Mark Hammer Stage Center reverb mods question

[Wild Zebra]

As a mechanical device, a spring reverb probably offers up more variation by adjusting the tone going into the springs than coming out of it.  Doesn't mean that a post-spring tone control will be of absolutely no use, but you may find that simply altering the bass response going into the springs will have a much more robust effect on tone than a subtle treble-trim control like the SWTC will have.

I've recommended it before and will recommend it again here:
1) Make the 470k feedback resistance in the op-amp feeding the springs variable.  Something like a 500k pot in series with a 220k fixed resistor is probably about right.  This will vary how hard you drive the springs, and elicit varying amounts of "sproing" out of them.  Note that with a 220pf feedback cap, a max resistance of 220k+500k will roll off the treble around 1khz, so you may want to drop the 220pf to 120pf or so if you use a variable drive-gain mod such as described. (Incidentally, in traditional Fender circuits, "Dwell" is actually how hard the springs are driven, NOT how much spring signal is mixed in with clean as the GGG schematic suggests).

2) The amount of "sproing" will depends on how much of that higher-amplitude bass you feed the springs.  The 22k/.02uf combo on the driver stage gives a bass rolloff starting around 360hz.  You can make it more sproingy by dropping that rolloff down with a .047uf cap, or make it more "air-ey" by raising the rolloff with a .015 or .012uf cap.  Note that by trimming off any bass signal fed to the springs, you will need a bit more drive.  Yet another reason to incorporate variable drive into the basic schematic.

3) Since the amount of pre- and recovery drive is bound to acquire some hiss, a variable treble cut control is probably a reasonable idea.  Here's how to incorporate an SWTC.  The mixer stage (IC1c in the GGG schematic) uses almost-equal input resistances (10k vs 4k7+4k7) to match the dry and wet signal levels.  Based on the roughly 10k input resistance, a 33k feedback resistor gives an approximate gain of a little over 3x.  The wet path is divided into two 4k7 resistors, with a .01uf cap to ground from their junction.  This provides one pole of lowpass filtering at roughly 3.4khz whilst still maintaining an input resistance similar to the dry path.  Readers of my posts will know that I often suggest something like this when people ask how to "warm up" the delay path in a delay unit.  Of course, what Anderton has done here is really the same thing as an SWTC, except its fixed rather than variable.  So let's unlock the variability.
     If the input resistance to IC1c is increased to 15k and the feedback resistance is increased from 33k to 47k, we end up with close to the same amount of gain in the final op-amp stage, but benefit by having an input resistance we can now use productively.  Replace R3 (10k) with a 15k resistor.  Replace the two 4k7 resistors from the wet level pot (R1, R2) to the mixer stage with a 2k2 resistor, 10k pot, and 2k7 resistor (in that order).  The wiper of the pot goes to a .015uf cap and then to ground.  With the 2k2 resistor on the wet-level side of the tone pot, the treble rolloff will go from roughly 880hz when the wiper is closest to the mixer-stage input (i.e., the junction between the 10k pot and 2k7 resistor), to roughly 4.8khz when the wiper is moved closest to the wet-level pot (the side where the 2k2 resistor is).

Inclusion of a bass-rolloff switch, a dwell/drive control, and a post-spring variable treble rolloff, in addition to the usual wet/dry level pots should

So I'm pretty keen on everything except  the end where we "unlock" the tone control via changing R3 to a 15K and replacing the two 4.7K's to the 2K2, 10K pot, 2K7 set up.  This is how I figure it from the Dwell wiper a 2K2 to the outside lug of a 10K pot and a 2K7 to the other outside lug to the 15K towards the Mix pot.Here's where I get a little lost it says .015 cap from the wiper to ground.  Are we replacing C2 the .01 with said .015?  If not where do I make the connection to C2 with the new resistor pot combo? 

Thanks  Schem link http://www.generalguitargadgets.com/diagrams/stage_center_reverb_sc.gif

[slacker]

Yeah that's right. You replace R1 and R2 with the 2k7,pot,2k7 combo. One end goes to the wiper of the dwell pot and the other end goes to where R1 and R10 join. The 0.015u cap replaces C2.

[Wild Zebra]

Cool, ready to go at this this weekend.  Now one or two more questions I thought of.  When adding the drive-gain mod (500Kpot+220K)  does it matter what lug I attach the 220K too?  Lug one or two and do I just run the other outside lug to ground? And speaking of grounds in the schem it has chassis grounds and circuitboard grounds.  Should I just put the chassis grounds to the tank?  Will that be fine? Thanks.

I will report back when i have it all together.

