Radioshack Realistic Pro Verb

[Passaloutre]

Ancient transistorized spring reverb unit here. I picked this up at an antique store. Can't find a schematic anywhere. I get a nicely boosted dry signal through it (reminds me of Echoplex preamp), and a very weak reverb signal with the knob all the way up.

I've attached some photos of the interiors. It's clearly been modified, three transistors have been changed (solder-spliced to the dead legs of the originals), and one (presumably original) four-leg metal-can transistor that I can't identify or figure out the pinout. The replacement transistors say "CBS 1232" on them, but I can't find a datasheet for that part, and I can't make out any markings on the four-leg guy.

The filter capacitors also appear unoriginal and the values seem to have been chosen at random as far as I can tell.

Anyone care to make sense of my schematic? I've double checked it with my meter, but I'm sure there remain errors. What's the deal with the four leg transistor? I'm more of an amp guy and don't have too much experience with transistors.

My draft schematic and photos of the guts are here: https://imgur.com/a/527Kr9i

It has a tiny tank inside, but I don't know how to measure the coils. I'd like to get it back to it's original shape, but if not I suppose I can build a different reverb circuit into the box, maybe with something like this: https://www.tubesandmore.com/products/reverb-tank-accutronics-amc2bf-2-spring







EDIT: looking through the JEDEC transistor package documentation, the metal can thing appears to be a TO-7 transistor, one of the legs would be a shield presumably.

[PRR]

There was another Tandy/RadioShack/Lafayette reverb on this board in the last week or two, "same" schematic. The poster reported that it had never been satisfactory.

EDIT: https://www.diystompboxes.com/smfforum/index.php?msg=1274156
Rodgre
L.R.E./Lafayette EchoVerb II
September 26, 2023, 06:49:03 AM
"... this EchoVerb has been in my collection for decades and I've yet to find a use for it in the studio, whereas the awful reverb in my Silvertone Twin-Twelve has ..."

Yeah, different store but they all went to the same Japanese Expos to buy "Your Name Here!" products. I never seen that schematic before; now I seen it twice in one week.

[Passaloutre]

Thanks PRR, will look around!

[Passaloutre]

I see the one you're talking about. Sounds like it may be a dead end then.

Maybe I'll gut it and build something with that mini Accutronics tank...

[Rob Strand]

I see the one you're talking about. Sounds like it may be a dead end then.

If you can hear some reverb it's working.
The circuit could probably do with some tweaks.

The pot resistance isn't shown.  I'm guessing some large value.

The small metal can is probably a transformer.   I haven't checked your tracing but from your schematic it looks like it's on the tank input/drive side.

If you disconnect the reverb input and output wires you could measure the DC resistance of both windings of the transformer and both coils of the reverb spring.  However, you don't need to do that to do some tweaks.

If the dry signal alone has a lot of gain through the unit then that's wasting gain which could make the reverb stronger.

Mods to push the unit into shape:

1) You can increase the 82k to reduce the clean gain then that will effectively make the reverb stronger.   Ideally you keep increasing the 82k until the dry/reverb balance is good.  Don't worry about the level dropping.

2) Once you have done (1) you need to see if the unit is less than unity gain.   The last transistor stage is pushing it's luck a bit for gain.  You can get more overall gain by reducing the 220 ohm emitter resistor on that stage.  Even shorting the resistor is OK.  Ideally it would be set so the overall gain is unity for the dry signal alone.

Here's some examples *without* factoring in the effect of the transformer.

RE = 220 (unchanged)   
Rdry was 82k now 180k
Result reverb could still be a little low

RE= short (just place a bridge on the back of the PCB)
Rdry  was 82k now 560k
Result reverb has decent level.

No doubt you will need to experiment.

Something else, you would probably be better off increasing the base cap on the last transistor from 100nF to 1uF.

