How to properly ground an input isolated/output grounded reverb tank?

[dwrockdoctor]

Hi all,

I've put together a DIY tube spring reverb based off of the lamington reverb. I'm having some ground loop issues somewhere in the recovery stage, when I dial in more wet signal the more hum I get, complete dry has no hum. I've even gone as far as using a hum loop blocker for all the signal grounds just like in the 63 RI fender verb but to no avail.

My reverb tank is of the isolated input/grounded output variety and currently I have the sleeves of both the input and output going to the same ground point connection, is this where my problem lies?

I can't get my head over how to best ground the RCA jacks for the tank, especially the isolated input.

Thanks for reading guys!

[Enzo]

OK, to be clear, you have hum when you turn up the reverb return level? That doesn;t mean it is the result of a ground loop.

Provide a schematic for what you built. Some circuits use a isolated input ground but don;t need to. The solid state drives in many amps DO need the input to be isolated. In those cases, the shield is NOT grounded, the shield connects to the isolated "ground" at the reverb pan input RCA jack.

Isolate the source of the hum. Unplug the drive from the INPUT jack on the pan. If the hum remains, then the input jack wiring was not involved.

Aside from hum, do you still hear the reverb spring sound? If you bang on the reverb pan to crash the springs, do you get the loud sound? Unplug the cord from the pan OUTPUT, still hum?

With cord plugged into output, pull the pan out of the cab and set as far away from the power transformer as possible. ANy difference?

The pan output end should be mounted away from the power transformer. The transducer can pick up the field.

Pull the cord from the OUTPUT jack, and measure resistance inside the pan at that end. Should not measure open, should measure something like maybe 200 ohms.

[dwrockdoctor]

Yep, hum as soon as I start mixing in more reverb. No hum when I set the mix control to all dry. I still get the spring sound, the reverb tone is actually really luscious and drippy so no dramas with it actually working, the springs crash when I give it a bash too, just a pervasive hum that I have issue with. Not as noticeable when I'm playing but still fairly apparent and I'd like to use it set to full wet sometimes for certain sounds. Inserting a graphic EQ into the wet sound and dropping out 100hz cuts a good chunk of the hum out.

When I unplug the reverb tank output I get god awful noise, from memory I don't think that happens when I pull out the reverb tank input but I'd have to double check that and get back to you. I also have the tank outside the main chassis on long RCA cables and moving it around doesn't affect the hum at all, it stays consistent no matter where the tank is situated.

Here's a schematic of the circuit, I didn't use the AC plugpack and stepup transformer, just a toroidal I had laying around that puts out voltage in the same ballpark. Also using a different tank then is in the schematic, it has the same input impedance but a lower output impedance than the suggested tank. Both tanks have the same input/output grounding scheme though.

[Enzo]

The return amp is a sensitive, high gain input, so pulling the cord from the pan OUTPUT can indeed pick up noise. The drive end is an output from the amp circuit, not an input, so pulling that cord ought not introduce noise. But if it did, it would be a clue.

You really ought to have a grid reference for the return tube, pin 7 of V1b. A 1 meg would do nicely. That explains the noise when the cird is unplugged there.

The schematic says 8EB2C1B for the reverb pan type. Considering the circuit, I would recommend 8FB2C1B. That is the next higher input impedance, which will make that drive signal from the plate of the pentode happier. I myself would prefer a 4FB2C1B or even a 9FB2C1B. Those are the full size pans rather than the shorty pan of the 8 series. But that is personal taste.

Also using a different tank then is in the schematic, it has the same input impedance but a lower output impedance than the suggested tank.

That doesn't sound right. All the reverb pans I ever see for sale have the same output impedance - the B letter - the maker has listed a couple odd impedances in their chart, but in 50 years of it, I have never seen one.

The input side - the first letter - is where they vary. The low impedance ones like Fender uses are A series. 4AB2C1B. F is the highest impedance, as in 4FB2C1B.

The schematic is unclear: It appears that both cables are grounded together. Is there a ground wire from the pan frame back to the circuit ground in the chassis?

This circuit clearly needs the jacks grounded at both ends, unlike a solid state drive I mentioned as a possibility.

[dwrockdoctor]

The tank I'm using is a Belton 3EA1C1B. 800ohm input 600ohm output. I pulled it from an old SS trace elliot combo I got given ages ago. The 8EB2C1B has the same input but the output is much higher at 2575ohms on the 8EB2C1B than the one I'm using so not sure if that will cause issues? If so, would there be a lot of changes I'd need to make to the recovery stage to accommodate what I have?

