ampage Tube Amps / Music Electronics	For current discussions, please visit Music Electronics Forum.

ampage archive

Vintage threads from the first ten years

Quack if you're stupid (long)

8/18/2000 2:24 PM
Mark Hammer	Quack if you're stupid (long) Pleased by the recent lengthy interchange Aron Nelson had with a newbie on Stompbox Forum about troubleshooting his effect, I figured I'd parade around in all my torn underwear glory here, and tell similarly too-ashamed-to-admit-it folks about how I got my Dr. Quack to work...finally. I had made several Dr. Q's in perfboard form, a few years ago, but they never really worked completely. I always thought there was something wrong with the schematic that Jack Orman had posted on AMZ, but the posted schematic was fine. Many consults later, this was abundantly clear to me, and lo and behold, I found that what prevented mine from working was a wire that had retained an intact outer insulation while being fractured inside. I replaced the wire and everything worked fine. Armed with the belief that I could successfully perfboard one of these, I tried it again two times this past spring. Both included more features than the original, and both were successes. I A/B'd them against an original in a vintage store and liked them more than the original. They kept the sound, but added more flexibility. Perfboarding is a drag though. I've pinched my fingers in the needlenose pliers more times than I care to think of, and finding out a little too late that you accidentally tied the feedback resistor to the non-inverting input when it should have gone to the inverting one (and having to disassemble) is exasperating. So is asking my little one to wait just 10 more minutes for story-time so I could be assured I didn't "lose my place" on the perf-board. Given the poor sound quality of the 1458 dual op-amp used in the original, I was pleased when Jack Orman designed the improved buffered/compensated Dr.Q which he called the Dr. Quack. I was even more pleased when RG Keen made up a PC-mask for it and posted it in PDF format so I could be assured that it printed out the right size. Clearly, the way was paved to throw one together in an evening. The parts were standard. The board was quick and easy to make. The design relatively idiotproof. So, two weekends back I got into a board-making frenzy and cranked out a bunch of different boards (detailed in the Music Electronics forum here on Ampage). Among them a Dr. Quack board. This past weekend I stuffed it and fired it up. Drum roll please. Nothing. Well, the bypass switch worked REAL good, so I got a terrific straight sound from it. Gotta love that true bypass, folks. Wed and Thurs I started debugging. Wednesday was a write-off. Still nothing. Given my past floundering with a busted wire, I checked some relevant voltages on the chip. I was kinda wondering why there was 9v going to "pin 1" but not to "pin 8". Finally, an epiphany. I realized that RG's PDF PC-mask was for toner-transfer, and was a MIRROR IMAGE of what the actual layout was supposed to be. I was using a different system altogether. This meant that all the pins on the dual op-amp were reversed. Yeesh! Fortunately, I had socketed the chip, so I pulled the chip and gently bent the pins in the other direction. The part number was now on the underside of the chip, and when inserted all the pins would be seated in their proper place in the socket. Ah, but a mirror image board means everything ELSE is flipped around too. While RG's Dr. Quack board is nice and compact, it is moderately symmetrical in some places...enough to throw an idiot like me off track. After flipping the chip, and a few "still-nothings" I then recognized that the leads to the various off-board components were also flipped around. So, out came the solder-sucker, and all those things got reinstalled. Decided to measure the AC signal voltage (200mv range on the meter) at the source pin of the FET. Nothing. A nice AC voltage on one side of the .05uf cap of about 150mv when I strummed, but nothing on the FET side of that cap. Something was wrong. I yanked the MPF102 that was in there (it had been cannibalized twice already), and stuffed the spot with a fresh J201....oriented properly. Bingo. We had signal coming out of the FET. Plugged it into the amp, fired it up to see what happens. Nothing. Then, in frustration a power chord. Wha. WHA??!! It worked! It just was VERY unresponsive. Check the schematic again. The original had a 2.2meg gain-setting resistor in the feedback loop of the envelope detector. This one had a 470k resistor. I had thought that maybe this was fine, given the FET buffer, but no. One thing you need to understand about my guitar is that is has precious little bass. With the Dr. Quack filter down VERY low in its sweep range the output from the effect, when insensitive and used with MY guitar, would be equivalent to virtually no signal at all. Crank up the ampo, and you'd hear a very faint guitar signal, but as far as I knew that could have been leakage...especially given that I had twisted a number of input and output wires together to organize them (bad practice, but what the hey, it's rock and roll). My guitar has an on-board pre-amp, and the sensitivity pot was turned up full, so I felt comfortable in assuming that 470k was not going to be enough. I swapped it for 2.2meg, and BINGO again. There was now enough gain in the envelope detector stage to drive the transistor into sweeping. It worked, and responded nicely. But it sounded like garbage (not like Shirley Manson, who actually sounds nice). Using some of my own advice, I realized that envelope ripple is a major source of distortion in these things, so I stuck a 330pf in parallel with the 2.2meg resistor. This would give the gain stage in the envelope detector a rolloff starting at about 220hz. Made a minor improvement, but not enough. Looked at the schematic and realized that the decay-setting cap for the Dr. Quack was 10uf. Too small. Subbed a 22uf cap. MUCH better. A bit slower decay time, but the smoothness in sound was a nice reward. Up until now I had been using a 1458, which I knew was less than ideal, although the original required it. Jack's improvements to the original were intended to make it possible to permit any dual op-amp to be used, so I stuck in a NE5532. MUCH better. More definition, and nicely quiet. I took out the 100ohm attack-setting resistor, and replaced it with a fixed resistor and 1k pot to make variable attack times. Since I had a long track record of goof-ups until now, why stop? I accidentally put in a 470ohm resistor instead of the intended 47ohm. Once detected, everything in the garden was lovely. So, now I have a functioning Dr. Quack that is cleaner than the original, has a somewhat wider sweep range, and is a little more responsive. Major insights: 1) MAKE SURE about the orientation of any PC-mask you are about to use! 2) 470k of feedback resistance in the envelope detector stage is NOT enough, unless you intend to drive it with line level signals. Go back to 2.2meg. 3) It WILL work with other op-amps. 4) I'm an idiot, albeit a very nice one.

