Reverb Wet/Dry Mix Problem

[Chuck H]

Ahh, the reverb hum thing. It can be tricky but with some testing and Steve Conners help I managed to implement a ground system for my "satellite" reverb amp that completely eliminated hum. But one issue at a time. eh?

[dchang0]

Well, sort of. Because of the difficulty of accessing certain components, it's best to try to fix multiple issues at once when "going in" each time...

So if you have any general tips to share before I pull the daughterboard out, please do.

Thanks very much!

[Chuck H]

Well... Your probably already aware, but reverb tanks are available with different ground schemes on the jacks. The ideal grounding for the input/output depends on the circuit and is important to avoid ground loops. You could try isolating the ground on the input of your tank. You would need to make modifications to the tank itself or order a tank with an isolated input ground. Then you can ground the tank however you like treating the input and output grounds individually. There are a limited number of amps that actually work well enough with the tank input and output both grounded to the spring case. So, this would be a combination of in chassis and out chassis (the tank itself) modification. I did have to isolate the ground in my project and I did it by drilling out the rivets that hold the input jack against the spring case and inserting a couple of hand cut cardboard gaskets, then using small nuts and screws to replace the rivets.

[dchang0]

Wow--that's pretty extreme! I have an Accutronics/Belton 8EB2A1B. It has some pretty fancy plastic isolation jacks, with just the output side ground tab screwed to the chassis. I confirmed that this is the case with a continuity check--both of my RCA cables are grounded to the ring on the footswitch tab (which is then grounded to the reverb pot's lower leg and the bus wire). Only the output cable's ground goes all the way into the reverb pan. Also, if I unplug the reverb return cable, it starts humming really loudly, and if I unplug the reverb send cable, it does nothing.

HOWEVER, I just noticed something while doing this test: if I remove the reverb tank from the amp cabinet and pull it away from the tubes, the louder 60Hz hum goes away, leaving a much quieter 60Hz hum "floor" underneath. If I move the pan near the tubes (especially the rectifier) or near the power transformer, the loud hum reappears.

So I need to come up with a way to properly shield the tank from the tubes themselves. I'm guessing it's the filament's AC that's doing this? Shouldn't the reverb pan's metal body, grounded correctly, shield it like a Faraday cage? Or does this mean that the ground is not good enough or that something needs to be decoupled further by an extra filter cap? I know for sure that the filament ground is very low on the bus wire--it is the next-to-closest to the transformer lug and just above the 0.047uF cap that shunts to ground off of one side of the fuse.

Moving the reverb pan away from the tubes isn't really an option, as the cabinet is very small. So I must do whatever it takes to properly shield the tank from the tubes' hum. I just read online that shielding the bottom of the pan with a metal cover works well, but it seems like the hum is going through the top of the pan anyway.

[Chuck H]

Since the tank is probably steel I don't think you'll find any better shielding (short of maybe a copper layer added to the steel like a Faraday shield). As strange as it seems you might try turning the tank 180 degrees. That and even in it's current position there will be quieter spots. With the tank plugged in and the hum going just move it around and find the quietest spot. Reverb tanks are really bad about picking up hum from transformers and power supplies. But these EMF's aren't all encompasing. There are usually voids in the field. Moving the tank around it's often possible to find a quiet spot. You must then mount it exactly there. I've even managed to put reverb tanks in head cabinets where if you moved the tank a half inch any direction the hum was unbearable, but almost silent where it was mounted.

[dchang0]

Thanks, man--you have definitely helped me figure out one of the biggest outstanding problems. I took the tank and placed it vertically, with the output side downwards, away from the tubes and especially the PT, and the hum dropped dramatically. I moved it back and forth a bit as you suggested and found a place where it's 99.9% gone.

Then there is still the very light 60Hz hum underneath...

I found this thread that has some good suggestions.

Reverb tank hum - diyAudio

The best choice is probably to drive the tank harder and dramatically reduce the gain on the recovery side (improve signal to noise ratio), but I will definitely try the other tricks (metal cover for open side of pan, etc.) Anything and everything that can help make this a recording quality amp!

Now, I've been reading that driving the tank with higher current is better than driving it with higher voltage gain, so I deliberately chose a load line for the triode-strapped-pentode 6BM8 that would result in cleaner tone and higher current. What do you think would happen if I traded current for voltage gain by choosing a lower load line, keeping a centered bias point at -14V? Never mind--I just looked at the graph, and the gain actually decreases a teeny bit.