I also found the schematic a bit unclear as to grounding the tank also. As I have mine hooked up, I have two separate ground wires running from the send and return RCA sleeves. Not using any shielded wire at this point.

And yes, it was the the RCA going to the recovery stage that created lots of noise when unplugged, I'll chuck a 1M resistor to ground off of pin 7 of V1b next chance I get.

Thanks for taking the time to help Enzo

[Enzo]

Hey, I believe you, it is a new one on me.

That 2500 ohm output is the common type we see on a million amps every day. All the Fender et al feed that right into a 12AX7 grid like here. 600 ohms is pretty low feeding that grid, though it might work OK.

Not using any shielded wire at this point.

Oh, well that there is a huge problem. The return wires need to be shielded. Just like a guitar cord needs to be shielded. That will be a huge antenna picking up noise if unshielded.

Use shielded cables for both jacks, grounding the shield up in the chassis. The run a separate ground wire to the reverb pan body to the chassis.

[Mick Bailey]

I would remove the RCA plug from the output of the tank and short it from hot to ground. Does the noise disappear with the reverb mix fully wet?

Edit; Just noticed you're not using shielded cable - you need to sort that first.

[dwrockdoctor]

Okay, great, that gives me lots to play with now and expanded my admittedly limited knowledge on the subject of spring reverb. I've basically made this out of mostly spare parts I already had without having to invest much extra money into. The circuit gets a really nice luscious verb going, first tube reverb I've gotten to play through properly, hopefully shielding the lines in and out of the tank will clean up most of the hum.

I'll let you guys know how I go. Thanks so much for sharing your knowledge, massively appreciated!

[dwrockdoctor]

Okay, I've tried the suggestion with shielding the RCA jacks that connect to the send and recovery circuits and experimented with different combinations of how to ground the tank and still can't kill the hum. Unplugging either or both send and return RCA jacks does nothing to kill the hum either. Not sure if the RCA cable I'm using is grounded or not though so maybe I should try a few different cables in case. Currently the tank is on long cabling separate from the main chassis so I can experiment with placement but no placement no matter how far away from chassis changes the hum.

I'm most likely going to tear down all the grounds and redo that end to see if that helps as I feel like I can improve that. At the moment I have all the grounds from the power filter caps up through all the other circuitry ground going to a hum blocker loop (two 6A diodes, a 600V cap and resistor in series to ground) to a star chassis ground. Mains earth is on a separate ground connection. I might try a bus bar ground system with everything connecting in correct order from power filter caps on up and see if that will clean things up.

Thanks again for all the help, its hugely appreciated

[Enzo]

It isn't just grounding, it is shielding. The drive side shouldn't matter.

Unplug the pan. You now have a cable from the chassis out to where the pan was. Bring your hand near it, does that increase the hum? Pick it up, don't touch the tip, increase hum? If yes to either, you are either not shielded or your shield is not grounded.

WHen I make RCA cables for reverbs I usually buy a plain old RCA cable like your stereo set uses. I plug one end into the reverb pan, and at the other end I cut off the RCA plug, and strip back the wire to solder into the chassis. If the chassis has RCA jacks on it, then plain old stereo patch cords work.

[dwrockdoctor]

Hey Enzo,

I finally got around to testing the RCA cables unplugged from the reverb pan but still connected to the circuit turn on. I was getting fluctuations in hum and noise on the reverb return RCA when moving my hand in proximity to the jacks ends. I shielded the internal wiring correctly going to the chassis RCA jacks on the chassis from the circuit reverb input and outputs but obviously the cables themselves aren't shielded (or not shielded correctly by me). They were also microphonic when I tapped them. I will run to the electronics shop and buy some good quality shielded cable to hardwire the jacks to the pan directly to the circuit with proper shielding protocol. Hopefully this will settle any hum left.

Thanks again Enzo and I will update on how I go!

[dwrockdoctor]

Back again! On a whim I tried a tube swap on the 12AX7 and that helped bump down the hum at 100hz to an acceptable level, must've been a partuclar microphonic preamp tube I had in there before.

I still think that I could get better performance out of the recovery stage due to the tanks output impedance mismatch (only 600 ohm when the circuit calls for a 2250ohm output tank). Is there any particular way I can bump up the impedance coming out of the tank before it hits the recovery stage, such as a transformer or something?

Otherwise I might just have to buy the recommended tank but this project was kind of about seeing what I can eek out of what I already have at hand for minimal cost.

Regards,

Dale.