6g15 buzz

[Stevem]

This layout works everytime just like it did for Fender.

[oxbow_lake]

If it were me, I would try doing things to mess with hum introduction. For instance, short that 22 ohm resistor. Did the hum change any at all?

Shorting the 22 ohm resistor resulted in significantly louder buzz.

Does moving the balance control from full reverb to full dry change the quality of the hum any? Perhaps moving it quickly end to end? That might make an audible change that could help point to the issues.

The hum stays the same, but sounds reverb-y when on the wet side. However, I am getting new noise now on the wet side, as described below.

Does the reverb tank have both input and output jacks grounded to cable shield? Or is it one that has the nominal cable ground open on either input or output jack? Reverb tanks come in all three variations.

This particular tank says it has an insulated input jack and grounded output, and looking in the tank itself this seems accurate.

There is new weirdness happening now. On the wet side, (dry signal not affected), the signal is *highly* dependent on the positioning of the RCA cable. In certain orientations I get lots of AM radio, in other positions I get what almost sounds like blocking distortion through the reverb, and sometimes I can get a 50% drop in reverb signal altogether. This is all consistently repeatable based on how I position the cable, and doesn't seem like there's a loose jack or other intermittent connection. Seems like the cable is picking up RF interference which is then causing oscillations and swamping the recovery gain stage? If the cables are shielded and the shields are appropriately connected to signal ground, I'm not sure why this would be happening. Could the fact that the tank jacks are not both grounded be contributing to this? I tried grounding the input jack on the reverb tank (so both in and out are grounded) and the RF interference is unchanged.

It is unclear to me if this interference in the reverb path might be contributing to the hum present on the dry signal, especially given that when I ground out the tank recovery jack the noise goes away in the wet signal but the buzz remains in the dry signal.

If you have another 47uF/high voltage cap, you could try temporarily connecting it across C1 to see if the hum decreases. If so, the ripple voltage is a contributing evil.

I didn't have a 47uF cap, but did have a 33uF and connecting it in parallel did not appreciably change the hum level.

I was thinking about adding .01uF snubber caps across the rectifier diodes in case that was adding to 120hz hum. I think I can get some later today and try it, can't hurt.

[oxbow_lake]

This layout works everytime just like it did for Fender.

Point taken, I may have been led astray by reports on the internet that the original had noise issues.

[Stevem]

Snubber caps only reduce noise hash not 120 hz hum.

[R.G.]

As I understand it, the 6G15 has a reputation for potential unpredictable ground loop interaction the the guitar amp.

When I built one, I did the same as you have, using isolated jacks, a single circuit ground point, etc, but also implemented Merlin's Hum Block Loop circuit that connects the reverb-unit circuit ground to the chassis ground. I included that feature right from the beginning, so I never ran it without, but I have no noise issues.

Might be worth a try as it is only a few components & easy to install.

Oxbow_lake's schemo and pics show he already has it in there.

Historical note: there have been many varieties of that same cap-resistor-diode circuit. I'd have to do some research to get a definitive history, but I believe it originated from a simple removing of the ground wire to a chassis, maybe in the 50s or 60s. Ground lifts were certainly in the audio field in those times.
Ground loops come from equipment having different AC voltages on their chassis from the AC mains connections and any differences in internal leakage. There's a picture in Guitar Amp Wiring Notes showing how this can happen. This caused LARGE circulating currents in the chassis involved, and in the ground/sleeve connection of the units. Breaking the chassis-to-AC-line connection stopped the currents, but left the issue of ground voltage offsets.
Of course, this isn't hum free; it was adapted to using a resistor of various values instead of an open circuit, which helped by reducing the circulating currents in "ground", and also limited the voltage excursions.
RF could still be a problem. The capacitor across the resistor helps shunt the RF between the different chassis involved, and this helps a lot with keeping RF-to-audio under control.
Putting impedance in the safety ground path raises safety issues. A pair of back-to-back diodes was added to limit the voltage that could build up between units. Ideally, this would limit the user-touchable voltages to +/- a volt or so. I suspect that the four-diode variant came about because 25A full wave bridge modules are easy to mount and cheap. This flirts with the limits on voltage caused by a line fault to ground in a chassis; this means that the resistor-capacitor-diodes lift circuit is maybe-probably-could-be OK, but might well not pass safety certification.

[R.G.]

Shorting the 22 ohm resistor resulted in significantly louder buzz.

Hmmm. Are both the reverb and amp plugged into the same duplex outlet?
With the 22R in place and both amp and reverb turned on but not cabled together, is there an AC voltage from one chassis to the other?

This particular tank says it has an insulated input jack and grounded output, and looking in the tank itself this seems accurate.

