The secret M117 flanger

[lars]

Sounded like a separate modulation underneath the main modulation to my ears.
Maybe they have one 4013/14013 running as divided by 2 from the first clock?

I have noticed this sound from my pedal as well. There is a subtle 2nd modulation effect taking place in the background. It would be amazing to be able to mod the clocks to really take advantage of the fact this is basically two flangers running simultaneously! Again, the unobtanium schematic for this version is the key.

[jwin615]

Looking at the Peavey DSC-4 dual clock chorus. It has 4 controls, 2x speed and 2x depth controls for each "clock". It really only has one clock, but 2x LFOs(072s) that look to be combined into a comparator(U3-B)and fed into the 3101. I would think this could confuse or glitch the 3101 but it makes for a very lush chorus.
Perhaps something similar is occuring in the MXR. Or it is somehow creating a decade of the LFO?

[Scruffie]

Did anyone consider the possibility that maybe it's just being used for something very dull, like switching? For that fancy new LED?

[museums]

Looking at the Peavey DSC-4 dual clock chorus. It has 4 controls, 2x speed and 2x depth controls for each "clock". It really only has one clock, but 2x LFOs(072s) that look to be combined into a comparator(U3-B)and fed into the 3101. I would think this could confuse or glitch the 3101 but it makes for a very lush chorus.
Perhaps something similar is occuring in the MXR. Or it is somehow creating a decade of the LFO?

There are a bunch of dual LFO choruses from that era and in practice the clock just sees a different pattern of rising and falling voltages over time than with a single LFO. The LFO with the higher depth control generally has more influence over the movement of the overall voltage. With both depth controls in the middle you can get interesting non-linear LFO movement.

IIRC the Peavey is a little different because it has two different LFO shapes.

[Scruffie]

Sounded like a separate modulation underneath the main modulation to my ears.
Maybe they have one 4013/14013 running as divided by 2 from the first clock?

I have noticed this sound from my pedal as well. There is a subtle 2nd modulation effect taking place in the background. It would be amazing to be able to mod the clocks to really take advantage of the fact this is basically two flangers running simultaneously! Again, the unobtanium schematic for this version is the key.

My guess is because of the MN3004's piss poor clock to gain ratio and the fact there's two almost certainly in series what you're both actually hearing is basically a tremolo along with the flanger.

If someone posts some decent gut shots it shouldn't be too hard to trace and confirm seeing as we already have the 117 schematic.

[jessenator]

If someone posts some decent gut shots it shouldn't be too hard to trace and confirm seeing as we already have the 117 schematic.

I'll PM you. Let me know if you need different angles than what I captured.

[Scruffie]

If someone posts some decent gut shots it shouldn't be too hard to trace and confirm seeing as we already have the 117 schematic.

I'll PM you. Let me know if you need different angles than what I captured.

Thanks, unfortunately to get a good look at the traces you'd need to unsolder the pots.

But, I thought of a much simpler solution... just test for continuity between pin 2 of both MN3004, if it beeps, they share the same clock.

If it doesn't beep, check pin 2 of one doesn't match to pin 12 of the other.

Still no beep? Then it's time to test the CD4013 for continuity.

[museums]

If someone posts some decent gut shots it shouldn't be too hard to trace and confirm seeing as we already have the 117 schematic.

I'll PM you. Let me know if you need different angles than what I captured.

Thanks, unfortunately to get a good look at the traces you'd need to unsolder the pots.

But, I thought of a much simpler solution... just test for continuity between pin 2 of both MN3004, if it beeps, they share the same clock.

If it doesn't beep, check pin 2 of one doesn't match to pin 12 of the other.

Still no beep? Then it's time to test the CD4013 for continuity.

Probed around in mine. One half of the inner 4013 handles generating the two phase pulses for each MN3004. The  outer 4013 is then the other half of the clock circuit although I believe based on photos earlier in the thread that there is a single oscillator feeding the two flipflops. BBDs are in parallel like the original design.

[Scruffie]

Thanks for that, so it's just using one half of each CD4013? What an odd design choice... Definitely not using the chip handling oscillator division in parallel? That might make a little more sense in lieu of a clock buffer, given the BBD's have equal clock capacitance to an MN3007.

