Humbucking Reverb Tank?

[Helmholtz]

I also thought about a notch filter. But it will mainly catch the fundamental. Don't know if that is enough.

Also if the passive notch filter is directly connected to the reverb transducer I'd expect some interaction with the mostly inductive impedance of the transducer.

[nickb]

I also thought about a notch filter. But it will mainly catch the fundamental. Don't know if that is enough.

Also if the passive notch filter is directly connected to the reverb transducer I'd expect some interaction with the mostly inductive impedance of the transducer.

The power transformer magnetizing current is quite distorted so there will be odd order harmonics present. That said if you hold the tank close to the PT there is a lot of fundamental there which leads me to believe that it would be quite effective. It needs to be tried out to see well it works. I'd be surprised if you could get even 20dB of improvement with a humbucker unless very carefully designed. The Twin T is very easy to try, cheap, and not dependent on orientation. A possible downside I see is that it does attenuate some of the low end, but without hearing it I cannot say if that is good thing or not.

The impedance of the inductance is negligible in comparison to the network values at the frequencies of interest. Any effects don't kick in until you get to the kHz, about 3db down at 4Khz. I suspect that is well outside of the tanks useful passband.

[jimboyogi]

The power transformer magnetizing current is quite distorted so there will be odd order harmonics present. That said if you hold the tank close to the PT there is a lot of fundamental there which leads me to believe that it would be quite effective. It needs to be tried out to see well it works. I'd be surprised if you could get even 20dB of improvement with a humbucker unless very carefully designed. The Twin T is very easy to try, cheap, and not dependent on orientation. A possible downside I see is that it does attenuate some of the low end, but without hearing it I cannot say if that is good thing or not.

The impedance of the inductance is negligible in comparison to the network values at the frequencies of interest. Any effects don't kick in until you get to the kHz, about 3db down at 4Khz. I suspect that is well outside of the tanks useful passband.

The notch filter is a good and interesting idea. My current plan is to investigate the dual coil hum bucking concept a bit further, but if it doesn't work or reduce the hum enough then I'll definitely trial the notch filter.

[jimboyogi]

Inductance will double only if the dummy coil has a similar core as the transducer. Using a dummy coil with core is advisable anyway as the core influences its hum sensing properties. A high permeability core focusses the interference field.

The impedance of reverb transducers is typically more than 10 times higher than the DCR, indicating that the impedance is mostly determined by the inductance.

Output transducer inductance may matter if "resonance tuning" with a parallel cap is used to improve high frequency response.

This is all such good information!
The reverb return amplification stage is a classic 1/2 12AX7 triode, using a 1M grid leak and 100K plate resistor. With this high impedance input, I don't think that a doubling of transducer impedance (from 2.25K to 4.5K) will make much difference. I can foresee a slight loss of HF, but this won't be a problem either in this circuit - I currently have a simple LP filter on the reverb tank output to attenuate HF, I could adjust or remove this if HF loss is too great.

R.G. - That's great advice about a simple way to use a dummy coil for hum cancellation. I am quite keen on the idea of generating 2x the reverb signal by having the 2 coils on the transducer, as this may increase the signal/noise ratio a bit further to help reduce the EMF noise. This implies connecting the coils in series, but I'll have to try and see whether (anti)series or (anti)parallel gives the best sonic results. But I will trial the dummy coil idea also!

[jimboyogi]

There's a unique approach! Good ol' Craig, he would think up something like that.

In comparison, many years ago I was asked by a studio owner if I could reduce/eliminate the hum in his expensive and presumably well-engineered Klark-Teknik "studio quality" spring reverb. I got nowhere. If KT had employed a method like this, or else a dummy coil, they'd be so far ahead of the game! In the end, I told him to find an electrically "quiet" place in his building where he could park his KT reverb, and run long balanced lines there and back.

Thanks Mark Hammer and Leo_Gnardo.
While searching the net about hum bucking reverb ideas I read of the existence of this rack reverb using the 2 tanks together. As I have a few reverb tanks, I gave the concept a try. I took 2 Accutronics tanks with the same code, and rewired them so that for both the input and output transducers, the coils were wired together in (anti)series. The concept was that each tank was 180 degrees out of phase with the other for both drive and recovery, and therefore in phase for the reverb signal output, but 180 degrees out of phase for the EMF signal. I then put them together and grounded both tank chassis to the one ground (output stage), and plugged it into the amp.
It seemed to have more reverb signal, but ~ 1/2 of the EMF signal at idle. The character of the EMF noise was different to the single tank also, while the single tank gave a "buzz" with lots of harmonics, the doubled (anti)series tanks gave a purer 50Hz hum.
So it worked a bit, but the EMF noise was still louder than I want. I wonder why the EMF cancellation wasn't more effective. Perhaps because the 2 output transducers were near to each other, but still picking up a slightly different EMF field? Or did I not connect them correctly for optimal noise cancelation?

[Helmholtz]

while the single tank gave a "buzz" with lots of harmonics

This indicates that it is probably not sufficient to suppress the mains fundamental with a 50Hz/60Hz notch filter.

