mecahnical delay line help...

[rx5]

hello,

any links or info about mechanical delay line??

I dunno if this can be done using common hardware....

[Peter Snowberg]

I know of three mechanical delay (leaving out echo chambers and special rooms). The first is done with piezoelectric crystals and a giant poice of quartz.... not too practical or useful for audio. The second is the good old "plate" reverb which uses a suspended steel plate with transducers at each end, and the third (and most common) is the timeless spring reberg tank.

Did you have something different in mind?

You can make spring reverbs, but the response is less than acceptable to me and it's a great deal of work. Spring tanks are pretty cheap for what you get out of them.

http://www.tubesandmore.com/
reverb tanks

-Peter

[rx5]

thats exactly it...the srping reverb...I had a drawing of it ..copied from an amp.... but I lost it..... wish I could have a diagram of it..... cant seem to find it on the web.....

[puretube]

maybe he`s thinking bout the good old Hanert`s scanner-vibrato`s delay-line as used in Hammond B3s....







//www.puretube.com

[rx5]

guys,

 any clear diagrams of the spring reverb you could possibly share with...


thanks.... :wink:




-Ralph

[Peter Snowberg]

Ahhhh... great! I love spring reverb.

Try the "Stage Center Reverb" at General Guitar Gadgets:

http://www.generalguitargadgets.com/v2/index.php?option=displaypage&Itemid=142&op=page&SubMenu=

You will need an Accutronics model 8DB2C1D Reverb Tank, or another tank with similar coils because the design uses an opamp to drive the springs. If you add an LM386 to the send jack, you can use a standard 8 or 10 ohm to 2.5K tank which is easier to get.

I like long tanks with three springs, but they can be harder to find if you need a higher impedance send coil as is used in the Stage Center unit (without a 386).

http://www.tubesandmore.com/new/scripts/silverware.exe/moreinfo@d:/dfs/elevclients/cemirror/ELEVATOR.FXP?item=P-R8DB2C1D

Best of luck with your build.

Take care,
-Peter

[rx5]

Peter,

ermm, I dont wanna buy it, I just wanna build it...the reverb tank  :wink:

wish I could still remember the insides of those tanks....seem to forgot about it... any picture you have there on the internals??


has somebody ever made those reverb tanks the "DIY" way??? :wink:  :wink:



Regards
-Ralph

[Peter Snowberg]

Peter,

ermm, I dont wanna buy it, I just wanna build it...the reverb tank  :wink:

It's a lot of work and the results will probably be very poor, but with that in mind....

You will need two coils, one about 8 ohms or and one about 2K or more, a spring with very small mass, and some rubber bands or thread to mount the spring.

The hardest part is probably getting the magnetic coupling between the coils and the spring. The pre-built units use a small transformer coil with laminations that have a gap. The end of the spring sits in the gap. There is also a small piece of magnetic material mounted on the end of the spring that sits in the middle of the gap to help with the transfer.

If you're making your own, I would recommend using small transformers (like an 8 ohm to 1K) and cutting the laminations to fit the end of the spring.

You need to mount the spring in a way that allows it to move freely while still holding it under a little tension. The commercial units use a small piece of spring bronze wire for this, but you could use thread. You just need to minimize the amount of contact between the spring and the mounting.

I wish you the best of luck, but like a watch, sometimes there is no substitute for something that took hundreds of hours to engineer and uses specially made parts.

-Peter

[Dan N]

This guy has some far out reverb builds!

http://www.angelfire.com/electronic/epeasant/projects/springs/springs.html

The accutronics home page may have photos or diagrams of more mundane reverb spring assemblies, as well as schematics.

[zachary vex]

dealer cost on a raw tank ready to connect to drive electronics is less than $20 for any model of accutronics type that new sensor sells.  they are basically very cheap.

