opamp driver for reverb tank, single supply

Status
Not open for further replies.
Jump to Latest
R

royalorchestra

Member
Joined 2014
#1
Hi!

I have a tape-echo / spring reverb unit. It is a Dynacord Echocord Super 76 from 1972. The tape echo sounds great. I replaced the reverb-tank with a 8CB3C1C. The difference was huge: from bathroom to church. Now i thought if I could improve it further and replaced the recovery stage with a +40db preamp i had around ( 40kΩ input ). better, less noisy.
The Echocord uses a ..709 opamp to drive the reverb. I wonder if I can improve the sound either by modding the existing circuit and changing the opamp to something newer or by building a new driver stage?

I found a circuit and some informations here:
Spring Reverb
Spring Reverb Unit For Guitar
reverb2

I tried to change the circuit to power it from a single +30V. (attached image).
Is it this easy? just add voltage dividers on the Inputs?
Should I use the mentioned NE5532? Or do I need a rail to rail opamp?
Is the BD139/BD140 output-buffer necessary?
Can the recovery stage provide +40db?

And what is the purpose of the 1k resistor in series with the input coil of the original Super 76 circuit?

My experiences with opamps are very limited. I tried some basic things with TL072 some years ago, but was not happy about the results.

So any Suggestions welcome!
best regards
Jens

complete Echocord super76 schematic:
http://www.peel.dk/Dynacord/pdf/Echocord%20Super%2076%20(1-0106)%20(Rev.%20d)%20(1972).pdf
 

Attachments

  • p34-f4_mod.gif
    p34-f4_mod.gif
    12.1 KB · Views: 718
  • Echocord_Super_76_reverb_part.gif
    Echocord_Super_76_reverb_part.gif
    98.6 KB · Views: 553
#2
The 5532 is fine but does need the buffer to drive loads below 600 ohms. The buffer output will be sat at +15 volts DC and so the cap C3 is needed. The lower opamp also has its output sat at +15 volts and so needs to be AC coupled to the next stage.

The 1k (R629) will be for both level matching and impedance matching because the 709 like any opamp has very limited drive ability.
 
#3
Thank you!

The 5534 seems to be pin-compatible when i stuff it at the good point in the DIP16 socket. So this could also work, perhaps with R629 lowered to 470Ω, compensation removed and the 22p cap connected from pin 5 to 8 ?
Is the 22p (and R622/ R627) necessary or useful in this case?

Then I could play with the caps and resistors in the feedback to the negative input to shape frequency response.
best regards
Jens
 

Attachments

  • Echocord_Super_76_reverb_part_5534.gif
    Echocord_Super_76_reverb_part_5534.gif
    98.3 KB · Views: 378
#4
I wouldn't have thought R622 was needed (with the proviso that these have all been fitted originally, probably to ensure stability... so always check its OK with a scope). R627, a series output resistor and the hf comp cap again are for stability. A low value (say 47 to 100 ohm) series resistor is common practice with many opamps. Stability issues show when driving low value capacitive loads, a few 10's to hundreds of pf.
 
#5
Considering single supplying with 30V you will need RC-lowpass-Filter for biasing the receive amp at least.
All in all the dynacord circuitry looks pretty good to me, the were clever in driving the coil with an ac-current source.
Coupling caps in both send and return paths should be reduced for a low cut of several 100Hz - to much low frequencies only mess up the reverb effect.
 
#6
You could use a buffer opamp instead of the many components of a transistor stage after the NE5532 or NE5534 such as the LM49600 or a LT1210 or something similar.

Or even replace it altogether with a higher output current opamp.

jer 🙂
 
Last edited:
#8
Thanks for your replies!
the dynacord circuitry is not bad at all and sounds good - I just had to replace the old rusty reverb tank. Now I want to match the driver to it and try to shape the sound of it.

I started with the accutronics site, but there is not much info which circuit would be best in my case.

The power supply has power resistors in series and filter caps to step down voltages. I hope I can hook up the buffer to the end without drawing to much current and blowing the 0,4A fuse.

What is a higher output opamp that can drive 200ohms - LA6510?
I saw Roland used a 3W power amp IC in their Space-Echoes and older Jazz Chorus Amps, around 1980 they changed the reverb driver to Op-amp + Buffer, similar to the schematics above.
I could try a TDA 7052a I have around (1W BTL power amp IC) while I'm waiting for the parts. But I think adding the discrete Buffer is the best option for me, most small power amp Ics want maximum supply voltages of 18V. I'd have to built myself a small heater...? and less possibilities to tweak.