The BBD's were in series in the original which it did look like was the direction it was heading in the gut shots Jessenator sent me, is parallel confirmed?

Are you able to measure the clock frequencies as well out of interest?

If someone posts some decent gut shots it shouldn't be too hard to trace and confirm seeing as we already have the 117 schematic.

I'll PM you. Let me know if you need different angles than what I captured.

Thanks, unfortunately to get a good look at the traces you'd need to unsolder the pots.

But, I thought of a much simpler solution... just test for continuity between pin 2 of both MN3004, if it beeps, they share the same clock.

If it doesn't beep, check pin 2 of one doesn't match to pin 12 of the other.

Still no beep? Then it's time to test the CD4013 for continuity.

Probed around in mine. One half of the inner 4013 handles generating the two phase pulses for each MN3004. The  outer 4013 is then the other half of the clock circuit although I believe based on photos earlier in the thread that there is a single oscillator feeding the two flipflops. BBDs are in parallel like the original design.

[museums]

There are two 4013s in the pedal, each with two flip flops. One of the 4013s dedicates one flip flop to each of the MN3004s to make the biphase clock pulses so each MN3004 has its own driver. Im honestly not sure this is even necessary considering the clock freqs (13KHz to 500KHz), which are roughly halved compared to an SAD1024 unit or generally any Panasonic BBD equipped flanger pedal. The other 4013 is the oscillator. Of the 4 flip flops available, I believe 3 are used.

I was reading the old schem wrong about the BBDs being in parallel. The input pins of the two MN3004s are connected in my unit and the clock phases are matched which IIRC is the differential operation shown in the SAD1024 datasheet. Maybe the folks at MXR looked at the performance of the Panasonic BBDs and traded super short delay times for better performance with differential configuration? The sound of the pedal is spectacular. Better than an SAD1024 MXR IMO.

The only DIY project I have seen that offered these alternate BBD configurations is the BYOC flanger, but Im pretty sure they were running parallel multiplex and calling it differential.

Side note - The lag here is like an hour or two for posts to appear. Never experienced that on another simple machines forum.

Thanks for that, so it's just using one half of each CD4013? What an odd design choice... Definitely not using the chip handling oscillator division in parallel? That might make a little more sense in lieu of a clock buffer, given the BBD's have equal clock capacitance to an MN3007.

The BBD's were in series in the original which it did look like was the direction it was heading in the gut shots Jessenator sent me, is parallel confirmed?

Are you able to measure the clock frequencies as well out of interest?

If someone posts some decent gut shots it shouldn't be too hard to trace and confirm seeing as we already have the 117 schematic.

I'll PM you. Let me know if you need different angles than what I captured.

Thanks, unfortunately to get a good look at the traces you'd need to unsolder the pots.

But, I thought of a much simpler solution... just test for continuity between pin 2 of both MN3004, if it beeps, they share the same clock.

If it doesn't beep, check pin 2 of one doesn't match to pin 12 of the other.

Still no beep? Then it's time to test the CD4013 for continuity.

Probed around in mine. One half of the inner 4013 handles generating the two phase pulses for each MN3004. The  outer 4013 is then the other half of the clock circuit although I believe based on photos earlier in the thread that there is a single oscillator feeding the two flipflops. BBDs are in parallel like the original design.

[Scruffie]

There are two 4013s in the pedal, each with two flip flops. One of the 4013s dedicates one flip flop to each of the MN3004s to make the biphase clock pulses so each MN3004 has its own driver. Im honestly not sure this is even necessary considering the clock freqs (13KHz to 500KHz), which are roughly halved compared to an SAD1024 unit or generally any Panasonic BBD equipped flanger pedal. The other 4013 is the oscillator. Of the 4 flip flops available, I believe 3 are used.

I was reading the old schem wrong about the BBDs being in parallel. The input pins of the two MN3004s are connected in my unit and the clock phases are matched which IIRC is the differential operation shown in the SAD1024 datasheet. Maybe the folks at MXR looked at the performance of the Panasonic BBDs and traded super short delay times for better performance with differential configuration? The sound of the pedal is spectacular. Better than an SAD1024 MXR IMO.

