Dual tank reverb

[Tubetec]

Ive recovered a few vintage Accutronics type 8 tanks(short/3 springs) from amps in recycle pile recently .
Previously Id run through various scenarios on the recovery side , the usual single ended arrangement ,
which works ok but its still very prone to nearby hum and rf noise sources even with FMJ screening .
I tried a single tank balanced arrangement , much improved rejection of 50/60hz and RF garbage ,especially SCR dimmer hash ,

I have two similar tanks with 2500 ohm o/p coils ,
I connected them in series(start-finish+start-finish) , with a mini toggle switch at the mid point of the two coils
to connect to ground/0V .
Im using the SSL2 hi-z balanced input on TRS , into 1Mohm ,
Both tanks are clamshelled to form excellent sheilding ,
the results are very very good , ,with electrically induced noise well below the noise level of the interface + transducer itself ,
The HF energy of the springs into 1Mohm with 2x50pf cable capacitance
extends well beyond 20khz where into 12 kohms line input it sound dull and lifeless in comparrison .
The other thing I noticed is twin core screened mic cable , at hi-z , is no good , as any movement of the cable results in a clicky static
noise . Two discrete lenghts of low noise instrument cable works well ,but its not so convienient to terminate into an XLR or Jack plug .
Traditionally reverb tanks are wired with cloth wrapped ,single core ,braided screen cable , so i'll try that option next ,but in a balanced/centre tapped config from the two transducers .

More to follow soon ,

[Tubetec]

I was able to get the Gibson cloth wrapped braided screen from an old buddy in the guitar trade ,
I had a pair of the vintage RCA connectors to hand which soldered to the cable remarkably effeciently ,also giving superb screening , plastic unscreened RCA plugs should be avoided completely at instrument levels.
At the far end a Neutrik TRS and as you can see Ive added about an inch of copper foil with heatshrink on the cable screen but making good contact with the plug shell where the strain relief boot screws in . There is a really large void at the back of any Neutrik jack or xlr where its only plastic between your signal conductor and the outside world , at mic or line terminating impedences this wont matter as much ,but at Hi-Z screening has to be meticulas or you can get noise ingress.
From a handeling noise point of view night and day difference between the cloth cable and any other plastic based screened Ive tried ,
Good quality low noise/ low capacity instrument cable is ok but still tends to show some sonic signature when the cable is wiggled or tapped , with the cloth wrapped I couldnt provoke any interference in the recovery stage whatsoever ,even with gain maxed out on the interface.

Input transducers are wired in series , the toggle switch ties the opposite ends of the two coils to to the shell , when feeding a balanced input it needs ground or the output of the springs tends to cancel ,but arranging the drive coils out of phase is another possible option .

Depending on load resistance and cable capacitance a HF bump forms across the reverb output transducers , were as well not to have that hump beyond 20khz , so a certain value of cable capacitance is desirable ,

My ear tells me Induced noise levels are extremely low , proper screening got me a 30 or more db reduction , balanced another large margin , then add the humbucking properties of two coils ,garbage isnt audible above the noise floor of the pre itself .
Next were going to have to don the tin foil helmet and power up the SCR Hash module
Its as well just to rely on REWs FFT screen at a high samplerate to confirm the results here rather that exposing your ears to square waves .

The precise grounding arrangements matter , tanks tend to come with either input or output coil tied to 0v/Chassis , we dont want to create any possibility of induction/Ground loops , we can get the benefit of galvanic isolation from the driving amplifier if we choose the floating drive coil arrangement .

[Tubetec]

I thought Id outline some of the reasons spring reverbs tend to have a bad reputation ,

1. Opperating tanks in close proximity to mains transformers(or SMPS) with unbalanced recovery and poor sheilding is a waste of time .

2. Amplifiers based on op amps alone dont have the abillity to drive reverb tanks cleanly ,neither do most standard tube drive amps which typically deliver a small fraction of a watt with monsterous levels of distortion .