8/18/2000 2:24 PM

Mark Hammer

Quack if you're stupid (long)

Pleased by the recent lengthy interchange Aron Nelson had with a newbie on Stompbox Forum about troubleshooting his effect, I figured I'd parade around in all my torn underwear glory here, and tell similarly too-ashamed-to-admit-it folks about how I got my Dr. Quack to work...finally.

I had made several Dr. Q's in perfboard form, a few years ago, but they never really worked completely. I always thought there was something wrong with the schematic that Jack Orman had posted on AMZ, but the posted schematic was fine. Many consults later, this was abundantly clear to me, and lo and behold, I found that what prevented mine from working was a wire that had retained an intact outer insulation while being fractured inside. I replaced the wire and everything worked fine.

Armed with the belief that I *could* successfully perfboard one of these, I tried it again two times this past spring. Both included more features than the original, and both were successes. I A/B'd them against an original in a vintage store and liked them more than the original. They kept the sound, but added more flexibility.

Perfboarding is a drag though. I've pinched my fingers in the needlenose pliers more times than I care to think of, and finding out a little too late that you accidentally tied the feedback resistor to the non-inverting input when it should have gone to the inverting one (and having to disassemble) is exasperating. So is asking my little one to wait just 10 more minutes for story-time so I could be assured I didn't "lose my place" on the perf-board.

Given the poor sound quality of the 1458 dual op-amp used in the original, I was pleased when Jack Orman designed the improved buffered/compensated Dr.Q which he called the Dr. Quack. I was even more pleased when RG Keen made up a PC-mask for it and posted it in PDF format so I could be assured that it printed out the right size. Clearly, the way was paved to throw one together in an evening. The parts were standard. The board was quick and easy to make. The design *relatively* idiotproof.

