Spring reverb problem

[HoodTube]

Hi folks. This one isn't about pedals as such, but wondering if anyone can give me any ideas.

I've got an old Polytone Mini Brute guitar amp (Mini brute IV, with reverb, distortion etc), and there's a problem with the spring reverb.

Basically, to get any useable reverb sound from it I have to crank the reverb dial up to maximum, but with that comes a load of horrendous hiss and hum. The amp is very quiet otherwise, when the reverb is all the way down.

I've just tried a new tank, so the problem must be somewhere else. The reverb dial is a dual pot with the distortion gain dial, for some reason. Could that be something to do with it? Quite honestly, turning the reverb dial all the way up sounds like I've switched on a drive pedal at high gain, there's so much background noise. Or is it more likely a problem with the reverb driver circuit? If so, any ideas how I could diagnose and fix it?

Cheers!

[duck_arse]

you will need to find and show a circuit diagram before the smart guys will help. also some photos of the stuff, for dopes like me to look at. please.

[Rob Strand]

A schematic would help.

The reverb driver on the Polytone's use an opamp driving two output transistors.  It's possible one of the output transistors has failed.    Probably the best way to check is to simply test the transistors.    (Large emitter resistors and the feedback loop around the opamp can make the circuit kind of work but it doesn't work well.  It limps along.  That sounds like your case.)

I'm assuming of course the replacement reverb tank is the correct one for the circuit.

I found this schematic,
https://el34world.com/charts/Schematics/files/Polytone/Polytone_preamp_mega_brute_mini_brutes_i_iv_1_.pdf

It has the transistors on the reverb driver (Q1 and Q2).

[Rob Strand]

Actually, keeping the possibilities open there could be a problem with OP5.
Maybe there's an open circuit or no light getting on the sensor.
As a test you could bridge-out OP5.

[HoodTube]

A schematic would help.

The reverb driver on the Polytone's use an opamp driving two output transistors.  It's possible one of the output transistors has failed.    Probably the best way to check is to simply test the transistors.    (Large emitter resistors and the feedback loop around the opamp can make the circuit kind of work but it doesn't work well.  It limps along.  That sounds like your case.)

I'm assuming of course the replacement reverb tank is the correct one for the circuit.

I found this schematic,
https://el34world.com/charts/Schematics/files/Polytone/Polytone_preamp_mega_brute_mini_brutes_i_iv_1_.pdf

It has the transistors on the reverb driver (Q1 and Q2).

Hello folks, thanks for the replies!

This one should be helpful, thank you. I had a look for schematics but couldn't really find anything that seemed to match my particular amp, so thanks for sharing this one.

Yesterday I had a poke around and seemed to find a quick and dirty solution to the problem for the moment anyhow, might be a "band aid" type of fix, but here's what it was. Basically I soldered a wire from the input sleeve to one of the suspension springs inside the tank (I know, I know). Soldering it directly to the tank body itself made it distort very easily.

But anyway, this has somehow resulted in glorious, massive amounts of reverb. Barely have to turn it up past one. Still get the hiss at 10 or nearabouts, but at that point it's just crazy metallic noise anyway, which is a cool side effect in itself.

No idea why this works as it has. If the thing is likely about to blow up because of this, please let me know.

Cheers again.

[j_flanders]

The drive section seems to be the same as described in figure 5 on  Elliott Sound Products' Spring reverb page:
https://sound-au.com/articles/reverb.htm




It clearly states that the input coil should be isolated. (sleeve NOT directly connected to the chassis/ground).
I guess there's some resistance from the suspension springs to the chassis (ground).
By soldering the sleeve of the input to those springs it is as if you added a small resistor parallel to R27 (R2 in Elliot's schematic), lowering its value, increasing the drive.

[Rob Strand]

But anyway, this has somehow resulted in glorious, massive amounts of reverb. Barely have to turn it up past one. Still get the hiss at 10 or nearabouts, but at that point it's just crazy metallic noise anyway, which is a cool side effect in itself.

What that is doing is increasing the drive signal to the reverb by bypassing the 27 ohm resistor R27.
(keep in mind if the  schematic doesn't match your amp R27 might be marked R27 on your PCB,
but it's likely to have a value around 10 to 47 ohm.)