[D Wagner]

Peter,

ermm, I dont wanna buy it, I just wanna build it...the reverb tank  :wink:

wish I could still remember the insides of those tanks....seem to forgot about it... any picture you have there on the internals??


has somebody ever made those reverb tanks the "DIY" way??? :wink:  :wink:



Regards
-Ralph

Are you talking about the "Slinky" spring reverb that uses speakers with dowels attached to move it?  I recall seeing that, but I doubt that I saved it to disk.  It was kind of far out there.  It was supposed to work pretty well (from the description), considering that the slinky is about 20' long when you stretch it out.

I'll  look today for the build article...you never know what I may have stored away.

Derek

[rx5]

Dan N,

wow that was cool.... another variation of the spring reverb....


Peter,

gonna find those springs...hehe... Dunno if I could find them... :roll:


could these types of reverb be emulated by electronic BBD chips??




-Ralph

[Peter Snowberg]

BBDs work great for echo, but not reverb. Picture a reverb as a HUGE number of echos in prallel, all with slightly different filtering and delay. There was a chip that was made to emulate a spring reverb, but it was only available for a few months and has since gone out of production.

I have looked for the right springs, but I have never found them outside of a real reverb tank.

That Slinky idea looks really neat, but it I'll bet it is not much better than a novelty. It should make a unique sound. If you build one, please post some pictures. I would love to see it!

-Peter

[brett]

I've been thinking about this for a while, and even built a unit out of 2 speakers joined with springs....but it didn't work very well.  MY latest thoughts revolve around using a box of material with send and receive tranducers on each end.  (King of a small physical model of a room, which is what you're trying to simulate).  I believe that reverb "tanks" are called that because some early reverb units were like fishtaks with a speaker at one end and a microphone at the other end.  That would probably sound cool, but is somewhat impractical.

Anyway, I've been thinking about appropriate materials for a "mini-tank".  Most materials are too dense, and transmit sound too fast, like really short reverb springs would (e.g. water transmits sound something like 4 times as fast as air, and steel is even faster).  An ideal medium is a gas with a high molecular weight (velocity of sound is proportional to the square root of the inverse of MW).  So carbon dioxide (MW=44) is slower than air (MW=29), but only by about 2/3.  If I find a safe gas with a molecular weight up in the hundreds, I'll be trying it in a tank reverb unit.

[rx5]

Peter,

I would love to post pics but Im having problems where to get those "slinky" springs.... all I could find is plastic....

I think the best way is to make multiple  electronic echo circuits in parralel...but would be too costly....I think... :wink:


I think Im gonna pass that one out.....

anyone got other site/s for other reverb tanks


Regards

-RAlph

[Arn C.]

What about the springs that are on screen doors?

Also, I think it was Ansil that has the one built out of a fish tank!  

Arn C.

[Arn C.]

http://www.geocities.com/cemaynard1134/jojo/waterverb.gif

Enjoy!
Arn C.

[Ansil]

What about the springs that are on screen doors?

Also, I think it was Ansil that has the one built out of a fish tank!  

Arn C.

yeppers.. it is currently dismantled right now.  no room for it in storage

[Mark Hammer]

I have proto'd parts of a home-brew "micro-spring" reverb using parts I picked up surplus and at the hardware store.  I need to get some other things out of the way first, but I promise I *will* finish this one and post it.  The sound is not Lexicon or EMT quality, but it is more than passable, and dirt cheap to make.  Best of all, you can fit it in an 8" x 4" x 2" box with ease.

I bought my springs at Home Depot (they're a big chain, but I don't imagine they make it out to the Philipines; still, there has to be some pinoy equivalent).  They carry many springs but only one type is compliant (unstiff) enough for this.  It is about 3/8" wide, and about 3" long with a loop at each end.  It will easily stretch out to 8" but you get a nice degree of sustained boing at around 5-6".  If you use some epoxy or cyanoacrylate or other glue that dries very hard, you can glue one of the end loops to a small speaker, like the ones used for beeping inside desktop computers or in very small transistor radios or walkie talkies.  That will serve as your driver transducer.

