How to change the decay time of a spring reverb tank?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
The BTDR 3 also has a strange modulated "pitch shift" when decaying if you listen closely. The chip is based on 3x PT2399 delays.

"[..]the associated patent, US8204240, shows that it uses three PT2399 delay chips, one of which is modulated at a slow rate[..]" Merlin Blencowe

Overall, though, the BTDR-3 is a fine piece for pedals and does a good job there, especially for guitars.
it's cool but it's nothing crazy.. i had done Impulse reponses of its H version: so you can have it in the computer!
 
The BTDR 3 also has a strange modulated "pitch shift" when decaying if you listen closely.
That's what made the success of Lexicon. It replaces the increasingly deteriorating tails with warbles that act like some dithering.
The chip is based on 3x PT2399 delays.

"[..]the associated patent, US8204240, shows that it uses three PT2399 delay chips, one of which is modulated at a slow rate[..]" Merlin Blencowe
Thank you for the link to the patent. I don't see how it makes it distinct from "prior art". The summing and feedback loops are implemented slightly differently, but only a novice would consider it as "new".
 
Last edited:
But you can't bang on a DSP and get that crashing thunder sound!
When I was managing an engineering group responsible for designing powered mixers and more, the modern DSP efx solution was a huge design time saver. Old school spring reverbs required a lot of tweaking to keep hum down and more.

JR
 
Cirrus Logic
WO2001082287A3 Richard Maslowski

Spring reverb simulation
Abstract
A process and implementing system are provided which simulate the response of a mechanical spring reverb device. The methodology includes a first part which simulates the characteristic 'chirp' sound that occurs immediately after the input signal is applied and typically decays to zero after several hundred milliseconds. The second part simulates the smooth reverberation that occurs immediately after the input signal is applied and typically lasts for several seconds. Both portions are combined to provide an output signal having a quickly decaying spring sound with a slowly decaying smooth reverb tail.
-------------------------------
When Richard started working on this I told him: dispersion. After a week of analysis he said: "Yeah, dispersion. In 1977 when I was working for 360 Systems Jim Kajiya,who was getting his Phd in signal processing at the University of Utah, suggested we use dispersion as a guitar effect. He synthesized some samples for us on the DEC System 10. His demo wasn't optimized so the effect was pretty disappointing. There is nothing new under the sun!
 
Cirrus Logic
WO2001082287A3 Richard Maslowski
Seems to go nowhere...
The methodology includes a first part which simulates the characteristic 'chirp' sound that occurs immediately after the input signal is applied and typically decays to zero after several hundred milliseconds.
This is actually the part most designers have tried to cancel.
Of course, an accurate simulation of a spring reverb should include them.
When Richard started working on this I told him: dispersion. After a week of analysis he said: "Yeah, dispersion.
I would think the effects of dispersion would be collateral; part of the sonic signature but not really wanted.
Of course for the sake of accuracy, they should be taken into account.

Is there a benefit to simulate spring reverb, including its most blatant defects, chirps, restricted frequency range?
For the sake of accuracy, yes; for operability, no.
Spring reverb is just another acquired taste. Its undisputed advantage is its density (as for plates and real chambers), which no electronic device can achieve. That is the reason why some musicians still prefer their spring reverb to a digital reverb. Bot have their place IMO.
 
Trying to replicate a spring reverb in dsp is very human.. People think that technology is what will save them, allow them to lower the cost, 'achieve the dream' without the complications.. The magic lies in imperfections with sound, the multiplication of these imperfections is what people call 'a vibe' and it's what gives personality to a sound/song etc..
 
Maybe Lenz law could be applied somehow. The spring movement is torsional but perhaps a closed loop could be hanging off the edge of the spring. Then put a tiny fixed iron rod through it with a coil. Energizing the coil induces current in the closed loop resulting in force on the fixed rod that should act to damp the spring. More current, more force, more damping.
 
Damping is based on resistive (as not reactive) reaction forces, i.e. when the force tends to counteract the movement by dissipating energy.
The force due to acceleration is purely reactive since it is stored as kinetic energy.
A constant force does not change damping, it just changes the equilibrium position.
In a spring tank, most of the damping comes from the mechanical suspension of the transducers, and any viscous material added.
It's analogous to how adding DC to a resonant RLC circuit does not change its resonance, except if it changes the nominal value of on or more of the elements.
 
A constant force does not change damping, it just changes the equilibrium position.
If this is in reference to the Lenz Law comment, the force applied by the magnet and spring loop would not be constant. It depends entirely on the movement of the spring loop. So no movement of the spring loop, no damping force. More movement of the spring loop, more force but proportional to the flux of the electromagnet.

