D & R Spring Reverb circuit

[StephenGiles]

http://www.d-r.nl/dnrsite/Service%20Parts/Reverb-Spring-Service.PDF
Could be interesting
Stephen

[A.S.P.]

information lost...

[Nasse]

I remember "constant current" drive opamp circuit mentioned

Edit Accutronics page has some chemos and text, and I have old 70´s National semiconductor Audio Handbook and idea is mentioned but the circuit in that book suggests it done with pre and post eq

[R.G.]

The spring driver in a spring reverb transducer is primarily inductive. If you apply a signal voltage that has a flat frequency response, the induced current in the transducer will fall off at -6db/octave because of the constantly rising inductive impedance.  The mechanical drive is proportional to the ampere-turns of magnetomotive force in the transducer, so if the current goes down, so does the mechanical drive. With constantly falling treble mechanical vibration, the treble recovered at the other end falls as well, because it's recovered as a voltage proportional to mechanical motion. Then you have to put in a massive amount of treble boost at the recovery end to get the treble back, and you have massive noise problems.

If you precompensate for this by driving the transducer with a flat *current* frequency response, then the stuff that comes out the recovery end is flat in voltage, and this makes your noise problems less.

There are two ways to do this. You can drive the send transducer with a constant current driver, or with a pre-equalized rising treble content. The CC driver ensures flat voltage response at the recovery end by directly forcing equal ampere-turns per volt of input signal into the send transducer. The voltage drive mode precompensates by equalizing treble constantly upwards at +6db/octave. This then works with the send transducer's -6db/octave to produce a net flat mechanical drive.

The two methods are actually the same. If you measure the voltage across a CC driver, you'd find that it increases at +6db per octave because the CC driver has to do that to get the current to be constant per volt.

Craig Anderton's Hot Springs reverb does this. There is a good explanation in more detail than I'm willing to type in the National Semi Audio applications book under the Floobydust section, and a dandy design for a stereo spring reverb with their low voltage power amps chips as drivers.

[puretube]

afaik, the pre-equalization works good for "ideal" inductors,
but rumour has it,
that reverb tanks have a very "hoppy" impedance-curve,
which makes a CC drive more advisable...

[R.G.]

afaik, the pre-equalization works good for "ideal" inductors, but rumour has it, that reverb tanks have a very "hoppy" impedance-curve, which makes a CC drive more advisable...

(a) yes, the world is full of ever-more-precise approximations. You never get the same answer twice until you're down to modelling individual atoms.

At the very least, the electro-mechanical nature of the unit will cause impedance variation, just like it does with speakers. Watching the voltage on a CC driven unit is one way to *find* the resonances, by the way.

(b) just like with speakers, user preferences will vary on individual sets of nodes/anti-nodes in the response.
(c) I'm not sure how you'd make a call as to whether CC or eq'd voltage is a better method. CC is for sure underdamping the resonances that certainly exist, voltage is a tighter grip, but not as precise in following the impedance curve. Variable impedance may be one way to go, but I think it leads you far off into the weeds tweaking.

[Dirk_Hendrik]

What I don't understand not being posted in this thread...

I enjoy the sudden attention for the company where I did my industrial training for half a year during polytechnics and thing the spring reverb is worth the attention. The ones who start browsing the D&R site a little further will find a lot more (older) schematics worth looking at, giving some good ideas in design of mixing consoles, very well sounding parametric EQ's, an enhancer/exciter (and beiing sewed by AKG at the time) etc etc.

[TELEFUNKON]

http://members.tripod.com/~roymal/index.html