[Mark Hammer]

Okay.  Here's the thing.  The SWTC relies on changing the resistance a cap-to-ground sees leading up to it.  As the resistance gets larger, the rolloff moves in a downward direction.

HOWEVER....where the wiper of that pot is can create some unintended side-effects, which the additional resistors are intended to fix.  What side-effects?  Well, if there was no fixed resistor between the pot and the op-amp input, when the wiper moved all the way over there, the treble-cut action of the cap would be applied to both the wet and the dry signal, since it would essentially be a cap from the "-" input to ground.  Not what you want.

Comversely, if you moved the wiper all the way over to the other direction, it would essentially be a cap from the wiper of the Dwell pot to ground.  That's confusing.  Why?  Because the way the SWTC works is that as the resistance leading up to the caop gets smaller, you introduce more treble.  But when the wiper moves to the output iof the Dwell control, it gets turned back into a treble cut again.  Oy!!  I'm getting dizzy already.

So, what you want to have are fixed resistors on each end of the SWTC: one to isolate it from the Dwell pot, and the other to prevent you from accidentally applying the treble cut to the dry signal too.

Why are they the values I indicated?  The SWTC pot, and added resistors form the overall input resistance from the Dwell pot to the mixer op-amp.  R3 is the input resistance from the Mix pot.  You can see that in the original you have 10k on the dry side, and 4k7+4k7 (9.4k) on the wet side.  The gain for each signal source is given by the feedback resistance (R5), divided by the input resistance.  So, to apply the same amount of gain to each signal, and make it easier to estimate where the control ought to be set to get a given tonal balance, you want the input resistances to be as close to equal as possible.

I selected added resistor values for each side of the SWTC pot that would be easily-obtained standard values and add up to a standard value to be used for the other side.  2k2+2k7+10k=14.9k, which is close enough to 15k.  If you wanted, you could replace 2k2 with 3k3, and 2k7 with 4k7, giving a total of 3k3+4k7+10k=18k, another standard value, and stick in an 18k resistor for R3.  The choice is entirely yours and won't disrupt the functioning of the 10k tone pot.  Just keep in mind that as the input resistance gets bigger, you will need to increase the value of R5 to be able to get the same amount of gain in that 2nd op-amp.  Stock, it applies a gain of just over 3x.  If the dry and wet input resistances were 18k, you would need to increase R5 to 56k to get the same approximate gain in that stage.

And yes, we are replacing the .01uf cap with a .015uf cap.  The reason is that, the original sticks a 4k7 resistor just ahead of the .01uf cap, providing a treble rolloff at 3.4khz.  We, on the other hand, are going to make that resistance variable, which means that we can move the rolloff upwards or downwards from 3.4khz.  With a 3k3 resistor and .015uf cap, when the wiper is moved all the way in the direction of the Dwell pot, the rolloff will start around 3.2khz.  As the wiper moves to the other side, the pot+fixed resistance adds up to 13k3, which, with a .015uf cap, yields a treble cut starting around 800hz.

Now that I think of it, if you stick with the suggested combined resistance of 15k (2k2+2k7+10k), use a .015uf cap (rolloff range of 870hz-4.8khz).  If you opt for a combined resistance of 18k (3k3+4k7+10k), use the original .01uf value (rolloff range of 1.2khz to 4.8khz).

Make sense?

[Wild Zebra]

Make sense?

yep, thanks much


How bout these here queries am I correct in saying this

When adding the drive-gain mod (500Kpot+220K)  does it matter what lug I attach the 220K too?  Lug one or two and do I just run the other outside lug to ground? And speaking of grounds in the schem it has chassis grounds and circuitboard grounds.  Should I just put the chassis grounds to the tank?  Will that be fine?

[Mark Hammer]

The suggested mod will provide a variable feedback resistance so that instead of being fixed at 470k, it will vary from 220k to 720k (or whatever the sum of the pot and fixed resistance works out to actually be).  Since gain increases as feedback resistance goes up, you will want their combined resistance to get larger as the wiper moves in a clockwise direction.

So, assuming you are looking at the knob of that pot, and the solder lugs are sticking out below, you will want to: a) solder the 220k resistor to the lug on the left, and b) run a wire from the wiper to the board.  Whether the 220k resistor or the wiper goes to the inverting input or output of that op-amp is immaterial.  Either arrangement still results in feedback resistance getting larger as the wiper moves away from the leftmost lug.

[Wild Zebra]

Thanks alot, I'm hoping to build this starting tonight, will post results.

