscillate, trem doesn't thump. Can't believe it.

However, I must've done something in the process because now my voltage clamp is regulating from about 420vdc to only 250vdc and won't allow me to clamp it any higher so my C and D nodes are grossly under powered. Must've fried something.

I think I'll go have a beer to celebrate the simplest details to overlook. Thank you, Tubenit for sticking with me. I'll keep y'all posted on this voltage issue now. pffffff

[tubenit]

discovered I used an isolated speaker jack... without grounding it to the chassis. Ugh. So, hooked up NFB and got squeal, replaced OT primaries how they were and with NFB, no squeal, reverb doesn't oscillate, trem doesn't thump. Can't believe it.

 

I hope you realize that it is likely that every experienced builder on this forum has made mistakes similar to that & somehow missed the "obvious" which isn't always so obvious.

I admire your perseverance and positive attitude!  Glad you got that piece resolved.  Celebrate that success and move on to the next challenge. 

If I am understanding what you've posted,  it sounds like the remaining issue is perhaps with the power scaling?   I have never used power scaling but have done a few cathode biased VVR amps.  However, I discovered I prefer the PPIMV approach now because of it's simplicity and reliability for me.

PLEASE keep us posted.  I'd sure like to hear how this turns out. I think you'll get this solved and have a nice sounding amp!

Respectfully,  Tubenit

[guitylerham]

Thank you for the encouragement and it won't be my last duh moment either.

Well, last night in a marathon, I eliminated the mixing stage so now I have two gain stages and the PI a la Sluckey's Deluxe Lite. I moved the insertion point for your reverb to just before the PI too. The amp really sounds good. What's interesting is I get to turn the volume up halfway for normal grit level and then max it for overdrive! I can see myself using much more of the volume than previously.

The reverb sounds great. At a bit before max reverb mix, it starts to run away... I do like how it's more moderate tone compared to the 1.5 tube fender verb!

Trem still thumps. I hope I can get that resolved with some lead dress perhaps.

I ditched the voltage clamp circuit as it stopped clamping, so I think something burned up but I guess I don't mind the 20vdc climb in the preamp as I dial back the power scaling.

I'm now focusing on the best master volume either as a dual concentric pot with the PS control or ideally, a well matched taper MV as a dual gang with the PS control, one knob. If it turns out I need a triple gang (i.e. one for PS and dual gang LarMar for MV) I can build one!

I set out to make this amp a one shot perfect build but once again after endless mods and substitutions, it's a mess inside. That's what happens when you're not just building a time-tested, verified layout and instead cobbling together a bunch of different ideas before you know they work together. That's the fun, I guess!

I'll get a recording of the amp so y'all can hear it. It's a nice sounding amp.

[tubenit]

Do you still have a 500p cap between the R and the dwell pot?   Thanks for sharing your success.

With respect, Tubenit

[guitylerham]

Do you still have a 500p cap between the R and the dwell pot?   Thanks for sharing your success.

With respect, Tubenit

Yep, I upped it to 0.001 as well.

I also installed a 250ka pot with a 100k off the wiper in place of the 1m in the input to the PI as a Bootstrapped MV. Still getting acquainted with PS and this particular MV. It's interesting how a Princeton Reverb has a nice warmth to it whereas this AB763 (whatever it is now without that 3rd gain stage) has a nice bark and slap but not the warmth, almost "paper-y" and "flatter", not in the sense of an EQ graph but flatter as in depth. Not a bad thing, just an observation. I'll have some recordings of both this weekend.

[ginger]

A Princeton Reverb is not an AB763 circuit . Just wondering if you simulated the missing trem ?

[guitylerham]

A Princeton Reverb is not an AB763 circuit . Just wondering if you simulated the missing trem ?

Of course, I've just recently built a PR with a similar topology and this current amp was going to be another one but I decided to build an AB763 instead. They're similar amps in the tone stack and output section except for the phase inverter (I attribute the biggest difference to the LTPI vs Cathodyne).

I left off the 47k to ground that used to simulate the tremolo intensity pot and I've omitted the 220k/220k voltage divider that was originally on SLuckey's Deluxe Lite. I might experiment there but without that 3rd gain stage, I can crank the amp and it's only mildly overdriven, which isn't a bad thing for me and I'd imagine it'd be even less so with padding down the signal. Does the 47K-to-ground tremolo resistor load down the last gain stage in a different way than attenuating the signal into the PI does?

I'm not quite sure how this reverb injection 150k resistor and the 47k act together. Does the 150k/47k become a voltage divider and dump a bunch of signal to ground?

[sluckey]

I'm not quite sure how this reverb injection 150k resistor and the 47k act together. Does the 150k/47k become a voltage divider and dump a bunch of signal to ground?

Yes it does.

[guitylerham]

I'm not quite sure how this reverb injection 150k resistor and the 47k act together. Does the 150k/47k become a voltage divider and dump a bunch of signal to ground?

Yes it does.

I am learning @_@

I'll experiment with a larger value than 47k there if I feel the need.

So a series resistance is basically invisible to signal AC until the current through it starts to create a noticeable voltage drop? The 150k's goal isn't to attenuate so much as it encourages some of (but not all) the signal on R end to go to the reverb Dwell pot, reverb then returns to the other side of at RV but not backwards through the 150k (which would result in some sort of feedback loop), instead goes to the PI...?


Edit: I’m reading a little known book by Merlin about preamps. I’ll cease all further questions until I get to the last page ;)