But I wonder if all the "buzz" is actually picked up by the transducer or if there is still another source like e.g. a grounding issue.

[R.G.]

There is a hum suppressor that is a bit more work, but that would probably turn in good results. It involves generating 60, 120 180, and 240 hz from the power line, then feeding a variable amount and phase shifted bit of that back into a mixer. This approach is to cancel the power line and its first three harmonics irrespective of where it's generated. Very flexible.

[Chuck H]

There is a hum suppressor that is a bit more work, but that would probably turn in good results. It involves generating 60, 120 180, and 240 hz from the power line, then feeding a variable amount and phase shifted bit of that back into a mixer. This approach is to cancel the power line and its first three harmonics irrespective of where it's generated. Very flexible.

Too nerdy. Two tanks has way more mojo

But seriously... Got a link or other reference? Sounds fun.

[jimboyogi]

Too nerdy. Two tanks has way more mojo

But seriously... Got a link or other reference? Sounds fun.

Too nerdy for MEF? Isn't that the point?

It's true that 2 tanks are big mojo, I wish my 2 tank experiment had worked better at eliminating the EMF noise.

I just spent 2 hours after work trying to cut laminations for a test 2 coil transducer, and failing. I'm cutting used laminations from an old expired EI transformer. I tried to cut the laminations together, as they are glued/rusted together. Introduce an electric hacksaw and they are no longer glued together, they are separated and bent out of shape Will try again tomorrow cutting one lamination at a time, then clamping together to true up with a file. That's the plan anyway. If that doesn't work, there's a wealth of other great suggestions in this thread to try!

[R.G.]

Too nerdy. Two tanks has way more mojo

True, but then nothing succeeds like excess.

But seriously... Got a link or other reference? Sounds fun.

Sure.

https://ethanwiner.com/filters.html

Built one. Works. The only issue with using this in general for hum-killing is that different situations of AC line, position of the amp, different guitars at the input, different cords, etc. will mess with the hum that you need to cancel, so it will have to be tinkered when you change things like that. But it ought to be killer at cancelling something like the hum inside an amp for just the reverb tank.

[nickb]

A completely different approach that is 100% immune to radiated fields and has a variety of springs reverb sounds and doesn't go CRASH if you whack it yet still costs about the same (or less) is this: https://www.pedalpcb.com/product/radiumsprings/

If you're a cheapskate a readily available PT2399 module can be hacked into a very respectable reverb for under $3. I used one to replace a cheap 9in reverb tank in a PA that someone had destroyed by connecting the input coil to the main output, perhaps to get more reverb, ***sigh*** .

[R.G.]

I also did a PCB to let a Beltone BTDR2 accept almost all coil-tank input levels and output signal acceptable to most tank recovery circuits. That is - a BTDR tank replacement.

[Chuck H]

I also did a PCB to let a Beltone BTDR2 accept almost all coil-tank input levels and output signal acceptable to most tank recovery circuits. That is - a BTDR tank replacement.

I considered purchasing something like that a couple of times when I've had issues with microphony, but having solved the problem the old fashioned way (pan isolation with heaps of foam) I never actually looked into it. I had assumed such a thing was already available?!? Are you saying you offer this in lieu of it being available otherwise? It seemed like such an obvious thing to have have a reverb tank plug in replacement "brick" considering the problems with acoustic tanks.

[R.G.]

It seemed patently obvious to me that a circuit block that met the input/output specs of a reverb tank but was solid state inside was needed. So I went and looked. It's been a couple of years since I did this, but I didn't find one then. So I designed one. It's a BTDR with one dual opamp for buffer/gain matching, two pots for signal level control, and some attention to power supply setup to get 5V@200ma from an amp. The idea was to find a dead mini-tank enclosure, gut it and keep the enclosure and phono jacks, and add a DC jack to it. Circuit worked OK on the proto I built up. The input sensitivity was variable to account for the variable sizes of drive on amps and the wide number of "input impedances" that real electromagnetic tanks have.

I didn't try to market it because I'm not a get-rich-selling-music-stuff kind of guy - just a DIY enabler.

[jimboyogi]

An update on my experimentation with the hum bucking 2 coil output transducer.

With my equipment and skills , I had no success cutting laminations to fabricate a new transducer core. So I hacked into an existing core, cutting one end from an "EI" output core, so it resembles an "E" with elongated upper and lower horizontal edges. I then stole an extra coil from another tank, and mounted 2 coils onto this core - 1 on the central limb, and 1 on the lower limb, both mounted near the limb ends. The magnetised coil end beads then sit between these 2 limbs. This should give these 2 limbs opposing magnetic fields as the beads move between them, giving the reverse polarity required for the hum cancelation.

I tried connecting the wiring in various combinations while listening to the EMF noise, both in series and parallel. There was a significant reduction in noise when connected in anti-series and in anti-parallel. However the noise reduction was not enough to solve the problem, quite disappointing. Also the actual reverb signal output was significantly reduced - I checked this in all wiring combinations to make sure it wasn't a phasing issue. I wonder if by opening up the laminated core, I lowered the efficiency of the transducer significantly, causing the loss of reverb signal recovery.

I still need to try the dummy coil idea.