The Unknown Spring Reverb

Started by razabri, September 07, 2020, 02:01:53 PM

Previous topic - Next topic

razabri

Quote from: anotherjim on September 10, 2020, 10:44:40 AM
1047: The 7 is misleading (is it a Z?). Probably 104 or 0.1uF aka 100nF. Not critical, it's the power supply high-frequency noise bypass/suppressor.
Vref, bias reference voltage for the TL082 opamp - you're missing a ground trace for the 47uF & 4k7 resistor at the bottom (may be outside the image border?).

The orange film cap values 10-90-an gives look right. I would trust them. Any film capacitors can replace those if necessary.

You're right! It is a "Z", so I'll input 0.1uF, as you said. Also, you're right about the grounds - I missed them drawing the circuit, I'll add those too.



Thank you for your input - I'm not sure about the orange caps, what you meant, but I'll have a look at it.

razabri

Quote from: duck_arse on September 10, 2020, 11:36:21 AM
if this board is from an organ, it might be a swell pedal output to the LM3080 is the missing control input thingy.

Sounds possible to me! Any idea if this could be useful in any way with adding the unit to a guitar amp? Thanks!

anotherjim

Just that there's nothing very special about them being orange, they are film type caps very suitable for audio. The important thing is that they are not ceramic disc types. The capacitance values will be correct (they read in uF).
The one in series with 1ohm on the power amp output is commonly found to be .22uF AKA 220nF. The one in the input to the power amp is .022uF (AKA22nF) and is I think a deliberately low value to reduce bass since bass doesn't work so well through a spring line.

Rob Strand

#23
QuoteI'm not sure if I understand this correctly - that no power should go to the TDA amp, but instead to power the board via first pin, to the 3080 with a 10k resistor? What would be the result of this?
Not quite.  So you need the 10k from connector pin 1 to +V to enable the audio through the LM3080.   In the default open pin state the LM3080 is cut-off.   You still need to connect power to run the circuit, as you did in your tests.

QuoteDrawing a schem.. may I ask of the value for this resistor? You didn't write it's value down in your trace...
Excellent job on the schematic.

Few small points:
- The transistor is pnp, so flip the arrow in the emitter
- The 10k + 220k gain  trimpot should wire the trimpot to the input and the 10k to the output.
   I think that missing value is 10k.
- The Vref network is missing.   4k7 + 4k7 + 47uF
- Output should come from pin 7 on TL082

So as far as using the circuit goes:
- The 2k2 on audio-in is likely to produce way too much gain, specially in combination with
  the gain of the TDA1904.
- Don't really need  LM3080, could take the output after the 0.1uF cap on pin 7 of the TL082.
  Lift input side of the following 68k to reuse PCB holes.
- Tweaks to a few caps to roll-off low end to reverb.
Send:     . .- .-. - .... / - --- / --. --- .-. -
According to the water analogy of electricity, transistor leakage is caused by holes.

11-90-an

Ahh this schematic drawing is fun...  :icon_biggrin:

Anyway, i was thinking, if the lm3080 was for reverb swells, OP could add a expression pedal with one end to either VCC or VREF and the other end to the "control input thingy"...

Updated schem with many wonderful colors.... 8)

flip flop flip flop flip

Rob Strand

QuoteUpdated schem with many wonderful colors....
Looks nice.

Maybe the 47k resistor from the tank is 4k7.


I was looking at how to merge the reverb into the amp.   It looks very easy since the amp is full of inverting circuits,
that makes mixing simpler.

https://elektrotanya.com/crate_tx-30_07s259_sch.pdf/download.html

However,  the amp has a number of inputs:  guitar, mic and external.   Also an audio out.

You have to ask yourself what signal do you want reverb  to be on and do you want reverb to be on the audio out.

Based on how most amps work I'd say:
- reverb on guitar
- reverb on mic
- no reverb on external
- reverb on audio out

That combination might not be easiest to do but we need to start somewhere.

It would be helpful if you can measure the DC resistance of the coils on the reverb.
It's not clear if the reverb drive coil is a low or high impedance type.  My guess is low.

I can see a few things which might make the reverb tone poor for guitar,
- TDA1904 is wired as voltage drive, not current drive.    This can make the sound muddy.
   A common way to compensate for this is to make the input cap to the drive circuit small.
- The output of the reverb tank drives 4.7k (I think).   This can cause the highs to roll-off and sound muddy.
- The inverting recovery amp for the reverb tank isn't the best choice for noise.
   You can probably get away with it though.

