MN3007 ADA Flanger Clone Questions

[derationalize]

I got this thing working, except for the dreaded threshold control. I was searching this thread, and it seems to me the voltages coming off the opamp by the threshold control (IC3) are out of whack. My voltages are:
IC3a Out: 0.1v
IC3a In - : 0.1v
IC3a In +: 7.3v
IC3b Out: 13.5v
IC3b In - : 6.8v
IC3b In +: 7.4v
Power and ground checks out fine. I have the threshold pot and the gate transistor replaced with leads to a breadboard to swap components around to try to get this working. It seems like fiddling with these does nothing. What's going on here?

[pacealot]

I wanted to bump this thread because I recently built the Flintlock PCB version of this circuit, and it works an absolute treat other than having something of the issue that Ed22 had a couple pages back with the low end range of the clocking. I built mine with all the prescribed Flintlock components, which indicates a 27pF cap for C26 (the cap between pins 1 & 3 of the 4047), but I'll be darned if I can get the low end to clock any higher than around 55kHz no matter how I chase the C-Range and C-Max trimmers around (I put a multiturn in for C-Max as the build doc suggests for better control). The only somewhat negative aspect of this is that it has an asymmetrical sweep, where it stays up in the "high" (short delay) range for a lot longer and comes down more quickly, and I can only partially mitigate this with the Auto trimmer — at full anti-(counter-)clockwise it's better, but still lopsided. Does anyone think that taking that 27pF cap down somewhat might help that issue, as reducing the 39pF seemed to have done for Ed22, or is that already around the limits of the circuit's tolerances? Any potential utility to either increasing the Auto trimpot value or else changing the proportions of the resistors around it (R45 & R46), or even swapping those two with each other? Thought I'd ask before I dismantle and start swapping components willy-nilly.

Schematic for reference:

[Kevin Mitchell]

I wanted to bump this thread because I recently built the Flintlock PCB version of this circuit, and it works an absolute treat other than having something of the issue that Ed22 had a couple pages back with the low end range of the clocking. I built mine with all the prescribed Flintlock components, which indicates a 27pF cap for C26 (the cap between pins 1 & 3 of the 4047), but I'll be darned if I can get the low end to clock any higher than around 55kHz no matter how I chase the C-Range and C-Max trimmers around (I put a multiturn in for C-Max as the build doc suggests for better control). The only somewhat negative aspect of this is that it has an asymmetrical sweep, where it stays up in the "high" (short delay) range for a lot longer and comes down more quickly, and I can only partially mitigate this with the Auto trimmer — at full anti-(counter-)clockwise it's better, but still lopsided. Does anyone think that taking that 27pF cap down somewhat might help that issue, as reducing the 39pF seemed to have done for Ed22, or is that already around the limits of the circuit's tolerances? Any potential utility to either increasing the Auto trimpot value or else changing the proportions of the resistors around it (R45 & R46), or even swapping those two with each other? Thought I'd ask before I dismantle and start swapping components willy-nilly.

Schematic for reference:

You're not alone on this. I'd have to pull mine out of storage and take another look but I'm pretty sure I got stuck at the clock calibration as well. I recall it mostly working but having a massive low end thud because I couldn't dial it in properly with the suggest values and didn't want to immediately muck with the timing caps.

[pacealot]

Thanks, Kevin – yes, I also didn't want to diddle around on the caps without receiving a bit more knowledgeable feedback first. For the moment (as I don't have any caps smaller than 27pF on hand yet), swapping the 68K and 82K resistors would be an easy thing to try, and reversible (as long as I go easy on them when desoldering), but it feels like that's not really the "correct" solution (though if it improves the sweep, I suppose that's all that matters).

Update: I did the resistor swap, and it did little besides kill the oscillation entirely on the lower half of the trimpot, and the sweep seems unchanged at the lowest oscillating setting. So back to stock it goes, provided I don't toast the resistors or something adjacent in the process (it was not the easiest of swaps, and I'm as hamfisted as it gets).

The good news is that, when the range is set to a more moderate sweep and the manual is around 3 o'clock or lower, the thing sounds just stunning, just as an A/DA should. It's only the more extreme sweep settings which really highlight the asymmetry – but of course those extreme settings are also ones I want to be able to use! 