[Wild Zebra]

Hello, hello, is this thing on.  That was my joke all day at band practice, anyway.  I built this, this here weekend.  Went together pretty smooth.  I'll tell you what it sounds pretty darn good. Definately spring for the right tank, no troubles, instead of trying to modify the circuit for some tank you found.  The "hammer mods"  are definately a nice add on.  Not as drastic as I thought they would be, but definately very usefull and worth it. Thanks Mark!.  I got a little hum and hiss with the last little turn of the dwell aka reverb amount pot, but thats about it, over all pretty quiet.  So time to find a home for this sucker.  Thanks again for the help and if anyones wondering about this circuit, get a tank do the mods and have fun.

[axg20202]

...definately spring for the right tank...

..jub, jub...

You'll love this build once it's up and running mate. I built one a year before last and love it. You can't beat real spring reverb (unless you've got real plate reverb).

[Mark Hammer]

Thanks for the feedback.  Ultimately, the springs will be the crucial factor in the tone, and there isn't a whole lot one can do to step around the tone they provide (or don't provide).  The suggested mods are simply a way of extracting their best behaviour.  Still, nice to hear that they provide added value in terms of doing that.

Note that the metal pan the springs sit in not only provides a solid structure, but also provides shielding.  The signal is at a particular low level and shaky (sic) juncture as it passes through the springs, which is why it needs to be boosted so much.  All that boost in the absence of shielding is bound to raise the spectre of hum.  Certainly keeping all leads to and from the pan shielded is a good idea, as is keeping them short.  You may also want to consider inserting a bass-cutting cap between the wiper of the Dwell pot and point D.  That way, any hum picked up en-route will be filtered out.  Reducing the value of C6 and including the suggested second cap can also be part of that same solution to hum.  In theory, you should be able to crank the Dwell pot up full if the hum is treated as just more annoying bass.  I've found in past, as well, that the airiness-vs-boxiness of reverb tone is partly dictated by how much bass you both push into the springs and recover from them.  Much like the bass content of the wet signal in a chorus, sometimes just cutting back the bass a bit makes the effect less in your face, and easier to appreciate.

[Wild Zebra]

Thanks for the feedback.

No prob, Yes when the dwell is all the way up besides the little bit o' hiss, it definately is pretty bass heavy.  So maybe a .01 cap off the wiper to point D?? What value do you recommend?  I'm looking forward to pumping My soundlab minisynth and sequencer through this soon.  I always liked alittle spring with my analog.

[Mark Hammer]

Start with .01-.022, and work from there.

[axg20202]

...another mod of sorts...the driver IC. I forget the details now but, thanks to old posts from Mark, I changed the driver IC to one with a difference impedence spec - my build was not useable without this change because it sounded like my guitar amp was in a cave.  I should note that I used the recommended spring tank that Smallbear supplies. the tank you use has a big effect on how well the circuit works and it needs to be tweaked accordingly. In short, the driver IC needs to be 'see' a certain ballpark tank impedance and it seemed that the Smallbear tank and the stock driver IC of the reverb circuit are not very well matched in this regard. The IC swap fixed it. I'm sure Mark will chip in here with the details as he can probably tell you what the replacement IC is off the top of his head. Otherwise, search some old threads - it's here somwhere. As I said though, it depends on the input impedance of the tank you are using.

Andy.

[Wild Zebra]

  Hmm, not sure about the IC, but I would like to know.  Mine sounds good as far as I know (but what do I know)

sounded like my guitar amp was in a cave

Isn't that what it's supposed to sound like?

[Mark Hammer]

...another mod of sorts...the driver IC. I forget the details now but, thanks to old posts from Mark, I changed the driver IC to one with a difference impedence spec - my build was not useable without this change because it sounded like my guitar amp was in a cave.  I should note that I used the recommended spring tank that Smallbear supplies. the tank you use has a big effect on how well the circuit works and it needs to be tweaked accordingly. In short, the driver IC needs to be 'see' a certain ballpark tank impedance and it seemed that the Smallbear tank and the stock driver IC of the reverb circuit are not very well matched in this regard. The IC swap fixed it. I'm sure Mark will chip in here with the details as he can probably tell you what the replacement IC is off the top of his head. Otherwise, search some old threads - it's here somwhere. As I said though, it depends on the input impedance of the tank you are using.

Andy.

The NE5532 and LM833 are dual op-amps that are comfortable driving the lower impedance loads of reverb pans.

[axg20202]

  Hmm, not sure about the IC, but I would like to know.  Mine sounds good as far as I know (but what do I know)

sounded like my guitar amp was in a cave

Isn't that what it's supposed to sound like?

Note really. Before changing the driver IC, the signal was so unbelievably wet that it was unusable. When I say cave, I mean huge, bottomless cavern. You don't want this. I believe I used the LM833 with the tank from Smallbear. Sounds great.

[Wild Zebra]

I'll have to try one sometime.