#102 "System" In a Box Build

[aflynt]

Here's a couple more gutshots.

I just tapped off the screen supply with another parallel filter node for the the D'Lator. The voltages are 280VDC on the CF plate and 255VDC on the return plate now using a 22k dropper and 100k plate resistors. It's about 20VDC higher than real D'Lators, but sounds fine, though. I initially built it using a 3.3k loop dropper and it sounded really dark and muffled with plate voltages in the low 300VDC range.

I'm going to have another iteration of the Reverb board fabbed since this one is more of a prototype with cut traces on the back and socketed parts. OSHPark is a great service for that. I had my relay supply, relay carrier, and bias supply / rectifier boards done by them too.

-Aaron

[aflynt]

When working with a faceplate I've had good success using a drill press and the faceplate itself to mark the chassis: Line up the edges of the faceplate carefully and transfer the hole locations to the chassis with a fine-point sharpie. Mark the centers with the help of a steel rule and drill 1/8-in pilot holes. Then change to a step-drill bit and enlarge the holes to size. Here's the trick: run the drill relatively slowly, and as you are enlarging each hole monitor it's location with respect to the circle made with the sharpie. If the center drifts you can bring it back by applying side-pressure to the chassis in the appropriate direction.

That's probably what I should have done. I did wind up using the step bit since that was the only one short enough to fit. It did occur to me to use the faceplates as a template once they came in, but by that time I'd already finished drilling the chassis. I'll definitely do pilot holes next time. Thanks!

-Aaron

[trane34]

Very clean. Looks great. I like the black boards too. I got mine from mcmaster. I can't seem to keep the flux from splattering all over it though and it looks messy up close.

[aflynt]

Excellent job Aaron! Some new stuff I've not seen before, very interesting. Incidentally you use black board material, so do I, think it shows off the parts better. Is that why you use it?

I'm working on a 102 myself right now, not as sophisticated as yours, basically traditional, classic PS not precision, but with an internal buffered loop and a third relay for the mid boost. However I've gone NOS wherever I could that makes sense, so OT, caps and pots and resistors etc are '60s to 80s stuff. 20 ga solid wire too.

Will post it when it's done shortly.

Thanks Dave! Looking forward to checking out your build. I used the black boards because I thought they'd be a better match with the purple fabbed boards I got from OSHPark for the relay carrier, relay supply, bias supply / rectifier and reverb boards. It does seem to show off the parts a bit better.

I intentionally did this build with all off the shelf, new-stock parts to see how it sounded. I won't be able to pass final judgement there until the amp is burned in, but so far I really like the sound. It's pretty close to my #102 with NOS parts, but a bit looser and softer feeling on the low end and ever-so-slightly brighter. I think the low end difference may be due to this one having the classic PS as opposed to the other amp having the precision. The brightness is probably due to it being brand new, but we'll see how it settles out.

Incidentally, I did a fixed voltage divider for the master with 560k / 68pF from input to wiper and 470k / 120pF from wiper to ground. 120pF matches the capacitance of the loop send cable I'm running in my other build. I've found that the return cable doesn't make as much of a difference, but there's plenty of coax in this build to tame things down either way.

Oh... And this one is quite a bit quieter hum and hiss wise that the other #102 I did.

-Aaron

[alvarezh]

Thank you Aaron, let me summarize so you can correct me:

A) "The reverb is positioned in the buffer circuit immediately after the return."

This means it's placed after the return 12ax7 section of the D'lator, right?

If and when you have time and if at all possible, could you be so kind as to post an schematic of this integration?

B) "It's in series with the signal. The mix control with the difference amp seems to work very well and I can't tell the difference between the mix down with the reverb on and the reverb bypassed at all."

So you placed a bypass switch to deactivate the reverb?

Very useful information for me. I apologize beforehand if i am constituting a hassle for you.

Lots of good thoughts and ideas put into this reverb, great work indeed.

Enjoy your amp!

[bluesfendermanblues]

aflynt, thanks for your reply regarding the reverb topologiens. Like Structo, i'm very interested in seeing a schematic,

Regarding using a (modest) 9v supply for the reverb circuit and having this after the dumbleator, which can produce large voltage swings, I would imagine that you would at risk of running out off headroom at gigging levels, Have you tried the amp at higher volume levels? Or does the PCB have an onboard voltage regulator (like e.g. the Suhr Minimix II) that turns 9v into 2x18v?

[aflynt]

Thank you Aaron, let me summarize so you can correct me:

A) "The reverb is positioned in the buffer circuit immediately after the return."

This means it's placed after the return 12ax7 section of the D'lator, right?

If and when you have time and if at all possible, could you be so kind as to post an schematic of this integration?

B) "It's in series with the signal. The mix control with the difference amp seems to work very well and I can't tell the difference between the mix down with the reverb on and the reverb bypassed at all."

So you placed a bypass switch to deactivate the reverb?

Very useful information for me. I apologize beforehand if i am constituting a hassle for you.

Lots of good thoughts and ideas put into this reverb, great work indeed.

Enjoy your amp!

Sorry. In hindsight, "Return" is pretty ambiguous. If you look at the schematic here: http://ampgarage.com/forum/download/file.php?id=14586 the reverb is placed in series between the wiper of 250k "recover input" pot after the return jack and the junction of the 10M, and 220k/500pF before the grid of the recovery triode. I actually replaced the 250k pot with a 220k resistor to ground since in practice I've always just run that pot wide open. So... My reverb input is connects directly to the return jack when the reverb is on. I think of reverb as the last effect in the loop conceptually.

I'm using a relay to bypass the reverb so it's foot-switchable. It's wired like a standard DPDT true bypass switch with the common lugs connected to the return jack and return grid network, the NC lugs tied together and the other two lugs going to reverb board in and out. It's wired backwards in the gutshot with the reverb board in and out connected to the NC lugs of the relay, so don't look at that.

-Aaron

[aflynt]

aflynt, thanks for your reply regarding the reverb topologiens. Like Structo, i'm very interested in seeing a schematic,

Regarding using a (modest) 9v supply for the reverb circuit and having this after the dumbleator, which can produce large voltage swings, I would imagine that you would at risk of running out off headroom at gigging levels, Have you tried the amp at higher volume levels? Or does the PCB have an onboard voltage regulator (like e.g. the Suhr Minimix II) that turns 9v into 2x18v?

See above. The reverb is effectively the last effect in the loop so headroom isn't really an issue unless you max the clean volume and send levels. In practice, the return stage of the loop seems to distort before the reverb circuit does when you crank the send. If I set the input high enough to distort, it distorts with the reverb on or bypassed. I was contemplating doing a higher headroom supply, but found it wasn't really necessary in practice. I'd need a higher headroom loop circuit to make it worthwhile. In retrospect I'm guessing that's why HAD put the recovery input control in there.

Of course, If you really wanted to overdrive the loop return for a different distortion sound you could always just bypass the reverb... Hmm... Maybe I should look at using a charge pump again. But then I'd need to knock down and recover the wet signal in the reverb circuit to avoid overloading the Belton modules... Maybe someday... For now this works perfectly for me... The amp was sparkly clean at really loud levels on the job the other night with the send on 4 or so.

-Aaron

[alvarezh]

Thank you Aaron for taking the time to explain.

I moded the schematic as I believe it to be your explanation.

I am I correct?

Thanks again.

[aflynt]

Thank you Aaron for taking the time to explain.

I moded the schematic as I believe it to be your explanation.

I am I correct?

Thanks again.

That is correct. You're quite welcome.

-Aaron

[aflynt]

Here's the schematic for the reverb itself if anybody's interested.

-Aaron

[alvarezh]

Great, now I have a clear pathway to follow.

Very grateful, thanks.

[bluesfendermanblues]

Here's the schematic for the reverb itself if anybody's interested.

-Aaron

Very nice. Thanks.

Do you know a source for a prefab PCB for the circuit?

[bluesfendermanblues]

aflynt, thanks for your reply regarding the reverb topologiens. Like Structo, i'm very interested in seeing a schematic,

Regarding using a (modest) 9v supply for the reverb circuit and having this after the dumbleator, which can produce large voltage swings, I would imagine that you would at risk of running out off headroom at gigging levels, Have you tried the amp at higher volume levels? Or does the PCB have an onboard voltage regulator (like e.g. the Suhr Minimix II) that turns 9v into 2x18v?

See above. The reverb is effectively the last effect in the loop so headroom isn't really an issue unless you max the clean volume and send levels. In practice, the return stage of the loop seems to distort before the reverb circuit does when you crank the send. If I set the input high enough to distort, it distorts with the reverb on or bypassed. I was contemplating doing a higher headroom supply, but found it wasn't really necessary in practice. I'd need a higher headroom loop circuit to make it worthwhile. In retrospect I'm guessing that's why HAD put the recovery input control in there.

Of course, If you really wanted to overdrive the loop return for a different distortion sound you could always just bypass the reverb... Hmm... Maybe I should look at using a charge pump again. But then I'd need to knock down and recover the wet signal in the reverb circuit to avoid overloading the Belton modules... Maybe someday... For now this works perfectly for me... The amp was sparkly clean at really loud levels on the job the other night with the send on 4 or so.

-Aaron

Prior to building 'real' dumbleators, I tried an early version of the Kleinulator with 9 volt. Worked like a charm at bedroomlevels, but not at gigging levels. In all fairness to the Klein design, the latest version (the Ceriatone version), use a Charge Pump design in order to up the voltage.

If 9v voltage is ok, why fix it.

[aflynt]

Here's the schematic for the reverb itself if anybody's interested.

-Aaron

Very nice. Thanks.

Do you know a source for a prefab PCB for the circuit?

I don't think anybody's selling prefabbed versions of this exact circuit with the two parallel modules (unless somebody happened to have the same idea as me), but there are a few people over at MadBean's board selling prefabbed versions of single board implementations. BOYC also has their own version too.

I'm not really planning on doing a big batch of boards and selling them, but If you're interested I can send you the Eagle .brd file for the schematic above. Then you could simply create an account at OSHPark upload the file and they'll send you a set of 3 boards in week or two. It's about $25.40 for a set of 3 from them. The latest board revision is unverified, though. The board I used for the amp actually has a few mistakes in it and cut / jumpered traces on the back. The new one fixes those, but I haven't had a chance to buy a lot and test it yet.

-Aaron