Balanced/hum bucking spring reverb recovery

[Tubetec]

Today I was able to make meaningful comparisons between a few different tanks ,
I was able to find a specific combination of tank ,local grounding and wiring arrangement that gets the capacitance seen from either end of the recovery transducer to ground down to an entirely insignifigant amount ,
I was looking at 2x80pF into 1meg with the all metal 80's era Peavey tank below ,
Im now looking at 2x6.5pF into 1Meg and the majority of that is due to the presence of the phono plug and proximity of +/- wires off the transducer .

I was able to get some low capacity Sommer cable from a good old buddy ,
Tomorrow I'll wire it all up and allow the ears hear the effect of the nearly 10:1 reduction in capacitance , I expect there to be night and day difference in how the spring behaves without the usual capacitance loading it down .

The hum-bucking element of this is self suggesting ,ie two recovery coils in as close proximity to the same physical space rejects interference better .....

TBC.

 

[Tubetec]

I know you want to combine a reverb tank /drive/recovery circuit and psu in the one box Rock .
Unbalanced reverb recovery is a non starter if you have a mains transformer anywhere nearby .
Unbalanced with hum bucking might be viable .

I'll try a few more tests this afternoon ,
Im going to ditch the TRS jack in favour of XLR ,
now that the capacitance of the pickup assembly has been reduced to a minimum , the unbalance a TRS jack puts on the signal is unwelcome in this case .
I have two short lenghts of low capacity cable . I'll only connect the screen at the XLR end , and create a seperate return wire for the tank enclosure to avoid creating a loop .

 

[rock soderstrom]

I know you want to combine a reverb tank /drive/recovery circuit and psu in the one box Rock .
Unbalanced reverb recovery is a non starter if you have a mains transformer anywhere nearby .
Unbalanced with hum bucking might be viable .

I'll try a few more tests this afternoon ,
Im going to ditch the TRS jack in favour of XLR ,
now that the capacitance of the pickup assembly has been reduced to a minimum , the unbalance a TRS jack puts on the signal is unwelcome in this case .
I have two short lenghts of low capacity cable . I'll only connect the screen at the XLR end , and create a seperate return wire for the tank enclosure to avoid creating a loop .

I am following your expedition with great interest.

 

[drtechno]

Today I was able to make meaningful comparisons between a few different tanks ,
I was able to find a specific combination of tank ,local grounding and wiring arrangement that gets the capacitance seen from either end of the recovery transducer to ground down to an entirely insignifigant amount ,
I was looking at 2x80pF into 1meg with the all metal 80's era Peavey tank below ,
Im now looking at 2x6.5pF into 1Meg and the majority of that is due to the presence of the phono plug and proximity of +/- wires off the transducer .

I was able to get some low capacity Sommer cable from a good old buddy ,
Tomorrow I'll wire it all up and allow the ears hear the effect of the nearly 10:1 reduction in capacitance , I expect there to be night and day difference in how the spring behaves without the usual capacitance loading it down .

The hum-bucking element of this is self suggesting ,ie two recovery coils in as close proximity to the same physical space rejects interference better .....

TBC.View attachment 119634

What is your resistance of the input? and what resistance is the output?
Because in a lot of cases all you would need is a dynamic mic transformer on the output, and a simple pad on the input.

 

[Migs 31]

Naive question here.
Would something like this work?
NOT my work. I got that on the net (Forsell or someone else).

"Merry New Year !!!"

M

 

[Tubetec]

I should have linked the preceeding topic for reference purposes ,
Microphonic Starquad ,in case anyone hadn't joined the dots ......


So with the most unfavorable tank shell ,sub tray ,inductor core grounding(0V) arangement and cheap cable Hi-z cable I observed a conductor to screen capacity of 2x 123pF +/-2%
A !:10 reduction was a little ambitious ,
the TRS jacks unbalance could no longer be tollerated ,
I ended up with very close to 2x45pF including coil self capacitance + 2x 25cm of hi-z cable(@80pF/M) +XLR into 1Meg balanced .

I'll give a proper check on the hum bucking inductor config tomorrow ,but a high margin of extra S/N reduction is to be expected from nearby sources of 50/60hz ,ie mains transformers .

Sources of mechanical noise in the reverb tank sub assembly itself are becoming more evident now .


Theres a bad buzz on my noise floor-
You wouldnt like me when Im angry .

 

[Tubetec]

I did comparisons between a couple of tanks into 1Meg balanced ,
A Belton with a 210ohm coil , 2250ohm impedence and the other unit I showed previously with a 370ohm coil approx 5000 ohms impedence .

