Trouble with 1-tube reverb

[Nathan]

I just got a homebrew up and running all except the reverb. It is a single 12DW7 circuit with a 8FB*** Accutronics tank, in essentially the same arrangement as a Gibson Scout. There's a cap driving it, and I checked the phase as it feeds back into V1 but to no avail. My tube amp tech couldn't get it running either. When I tap the reverb tank, it clearly makes the characteristic crashing sound through the speaker.

Any tips on troubleshooting this kind of thing? I'm somewhat of a newb, and it's my first attempt at DIY reverb. The only other reverb job I did was from an Allen kit (paint by numbers).  

[sluckey]

Show us a schematic.

[phsyconoodler]

Usually if the tank splashes the recover is where the culprit is.RCA cables and jacks are not created equal.Cheap ones cause lots of troubleshooting headaches.

[tubesornothing]

When I tap the reverb tank, it clearly makes the characteristic crashing sound through the speaker.

You are not getting a signal into the reverb tank.  As the other (wise) members say, check cables and connectors, failing that post a schematic, and perhaps a gut picture.  You are almost there!

[tubeswell]

A schematic is going to help everyone understand where your amp is coming from. With reverb its important to get enough juice to drive the RT/pan, as well as 'matching' the gain structure/signal strength from the recovery/wet side with the dry side.

By your description about the pan-tapping trick, you sound like the recovery side is working, so now you need to solve the reverb send side. So break it down to the various bits - Is the pan input transducer working?, Does the RT work? is the driver triode working? Is there enough gain going into the driver triode to get a decent signal into the pan? etc

[Nathan]

Thanks everybody -I'll work on getting a schem up.



How's this? I *think* everything is right. The tank is an 8FB3C1C which means Input Insulated and Output Insulated and vertical connectors up. I should mention that I have no idea what that means for hookup. I am using RCA cables sold by Allen and meant for reverb tanks.

Currently the reverb send (V2 pin 1 which is the 12AU side of the 12DW7 tube) is reading ~740 ACV and ~308 DCV right before the 0.68uF cap which is driving the reverb tank. I also don't know if this is good or bad. I do believe the cap should be blocking the DC.

Thanks all for helping!

[tubesornothing]

Well given that:

a)  your guitar signal works OK
b)  you hear a sproing when you knock the tank

I would check the following:

a) triple check the components and connections from the take off at the tone control (yellow hilighter and schematic work)
b) are you using the correct side of the DW7 for driving the tank (pins 1,2,3)
c)  are you using the right tank?  It needs to be high impedance
d) chopstick all connections

Failing all that, try another tube, then try another tank

[Nathan]

Right on -thanks! The problem I have is that I never did find a schematic for this type of reverb. I borrowed it from the GA-17RVT, but this idea of using a 12DW7 is from Swart. Basically I have replaced the reverb transformer with a cap a la Traynor. It all SEEMS ok, but clearly it's not working. Would the value of the drive cap make a big difference? I thought it was just blocking DC.

Yes, this is the highest input impedance tank Accutronics makes. 

I thought it might be a phase issue feeding back to V1b somehow, but I tried it without the cap and no dice.

I hope the input transducer is not dead. I had to pay twice for the shipping on this tank (long story)! I'll report back when I try a few things. Please advise if you see anything wrong though!

[tubesornothing]

Hold the presses - I see the problem.

Its a little hard to see on your schematic, but:  1M on the cathode?  82R on the plate?

This is a little wonky. You mentioned swart.  On the AST, he uses a 1k/25uF cathode (pin 3) and  10k on the plate.  The coupling cap of 0.22uF.  The rest looks OK.

[Nathan]

Do you think that's enough to kill the signal? I will take your word for it and try when I get home.

[tubesornothing]

Well, it could be some kind of cathode follower.  Those values are way different than I have ever seen.  However PRR or sluckey might have a comment.

[sluckey]

No need to say anything until he changes those resistors. I'm pretty sure the values shown on that schematic are show stoppers.

[PRR]

> it could be some kind of cathode follower.

No signal is taken from cathode. ipso fractal, it aint no cathode follower.

> essentially the same arrangement as a Gibson Scout. There's a cap driving it

I don't see any part of this well-copied from a Scout. You've replaced the efficient transformer with a much too small resistor, got the cathode bias 1000X bigger than the Scout's. The tube is "starved" (teeny current) and what little it squeaks out would rather go through the plate resistor than the tank.

If you can't calculate it, copy it. That includes the number of zeros on the parts. I'm not a plumber, and I got pipe-trouble. I see that plumbers use 1/4" tube for an ice-maker and 4" pipe for the sewer. There's probably a reason. So it might be unwise to extend a waste-drain with 1/4" tubing. In this case I can picture poop flowing (NOT!) in a thin hose, and stagnant water if I ran 4" pipe to an ice-maker. In electronics we can't see the poop, but we do want to know or copy "reasonable sizes".

> Would the value of the drive cap make a big difference? I thought it was just blocking DC.

If "DC block" was your ONLY goal, you'd just cut the wire.

The thing is: you want audio to get through. And the tank is a heavy load. You need a fairly big cap. Just like plumbing the toilet. While we vent so the steady ("DC") stream of air from septic to roof does not come up the toilet, we do need to pass the "interesting" stuff from toilet to septic. You need a handful of pipe or capacitor. (Can you tell how much fun I had today with poop-pipes?)

8FB3C1C's important specs are;
Type: Accutronics Type 8
Input impedance: 1,925 ohms
Output impedance: 2,575 ohms

OK, that's as good as you will get for a plate-loaded reverb driver.