The most interesting part in this unit for me right now is the dual pot to blend between tape echo and reverb. (schematic above) It has taps in the middle of both 250k pots connected so the wet reverb bleeds to the tape input to produce a nice sounding mix in the end. I wonder if there is also feedback from the reverb output to it's own input?
- When turned to full reverb it gets noisy, it is not caused by the recovery stage and propably(?) not by the driver. When the units input is turned down the noise is gone, when turned up it sounds like the guitar's singlecoil hum is double, or like only a cable is plugged in a preamp . I will try to disconnect terminal 19 temporarily to hear what happens.

Thanks again, I learned a lot- now it is time to solder.
best regards
Jens
 
Last edited:
#9
The LA6510 is a dual poweropamp designed for BTL configuration.

To fully drive a 200ohm load with a max. 15v the opamp would need to be able to supply up to 75ma of current swing.
Most opamps are only rated to 10ma for a non-distorted signal and there are some that are in the 30ma to 50ma range such as the NE553x's (38ma shorted current limit).

A small chip amp is a Higher current output opamp as well.

The LME 49600 seems to be the easiest that could be adapted to the existing circuit.
It is a unity gain buffer capable of +/- 250ma of signal current.
It is a unity gain buffer that is just inserted at the output of the existing opamp and feedback resistor.
Its internal construction is basically the same as adding all of those external transistors and diodes.

Here is the LME49600 data sheet,

http://www.ti.com.cn/cn/lit/ds/symlink/lme49600.pdf

here is a particular example,

http://www.tech-diy.com/Preamplifiers/CompoundOpAmp/Compound_Opamp.gif

found here,

https://www.google.com/search?q=lme...162104-compound-opamp-w-lme49600.html;673;421

As for the noise you are getting it is hard to make any assumptions without seeing the complete schematic.

This is something that needs to be traced and could be caused from anything from a bad pot too an opamp or transistor.

Or just an age old circuit design that is already noisy, particularly the reverb output buffer amp.
Check that any of those transistors may have become leaky or excessively noisy.

FWIW

jer 🙂
 
#11
"Or just an age old circuit design that is already noisy,..."
This seems probable 🙂
I will see what I can get. Mouser has LME49600 but it is a bit expensive because of delivery here in Germany. (20€ or order random things for 60€ then shipping is free...).
Reichelt has NJM4556 from the New Japan Radio Company. I'll get one.

Already tried a bit more- the pot circuitry is not causing the buzz.
I will replace driver-/ recovery-stage and compare what sounds best.

Maybe the circuit is fine, but a springtank is just not a noise filter tool by itself...
thanks again and best regards
Jens
 
#13
I replaced the driver circuit with minor changes, i used a NE5534 and decreased R4L a bit. It worked great from the first test and removed the noise-problems i mentioned earlier.
I could not get any non-inverting recovery stage to work good. (Just got astonishingly clear radio receiving at one point when i was swapping components on the breadboard...) It must relate to the following circuitry of the unit.
So I tried the inverting configuration with 22k / 2M (~90x). This works great now!
I powered the opamps from different points of the RC-line to prevent motorboating.

It seems I don't understand something obvious about the difference between the output of the inverting / non-inverting opamp. I also tried to split the recovering amplification between 2 inverting opamp-stages to see if this makes a difference somehow. But the result was much more quiet than what i expected. wrong impedance-bridging?
Putting caps (tried 500pF - 100nF) parallel to the recovery coil did not make a really audible improvement to my ears ( accutronics mentions a 22nF cap to "improve high frequency response").
The unit has a switchable tonestack behind the recovery circuit anyway. It is useful to increase highs or decrease lows if necessary. When it is switched off, the wet reverb signal is routed through a 4.6nF cap now. This way it is possible to cut lows of the reverb, leave the tape-delay unchanged and combine full volume delay with reverb. (not possible in the original circuit). I use the footswitch as the input so the unit does not output any dry signal.
It is also possible to swap the two output wires of the preamp to get full reverb and adjustable dry level.
I can strongly recommend this changes (take care of mains voltage very close to the reverb tank!). If someone is interested I can also draw a schematic or take photos of the mods.
Thanks again for all the help! Great forum!
jens
 
Last edited:
Status
Not open for further replies.
Top Bottom