The other end can go to several things.  I tried glueing the other end to another speaker but that provides too low an impedance to the input of the recovery amp, so I stuck a 1.5k:8R transformer (the little Radio Shack one) between the speaker and the op-amp recovery stage (the 8R end is connected to the speaker.  This provides decent impedance matching and rasies the reverb signal up to around almost 20mv or so.

Better than this is soldering the receiving end of the spring to  a ceramic/piezo disc.  This has a few benefits.  First, it is much more rigid  because of the use of solder rather than glue, and the presence of a hard surface.  This conserves more of the reverberant energy and hangs onto more high end.  It is also easier to secure.  It is also much easier to make a high input impedance buffer out of FETs or whatever, to cope with what the disc delivers than it is to get a suitable impedance matching transformer and increase the load the recovery stage sees.  With a speaker and transformer, I get about 20mv.  Wih the spring soldered to a piezo mic cartridge/disc, I get about 60mv no problem, without even any additional circuitry.  This is good because it means you don't need as much recovery-stage gain.  Low gain requirements means less noise overall.

A clever idea that Peter Snow suggested to me, is the use of a Y-shaped configuration.  The speaker drives one spring, which is soldered at the end loop to two other springs, each going to their own transducer.  This has several advantages.  First, it makes the effective spring length longer which adds to delay time without adding all that much to the package size.  Even though it helps to make the effective spring length longer, it does so with less sag because the entire assembly is suspended at 3 points rather than 2.  Second, the two output paths are not going to be identical (and you can make them even less identical by spacing out the receiving tranducers in an asymmetrical way), increasing the diversity of arrival times which provides less nasty boing.  Third, as the PAiA/Anderton Hot Springs reverb makes good use of, you place the two transducers in or out of phase with each other, and use the cancellation to get rid of large amplitude mechanical noise or simply to just produce a different sound.  Heck, if you make the two receiving transducers different in what they pick up, you can even stick in level controls for each and mix them for different sounds.

I'll be driving the speaker with a 386 and some frequency-tailoring.  If you use a cap noticeably smaller than the 10uf feedback loop cap shown in the datasheets for the 386, you can provide high-end boost to compensate for the high end that gets lost in the spring itself.  For starters, I'll use a FET or MOSFet clean booster with modest gain to recover the signal from the mic cartridges.

You can probably use the speaker as is, but what I did was stiffen the cone by painting on some diluted carpenter's glue and letting it dry.  Then I took a small X-acto knife and cut away some of the cone area so that he voice coil was appropriately supported by the cone itself wouldn't move quite as much air.  That makes the unit itself quiet enough that you can't head the speaker once the lid is on.

I have no idea about the specific current requirements, but my guess is that trying to run it off a 9v battery will be an exercise in frustration.  I'll try it, though and see if a standard 9v is adequate.

[Peter Snow]

Another idea to consider for the recovery module is a turntable cartridge.  You would have to match the recovery preamp to the output of the pickup, which is normally quite low for a moving coil unit (actually, moving magnet, I think ).  Then you could mount it up side down with the stylus just touching the bottom of the spring.  A dab of glue would hold it in contact with the spring.  If that cartridge happens to be stereo you could just duplicate the recovery circuit for a stereo effect.  Not sure how that would sound given that the signal would come down the same spring rather than two separate springs...?

This idea comes from a reverb circuit I built when I was 16 which used two piezo (crystal) pickups as the drive and recovery modules. These things were commonly used in cheap "record players" back in the late 50's early 60's. The whole thing was tube driven and had a neat little chopper circuit (tremolo) on the reverb signal that gave it a pseudo echo effect.  Any of you older guys in the UK may remember other related circuits in Practical Wireless, developed by a gentleman called F.C.Judd.  

The whole thing never really worked well because I don't think the tubes were driving the first pickup hard enough.  If I had been more proficient I would have figured out how to fix that but instead I substituted a moving coil microphone for the recovery module which was marginally better, but then it became very sensitive to external noise.

Some of these ideas may be easier to implement today given the "better" technology availble now.  

Cheers,

Peter