[pacealot]

(Final?) update: I got the clocking to the prescribed specs by actually adding slightly to the capacitance rather than decreasing it. A friend gave me a handful of very small pF ceramic caps to play with since I'm light on those, and I messed around in both directions until the numbers started getting in line. If I'd had a 33pF that probably would've been perfect; as it was, I tacked a 4pF onto the original 27pF which tested a little high (around 29pf). I did enough clumsy solder-related damage to some of them in the process to warn me off of swapping back the 68K and 82K resistors, but since I'm trimming the Auto as far anti-clockwise as I can before the osc collapses anyway, it ought to be fine for now. If I ever end up rehousing it in future, I might get in there and put in a fresh pair of 75Ks in just to still have a little extra room to trim in that direction, but for now, I'm calling it good/done. I swear the sweep still shoots down audibly faster than it goes up, but since the clock is solidly where it's meant to be now, I guess that's just how it is. It's still the most powerful and amazing sounding flanger circuit I've heard yet, so I'm calling it a win... 

[pacealot]

With a lot of BBD discussion around here lately, I was inspired to re-investigate my asymmetry issue with this pedal, as it's still been annoyingly evident. I noticed that, both on the Moosapotamus and Lectric-FX schemos, other ICs are indicated in addition to the 1458 for IC4, the LFO generator. So I did a simple swap of first a TL062 and then a TL072 for that 1458, and I found that, while it still doesn't want to be truly 100% symmetrical, and the closest it will get is just at the point of dropping the "auto" trimpot voltage divider as low as possible before collapsing the LFO entirely, I did get significantly better results with both chips than with the 1458. The TL072 got just a hair closer to my ear, so I'm leaving that one in. I also needed to leave the 68K and 82K resistors (R45 & R46) in the "wrong" places in order to get that voltage low enough (it reads ~6.07V before the LFO collapses). I still don't understand enough to know why it has to go so far out of spec in order to get the LFO even close to symmetrical, but at this point, I think I need to be actually done for real this time. I just wanted to update this thread in case anyone in the future has similar struggles.

[Paul Marossy]

With a lot of BBD discussion around here lately, I was inspired to re-investigate my asymmetry issue with this pedal, as it's still been annoyingly evident. I noticed that, both on the Moosapotamus and Lectric-FX schemos, other ICs are indicated in addition to the 1458 for IC4, the LFO generator. So I did a simple swap of first a TL062 and then a TL072 for that 1458, and I found that, while it still doesn't want to be truly 100% symmetrical, and the closest it will get is just at the point of dropping the "auto" trimpot voltage divider as low as possible before collapsing the LFO entirely, I did get significantly better results with both chips than with the 1458. The TL072 got just a hair closer to my ear, so I'm leaving that one in. I also needed to leave the 68K and 82K resistors (R45 & R46) in the "wrong" places in order to get that voltage low enough (it reads ~6.07V before the LFO collapses). I still don't understand enough to know why it has to go so far out of spec in order to get the LFO even close to symmetrical, but at this point, I think I need to be actually done for real this time. I just wanted to update this thread in case anyone in the future has similar struggles.

That's interesting. As I recall, on my early version of the Moosapotamus build, I was never able to get the clock frequency quite up to spec but to my ears it sounds fine. I never really use mine for heavy flanging but rather for a cool pseudo-chorusing sound that reminds me of the sounds on the Andy Summers-Robert Fripp "I Advance Masked" album.

[pacealot]

That's what's so great about this circuit — when it's dialled in, it can do every flanger pedal sound I've ever wanted, from that milky Andy Summers-y Mistress-type sound you describe, to the whooshiest (now that's a word) Pat Travers/Paul Gilbert/EVH jet plane you've ever heard, and anything in between. It even does straight-up chorus perfectly well at the longer delay settings — nearly modulated slapback, even. It just takes some perseverance to get it there.

I do wish I was capable of understanding the circuit properly, but I can only just wrap my brain around anything even slightly more complex than a fuzzbox, so this was a complete paint-by-numbers job on my part — in well over my head here...

[Paul Marossy]

That's what's so great about this circuit — when it's dialled in, it can do every flanger pedal sound I've ever wanted, from that milky Andy Summers-y Mistress-type sound you describe, to the whooshiest (now that's a word) Pat Travers/Paul Gilbert/EVH jet plane you've ever heard, and anything in between. It even does straight-up chorus perfectly well at the longer delay settings — nearly modulated slapback, even. It just takes some perseverance to get it there.

I do wish I was capable of understanding the circuit properly, but I can only just wrap my brain around anything even slightly more complex than a fuzzbox, so this was a complete paint-by-numbers job on my part — in well over my head here...

lol having built stuff off and on since 2000 there are still a myriad of things I don't quite understand.