Spring tank recommendation

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

electric_diaries

Well-known member
Joined
Apr 8, 2012
Messages
82
Location
Germany / New Zealand
I am looking for some recommendations on spring tank selection.

As I see it, there are 3 main factors that affect the sound of a tank:
- Physical Spring/Tank length (short or long)
- Decay time (short, medium, long)
- amount of springs (2 or 3)

After a while of searching on the internet I get descriptions like “thick, focussed and wet bite” which don’t really mean anything to me. I am looking for a spring tank that has more of the characteristics of a hall reverb than a classic spring sound. Very spacial and ambient with as little as possible of this springy boingy slapback sound.

At a guess I would imagine that I should first test a tank that has: Long tank, Long Decay with 3 springs. Would you guys agree or offer anything else?

I can take care of the other factors (eg, impedance, mounting style etc)

I know that I won’t ever get a spring to sound like hall reverb. This is for a specific project with limitations, so while I appreciate all feedback, I'm not looking for suggestions to use different types of reverb.

THANKS!
 
electric_diaries said:
I am looking for some recommendations on spring tank selection.

As I see it, there are 3 main factors that affect the sound of a tank:
There are more variables than that, including mechanicals, like mounting orientation (vertical or horizontal), transducer impedances (drivers and pick-ups), spring diameter/media, etc.
- Physical Spring/Tank length (short or long)
spring length generally affects the initial slap in the typical range of 25-35 mSec
- Decay time (short, medium, long)
this can be affected by mechanical dampers for faster decay.
- amount of springs (2 or 3)
more springs create a more "diffuse" sound, but at the end of the day it's still a spring with only a couple finite repeat intervals (thus the "boing" sound). 
After a while of searching on the internet I get descriptions like “thick, focussed and wet bite” which don’t really mean anything to me. I am looking for a spring tank that has more of the characteristics of a hall reverb than a classic spring sound. Very spacial and ambient with as little as possible of this springy boingy slapback sound.
one (very) old studio trick is to add some pure delay in series with the send to the spring. This will push out that first slap giving it the sound signature of a larger space(?).
At a guess I would imagine that I should first test a tank that has: Long tank, Long Decay with 3 springs. Would you guys agree or offer anything else?
Back in the 70s/80s I had a friend winding his own springs and doing other tricks I was sworn to secrecy about.
I can take care of the other factors (eg, impedance, mounting style etc)

I know that I won’t ever get a spring to sound like hall reverb. This is for a specific project with limitations, so while I appreciate all feedback, I'm not looking for suggestions to use different types of reverb.
OK...  but maybe temper your expectations.

Do you have an existing spring sound that you like and want to target? Perhaps reverse engineer one of them.

JR
 
The "slapback" sound of a spring reverb is caused by putting too much energy into the spring. If you limit power by cutting low frequencies, using a limiter and by carefully adjusting the level going into the spring, you can significantly improve results. But then the problem is noise because if you attenuate what goes in you have to just amplify the output that much more. Spring reverb tanks are notorious for picking up nearby EMI and for being generally noisy. Mains power nearby in particular will induce hum. The best way to minimize that is to simply put it physically far away from any electronic gizmos that are known to emit EMI (like a wall socket with something drawing a lot of current) and away from vibration. Even a tiny vibration will be picked up and raise the noise floor. That is why the old spring reverb like EMT AKG units have lots of shock absorption, insulation and shielding. Other stuff like noise cancellation of transducers and recovery amplifier performance is much less effective by comparison.

I have thought about doing a spring reverb unit for a while. The current idea is to use shielded Cat6 cable to deliver high level signal to a remove steel enclosure and then step down through a ~600R:4R transformer with two 10R tanks wired in anti-parallel and then do recovery through phantom powered circuits (like the usual Schoeps circuit that lots of mics use maybe). This sort of arranegment allows one potentially long cable can deliver power and signal and possibly two return signals which permits placing the unit in a possibly remote low EMI / low vibration location.
 
Wow JR, thank you so much for the explanation!!

JohnRoberts said:
There are more variables than that, including mechanicals, like mounting orientation (vertical or horizontal), transducer impedances (drivers and pick-ups), spring diameter/media, etc.

I haven’t gone super deep into testing all the available variables yet (which is why I’m asking here before ordering more tanks) but to me, the other variables that are available to specify when ordering like mounting orientation and impedances didn’t seem to change the sound very much when handled properly (eg driving/receiving with correct impedance and mounting in the direction specified for that model). I have my preferred impedances and mounting direction.

spring length generally affects the initial slap in the typical range of 25-35 mSec
So a longer tank will have a more delayed slap, but it will still have the same amount of slap (the sound I want to avoid)?

more springs create a more "diffuse" sound, but at the end of the day it's still a spring with only a couple finite repeat intervals (thus the "boing" sound).
Ok so more springs is probably better for my purposes?

one (very) old studio trick is to add some pure delay in series with the send to the spring. This will push out that first slap giving it the sound signature of a larger space(?).