*** A caveat to these values is I'm assuming the inputs are open when not in use.   Your schematic doesn't show the details.  It might not be a correct assumption.   You can get a situation where the mic input shorts to ground with nothing plugged in, which creates a divider at the input then you need more gain later to compensate.

[Passaloutre]

Thanks for the help. The inputs are open jacks.

I replaced the dry mix resistor with a 500k pot and jumpered the emitter on the last transistor like you said. It's definitely improved, but the reverb is still all in the last 10% of the reverb pot. Not sure about increasing the last coupling cap, it's very bass-heavy as is. I could probably decrease the cap going into the tank...

Unrelated to the tone, I did trip my breaker plugging to two-prong in the wrong way. Maybe time for a grounded plug...

[PRR]

The small metal can is probably a transformer.

It fits with a Germanium, ATT still readily available on the Japanese market. A 4th lead would be the metal can, shield. Not always needed in audio but often needed in radio, and "Japanese Pocket Radio" was the iPhone 13 of the day.

The Lafayette ECHO-VERB from the other thread clearly shows a PNP, nominally 2SB347, which is 3-leg; but everybody had "equivalents" and if they had the 4th leg they either soldered it or cut it.

The drawing for the The Lafayette ECHO-VERB is wonky, patchwork. I have re-pasted it, maybe more sensibly? TR1 is a gain stage, maybe 10X (TR2 loads it), also input mixer. TR2 and TR3 is a low voltage gain amplifier with significant (few mA) output current. This flows in reverb tank input coil. DC in coil is not Best Practice but certainly Good Enuff for RS/LRE. The TR3 current in the DCR of the coil gives DC feedback to TR1 base. Coil DCR must be "hundreds" or else the drop in the 1k in the supply filter implies overvoltage on the first filter cap. The lack of recovery amp after the tank also points to low expectations. If the tank fits inside the case shown at the top of this thread, then if Hammond tanks are clarinets, this is a kazoo.

[PRR]

I did trip my breaker plugging to two-prong in the wrong way.

That sounds like a very sick cord or an internal transformer short. 3-pin is not a proper fix. (It may seem that way on all correctly wired outlets, but US wire-workers don't know black from white.)

IF you can get it right once, measure the AC Volts on the secondary, and the core width, find something new to replace it.

[Rob Strand]

It fits with a Germanium, ATT still readily available on the Japanese market. A 4th lead would be the metal can, shield. Not always needed in audio but often needed in radio, and "Japanese Pocket Radio" was the iPhone 13 of the day.

The Lafayette ECHO-VERB from the other thread clearly shows a PNP, nominally 2SB347, which is 3-leg; but everybody had "equivalents" and if they had the 4th leg they either soldered it or cut it.

That makes more sense than a transformer.   I was ignoring those connections to DC for now.  After looking at that schematic the connections do line up with a PNP.

The drawing for the The Lafayette ECHO-VERB is wonky, patchwork.

The original is a very unnatural way to draw that ckt - yours looks much better!

The fine details of reverb spring might allow a better estimate for the gains.   Also the possibility of cranking up the drive current instead of the just the gain.   No transformer cluttering things up is a bonus.    The input jack switching, if any, is an unknown though.

Not to worry, I don't think that details affect the idea behind the mods.   Even if it cannot be made "perfect" there's definitely room for improvement.

[Passaloutre]

There is no switching on the input jacks

Thanks again for providing clarity

Kazoo is the right word. The tank is about 6 inches long. Sounds like yelling into a trash can.

Reducing the dry mix resistor to near nothing turns it into a decent fuzz though.

[Rob Strand]

Kazoo is the right word. The tank is about 6 inches long. Sounds like yelling into a trash can.

It might be helped by a better drive circuit plus a bit of high-pass filtering drive side.   A lot of reverb circuits are way off the mark for the drive circuit and the recovery circuit.


The drive circuit is a bit different to the Lafayette.  The drive stage is AC couple in the input and DC coupled on the output.  I'm not even sure that funky metal can transistor is biased correctly!