So, two weekends back I got into a board-making frenzy and cranked out a bunch of different boards (detailed in the Music Electronics forum here on Ampage). Among them a Dr. Quack board. This past weekend I stuffed it and fired it up. Drum roll please. Nothing. Well, the bypass switch worked REAL good, so I got a terrific straight sound from it. Gotta love that true bypass, folks.

Wed and Thurs I started debugging. Wednesday was a write-off. Still nothing.

Given my past floundering with a busted wire, I checked some relevant voltages on the chip. I was kinda wondering why there was 9v going to "pin 1" but not to "pin 8". Finally, an epiphany. I realized that RG's PDF PC-mask was for toner-transfer, and was a MIRROR IMAGE of what the actual layout was supposed to be. I was using a different system altogether. This meant that all the pins on the dual op-amp were reversed. Yeesh! Fortunately, I had socketed the chip, so I pulled the chip and gently bent the pins in the other direction. The part number was now on the underside of the chip, and when inserted all the pins would be seated in their proper place in the socket.

Ah, but a mirror image board means everything ELSE is flipped around too. While RG's Dr. Quack board is nice and compact, it is moderately symmetrical in some places...enough to throw an idiot like me off track. After flipping the chip, and a few "still-nothings" I then recognized that the leads to the various off-board components were also flipped around. So, out came the solder-sucker, and all those things got reinstalled.

Decided to measure the AC signal voltage (200mv range on the meter) at the source pin of the FET. Nothing. A nice AC voltage on one side of the .05uf cap of about 150mv when I strummed, but nothing on the FET side of that cap. Something was wrong. I yanked the MPF102 that was in there (it had been cannibalized twice already), and stuffed the spot with a fresh J201....oriented properly. Bingo. We had signal coming out of the FET.

Plugged it into the amp, fired it up to see what happens. Nothing. Then, in frustration a power chord. Wha. WHA??!! It worked! It just was VERY unresponsive. Check the schematic again. The original had a 2.2meg gain-setting resistor in the feedback loop of the envelope detector. This one had a 470k resistor. I had thought that maybe this was fine, given the FET buffer, but no. One thing you need to understand about my guitar is that is has precious little bass. With the Dr. Quack filter down VERY low in its sweep range the output from the effect, when insensitive and used with MY guitar, would be equivalent to virtually no signal at all. Crank up the ampo, and you'd hear a very faint guitar signal, but as far as I knew that could have been leakage...especially given that I had twisted a number of input and output wires together to organize them (bad practice, but what the hey, it's rock and roll).

My guitar has an on-board pre-amp, and the sensitivity pot was turned up full, so I felt comfortable in assuming that 470k was not going to be enough. I swapped it for 2.2meg, and BINGO again. There was now enough gain in the envelope detector stage to drive the transistor into sweeping. It worked, and responded nicely. But it sounded like garbage (not like Shirley Manson, who actually sounds nice).

Using some of my own advice, I realized that envelope ripple is a major source of distortion in these things, so I stuck a 330pf in parallel with the 2.2meg resistor. This would give the gain stage in the envelope detector a rolloff starting at about 220hz. Made a minor improvement, but not enough. Looked at the schematic and realized that the decay-setting cap for the Dr. Quack was 10uf. Too small. Subbed a 22uf cap. MUCH better. A bit slower decay time, but the smoothness in sound was a nice reward.

Up until now I had been using a 1458, which I knew was less than ideal, although the original required it. Jack's improvements to the original were intended to make it possible to permit any dual op-amp to be used, so I stuck in a NE5532. MUCH better. More definition, and nicely quiet. I took out the 100ohm attack-setting resistor, and replaced it with a fixed resistor and 1k pot to make variable attack times. Since I had a long track record of goof-ups until now, why stop? I accidentally put in a 470ohm resistor instead of the intended 47ohm. Once detected, everything in the garden was lovely.

So, now I have a functioning Dr. Quack that is cleaner than the original, has a somewhat wider sweep range, and is a little more responsive.