The only DIY project I have seen that offered these alternate BBD configurations is the BYOC flanger, but Im pretty sure they were running parallel multiplex and calling it differential.

Side note - The lag here is like an hour or two for posts to appear. Never experienced that on another simple machines forum.

Thanks for that, so it's just using one half of each CD4013? What an odd design choice... Definitely not using the chip handling oscillator division in parallel? That might make a little more sense in lieu of a clock buffer, given the BBD's have equal clock capacitance to an MN3007.

The BBD's were in series in the original which it did look like was the direction it was heading in the gut shots Jessenator sent me, is parallel confirmed?

Are you able to measure the clock frequencies as well out of interest?

If someone posts some decent gut shots it shouldn't be too hard to trace and confirm seeing as we already have the 117 schematic.

I'll PM you. Let me know if you need different angles than what I captured.

Thanks, unfortunately to get a good look at the traces you'd need to unsolder the pots.

But, I thought of a much simpler solution... just test for continuity between pin 2 of both MN3004, if it beeps, they share the same clock.

If it doesn't beep, check pin 2 of one doesn't match to pin 12 of the other.

Still no beep? Then it's time to test the CD4013 for continuity.

Probed around in mine. One half of the inner 4013 handles generating the two phase pulses for each MN3004. The  outer 4013 is then the other half of the clock circuit although I believe based on photos earlier in the thread that there is a single oscillator feeding the two flipflops. BBDs are in parallel like the original design.

Ahhh, this is making much more sense, so they've parallel multiplexed (I assume unless there's an extra op-amp to achieve differential operation) and halved the clock to achieve the same delay times as the original probably, as you note regards SAD vs. Panasonic, to avoid the MN3004's heavy signal drop off at those higher clock frequencies and also so there's not a heavier high frequency roll off from running them in series.

No, parallel clocks was probably a little overkill but suppose it doesn't hurt, especially with the longer clock traces.

Also helps makes sense as to why the later version just uses a single MN3204 but with heavier noise reduction circuitry.

Think the mystery is solved, still, might be interesting to trace and see if there was any other alterations to the circuit... I do notice an extra diode in the astable section area.

Yeah, the forum does appear to be having some issues, there's a note on the homepage up at the top right.

[jessenator]

Yeah, the forum does appear to be having some issues, there's a note on the homepage up at the top right.

Yeah, it's been doing that a while now. Occasionally I get someone else's account on the main page...

Think the mystery is solved, still, might be interesting to trace and see if there was any other alterations to the circuit... I do notice an extra diode in the astable section area.

This is seriously cool. LMK if you need a donor. Not that there are any 3004s to build anew 

[museums]

Yeah, the forum does appear to be having some issues, there's a note on the homepage up at the top right.

Yeah, it's been doing that a while now. Occasionally I get someone else's account on the main page...

Think the mystery is solved, still, might be interesting to trace and see if there was any other alterations to the circuit... I do notice an extra diode in the astable section area.

This is seriously cool. LMK if you need a donor. Not that there are any 3004s to build anew 

The thing is you wouldn't need MN3004s because this version is clocked low enough that you could use MN3007s and clock it at normal freqs and get the same results.

[Scruffie]

Yeah, I'd only really trace it just for completion, as you say, no MN3004 and I don't think there's anything majorly exciting hidden in there, just an interesting little piece of MXR history.
If you're up to de-soldering the pots I'll have a look over though when I get a chance, not impossible they've tweaked here and there.

The MN3004's have a minute amount of extra high frequency content over the MN3007 but then you're just trading benefit in one area for some negatives in another... at the time Panasonic had only stopped producing BBD's a few years earlier so the choice of chip may have just been what was practical to source.

Plus, if you're really missing said high content, you can make R22 & R25 6k2 on the madbean collosalus and it should more than compensate for the difference.

[lars]

Plus, if you're really missing said high content, you can make R22 & R25 6k2 on the madbean collosalus and it should more than compensate for the difference.

Thanks for the suggestion! I'll have to try that out on my Collosalus (2022 version) build. I'm still not quite happy with the performance with it. I've noticed that Q2-Q5 have a huge effect on the sound, so I've been swapping several different transistors types in those positions. Right now it sounds pretty good, and has the most eclectic combo of transistors in those positions.