Lots more level from the 370 ohm tank , but also much more sparkly highs , despite the much larger incidental capacitance compared to the other due to coil ,core and grounding arrangements .

I was wondering , would it be possible to make a hum bucking pick up coil for a reverb tank and also reduce the coils self capacitance ?
Would a dual chamber bobbin like what you see below help me ? ( if I could find one with the correct dimensions)

 

[EthanSpaulding]

Are you doing these measurements into 1M to exacerbate the capacitance? For an example, mesa boogie used 220k grid resistor for return for a long time, then switched to 22k using, apparently, the same tank.

I could see the dual chamber bobbin being useful for adding a dummy coil (reverse wound coil) to simply cancel the hum. I think this solution would work well as these coils are small in geometry, the inductance will be small, and the capacitance will be small. My guess is that any effects will be above the frequency response that you are looking for.

Possibly a more exciting solution would be to have a hum bucking coil arrangement where there is a coil on either side of the magnet. So, one coil sees the South Pole of the magnet, and the other coil sees the North Pole of the magnet. Then connecting one coil in reverse puts the signal in-phase, and puts the noise/hum out-of-phase. (My apologies for re-stating what you already said above about the hum bucking coil)

 

[Tubetec]

Thanks Migs and Ethan ,
Well the idea with the single coil balanced pick up arrangement was to help reduce induced noises from nearby transformers etc ,which it does very well .
I tried varying the matching from 1Meg down by placing a pot across the transducer , it didnt have much effect at all until the pot created a short circuit across the coil and levels dropped away completely .

What I was thinking about was two tanks in series ,as the output from each is uncorrelated the signal adds which ever way you connect the coils , but in one config external noises are effectively reduced .
Coil placement and orientation may well have a bearing on how good the noise rejection is ,

Two coils on a single transducer sounds like a plan , but thats difficult because the coil normally fits on the centre limb of the core and thats the only way it could work with the lams involved .
A dual chamber bobbin might work though .

 

[Tubetec]

I reviewed a few previous posts on reverb tanks ,
most modern tanks use a tiny blob of glue to hold the connection from the spring to the transducer stable .
My go to method for removing this is a very fine tipped soldering iron , no chemicals involved ,
you can use a tiny drop of epoxy or cynao-acryl gel applied from a pin head to prevent any chance of metal on metal noises after you've made your mods .

Uncle Dougies video style is somewhat pedantic ,but I sat though the most of it and learned a few nice tips ,cheers

 

[EthanSpaulding]

So, if I'm understanding correctly, you don't gain any noise reduction by lowering the grid resistor from 1M to 220k (and even lower). I have seen schematics where the return stage has no grid resistor and the ground reference is simply through the reverb tanks coil. Which, I suppose makes sense.

This is a really interesting idea. I saw no easy solution after examining a few reverb tanks I have here. UI laminations could do the trick. Or just build a separate transducer to sense the coil directly. Like you would build single string humbucker for a bass guitar.

 

[Tubetec]

One interesting thing about the Uncle Doug video is it shows the amount of metal between the mains transformer and the tank in the Fender reverb , First you have the chassis , then the pan for the smoothing caps then the tank enclosure itself , all helping cut down hum entering the output transducer .

So the input resistance of the SSL2 on Hi-z is 1Meg , I put a 1Meg rheostat wired pot in parrallel , so resulting load on the tank is 500kohms , theres very little difference in levels from the spring until the pot reaches almost the end of its travel , when the sound cuts abruptly .

Whats confusing me now is that the tank with a lot of extra capacitance and 370 ohm coil sounds brighter than the one with less capacitance and only a 210 ohm coil , its a bit of a head scratcher .

The other side of the equation I havent mentioned yet is drive impedence ,
Ive tried out various tanks with different drive impedences , higher values do ok from a line output lower impedences like 100-200 ohms can be driven with a headphone output , but by far the best results Ive got have been using an 8 ohm drive coil and small power amplifier .
If the unit is to be used with a guitar amp you need an attenuator in the form of a wirewound pot , I usually choose 100-200 ohms . this works very very well , you dial in your sound on the amp , then attenuate the speaker output to the tank to prevent gross distortion in the drive transducer .

I did see Waynes four spring stereo reverb , I might try and find the documentation and schematic to see what he has going on inside .

 

[KA-Electonics.com]

MasterRoom XL-305 clone: MicMix MasterRoom XL-305 Spring Reverb Clone - Pro Audio Design Forum

The MicMix XL-305 had 6 lines per channel using Accutronics Type 9 three spring assemblies.

In the clone version I made some changes in drive and recovery.

