Single tube Reverb headache

[shaun]

Hi EL34 World,
I've built the Hoffman single tube reverb circuit successfully in the past from a 5e3 Hoffman schematic, and it sounds very good to my ear. Schems attached.

Now I'm trying to build the same single-tube reverb circuit in a 50w amp, but I'm getting loud and angry hum in the return side of the reverb circuit. I've checked and rechecked every component as well as all my wiring, plus I've pulled out large tufts of hair (which usually helps), but I can't track what's going on. Thus....heeeelp!

The main difference in the two designs is that the Hoffman design goes into the cathodyne PI, while I'm trying to send it into an EQ recovery stage.

When checking voltages for things like stray DC on the Reverb send and return pots, I get very loud scratchy hum when simply touching the meter lead to the return pot lugs, which I can track through to the input grid of V4a. I realize I'm messing with the input signal going to the signal grid of V4a, but even so, the wiring seems "hot" and over-reactive compared to when I usually test voltages; I apologize for the amateur way of expressing things, but when a circuit seems hyper-sensitive, I sometimes see it as a sign of things awry in the circuit. It's been weeks of testing and retesting, my conclusion being there's an issue with my design. Your help and expertise is much appreciated.

(Just on a side note, I posted about this build recently regarding having too much B+ due to an over-powered PT. In the end, I simply bought a new toroidal PT, using the Antek AS-1T175-100v instead of the AS-1T200-100v. I now have a comfortable 450v on the 6L6 plates instead of 500vdc.)

[acheld]

I'm not the expert here, but a couple of thoughts:

Your insertion of reverb back into the circuit differs from a Deluxe Reverb ab868 which uses a large resistor in between the input leg and the insertion leg, I think something like 3m or so with a parallel cap. 

I have had a couple of instances where I improperly grounded the reverb tank.  The grounding is on the input side, I think, not the output side.  If you ground the output cable through your chassis, it will hum.   

These are "off the shelf" comments, not using AI, but you would be wise to check your sources anyway!

[shaun]

Interesting - yes, I studied the 3.3M type typical to many of the Fender reverbs, and this circuit is different. That monster resistor does all sorts of interesting things to the signal, but it's not needed for this circuit. And also yes to the grounding on the single reverb RCA. I'm starting to wonder whether I have a bad tranny - the 1750a - but the DC resistance reads okay on all taps.

[tubenit]

A couple of thoughts ..............   


Have you tried having the reverb insertion point be just prior to the LTPI?  That's where I typically have mine. You do have a gain stage after your reverb prior to the LTPI.


Does it hum with both your clean and more overdriven channels? 


Have you tried a different tube for the reverb?   Have you tried a different reverb tank?  Are you using shielded wiring to the reverb tank that is making proper grounding with the shield?


In an amp chassis where there is ALOT stuffed into a chassis, particularly higher gain amps, the layout of the reverb is important. Can you please post some pictures of your amp chassis interior?


With respect, Tubenit

[ac427v]

The blackface tone stack cuts signal far more than the tweed deluxe tone control in your successful build. You must have a tone stack recovery stage before the reverb send. Your design sends an extremely low signal to the reverb circuit in comparison to the inherent hum signal added by the reverb circuit. This noisy mix is amplified by the reverb recovery stage and the reverb mix stage.
No, your parallel input channel cannot boost the signal enough to compensate for the absence of a proper tone stack recovery stage. Sorry

[shaun]

Hi Tubenit, and thanks. I've checked everything on your list about three times over. The chassis IS pretty tight - a lot packed into one box, so I've experimented a lot with lead dress etc. I'm aware of the more obvious pitfalls and relocated the 1750a trans accordingly.

To your point, ac427v, I wondered about that but my knowledge is limited, so the wondering was inconclusive - until now :).

I initially had a stand-alone make up stage after the Tone stack, and at that time I had inserted the reverb after the make up stage and before another 12ax7 gain stage, but boy, I had so much gain it was unmanageable - every little hiss was magnified through the roof - quite an interesting experiment actually, but I could not control the gain. So I had to remove that extra gain stage (used the extra stage by converted V1 into parallel wiring). So I relocated the reverb insertion point but wondered whether I'd run into problems inserting it right after the TS. You've dialed in on a circuit design issue that I couldn't get my head around - many thanks. I've come across a few schematics that insert the reverb - as you said - right before the long-tailed pair, so that will be my next move. If anyone has a schematic they like for this, please lemme know.

Pics of the build attached. Onward and upward.

[tubenit]

I truly wish I could give you a different response.  I'm thinking your layout is problematic for the reverb.

Is this pot that I've circled the reverb return pot?  If so, that's awfully close to the power section of the amp & if that's the case, I can easily see this being problematic.

And is that tube that I circled the one tube reverb?  If so, that's awfully close to the power section of the amp also.


Where are the RCA plugs for the reverb cables?  Are they closer to the power section than the preamp section?  That might be a problem also. 