3. Reproduction of the reverb signal via a guitar amp/speaker with the tank housed in the amp cabinet is a worst case scenario.

Now ways to improve matters ,

1. Situate the tanks remotely , using an attenuated speaker output from the guitar amp provides more than adequate drive levels and no problem feeding a long cable .It also means its the output sound of the amp itself thats effected , not a signal tapped off in the pre-amp stage .

2.A balanced/humbucking recovery stage that combines the outputs of two tanks vastly reduces susceptabillity to induced noise while creating a more dense ,smooth pattern of reflections.

3.Lower noise levels in the recovery stage means the spring doesnt need to be driven nearly as hard on the input side to overcome backround noise , so much less of that clangy metalic twang we associate with tanks when the input transducer starts to saturate.

4. Reproducing the reverb signal over a set of monitors completely separate from the direct sound of the amp adds an extra spatial dimension , also making it very easy for the engineer to add other time based efx in parallel ,like delays and modulations .

[Tubetec]

Heres a link to a teardown of the Behringer 636 ((Grampian style) reverb with some examples of how it sounds .

https://youtu.be/mAOIbjemDNQ

This guys tank arrived with one of the suspension springs detatched , the floating sub assembly appears to be slightly bound up even with everything as it should be .

Uli went with one of the small Belton two spring tanks , its looks like an easy upgrade to put in a type 8 three spring unit , but an alternative mounting arrangement would need to be found .

I read a few reviews and repair threads about the Great British Spring , I like the idea of using the standard 4 inch soil pipe but dont like the way the tanks are rigidly mounted inside there , The type 8 tank is around 3 inches wide and a little over an inch high , so there should be space to mount a pair in the pipe while allowing them remain acoustically isolated .

[emrr]

Building a studio spring was on my list forever, not sure I’ll ever get to it….

[mediatechnology]

All you have to do is cut your springs equally-tempered.
The math works whether you use 6 or 12 lines.
I have the spring turns count numbers for the Bell/Accutronics.

Uli's spring has a lot of flutter echo.

viewtopic.php?p=12058#p12058

[mediatechnology]

All you have to do is cut your springs equally-tempered.
The math works whether you use 6 or 12 lines.
I have the spring turns count numbers for the Bell/Accutronics.

Uli's spring has a lot of flutter echo.

viewtopic.php?p=12058#p12058

[Tubetec]

I cant get caught up in the weeds with cutting springs Wayne , its too niche ,

Lets give a bow to the GBS with the 4 inch PVC shit pipes readily availabillity,in a variety of colours, from your local plumbing/hardware ,
but add a four pole HPF to external sources of vibration ,
As I said, the FFT shows the dominant source of noise after screened/balanced/humbucking coil into Hi-z (symetrik) is physical vibrations ,

I found a way to suspend a pair of tanks inside a standard 4 inch pipe and associated fittings , fourth order LF roll off , without a single electronic component added !
you can have your cake and eat it
the juicy details will follow in due course .....

[Tubetec]

A double width layer of self adhesive draft excluder foam is layed on the inside edge of the brackets , then the pipe assembly is lightly clamped into place on a baseboard or plinth, with isolating rubber shockmounts/feet of the appropriate load bearing capacity between each stage ,

We end up with several stages of decoupling of the reverb springs from sources of external physical vibrations ....

[mediatechnology]

I found a way to suspend a pair of tanks inside a standard 4 inch pipe and associated fittings , fourth order LF roll off , without a single electronic component added !
you can have your cake and eat it
the juicy details will follow in due course .....

That's what we did at MicMIx 50 years ago.

You can get pretty close by measuring their length rather than counting turns.
Works out to be 54.5 us/turn on the Bell/Accutronics.

Reverb_6_Lines.jpg (8.45 KiB) Viewed 2894 times

Place the odd lines on the left, the even on the right to interleave them and provide a nice stereo image.