Major insights:
1) MAKE SURE about the orientation of any PC-mask you are about to use!
2) 470k of feedback resistance in the envelope detector stage is NOT enough, unless you intend to drive it with line level signals. Go back to 2.2meg.
3) It WILL work with other op-amps.
4) I'm an idiot, albeit a very nice one.

8/18/2000 3:17 PM

Eric H

quote:
"1) MAKE SURE about the orientation of any PC-mask you are about to use!"

I've heard that a pcb that's reversed (mirror-imaged) works great if you surface-mount the components on the solder side. This is what I've heard. This method requires discovery of the incorrect orientation of the board BEFORE you stuff it, and I haven't been that lucky

-Eric

8/18/2000 8:43 PM
frank clarke	Reversed PCB's do work, and are great for the drilling-challenged (me). Even with a non-reversed RTS layout, you just have to bend the pins on the IC's, everything else is fine. I don't think I could build anything significant that worked first time. Fender amps have little CP (check point, I think) numbers over the PCB's, and corresponding voltages on the schematics. It might be smart to have little diagnostic triangles on the copper traces and write down the correct voltages in advance. This would be for pcb's you give to other people. I still expect my own work to fly first time.

8/18/2000 8:43 PM

frank clarke

Reversed PCB's do work, and are great for the drilling-challenged (me).
Even with a non-reversed RTS layout, you just have to bend the pins on the IC's, everything else is fine.
I don't think I could build anything significant that worked first time. Fender amps have little CP (check point, I think) numbers over the PCB's, and corresponding voltages on the schematics. It might be smart to have little diagnostic triangles on the copper traces and write down the correct voltages in advance. This would be for pcb's you give to other people. I still expect my own work to fly first time.

8/18/2000 4:53 PM

Daniel R. Haney

Mark Hammer wrote:

quote:
"4) I'm an idiot, albeit a very nice one."

You're also a very industrious one.
Hats off to you, Sir.

I often rediscover that something works better
if you plug it in.

-drh
--

8/18/2000 7:08 PM
moocow	Maybe PC board artwork should contain something to warn you when you've inadventently flipped it. How about: "If you can read this, you've reversed the board !" only printed mirror-image.

8/18/2000 10:51 PM
brent	thanks mark, i was planning on building (or trying to build) the quack relatively soon. Your words should aid me in my attempt. brent

8/20/2000 1:55 AM
Mark Hammer	Oops, I did it again... I made the Small Stone board from the layout on JD Sleep's site, and made the same blasted mistake again...I made the board in mirror image. I had stocked the board over the week, and soldered in the off-board components this morning, anxiously awaiting to be "stoned". Nada. Garnicht. Bupkes. Rien. Measured the voltage on the appropriate pins on the CA3094's, and they were wrong. Checked the schematic and the board, and once again realized that the board was the mirror image of what I really wanted. Unsoldered all the chips, flips the pins around...carefully...with the needlenose pliers, and resoldered them. It worked! The sound wasn't as dense as I was expecting, and the LFO goes from a bit too slow to the audio range for some ring mod like effects, and it also has that annoying clicking, but it works. Memo to PC layout designers: PUT SOME TEXT ON THE BOARD!

8/20/2000 1:55 AM

Mark Hammer

Oops, I did it again...

I made the Small Stone board from the layout on JD Sleep's site, and made the same blasted mistake again...I made the board in mirror image. I had stocked the board over the week, and soldered in the off-board components this morning, anxiously awaiting to be "stoned".

Nada. Garnicht. Bupkes. Rien.

Measured the voltage on the appropriate pins on the CA3094's, and they were wrong. Checked the schematic and the board, and once again realized that the board was the mirror image of what I really wanted. Unsoldered all the chips, flips the pins around...carefully...with the needlenose pliers, and resoldered them. It worked!

The sound wasn't as dense as I was expecting, and the LFO goes from a bit too slow to the audio range for some ring mod like effects, and it also has that annoying clicking, but it works.

Memo to PC layout designers: PUT SOME TEXT ON THE BOARD!

Page 1 of 2

Next>

Last Page>>