I'm not sure you can "get there" to your goal with that layout?

Respectfully,  Tubenit

[tubenit]

I love new ideas and innovation in approaching amp building.  So, my hat's off to you in attempting something interesting. It appears that you've been somewhat inspired to have a Dumble-ish type overdrive channel inserted between the 1st and 2nd gain stage of your clean channel amp design?


Your approach is interesting in that it allows two different treble, bass and mid tone stacks. However, you can use a 5E3 type tone stack for the OD section instead and still have tone shaping on the OD channel.  OR use a pot that drops treble to ground.   

I've made two attempts over the years to try inserting an overdrive section between the 1st and 2nd gain stages of a clean channel. I was not happy with either of them & found I had unwanted noise.

Sometimes borrowing from proven layouts is useful. 

I'd like you to compare your schematic with a Tweed Bluezmeister I did.  Note that the reverb tube is prior to the LTPI tube and further away from the power section of the amp.

As a side note .......... since so many of my builds are kind of an experiment,  I like using parallel turrets as a layout platform. This approach allows me to make significant changes in design without having to build a new board. Just a thought here.


I have attempted builds that I wasn't happy with and gutted most of the amp and started over before. I sure didn't like doing that, but I knew I would never be pleased with what I had before me and knew it needed a new design or layout.   

With respect, Tubenit

[ac427v]

I suspect your first design distorted badly because the reverb splitting resistor was too small. Fender used 3.3 Meg to force most of the clean signal from the first two stages to drive the reverb tube. The result was reverb level far above the hum level. The 3.3 Meg value also greatly reduced the dry signal going to the mix stage. Voila! No nasty distortion.
Finding the right combination of values for the dividing resistors depends on the circuit topology. If you put the mix stage back in front of the PI and experiment with 3.3 Meg and other values of resistor you may find reverb nirvana. Many successful amps drive the reverb springs with only one triode. It is doable!

[shaun]

Fender used 3.3 Meg to force most of the clean signal from the first two stages to drive the reverb tube. The result was reverb level far above the hum level. The 3.3 Meg value also greatly reduced the dry signal going to the mix stage.

Exactly! Thanks for that. I just remembered how I ended up in this spot (the build has been going on for 4-5 months at this point, and I forget the various paths I went down initially). I was kinda following the PR reverb design, but because I did not include the 3.3M res in the single-tube reverb design I employed, I was getting huge signal. This speaks to acheld's query early in this thread. So I had a choice to either add a 3.3M res - which I decided not to do in an attempt to leave the signal path alone as much as possible - or remove the extra make-up stage. I did the later, especially as by then I'd added phase coupled neg bias to the LTP for max clean gain. The result was that I inserted the reverb right after the Tone stack because I have the tremolo right after the final 12ax7 gain stage, and I worried that the reverb and the trem circuit might interact in unhappy ways if I put the reverb in/out right before the trem circuit. But maybe it would be fine, and perhaps I should give it a try.

I've seen schems that have single-tube reverb in other Fender style designs with LTPS, so I'll go hunting for one of those and let you know how I get on. 

[shaun]

"I love new ideas and innovation in approaching amp building.  So, my hat's off to you in attempting something interesting."

Thanks Tubenit. I just sorta stumble my way forward, but when something gets dialed in, the reward is hard to beat. Still a way to go with this build. I haven't looked too closely at Dumble, but I thought about using a single pot TS as opposed to a TB for both channels. The guy I'm building this for is an excellent guitarist and is picky about tone plus a desire for max clean headroom, so I want to give him as much sound-shaping as possible, plus I'd never built an OD channel using a TB tonestack before and wondered what the result would be. Not sure yet, but I'll post some audio if I ever get to that stage.

"I'd like you to compare your schematic with a Tweed Bluezmeister I did.  Note that the reverb tube is prior to the LTPI tube and further away from the power section of the amp."

Thanks again, Tubenit, I'll look for that one. 

"As a side note .......... since so many of my builds are kind of an experiment,  I like using parallel turrets as a layout platform."

Sounds interesting but I can't find a pic online to see what that looks like. First time I heard of it and always keen to learn.

"I have attempted builds that I wasn't happy with and gutted most of the amp and started over before. I sure didn't like doing that, but I knew I would never be pleased with what I had before me and knew it needed a new design or layout."

Yes indeed - close enough isn't good enough because the amp will never be what I hoped for, and my knowledge will be curtailed as a result. So I'm with you on that! This is the second chassis I've made for this build (I use aluminum sheet and a bending brake - they take a long time to manufacture), and I'm prepared to make a third if this one doesn't work properly. But I'm feeling optimistic.

As for the circled pot, that's the cap-coupled neg bias pot, and the circled tube socket is the 12AT7 PI. The reverb is two to the left, and it has two tubes before it - one for clean input and one for the cold clipper OD channel, a la Marshall style. On the front panel, there are two large concentric pots for send and return of, first, the reverb, and on the far right is the Trem send return.