 
If this is in reference to the Lenz Law comment, the force applied by the magnet and spring loop would not be constant. It depends entirely on the movement of the spring loop. So no movement of the spring loop, no damping force. More movement of the spring loop, more force but proportional to the flux of the electromagnet.
TBH I don't figure out the arrangement you propose.

I don't see how it would pertain to a spring reverb. The force that occur in the experiment is transformed in mechanical potential energy (mass x height of displacement). No force is dissipated.
 
Last edited:
Like this:

IMG_4837.jpg

Although with this particular arrangement there are issues related to the movement and weight of the loop and so on. This is sufficient to illustrate the idea but probably wouldn't work in practice.
 
Lining the aluminium sub tray with a strip of felt is another thing Ive done to take down metalic overtones a bit , it also changes how the spring responds to physical vibration , kicks or whatever , less metal on metal and more a deep low end boom with an initial metalic attack when the magnetic slug makes contact with the core . The felt naturally damps air movement in the vicinity of the spring but also provides a soft landing for crashes bangs and wallops , the other thing I did was add gromets to the locator holes in the alloy sub tray , again it helps lessen any tendency for metal on metal type sounds under any circumstances .

In the long spring tanks , two pairs of springs wound in opposite directions join , theres what looks like a 'ball end' from a guitar string riveting both springs together . Maybe placing the usual magnetic slug and a third modified drive transducer (coil and lams) in the middle of the spring would be the easiest way .

I added a small coil to the end of an Accutronics spring , with two very powerful n-dym slugs set in a drilled timber block , it worked as either drive or pickup transducer , but the extra mass of the coil and lead out wires damped the highs too much . Coil alignment with respect to the magnets played a big part in the level I was able to get out .
For purely magnetic damping as opposed to the active electronic kind it would a lot simpler to implement small magnets on a thumbscrew that damp the movement of the spring depending on distance . The phenomenon is found with guitar pickups to some extent , if the pole pieces themselves are too close to the strings some of the sustain can get choked out , same with a spring .

The principle used in the EMT plates is mechanical damping where either via a motor remotely or manually , an acoustic damping panel is moved closer or further from the plate , obviously this method doesnt lend itself as effectively to a spring . I have used dampers in contact with the spring itself , soft dampers kill the high end , a metal contact point against the spring gives some interesting variations on the usual tone , you can tune in an early reflection by moving the contact point longitudinally across the spring , a bit like a slide guitar .
Maybe a tank with a bottle neck would be cool .

End of day theres a lot you can do physically to the spring to modify its decay/frequency , dampers made of various materials with adjustable height and contact with the springs , its definately tone control ,but theres no electronics required. Repeatabillity you wont have , but you can tune it to source and record it as a seperate channel in the DAW .
 
Actually, I've many times considered that the Dwell control, were it made electrically variable, could be a good addition to my Fender Standalone verb. I definitely think that manipulating the Dwell control during performance could result in more interesting reverb results. (And of course, you all know of Neil Young's 'Whizzer' controller for his Tweed Deluxe? It's electromechanical, the knobs are actually turned, but perhaps a tremolo circuit or a different type of pulsing modulator circuit could electrically control the Dwell in the Fender unit. Perhaps a compressor/gate/limiter type circuit that paralleled the input that would adjust the Dwell control as the player's attack changed. I think this would be useful and would also create a different spring sound, as well. I'm not saying this would create what the OP wanted but it's at least more achievable than what I've read here.

As to IR use in a DAW, the SIR3 program with the Bricasti M7 or Penquin impulses are as good as it gets.
 
Interesting, especially since the short (decay1) reverb tanks are very hard to come by.
Can you please explain this a bit more?
If the "sizzle" in a particularly long spring is too much, try dampening them by bending. It will then be less lively (like when you tap on the side of the tank and it goes crazy). But I am not condoning wrecking any springs! Other manufacturers would sometimes put foam inside the springs or plastic tubing along the ends to tame them.
 
Maybe a small squirrel hair paint brush as an adjustable damper would work well .

Tank orientation is vitally important , if its wrong the magnetic slugs will make contact with the lams and you'll get a glitchy ratchety output signal ,springs can also stretch from being bounced around in the back of a van for thousands of miles , I have shortened springs to get proper tension back and stop that infernal clanking sound .
 
I can only repeat myself, thank you very much for the numerous ideas and information! There is a lot of exciting stuff in here.👍

My experiments so far have shown that electronic control (signal processing) of the decay is effective but also changes the original sound. I see that as an useful add-on.

As a basis, I will rather favour corresponding tanks with different decay lengths, possibly also mechanically modified.

The topic of the position of the springs is also interesting and deserves its own thread.

My current test setup switches the individual tanks, which works well, but my next design will make the individual tanks variable mixable. That brings a lot of advantages.
 

Latest posts

Back
Top