I suspect the first two issues need to be tweaked as they both contribute to a muddy reverb.

Send:     . .- .-. - .... / - --- / --. --- .-. -
According to the water analogy of electricity, transistor leakage is caused by holes.

willienillie

Circling back around to this thread, so I was wrong about the LM308.  I thought I saw CN but it was 0N.  Oops.

11-90-an

Quote from: Rob Strand on September 10, 2020, 09:46:10 PM
https://elektrotanya.com/crate_tx-30_07s259_sch.pdf/download.html

However,  the amp has a number of inputs:  guitar, mic and external.   Also an audio out.

You have to ask yourself what signal do you want reverb  to be on and do you want reverb to be on the audio out.

Based on how most amps work I'd say:
- reverb on guitar
- reverb on mic
- no reverb on external
- reverb on audio out

That combination might not be easiest to do but we need to start somewhere.

It would be helpful if you can measure the DC resistance of the coils on the reverb.
It's not clear if the reverb drive coil is a low or high impedance type.  My guess is low.

I can see a few things which might make the reverb tone poor for guitar,
- TDA1904 is wired as voltage drive, not current drive.    This can make the sound muddy.
   A common way to compensate for this is to make the input cap to the drive circuit small.
- The output of the reverb tank drives 4.7k (I think).   This can cause the highs to roll-off and sound muddy.
- The inverting recovery amp for the reverb tank isn't the best choice for noise.
   You can probably get away with it though.

I suspect the first two issues need to be tweaked as they both contribute to a muddy reverb.



Why not just place the reverb after the summing amp? just makes everything simpler... maybe a switch for on/off reverb and a pot for wet/dry mix?
Maybe OP should test putting the tank first, then change the resistor values when it's audibly muddy... something bad might happen along the way..  :P
flip flop flip flop flip

Rob Strand

QuoteWhy not just place the reverb after the summing amp? just makes everything simpler... maybe a switch for on/off reverb and a pot for wet/dry mix?
Maybe OP should test putting the tank first, then change the resistor values when it's audibly muddy... something bad might happen along the way..  :P
Not sure which summing amp? 

We need the reverb board before at least one summing amp so we can combine the Dry and Wet.    The amp has more than one summing point we can use.

The switch idea will work, perhaps two options.   Could even tweak them with an internal trim pot (s).

As for something bad happening, that's possible for sure.   I guess you can always fix any unintended damage.
Send:     . .- .-. - .... / - --- / --. --- .-. -
According to the water analogy of electricity, transistor leakage is caused by holes.

11-90-an

#29

So it's actually a power amp?... (Don't really know... :icon_redface:)

So it must be this one...


Something like this?

Of course, "Reverb Tank" means all associated components... Including an wet/dry simple op-amp mixer that can be demonstrated by the other forumites... ;) ;)


If you look closely, you can see a vactrol which controls the gain of that op-amp... That's somewhat like a limiter, i suppose... :icon_biggrin:

flip flop flip flop flip

Rob Strand

QuoteOf course, "Reverb Tank" means all associated components... Including an wet/dry simple op-amp mixer that can be demonstrated by the other forumites... ;) ;)


If you look closely, you can see a vactrol which controls the gain of that op-amp... That's somewhat like a limiter, i suppose...


That way will work.   So you resistively mix the dry and reverb, a but like R20 R21 on this circuit,
https://elektrotanya.com/PREVIEWS/63463243/23432455/marshall/marshall_valvestate20_20w_8020.pdf_1.png

Yes the Taxi is supposed to have a limiter, it stops those small amps farting out.
Send:     . .- .-. - .... / - --- / --. --- .-. -
According to the water analogy of electricity, transistor leakage is caused by holes.

anotherjim

Quote from: willienillie on September 10, 2020, 09:57:04 PM
Circling back around to this thread, so I was wrong about the LM308.  I thought I saw CN but it was 0N.  Oops.
I thought it might be QN which I'm sure was a code on some NS chips, but no  - there is no Q version of the 308.

Really good reverse engineering work here though. I don't really have the stomach for it myself - only do it when I must!

razabri

Wow! Great input from everyone - much appreciated.