[duck_arse]

Reducing the dry mix resistor to near nothing turns it into a decent fuzz though.

that looks like a fuzzface with Q1 feeding a Heatkit/Bosstone as Q2.

[Passaloutre]

For the sake of posterity, I've redrawn my schematic here with the metal can as a PNP transistor. Hopefully this can help someone else in the future.

The parallel base-collector resistors intrigue me. They both look original, hard to believe Radioshack ran out of 47k resistors.

[Rob Strand]

The parallel base-collector resistors intrigue me. They both look original, hard to believe Radioshack ran out of 47k resistors.

You will probably find the original documentation has a fixed value then second adjust on test (AOT) resistor in parallel.   That allows units which don't pass some manufacturing criteria to be reworked/adjusted.   However, what sometimes happens is they find a particular AOT value works and they just leave that value in there.   You can see both resistors were put in from the start.  Whereas in other cases it's clear someone has tacked-on a parallel resistor later.

Another way it's done is to snip-off an AOT  resistor.  So in this case the resistor has been left in.

[PRR]

redrawn my schematic here with the metal can as a PNP transistor.

Thanks. But I doubt the ground is on Collector; more likely the shield/can.

The parallel base-collector resistors intrigue me.

Both in || is 16.6k. If you get a high-gain transistor and high transistor current (low B+ in power supply), snip-out the 22k and leave the 68k. Since there is NFB and the exact current is not critical, a 4:1 change of bias resistor is close enough for candy-boxes.

The mixer network IS forked-up. A 100k pot and 3.9k or 6.8k series resistor, "all the action" is in the last 10% or so. Nothing is wrong except you (some pre-you back in 1971) didn't pay for another stage or two of buffer and boost. Just crank it.

[Rob Strand]

For the record as well.

With the PNP germanium added.  The mod values recommended above change:
- leave 220 ohm emitter resistor on the output stage transistor
- change 82k to 560k for unity gain but reverb level high
     or,
  change 82k to 330k for non-unity gain but reverb level about same level as dry
- to decrease the overall gain increase the 220 ohm emitter

There's some assumptions about the sensitivity of the reverb spring here as well.  If anything, the actual output from the spring might be lower than my assumption meaning the mod might favor 560k vs 330k.   That difference is minor in the scheme of things as the existing 82k is way off and will make the reverb weak relative to the  dry signal.

I haven't given any mods to add high-pass filtering to the reverb path.  It might help this unit sound better.

Something else about this unit is the reverb signal is out of phase with the dry signal.   You can sometimes get away with it but maybe it doesn't help get the best sound.  If the tank input isn't grounded you could flip the phase there - maybe it's already done?  Worth a try anyway.

I noticed on the Lafayette Echo-Verb II the dry signal and reverb are mixed at the output socket which removes the phase issue.   The previously given Lafayette Echo-Verb schematic is for Lafayette Echo-Verb "I".   The Lafayette Echo-Verb II has a similar circuit except resistor mixing of many inputs at the input side and mixing of the reverb and dry signal at the output terminal - no online schematic.

I didn't look at level adjustment mods for the Lafayette Echo-Verb in the previous thread,
https://www.diystompboxes.com/smfforum/index.php?topic=131193.0

[PRR]

the reverb signal is out of phase with the dry signal. 

Even cheap tanks have path-delay longer than an audio cycle; actually a complicated delay varying with both time and frequency. Just like a Real Room, it will null some frequencies and reinforce others. I have never found a real difference in reverb polarity.

[Rob Strand]

the reverb signal is out of phase with the dry signal.

Even cheap tanks have path-delay longer than an audio cycle; actually a complicated delay varying with both time and frequency. Just like a Real Room, it will null some frequencies and reinforce others. I have never found a real difference in reverb polarity.

It's something to try.

The ease with which the stages overload seems to be a problem on this unit, give it a bad sound.   Increasing the 82k is definitely worth trying.