The plate load resistor is a compromise between the actual load (around 2K) and the tube's plate resistance (about 7K for 12AU7 or the fat side of a 12DW7). 82 is not a compromise. 6.7K is optimum but 10K will be easier to find in the required 5 Watt size.

Note that this is 100 times bigger than you seem to have.

The cathode resistor will typically drop "a few" volts at the design cathode current. Call it 3V for starters. You are trying to drive a low-Z load with inefficient coupling, you need big current. Your little tubes run 1mA (200K plate resistor dropping about-half of 300V supply) and drive silly little pots etc, your big 6V6 suck about 40mA each and do real work (a speaker), you should be thinking like 10mA. To drop 3V at 10mA you figure 3V/10mA= 3V/0.010A= 300 ohms. Use 270 ohms. 1K may work, and save power; seems thin to me.

Note that this is 1000-3000 times smaller than your 1Meg.

The combination of plate-side 100X off and cathode-side 1000 times off suggests that driver gain and power are maybe 100,000 times smaller than they could be. Since it takes more parts than this to get BIG reverb, and you have to trim 2 triodes and a transformer to get even modest reverb, 100,000 times worse is "no" reverb.

You are starved for signal, you do NOT want the resistor dividers at the grids.

You don't want bass in the reverb (mudd) and you don't want to load your preamp and vol/tone network, so you steal the 500pFd cap and 1Meg from the nearest Fender (it works).

Bypass the cathode resistor with any handy electrolytic.

I can't make  out what size tank coupler you used. Ampeg and I tend to go ~~0.5uFd, a 0.22uFd is mentioned, whichever, but something in that range. A 0.01uFd, fer example, would (in this low-Z case) only pass the tizz, not the body and voice of the guitar.

[Nathan]

That's awesome. I will need to read this over again and again before I understand everything, but I am really grateful for your help!

First off, I did as tubesornothing said and it worked! For 10 minutes. I only had 15 minutes last night due to family duties, but I was extremely stoked to hear reverb. And it was good too! The amp has a couple of other issues though (it's very loud/hot and is noisy). For some reason, the 'verb crapped out. Could it be that I need a 5 W plate load resistor? I used 1 W but Swart looks to have used 1 W too. My cathode resistor is 360 ohms and the coupling cap is 0.68uF.

I'll troubleshoot when I have some spare time. I typically come home ~5:30, feed the kids, bath and put them to bed ~7:30 and then it's too late for guitar stuff. The 30-40 minutes between getting home and changing clothes and supper is all the time I get on a weekday!. I need a house with a garage or basement...   

Second, I once brought one of my homebrew disasters to a tech to fix my mess and he used the plumbing analogy. I'm an engineer by trade, so he knew I'd understand the electronics in terms of pressures and flows. That clicked with me, but I'm still not calculating anything yet, just copying schems.

Lastly, this project came about from gut shots of a friend's Swart AST. In doing my best to fill in the blanks, I clearly misread some of the color codes! I am just so glad it's nearly done.

If you guys know a fix, my tremolo is "ticking".   

[HotBluePlates]

If you guys know a fix, my tremolo is "ticking".

Ticking is often clipping in the oscillator tube. You have a 10k plate resistor drawn on your schematic for the 12AX7 oscillator. That seems 1/10th as big as it should be (3rd band on the resistor orange, instead of yellow). Try a 100k resistor instead of the 10k.

There is already a resistor and cap to ground from the 12AX7 plate, which looks to me like a way to shunt off high frequencies created when the oscillator clips. Between that and the change from 10k to 100k plate load, your ticking problem could be solved.

[tubesornothing]

I think the trem plate resistor is off to the right side.  And it looks like it is 900k, but is hard to tell.  Anyways on like last AST I looked at the trem plate is 220k, however, I think yours is an earlier model.  Current models have a 430R power tube cathode.

Also, what are your cathode resistors for your two gain stages?  Its hard to read your schematic - 15K?  They should be 1.5k.

[Nathan]

Alright -I got 2 new resistors in because I somehow did not even follow my own schematic and I feel like an idiot. The R25 900k was actually 100k and the other (R39) was 10k.

So I popped in a 1M for R25 in the absence of 900k and used 100k for R39. I have to admit I'm not sure which is the plate resistor and which is the oscillator. It just looks to me like the 2 resistors form the outer lugs of a pot and the plate would be the center lug. But I digress.

I had the same problem as before, getting no sound, and the standby switch would cause a loud electronic swooshing sound like something discharging if I flip it off.

So I retubed everything and got sound. Still no reverb, no tremolo, and it was too loud with lots of scratchy noise, just like when I started this thread. Now I have worked on this thing in my spare time for way too long to give up. I know I am on the right track since it worked fully last week except for the minor problems.

At this point, I would take it to a tech if there was one I trusted in my town but there isn't. Can you guys keep pointing me in the right direction? I am deeply grateful for the help!

[tubesornothing]

Hi Nathan, can you pop up a better schematic as you have it currently built it?  I can't read the values not reference numbers well enough.

thanks

[Nathan]

Ok all,

I have the thing working about 95% right. I incorporated the suggestions but there's something I need to figure out with the reverb:

It does work now, but it only comes on when I hit the strings harder. If I pick lightly, no verb. It's pretty weird but it feels like there's a gate on the tank.

I'm not sure what would cause that. Could it be the tank coupling cap? I have a 0.68 uF because it was there, but Traynor and Ampeg usually have 0.1 uF. Is it too much resistance? Pot value?

Thanks if you know!