I have to say that I can understand most of what you're suggesting, but since I'm a newbie in understanding circuitry I'll have to refer to a drawing I've made to make those applications - just so I can know where I'm at, before I learn and understand the language. So, here's the drawing with those details:

-enabling the audio 1+8 via 10k, for test purpose
-2k2 possibly needs to be of higher value, I don't know - 10k?
-mod to bypass the 3080
-as for tweaking the caps, I guess this would come when we have something audible to work with



Thanks for getting into connecting with the amp as well - I would need to decipher the suggestions on site, and to see what's what and how to approach this. Before, I was thinking this could be done by just soldering the IN/OUT/GROUND of the effect unit on the inside of the INSERT jack, as it acts as a loop effect and it's a stereo jack with mapping of the usage given in the amp's manual:



I was thinking that no more than a output volume pot would be necessary between the two, to control the mix, but I understand that a reverb pot would be needed to control the level of the reverberation.

Also, the amp has it's own power plate circuit that's separate from the rest and I don't know if powering the effect would be better to run from there, than to just tap into the battery, with an OFF/ON switch.

11-90-an

Uhh... sorry to rain on your parade, but there isn't any reverb pot that controls reverbation...
If we would add a "reverb" pot, it would just be the wet/dry mix... :icon_biggrin:

Are you sure you don't want an expression pedal controlling reverb swells? :icon_mrgreen:

Yes, that INSERT jack would be much better... (would avoid ruining your amp..)
If I may ask.... do you have any electronic components at hand? Say, any op-amps?
flip flop flip flop flip

razabri

I don't mind the rain ;D , as long as my feet are dry and insulated by a thick rubber sole  :D
Guess I was thinking more like a dwell control or something, if that makes sense, or to have a choice of quick, slap back and long, deep one...
The use of swells does intrigue me and I'd be glad to try it out, but maybe more like a auto controlled, like auto wah or something... Rather than complicate the thing with an expression pedal.
I have some stuff from previous builds and I could always get some in a local store. Right now I have a TL072 and TL082, KA358 as well, but that's just another 072...

11-90-an

I don't think you can change the time constants on a reverb tank... not really sure though, maybe adjusting the "pre-gain" before entering the spring-tank..

In reality, the expression pedal will be about 5x easier to implement than the envelope follower... ;)

Good to see that you have some op-amps, and even a local store that's (presumably) open.. :icon_eek:


flip flop flip flop flip

anotherjim

In traditional gear, "Dwell" is really a drive level control into the spring. Easy to do, it's just a volume control before the power amp.

razabri

It's alive! Again, thanks to Rob's suggestion - I've connected the 10k to the pins 1 and 8 and got the sound of it. The sound of the springs when touched got immediately much louder than before and I also did a little test with the guitar. Here's the clean (LP Jr straight into the Reaper):

https://drive.google.com/file/d/18bKUMfEZzLxyDr-XNgeDB3iQvUNzfz4f/view?usp=sharing

And here's the 'verb:

https://drive.google.com/file/d/17JFoRakT7A4yh5ai4lo4VvgkvMnLYrd-/view?usp=sharing

I'm not sure on how to improve the sound of it, what exactly does it lack. It certainly distorts at times and that should be taken care of. I did roll back the trim pot a bit so it doesn't get nasty - all the way it's just too much distorting, but the signal is more audible.

Also, the question remains about the mixing and how to apply that passive mix - I'd probably need to build a sister board  of the mixer and I'd need a vero diagram for it, as I'm not skilled in making it out a schematic. I'll try to find a stereo jack and check if I can hook it up via INSERT jack.

11-90-an

Now that I think about it, I think the LM3080 could be used as a makeshift compressor to limit gain...
@razabiri does the distortion come from the reverb tank ckt or from the amp being too loud?

Passive mix you say...  ::)
flip flop flip flop flip

razabri

Nooo, I meant resistive mix like Rob said, but of course, I have no idea what I'm saying half of the time so there you go  :)
Quote from: Rob Strand on September 11, 2020, 01:25:33 AM
That way will work.   So you resistively mix the dry and reverb, a but like R20 R21 on this circuit,
https://elektrotanya.com/PREVIEWS/63463243/23432455/marshall/marshall_valvestate20_20w_8020.pdf_1.png

Yes the Taxi is supposed to have a limiter, it stops those small amps farting out.

The distortion is from the tank, as I didn't record it via amp at all, just stuck it to the DAW directly, so that's the Reaper I'm referring to. Guess I'll try to replace 2k2 resistor, give it a higher value, as Rob suggested as like this it has too much gain - and there I made a mistake on the drawing - I marked a different 2k2 as an input resistor, here's the corrected drawing: