Losing sparkle with Dumbleator...

[Tonegeek]

Losing sparkle with Dumbleator...

I recently noticed my built in Dumbleator loses something on the very top end of the signal when in circuit. I have A/B'd this thing extensively using a bypass switch. I don't have any cables hooked up to it, it is just routing signal in, through the tube and back into the PI. You might not notice this loss using an HB guitar but with a Strat it is noticeable. I would like to flatten the response of this circuit if possible.

Tube is a EH 12ax7

Heres what I have tried:

changing the 220k grid resistor on the recovery stage to 470k. No help.
adding bypass caps on the pots (like the real deal). i can see using these to compensate for cable losses, but I should not be seeing losses without cables so this is a last resort. It does not work that well anyway.

Here is the exact circuit I am using with voltages:

[Structo]

Do you have a bright cap on your master volume?

If not, try a 47pf there.

[Tonegeek]

Do you have a bright cap on your master volume?

If not, try a 47pf there.

I don't have any bright caps except on my input volume with the bright switch which I never use. I did try bright caps on the sends and return pots of the loop but it did not have the desired affect. I am hoping to fix this in the loop as that is where the losses are occuring. the rest of the amp is OK, but if I can't flatten the loop then I will start trying to compensate elsewhere in the amp. I have a clean and OD master so i could try caps there as you suggest. I would really like to understand why the losses are occurring in the first place. Is this Miller capacitance losses you think? I wonder if I should just try a brighter sounding tube like a Sovtek 12ax7wa.

[odourboy]

Try a bright cap on the 'lator Return Level pot. I've noted treble attenuation in that part of the circuit with a 'scope in every 'lator I've built. You may need to go as high as 220p to fully restore the highs (though most tend to use the loop to tame some highs).

[erwin_ve]

I use a different schem (as posted by Brandon). Sparkle enough for me in that build.

[Tonegeek]

I tried a different tube (GT) and it helped a bit. I put 47p on both dlator pots and it helps a little more. I shall keep bumping up the values and see. I think there is something else going on that is more than just the high end affecting the character of the sound too. I think it is just a little smoother sounding without the loop. It is very subtle but it is noticeable when I A/B it.

About taming highs: Originally I liked the loop just because it did mellow things out a little. Now the amp has broken in (or my ears have) and I just want the loop to be totally transparent (at least with no cables plugged in).

I tried the .001 on the send pot and it is way too much, even with cables plugged in. I am thinking 100p - 200p ought to do it, with one on each section.

[Tonegeek]

Try a bright cap on the 'lator Return Level pot. I've noted treble attenuation in that part of the circuit with a 'scope in every 'lator I've built. You may need to go as high as 220p to fully restore the highs (though most tend to use the loop to tame some highs).

I have a scope but I would think you need a spectrum analyzer to see something like that. How would I set my scope up to see the attenuation? Do I need a white noise source? thanks

[butwhatif]

Brt caps on either the send or ret pots were always way over the top for me. Can't imagine ever using them. I would investigate the small brt cap on the master vol , properly selected it works. With a 1KZ sine wave thru the OD, the resulting clipped wave in mine has a very slight peak on the front end.

Re the loop loss, a square wave thru the amp/loop in clean mode will show u what's happening, but i'll bet
the loop doesn't lose anything. Try 1K or 800HZ. Other hi freq sq w will also show u if there's hi attenuation.

[Normster]

I'm not sure what layout you're using, but I have noticed a loss of "sparkle" with the real-deal layout. (8" x 17" chassis) I assume it's because of the amount of shielded cable. The most transparent buffer I've built is my early point-to-point prototype. Another, built into a 5" x 13" box had a bit more coax in it but was still pretty transparent. Overall, the 5" x 13" is the one I stuck with.

[butwhatif]

Mine's internal, no cables=no loss, no extra bs. Trimpots for send+return. External is hardware i don't need.

[Tonegeek]

Mine's internal, no cables=no loss, no extra bs. Trimpots for send+return. External is hardware i don't need.