That could be compensating many issues:
- R27 is faulty or open.
- The R27 value doesn't suit your reverb.   If the original reverb did exactly the same as the old one
  then you can only assume the marked valued on R27 is correct.
- The increased drive is compensating for OP5 being higher than normal resistance.
   The easiest way to determine a fault with OP5 it to bridge it out.   If there is a large
   increase in level it means OP5's resistance is probably too high.   Then work out why.
   (I mentioned a few things above.)

[HoodTube]

Ah I see, I could try these suggestions out sometime soon, but in the meantime, is what I did likely to damage anything? I actually really like the sound I'm getting at the moment with it.

[Rob Strand]

Ah I see, I could try these suggestions out sometime soon, but in the meantime, is what I did likely to damage anything? I actually really like the sound I'm getting at the moment with it.

You might get away with it as is but with +/-15V supply rails there's a risk of stressing out the reverb drive coil.    If you temporarily add about 10 to 22 ohm in series with the drive coil (or C20) it would lower that risk.  It will lower the amount of drive a bit too but not an enormous amount. (I'm assuming the reverb tank you bought is an 8 ohm input type.)

However if you are going to do that you might as well do it the correct way.  Remove your shorting wire to ground and lower R27 (27 ohm)  - as j_landers already suggested.

[HoodTube]

Ah, I see. Well I'd rather avoid any problems and do a decent long term fix.

I should've probably mentioned at some point, I'm still a bit of a novice and only attempt these things a couple of times a year, or when something goes wrong that seems fixable with my limited abilities. But I think I'm (somewhat) understanding you. I hope.

In the most basic terms possible, am I supposed to remove the resistor from R27? Do you know what value to replace it with so that I'd still have a high reverb output level after removing the wire I put inside the tank?

Sorry for the possibly idiotic question. And cheers for getting me this far regardless.

[Rob Strand]

In the most basic terms possible, am I supposed to remove the resistor from R27? Do you know what value to replace it with so that I'd still have a high reverb output level after removing the wire I put inside the tank?

The idea is to reduce R27.  So you can you that by replacing it, or, since this is a temporary fix you just solder a lower resistor value resistor across R27 on the back of the board.   Something like 10 ohm would be a good place to start.

The main thing to realize is this fix might not be fixing the true cause of the problem it might only be compensating for it.   The fact it works from outside of the box gives you the wrong impression the fault is fixed.

I'm not sure just how low the level was without you mod.  If it was *WAY* off then there's a good chance there is a real fault.   However if it was mildly off in that setting 10 still didn't suit you taste then perhaps there's no fault and you just need a "mod" to increase the level to suite you taste.   Dropping R27 would be an acceptable mod to a point.

Some possible faults it could be:
1) R27 is open or damaged.  In which case replacing R27 *is* fixing the cause.
     You can verify that by measuring the exiting R27 with a multimeter.
2)  Q1 or Q2 blown and/or one of the 10 ohms resistors (R30/R31) blown.   
     You can check this by measuring the voltages to ground
     on each side of the 10 ohm resistors and/or by testing the transistors directly
     That's four voltages.   You could also measure the voltages on the bases of Q1 and Q2.
3) C20 faulty or open.   Still possible but it seems OK based on you current mod.
4) Something wrong with OP5, as mentioned previously.
5) You new reverb tank, and perhaps the one in there originally,  is not the correct type.
      For example if you put in an 8 ohm type and the original was higher then you would
      need to mod the circuit by reducing R27 to suit the 8 ohm.
      (I have not checked what these amps normally use the thing that makes me suspicious
      is the relatively small value of C20 for an 8 ohm reverb tank.)

As you can see you can hide plenty of faults by making a radical mod like you have at the moment.

[Rob Strand]

Can I ask what reverb spring you have.

When I looked here,
https://amprepairparts.com/reverb.htm

It states,

Revisit™ Reverb Tank, 9" long, 3 spring, equivalent to Accutronics 8DB2C1B.  Short Tank, Medium Decay, input impedance 310 ohms, output 2575 ohms, insulated input and grounded output.   Mounts horizontal, open side down.   Primarily used by Marshall® VS100, Marshall® VS265, a few versions of the Marshall JCM900 series, Polytone® Minibrute I through Minibrute IV, Polytone® Megabrute, Fender® FM212R, FM210R, FM65R, Frontman 212R, Frontman 210R, Frontman 65R.

So that's a 310 ohm input impedance spring.

That seems to explain the low 10uF value of the C20 cap.

If yours is an 8 ohm spring you will need to mod the circuit for sure.  (In fact I'm not sure the output transistors will like driving an 8 ohm reverb spring.)