Adding a couple more snaps for reference.

[tubenit]

The reverb is two to the left, and it has two tubes before it

Ok, great! Then disregard my concern that reverb components may be too close to the power section.


With respect, Tubenit

[shaun]

So...after looking at the Bluezmeister Tremverb, I inserted the reverb right after the coupling cap after V4a - updated schem attached.

The results were good, although tweaking will be needed. The reverb works fine now - yay! And thanks for the help. No hum, but the actual reverb itself could sound better, so that's next on the list. The trem also works fine with the reverb right before it, although the intensity is a bit shallow. I've read somewhere recently that adding a resistor (470k?) between the signal and the intensity pot can ameliorate this, so I'll try that.

Regarding the OD channel, I'm happy to report that it really kicks ass. I've had previous success with the Marshall cold clipper circuit a la Rob Robinette, although I like the Rivera upgrade and find that going up as far as a 47k cathode resistor on V2a - rather than the Marshall 10k - makes the OD very manageable. The 10k allows a huge volume boost, which the larger resistor tames in a very nice way. The OD channel does sound like there's a ground loop in there, so I'll have to figure out that hum.

If I get all that sorted out, all I'll have to contend with is the three-way footswitch. I've found some great diagrams on this very forum to aid with that, but sometimes I find that the simpler the problem, the more confused I get! And switching often seems simple while actually being quite tricky to get right.

Anyway, Thanks again. When the amp is getting closer to completion, I'll post some audio. I'm matching it with a 12" Jensen Tornado Special Edition (like the Stealth but without the 3-5dB dip at 1500h). At 100w, I'm hoping for minimal break up. Neodymium can be a bit sterile compared to typical guitar spkrs, so I'm hoping it'll warm up as it break in.

[ac427v]

I have had success by disconnecting the ground connection on the trem Intensity (Depth) pot.

As built, the 50k pot load allows much of the preamp signal to drain to ground instead of being sent to the trem circuit and the PI.

In contrast, Sluckey's grounded pot version works well with the stock fender reverb. Omitting the reverb mixing stage drastically reduces the signal level to the PI when using your blackface channel. You need all the signal you can get! The downside: You may need to tweak your OD channel to reduce gain hitting the reverb drive tube.

If the 50k pot does not allow you to turn down the trem intensity, increase the pot value. 250k reverse audio or linear pot worked for me.

[shaun]

Thanks ac427v - great info and a good starting point.

[shaun]

Well, the reverb headache continues. I wasn't getting the required volume back from the reverb circuit when it was simply inserted into the signal line, so I inserted the return straight into the LTP, as I've seen Mr. Luckey do in his AC15 rev design. I built it once before on a different amp and it worked fine.

This time, it increased the reverb signal and the overall dynamic of the signal in very good ways, that's a big plus. But with the increased volume came the return of the noise. I am simply unable to troubleshoot the loud buzzy hum I get in the reverb return. I've tried a slew of grounding scenarios (including star), and I've tried a variety of ways of feeding the B+ to the reverb trans to isolate it from other B+ lines. But here's the result that is always there - a buzzy hum that increases when the return pot is increased. As you'll see in the pic, the trans leads enter the chassis very close to their destination. On the flipside of the chassis, the tubes all have shields.

I've narrowed it down as best I can to this: when I disconnect the return RCA signal from the tube input (in this case, the very short yellow lead in the pic) the buzzy hummy noise is replaced with that very identifiable open-input tone we get when a tube input is left ungrounded and with no signal. So the tube seems fine (yes, replaced it and replaced all surrounding components more than once). This loss of buzz occurs only when the yellow wire is disconnected at this point - for example, I have disconnected every other possible lead in the reverb circuit to test each connection, and the buzz always remains. Like the raven over my door - it seems to be there evermore, rather than nevermore. I'm feeling as dark as Poe. The noise is there even when the tank is disconnected, so it doesn't seem to be the tank. On the plus side, I'm getting all the reverb I could wish for. I've tried reducing the B+ from between 200v-330v - the buzz always remains.

So I probably need to reconsider the layout of the RCA jacks and maybe even the location of the rev trans. Which might mean building a new chassis...again. Sigh. I really bit off way more than I could chew with this build. I suppose I'm learning a lot, especially regarding how much I don't about high gain amp building. I can only think it must be poor layout design.

Latest schem and pics attached. Thanks for any feedback.

[tubenit]

It's hard to tell from the pictures, but this has the appearance of a high voltage wire either between or right next to the heater wiring??   Is that the case?   

What happens when you use a non-conductive stick like a bamboo chopstick to move the heater wires around there OR the high voltage wire?

Can you quadruple check your wiring there please?  Does it accurately match the schematic? 


Is your 220p cap going to the ground of the reverb return RCA plug or the reverb send RCA plug?  Not saying you've wired this wrong, I simply can't tell from the photo.  Need a better overhead photo.

Can you post a direct overhead look photo of this section please?  It might help us help you?

With respect, Tubenit