Very interesting! Adding a short delay would probably be possible in this project, I will try it out. Does delaying that first slap do anything to reduce it’s presence though?

OK...  but maybe temper your expectations.
Do you have an existing spring sound that you like and want to target? Perhaps reverse engineer one of them. 

Expectations tempered. I have a working model which is “good enough”. Adding a filter definitely helped to reduce some of the boingy sound. I don’t really have a target, but I wanted some advice before ordering my next tank since there are a million options.

You have already been a great help!
 
squarewave said:
The "slapback" sound of a spring reverb is caused by putting too much energy into the spring. If you limit power by cutting low frequencies, using a limiter and by carefully adjusting the level going into the spring, you can significantly improve results.

Thanks for the tip! I will try it out again, but I think I was getting the opposite results. If I recall correctly, as I drove the input transformer towards saturation (easy to do), I noticed that the natural compression seemed to remove a lot of slapback caused by the dynamics in the music. The problem of course was that It very easily became too distorted. Maybe your idea of a limiter before the input transformer could help with that though.
 
squarewave said:
The "slapback" sound of a spring reverb is caused by putting too much energy into the spring. If you limit power by cutting low frequencies, using a limiter and by carefully adjusting the level going into the spring, you can significantly improve results. But then the problem is noise because if you attenuate what goes in you have to just amplify the output that much more.
I've never heard that called "slapback" but there are clearly audible transient artifacts from overdriving a spring. Perhaps related to saturating the driver magnetics, but I never had to look into this personally because I had junior engineers doing this when my topbox powered mixers had spring reverbs in then... By the turn of the century we were all digital efx in the topboxes.
Spring reverb tanks are notorious for picking up nearby EMI and for being generally noisy. Mains power nearby in particular will induce hum. The best way to minimize that is to simply put it physically far away from any electronic gizmos that are known to emit EMI (like a wall socket with something drawing a lot of current) and away from vibration. Even a tiny vibration will be picked up and raise the noise floor. That is why the old spring reverb like EMT units have lots of shock absorption, insulation and shielding. Other stuff like noise cancellation of transducers and recovery amplifier performance is much less effective by comparison.

I have thought about doing a spring reverb unit for a while. The current idea is to use shielded Cat6 cable to deliver high level signal to a remove steel enclosure and then step down through a ~600R:4R transformer with two 10R tanks wired in anti-parallel and then do recovery through phantom powered circuits (like the usual Schoeps circuit that lots of mics use maybe). This sort of arranegment allows one potentially long cable can deliver power and signal and possibly two return signals which permits placing the unit in a possibly remote low EMI / low vibration location.
I was glad to be done with them... too much engineering bench time was spent getting hum and noise out of springs inside tight powered mixers.  The digital replacements made life a lot simpler for new designs.

JR
 
JohnRoberts said:
I've never heard that called "slapback"
I have not either but electric_diaries said "boingy slapback" and I think we all know precisely what that means. I don't really know the details of the source of that effect. You're saying it's because a spring has "finite repeat intervals" and that makes sense to me. But the effect is more pronounced with more signal. The large initial transient of a guitar pluck in particular can really torture a spring reverb into making that "boingy" sound which would suggest maybe a fast leveler circuit would help minimize it.
 
squarewave said:
I have not either but electric_diaries said "boingy slapback" and I think we all know precisely what that means.
I do not recognize that as a clearly defined term.
I don't really know the details of the source of that effect. You're saying it's because a spring has "finite repeat intervals" and that makes sense to me.
The characteristic "spring" sound is due to short, finite, fixed repeat times.
But the effect is more pronounced with more signal. The large initial transient of a guitar pluck in particular can really torture a spring reverb into making that "boingy" sound which would suggest maybe a fast leveler circuit would help minimize it.
Perhaps, but the overload transient artifact is distinctly different from typical spring sound. Also different from typical signal clipping (voltage saturation). My suspicion is perhaps magnetic saturation in driver coil (this is mostly a guess). 

Perhaps a smart DSP limiter that understands the mechanism well enough to model it could work well, but most manage to work around it without DSP.  It is a little late in the game to throw much new technology at this narrow aspect.

JR
 
squarewave said:
That is why the old spring reverb like EMT units have lots of shock absorption, insulation and shielding. Other stuff like noise cancellation of transducers and recovery amplifier performance is much less effective by comparison.

I'm guessing you really mean AKG units rather than EMT.

The AKG spring reverbs sound great.  If you want a spring reverb that can be made to sound anything like a hall, that is what you need.
 
I have a orban 111b reverb that is really nice and several spring reverbs for guitars.  The Orban is a different beast that can be used as more of a general reverb  than the tube amp reverbs. The tank is not the main difference, it is circuitry around it. 