[Rob Strand]

If I can assume you do in fact have an 8 ohm tank, if don't want to change the 10uF cap C20 to a larger value you pretty much have two options.

I'm not 100% sure on the exact details of what you have done but something along those lines will in fact work.

Option 1: (voltage drive)

On the schematic you want to wire the tank side of E16 to *circuit* ground.   Then the circuit side of E16 is left open.  You don't want to ground the *circuit* side of E16, either directly or indirectly via wiring, as this will bridge out the 27 ohm resistor (R27) and the gain will be very high - I think you already did this in some form when you first did the mod.

The problem is there's no ground point at the PCB to connect the tank side of E16 wire to the circuit ground.  You would have to find one.

The undesirable option is to connect the sleeve wire of the tank in to the ground of reverb out.  In this case the E16 wire is essentially unused.   The reason why this configuration is undesirable is the drive currents of the tank input coil flow down the same ground wire as the sensitive tank output signal.

If I can guess how you have it now it's probably this last case.  It will work but with undesirable grounding, although convenient to wire up.

The reason you can get away with what you have done is because C20 is small.


Option 2: (current drive)

Wire the reverb exactly as it originally was.  No funky mods to the reverb tank or grounds.

Solder a 10 ohm resistor in parallel with the 27 ohm resistor (R27).   That makes R27 look like about 7 ohms.

[HoodTube]

Hey again Rob! Sorry for the late reply, time differences I think.

I've actually put the original tank back in (Accutronics 1DB2C1C) after discovering that that wasn't the root of the issue. It has a much better sound to my ear than the replacement did (MOD 8DB2C1D), which I've now returned.

I reckon I'll give option 2 a try first of all. Seems fairly straightforward for even someone like me (famous last words maybe).

Out of curiosity, how long do you think I'd get away with it if it were left as it is now with my dodgy wire the tank ground to spring solution? Is something just waiting to blow or is probably not being driven enough to bother anything?

Cheers again for all your help man. Really is very much appreciated. 

[j_flanders]

The image below shows what your 'mod' does.
The path in red is the path from the input sleeve to ground. It connects to ground through r27 which is 27 Ohm.
The path in blue is your mod.
Because of the 'imperfect connection of the suspension spring to ground you put some additional resistance parallel to r27.
Suppose the resistance of that imperfect connection is 10 Ohm, then the total resistance is: 27 Ohm || 10 Ohm.
This online calculator says: 27 Ohm parallel to 10 ohms gives: 7.30 Ohm
https://www.allaboutcircuits.com/tools/parallel-resistance-calculator/
But that 10 Ohm is just my guess. Could be 2 Ohm or 100Ohm.

Out of curiosity, how long do you think I'd get away with it if it were left as it is now with my dodgy wire the tank ground to spring solution? Is something just waiting to blow or is probably not being driven enough to bother anything?
Cheers again for all your help man. Really is very much appreciated.

The connection of that suspension spring to the chassis could worsen over time because of corrosion, resulting in more resistance parallel to r27, lowering the drive again.
The connection of that suspension spring to the chassis could also improve by 'biting' into the metal chassis after moving the amp around, kicking it for the crashing springs effect, resulting in even less resistance parallel to r27 increasing the drive more than you'd like..

It is an unpredictable and unreliable solution. You're depending on two metal parts touching eachother, having an imperfect conductive joint, giving you a bit of resistance. You can do the same by adding a small resistor with a reliable and known value parallel to r27.

[HoodTube]

Hey man, thank you for taking the time to explain that. It's sort of making sense now.

I think I'll have to bite the bullet and open it up again and give this a try. I'll report back soon.

Cheers!

[Rob Strand]

reckon I'll give option 2 a try first of all. Seems fairly straightforward for even someone like me (famous last words maybe).

Option 2 is essentially what was there in the original circuit.

What you are doing by putting a resistor in parallel with R27 (27 ohm) is boosting the current through the reverb tank, which obviously makes it louder.    This type of mod is fine.

I'm not sure *exactly* what have done in you mod and the way you have it may not correspond to 10 ohms.    Nonetheless you can experiment with different values: 10 ohm, 22 ohm, 33 ohm.

[HoodTube]

Hello again folks. I've done as suggested and added a resistor in parallel with R27. I used 5.1 ohm, and it's worked like a charm. Hopefully have a reliable reverb for a long time now.

Thank you very much again for all the advice, it truly is appreciated!