There is new weirdness happening now. On the wet side, (dry signal not affected), the signal is *highly* dependent on the positioning of the RCA cable. In certain orientations I get lots of AM radio, in other positions I get what almost sounds like blocking distortion through the reverb, and sometimes I can get a 50% drop in reverb signal altogether. This is all consistently repeatable based on how I position the cable, and doesn't seem like there's a loose jack or other intermittent connection. Seems like the cable is picking up RF interference which is then causing oscillations and swamping the recovery gain stage? If the cables are shielded and the shields are appropriately connected to signal ground, I'm not sure why this would be happening. Could the fact that the tank jacks are not both grounded be contributing to this? I tried grounding the input jack on the reverb tank (so both in and out are grounded) and the RF interference is unchanged.

It is unclear to me if this interference in the reverb path might be contributing to the hum present on the dry signal, especially given that when I ground out the tank recovery jack the noise goes away in the wet signal but the buzz remains in the dry signal.

RF issues is a plausible theory, especially given that the position of the cable repeatably causes it to change. I would be tempted to do a slam dunk on RF issues if it were mine.
Do you have small (0.01 or so) ceramic caps from the sleeve contact of the input jacks to the chassis, within maybe ½” of lead wire?
Can you add a 100 to 1k resistor in series with the input signal right at the tube grid, with a 0.01uF ceramic to the local tube ground for each input? I would be tempted to add a 100R in series with a 0.01uF across the send jack to the reverb tank.
There are other more extreme measures that can be done if any of these change anything. None of these will degrade the sound and can be left in whether or not they help.

I didn't have a 47uF cap, but did have a 33uF and connecting it in parallel did not appreciably change the hum level.

OK. So 120hz ripple isn’t the main issue then.

I was thinking about adding .01uF snubber caps across the rectifier diodes in case that was adding to 120hz hum. I think I can get some later today and try it, can't hurt.

You’re right, it won’t hurt. There is a degenerate situation where SS diodes slam off abruptly and cause RF ringing on the connected wire leads that can broadcast blips of RF into the other circuits to be detected and sound like 60 or 120Hz or their combinations.

[B Ingram]

This layout works everytime just like it did for Fender.

No, it does not.

The original 6G15 (I have one from 1964) uses a 2-prong power cord, and thus does not create a ground-loop with the amp by way of 3-prong cords on both devices.

The Reissue Reverb Unit uses a ground-lift (CR5, CR6, R23) just like Oxbow_Lake is trying to implement, as well as a second resistor to isolate the pots (R22) all visible over in the power supply.

This was a necessary change because Fender's Reissue Reverb, and their amps, all have 3-prong cords to ground all chassis for safety. One could break the resulting ground-loop with an audio transformer (say, at the amp's input jack), but Fender implemented the less-expensive ground-lift on the Reverb Unit instead to resolve the issue.

[oxbow_lake]

...as well as a second resistor to isolate the pots (R22) all visible over in the power supply.

Thanks for the link, that is useful! R22 is interesting. For non-pcb mount pots, would you just solder a "ground" wire along the pot backs and connect that to signal ground through the 15R resistor?

Also interesting that the filter caps start at 22uF and end with 47.

[R.G.]

Thanks for the link, that is useful! R22 is interesting. For non-pcb mount pots, would you just solder a "ground" wire along the pot backs and connect that to signal ground through the 15R resistor?

For pots that are enclosed-back style, the enclosed back is an extension of the chassis, so the signal lugs should not be tied to the pot backs, if that's what you meant. If you mean a ground buss bar all along the pots that the signal lugs attach to - that's one way to do it. Not as predictable for hum/noise performance as star grounding, but usable.

The 15R grounding hack may work in certain situations, but it's like all non-star-ground systems: it needs tinkered.

[oxbow_lake]

Well, the snubbers (0.01uF over the PT secondaries and across each rectifier diode) helped, the hum is almost tolerable with the reverb going into the front of the amp. The tank/RF noise/distortion is still problematic, so I will work on RF-proofing it over the next couple of days as free time allows.

[R.G.]

Well, the snubbers (0.01uF over the PT secondaries and across each rectifier diode) helped, the hum is almost tolerable with the reverb going into the front of the amp. The tank/RF noise/distortion is still problematic, so I will work on RF-proofing it over the next couple of days as free time allows.

Congrats! You nailed a hum source. I feel slow for not noticing that the rectifiers are a big bridge assembly. These usually involve the worst diodes in terms of slamming off and needing snubbing.

Caps across rectifier diodes help, usually, but an even better solution would be to replace the diode bridge with a tag strip to hold a couple of UF4007 diodes. The "UF" means ultra-fast recovery and soft recovery transients. These would cut down on the source of the RF blasts even more.

Page 20 of Guitar Amp Wiring Notes also lists a trick I found during writing it up. Making the cap you have across the transformer secondary be instead a 1000pF in series with a 1K resistor is supposed to swamp rectifier slam-off noise even better than caps across the secondary and the rectifiers. Might be worth a try.

[oxbow_lake]

Good news! With the following alterations the reverb performs in a satisfactory manor, with very much tolerable hum even when the mix is set entirely wet and the amp gain is cranked up (which sounds like hell anyway, so not typical usage).

1. 1k grid stopper with 470pF cap shunting it to ground on the reverb recovery stage (U2B on the schematic)
2. 10k grid stopper on the dry signal cathode follower (U2A on the schematic)
3. 0.01uF snubber caps on the PT secondary and rectifier diodes as previously described

Looking in Merlin Blencowe's preamp book he shows the two 100 ohm resistors going to the power tube cathode as a legitimate way of making an artificial center tap and elevating the voltage, so I put it back that way just to make the wiring a little cleaner. Not sure it made much difference in terms of hum, but it's not worse.

Thank you all so much for taking the time to help me with this. Every time I post here I learn a ton, and so far have always ended up with my problem solved.

[oxbow_lake]

Caps across rectifier diodes help, usually, but an even better solution would be to replace the diode bridge with a tag strip to hold a couple of UF4007 diodes.

Page 20 of Guitar Amp Wiring Notes also lists a trick I found during writing it up. Making the cap you have across the transformer secondary be instead a 1000pF in series with a 1K resistor is supposed to swamp rectifier slam-off noise even better than caps across the secondary and the rectifiers. Might be worth a try.

I used the diode bridge because it's what I had lying around, but this is an easy enough thing to change that next time I'm ordering parts I'll give it a try. Guitar Amp Wiring Notes was what made me think the snubbing would even be worth a shot, so thanks again!

[R.G.]

I used the diode bridge because it's what I had lying around, but this is an easy enough thing to change that next time I'm ordering parts I'll give it a try. Guitar Amp Wiring Notes was what made me think the snubbing would even be worth a shot, so thanks again!

Good on you!! And thanks for the real world feedback. I would be very interested if the 1000pF/1K winding snubber works for you. Works great in simulation, but I haven't done a build since I found that trick. The technique was developed by a batch of hifi builders trying to get rid of rectifier buzz in their high end amps.

[oxbow_lake]

Turns out the RF interference (?) is still there, it's just harder to notice if the tone knob is turned below about halfway. The wet signal has a constant crackling sound and some hum/buzz. Additionally, I get radio interference and sometimes full-on oscillation at the output jack of the reverb unit (~100mV at ~12kHz) depending on the positioning of the RCA cable.

One thing that consistently reduces the volume of all of the noise is if I am physically holding the RCA cable and also touching the chassis. This works if I touch the chassis of the reverb unit, amp, or any grounded equipment plugged into the same power strip. I feel like this is a big clue, but I'm too thick to know what it means.

I feel like this is a grounding issue. I checked all my isolation washers, everything that is meant to be isolated seems isolated.

Chopsticking and pushing leads around made no difference.

Swapping tubes out made no difference.

Following some of the previously posted suggestions:

Hmmm. Are both the reverb and amp plugged into the same duplex outlet?
With the 22R in place and both amp and reverb turned on but not cabled together, is there an AC voltage from one chassis to the other?

Both are plugged into the same power strip. There is no AC voltage between the un-coupled chassis.

Do you have small (0.01 or so) ceramic caps from the sleeve contact of the input jacks to the chassis, within maybe ½” of lead wire?

Do you mean the RCA jack coming back from the tank, and the guitar input jack? And you're saying go from the signal ground to chassis at these two points? If so this had no effect.

Can you add a 100 to 1k resistor in series with the input signal right at the tube grid, with a 0.01uF ceramic to the local tube ground for each input?

10k grid stoppers with each grid shunted to local (signal) ground through a 470pF ceramic cap has not helped things when I put them on the input stage of the 12AT7 (U1B, pin 2) and the tank recovery 12AX7 (U2B, pin 2).

I would be tempted to add a 100R in series with a 0.01uF across the send jack to the reverb tank.

Also no improvement.

A 470pF from plate to cathode of the recovery stage also did not reduce RF noise or oscillation.

The noise is present regardless of what amplifier I plug into. The noise is actually still present going into the effects loop return of the amp, just quieter - if I put a boost the signal coming back with a pedal I get the same noise. So I don't think this has to do with the preamp of the amplifier anyway.

I have a new RCA cable coming. I have UF4007s coming to replace the diode bridge.

My last thought was that when I first put the amp together I had the rectification scheme wrong and was producing ~550V DC rather than 310V. It took a couple of seconds (literally) before I realized this, so the filter caps (all rated 450V) were exposed to that for a brief period. Their ESRs are all normal, and jumping the caps with that new 33uF electrolytic doesn't resolve it, so I don't think this is a case of bad filter caps, even though they saw some brief abuse.