The Orban uses a 6 spring reverb tank (not sure of the brand) and has some frequency shaping options and limiting
The manual gives interesting details and you can find the schematic online. Not as complex as the AKG but quite a bit more than the Fender 6G15

From the manual:
"The principle problem is a coloration sometimes called "spring twang". This is a peculiar "boing" or "twang" sound that the spring makes when driven by sharp transients, like drum beats etc. The 111b contains a "floating threshold limiter" which recognizes such transients and automatically ducks down their level..."

 
An additional thing to mention is how the spring is driven: voltage or current steering.

For the former, the spring tank coil impedance will limit upper freq-response, which can be a good thing for guitar-applications.

For studio-reverb applications (as far as these can be addressed by a spring) current steering is usually used, to extend bandwidth.

Old example circuit:
https://www.d-r.nl/assets/reverb-spring-service.pdf
(see page 2, opamp 2 & the 100 Ohm resistor)

... and you'll have seen the mighty ESP-site already, also mentions the current drive topology:
http://sound.whsites.net/articles/reverb.htm

Bye


 
craigmorris74 said:
Could the finite number of reflection be increased by having a spring tank with different length or type of individual springs, creating a more diffuse sound?

Just the other day I was wondering why (spring) reverbs usually don't feature a feedback-control, whereas all/most delays have one. I figure there'll be a reason for it, or it might not be practical (harder to keep control vs a delay-feedback-control).
But if it can be done, perhaps this can be effective for what you're looking for.

The info from JR earlier in this thread learned that tank-length isn't necessarily the thing that defines the length of the tail, didn't knew.
 
craigmorris74 said:
Could the finite number of reflection be increased by having a spring tank with different length or type of individual springs, creating a more diffuse sound?
Using several tanks of different characteristics reduces the impact of "boing", replacing one big "boing" by several less loud and time-spaced. IIRC, the Fairchild reverb unit had three tanks in parallels.
The use of a limiter is recommended too, but not just any limiter, since it's transients that trigger the boing; in fact a de-esser is more appropriate.
 
I recently built a spring reverb unit for general studio reverb duties. I used two tanks in parallel (type 9, short decay tanks).  The decay on those still sounds reasonably long to me. The input/output circuitry I used is based on the circuits laid out in that ESP Sound article posted above.

I'm still working on fine tuning it but it sounds good. I'm fiddling with adding switchable high pass and low pass filters to the input, and adjusting the type of lamp limiter to avoid that percussive boing sound. Also +1 to finding a good isolated place to keep the tanks. I put mine in their own dedicated enclosure and they're still pretty noisy when racked up with everything else, but much better when moved to a closet on the other side of the room.
 
Matt C said:
I'm still working on fine tuning it but it sounds good. I'm fiddling with adding switchable high pass and low pass filters to the input, and adjusting the type of lamp limiter to avoid that percussive boing sound.
A lamp limiter is way too slow to avoid boing; it helps protect the transducer against overpower, though.
 
Hi

The AKG approach about "boing" sound is perfect...
Send the same signal at each end of the tank with polarity reverse... unfortunately you need 2 coil at each end which is no usual...

best
Zam
 
zamproject said:
The AKG approach about "boing" sound is perfect...
Send the same signal at each end of the tank with polarity reverse... unfortunately you need 2 coil at each end which is no usual...

The PAiA Hot Springs does an alike trick (see pic), no idea who got inspired by whom. Simply uses two separate tanks. I figure the AKG is older, but seemingly uses a more elaborate custom kind of spring tank?

https://www.paia.com/proddetail.asp?prod=6740K

Basically the idea relies on two spring-mechanisms being essential identical w.r.t. the initial boing (so the kind of cancel), but are 'later in time' different enough to still have resulting 'woosh'.

Happened to see the GuitarPlayer article (Nov 2018) by Craig Anderton where he refers to (t)his old design (Stage Center Reverb), and then he goes on to recreate it in software (Helix in this case).

Back to the PAiA schematic, could be nice to spend an additional opamp for stereo-out* applications (you loose the boing cancelling). An alternative could be to sum the signals after these opamps (spend one more opamp) for maintaining a mono out, with trimming-possibility of fine-tuning the boing-reduction.

*: The D-R unit mentioned above is a mono-in, stereo-out spring reverb, using two long springs and sounds pretty tasty.

Bye
 

Attachments

  • hotspsch.jpg
    hotspsch.jpg
    49.7 KB · Views: 15
clintrubber said:
The PAiA Hot Springs does an alike trick (see pic), no idea who got inspired by whom. Simply uses two separate tanks. I figure the AKG is older, but seemingly uses a more elaborate custom kind of spring tank?

Yes, the AKG (I'm talking BX20) have a way more complex spring tank, each spring is about 1.5m with various spring type in series.
Both end have two coils, -sender- and -receiver- and audio is feed both side phase fliped so you can perfectly null the boing as audio pass to the exact same mechanical spring, which is not the case at the schemo you show.
All this is for one side (it's a stereo unit)

Still the Paia implementation of this idea should work too.

Best
Zam
 
Back
Top