I find that there is a sweet spot associated with the loop (or there used to be before I converted to the BM, not sure now) so I have to have the pots external, otherwise the trimpots are a good idea, one that I may try sometime. The return pot is my global master so must have that.

Brt caps on either the send or ret pots were always way over the top for me. Can't imagine ever using them. I would investigate the small brt cap on the master vol , properly selected it works.

I don't usually like bright caps anywhere, but some of the experiments with the bright cap on the loop return (master for me) are showing some promise.

Re the loop loss, a square wave thru the amp/loop in clean mode will show u what's happening, but i'll bet
the loop doesn't lose anything. Try 1K or 800HZ. Other hi freq sq w will also show u if there's hi attenuation.

My loop is definitely losing something but as I said earlier, I don't understand why. There is no shielded cable and the runs are extremely short, except for the input and output wires which are about 7" each factoring in that they go from the board to the bypass switch, then to the jacks. I will drag my scope out soon and try the square wave thing.

I'm not sure what layout you're using, but I have noticed a loss of "sparkle" with the real-deal layout. (8" x 17" chassis) I assume it's because of the amount of shielded cable. The most transparent buffer I've built is my early point-to-point prototype. Another, built into a 5" x 13" box had a bit more coax in it but was still pretty transparent. Overall, the 5" x 13" is the one I stuck with.

See above- no shielded cable, very short runs as most components are direcly soldered to the tube. Let me just say what I am hearing is very subtle, it is an ever so slight dip in the high register. I see it as much as a loss of high harmonics as actual high fundamentals. Its just enough to nag me into trying to fix it.

[Normster]

Oops, missed the "built-in" part of your post. In looking at your schematic, I noticed that the cathode bypass cap is 4.7uF. I believe Brandon indicated that it's actually a 47uF. Of course, I doubt that's doing anything to your top end, but thought I'd point it out anyway.

[odourboy]

Try a bright cap on the 'lator Return Level pot. I've noted treble attenuation in that part of the circuit with a 'scope in every 'lator I've built. You may need to go as high as 220p to fully restore the highs (though most tend to use the loop to tame some highs).

I have a scope but I would think you need a spectrum analyzer to see something like that. How would I set my scope up to see the attenuation? Do I need a white noise source? thanks

A 'scope with a function generator so you can manually sweep a sine wave through the 'guitar' spectrum. Two of my loops are 'internal' without long cable runs to dull the highs and I still found they needed help.

[Tonegeek]

Oops, missed the "built-in" part of your post. In looking at your schematic, I noticed that the cathode bypass cap is 4.7uF. I believe Brandon indicated that it's actually a 47uF. Of course, I doubt that's doing anything to your top end, but thought I'd point it out anyway.

Wow! I revisited what I guess is Funk's schematic and it is 47uF. I will test the correct value later today and see if that changes the character of the bottom end, (although lack of bass has never been my problem). As you stated, it probably won't help the highs. You may have also noticed I used one 22m resistor in the local neg. FB circuit instead of the 2x10m in the real deal, but doubt that has much affect. Actually I disconnected the FB circuit as part of my testing and it did not seem to affect the highs at all. I am guessing the feedback would have more affect of slightly squashing all but the very low freq. equally.

[Tonegeek]

A 'scope with a function generator so you can manually sweep a sine wave through the 'guitar' spectrum.

So, I imagine you run something like a 1k signal into the amp, and adjust the amplitude so that it is exactly the same with the loop in as with it out. Then run it up to say 5k-8k and do the same comparsons until you see where it starts to fall off. Am i on the right track? I guess with a good signal generator, it would be calibrated to put out equal voltage at all frequencies. My home made one won't do that, so I have to be creative in these types of applications.

Two of my loops are 'internal' without long cable runs to dull the highs and I still found they needed help.

My experience exactly. When my amp was new, I liked the affect the loop had on it, now I need that high end back.