I guess I'll have to start biasing hotter where the curves are farther apart to get more gain AND keep current close to where it was...

Crap--it's 1AM--time to sleep! This stuff is like a drug!

[Enzo]

Hum can come from a million places, and EACH source of hum will have its own cure. Whatever you do to lower ground related hum, for example, will have nothing to do with power supply ripple hum, and vice versa.

Before you start throwing hum cures at it, you really need to determine just where all these hums are coming from.

Just knowing what hum you are hearing helps. 120Hz hum is from power supply ripple one way or another. 60Hz hum is from something else. Both are the same tone, but the 120 is an octave higher than the 60.

For example, hum in the reverb tube circuit may not be reverb related hum at all. That recovery triode will certainly amplify any hum picked up by the reverb pan, but it also can be susceptible to heater induced hum. In that case, going to DC heaters or elevating AC heaters to a positive DC voltage will fight that, but doing anything to the reverb pan will have no effect.

And a wire can pick up hum from the flourescent lights in the room, or it can pick it up from the power transformer field. So in the case of a wire like that, shielding the chassis will help block the flourescent noise, but won;t help the magnetic field noise.

[dchang0]

Hi, Enzo--

It's definitely 60Hz hum and sounds most like a light ground loop. But if you have any tips on how to narrow down the source of the problem by process of elimination, that'd help.

99% chance that it's coming from the reverb circuit, as the amp did not hum even when dimed before the reverb circuit was installed.

It's not likely the reverb driver, because I can unplug the reverb driver cable and the hum remains. But it still could be a grounding problem in the driver circuit.

The recovery side is definitely too gainy, and I will be modifying the circuit to lower it. Hopefully, it'll reduce the causes of hum you mentioned. If it helps to diagnose the problem, as I turn the reverb pot up, the reverb signal ramps up at a steeper slope than the accompanying hum. Here's the latest graph:



As you can see, the noise is much better.

Ah well, not much I can do until the replacement 6BM8 arrives. The microphonics are so problematic that they get in the way of further diagnosis.

[dchang0]

Good news, guys. After reviewing the G-12 and 500 R2 schematics, I dropped the grid leak resistor on the recovery stage dramatically (any value below 500K seems to work well, though the other amps use 22K and 27K). I also increased the plate resistor from 180K to 220K, also inspired by the G-12 and 500 R2.

The result is that the 6BM8's microphonics are greatly reduced (though still present), and the reverb effect no longer distorts when turned past 8/10. That basically means that the amp is in a perfectly useable configuration, excepting the tiny bit of ground hum that I know I can remove by rearranging the ground buss.

I don't get the crazy surf levels I had before, but the trade off is worth it for clean reverb tones all the way up the dial. The only overdrive comes from the volume pot, as it should be.

So, I'm pretty much done, though I could try to improve the mix further for fun.

Here's the schematic for what is working right now.



Thanks very, very much to each of you who helped me out so much. I've learned a ton from you guys and now have a great amp to show for it. I'll rebuild it one more time with a more compact and properly-arranged layout and ground scheme, and then it becomes my main amp. Cheers!

[km6xz]

One question, did you check for leakage with the .022 cap coupling from the recovery amp plate. My calculations suggest that the tube, reverb pot and summing resistor values should not be dropping the dry signal as you have. If there is a little DC on the reverb pot, it would explain a lot.
If you do not have a hi z meter or scope, you can see the effects of operating point of the V1B plate with any meter while rotating the reverb pot. As drawn they circuit should do what you intended. Something is wrong with the unit, not the plan.

[dchang0]

I haven't checked for DC yet but will certainly do so! I have another .022 on order just in case. Thanks very much for double-checking my "plan." I'll get back to you in a bit with the results of the DC test.

[dchang0]

I checked the DC voltage at the junction between the .022uF coupling cap and the center wiper of the reverb pot. It's -0.009VDC, probably not a problem.

[dchang0]

All done, guys! Replacing the 500KA pot with a 100KA pot allowed me to go down to a 100K mix resistor and no reverb send leg resistor. In other words, the mix network got really, really simplified. See the final schematic at this thread (don't want to post in two places--might cause confusion).

http://music-electronics-forum.com/t22535/

Thanks to everyone for their help! When I rebuilt the eyelet board from scratch, I solved the noise issues at the same time. The amp is SUPER quiet, making a very light hum only when either pot is cranked past 9/10.