The drive coils are 10Ω (1 kHz) wired in anti-parallel for hum-bucking in the receive pickup coils.
(Maintaining a low drive impedance also helps bass response.)
The drive amp is equalized, very low impedance output, non-constant current to provide damping and prevent cross-communication between transducers.
(Constant-current drive with its high output impedance sounds less clear and defined. That was a significant departure from the MicMix MasterRooms. In addition MicMix used series/parallel transducer connections which allowed them to back-feed one-another.)
There is also a HF rolloff to prevent driving the springs past the point they have output and reverberate.
The pickup coils are 2575Ω wired in series anti-polarity for hum-bucking. The pickups have a mu-metal wrap added.
The recovery amp is a fully-balanced instrumentation amp with a common mode rejection stage similar to my MM phono preamp.
An external switching power supply is used.

And, most importantly, the timings of the springs follow an "equally-tempered" logarithmic progression with the left and right channel timings interleaved to create a stereo image.

The clones:
AudioScape XL-305R: AudioScape XL-305R Equally Tempered Stereo Reverb — Vintage Pro Audio Equipment | AudioScape Engineering Co.
AudioScape now has a XL-305 plugin: XL-305R Equally Tempered Stereo Reverb Plugin — Vintage Pro Audio Equipment | AudioScape Engineering Co.
Fuse Audio VREV-305 plugin: Fuse Audio Labs | Audio Plugins and Mobile Audio Apps

For the record I don't receive any royalties from the sale of plugins or hardware but do work very closely with Chris at AudioScape from time-to-time.

 

[abbey road d enfer]

What I was thinking about was two tanks in series ,as the output from each is uncorrelated the signal adds which ever way you connect the coils , but in one config external noises are effectively reduced .

That's about what I did in the 80's, except I wired them in parallels and it worked quite well.
Actually reducing capacitive loading is counterproductive. Accutronics used to publish recommanded capacitive loads, which produced a HF hump that help cancelling the natural "darkness" of spring elements.
I used the tanks with the 2kohm pickups, with a 5 meter cable between the electronics and the tanks, for a capacitive load of about 700pF.

 

[Tubetec]

Thanks for that Abbey,
helps explain what Ive been seeing ,

It seems the higher the impedence of the coil the bigger and higher up the hump goes , into a given load/capacitance .
I dont have any of the really high impedence coils , 10-12kohms which measure around 800 ohms on the DMM , thats a lot of turns of very fine wire , as noted before the transducers with the plastic mini plugs are very prone to getting damaged and are best avoided .

I very much like the idea of trying the fet mic booster type circuit for tank recovery , perhaps with a few positions HPF and attenuation .The audio section of the Sennheiser MZA-14 PSU ticks all the boxes but an XLR barrel adapter would be a lot neater , supplied with externally generated 48v ,of course .

 

[abbey road d enfer]

It seems the higher the impedence of the coil the bigger and higher up the hump goes , into a given load/capacitance .

It shouldn't be so.
The higher the nominal impedance, the higher the inductance, which results in lower resonant frequency and lower Q (smaller hump).

I dont have any of the really high impedence coils , 10-12kohms which measure around 800 ohms on the DMM ,

The 10-12k model is only for short cable length, as is typical in a tube combo amp.

I very much like the idea of trying the fet mic booster type circuit for tank recovery , perhaps with a few positions HPF and attenuation

I've found that any filtering is best done on the drive side, not on the recovery side.

.The audio section of the Sennheiser MZA-14 PSU ticks all the boxes but an XLR barrel adapter would be a lot neater , supplied with externally generated 48v ,of course .

That would be a neat solution, albeit quite extensive, in teh sense it requires an actual mic preamp with phantom power to receive the signal.
I never had any noise problems even with 5-7 meters of cable, as long as I used the tanks with the floating coils and a separate ground wire.

 

[EthanSpaulding]

That would be a neat solution, albeit quite extensive, in teh sense it requires an actual mic preamp with phantom power to receive the signal.
I never had any noise problems even with 5-7 meters of cable, as long as I used the tanks with the floating coils and a separate ground wire.

With the floating coils, are you modding a stock reverb pan and swapping the RCA for TRS jack or...?

 

[abbey road d enfer]

No. I used the floating version and connected the tanks to 5 and 6 pin XLR's. 4wires+ground . The 6-pin was just for avoiding confusion.

 

[KA-Electonics.com]

Accutronics transducers:

 

[abbey road d enfer]

Accutronics transducers:

Wow! Thanks.
Did you measure inductance? I know I can use Spice to get it, but the results are somewhat inaccurate.