Egregiously Expensive Cruch - Custom Cornish CC-1 Clone, take 3

MattG

Well-known member
Build Rating
5.00 star(s)
Says the rest of the forum, "Sheesh Matt, haven't we seen this a few times already?"

Yes, but I finally got it right! I started with the Aion Cepheus. Then I made a custom PCB version of the circuit that had all kinds of problems, and though I got it working, it was too nasty to post a build report. I did post my second try, which also had issues, but not nearly as many as the first go.

So here's the most recent version, which worked perfectly on first power up!

...almost. The circuit is perfectly fine, but when I did the enclosure artwork, I forgot that I swapped the Bass and Treble knobs in the most recent PCB revision! I was playing through it, putting it through the paces, and thought, "Where's my treble? Did I make another circuit mistake?" I kept turning up the "Treble" knob, but definitely wasn't getting more treble... then I turned up the "Bass" knob - there's the treble, d'oh! But I consider that a cosmetic error, so no need to revise the PCB.

Anyway, this build has a few features that are either new to me or somewhat different from typical builds:
  • The stock circuit, as traced by Aion is in my opinion kind of weird in that it uses a single transistor buffer, a dual-opamp and a single-opamp. That single transistor buffer is used in many (maybe all) Cornish designs, it's been discussed here and on other forums. I don't have the tech chops to weigh in on the engineering behind its design, but all the other experts seem to agree it's nothing particularly special, just a bootstrapped emitter-follower. It's cheap to implement, but a fair number of parts. So my thought was, why not use a good ol' opamp buffer instead? Then we can use a dual-opamp in place of the single for no change in required PCB space, and an opamp buffer uses fewer support components than the transistor buffer, for a net saving on components and PCB space.
  • If you look closely at the top of the PCB, you can see a large-ish transistor with a metal backing. This is the Infineon IRFU5305PBF, a power P-channel MOSFET used to implement the polarity protection described here by RG Keen: Advanced Power Switching and Polarity Protection for Effects. This scheme is indeed more complex than the classic series 1n5817 that we typically use. The downside of the series 1n5817 is the small amount of voltage drop. Using the P-MOS scheme has no fixed voltage drop like a diode, only some resistance-based voltage drop due to RDSon of the transistor. So to minimize voltage drop, we want the lowest possible RDSon. I spent some time looking at P-channel MOSFETs on Digi-Key, and getting past the TO-92 package, into the MOSFETs I assume are mostly intended for high power/high current applications is where we find the best RDSon numbers. This IRFU5305PBF has an RDSon of 65 milliohms! It's capable of handling 30 amps - the metal backing is for mating it to a heatsink. It's in a TO-251 package, which actually isn't that much bigger than TO-92. File under deliciously overkill!
  • Since I saved approximately $600 by DIY'ing this circuit, I figured I had room left in the budget to splurge on a fancy footswitch. So when Antique Electronic Supply (aka Tubes and More aka @Amplified Parts) had a sale recently, I grabbed this Lehle Long Life SPST Footswitch. This is a ludicrously overbuilt switch, but I'm a sucker for overkill!
  • This is the first time I've used "teardrops" in the PCB design (see here for KiCad docs on teardrops).
  • The engage/bypass circuit is integrated on the board. This isn't new, but you can read about it here: Buffered Electrical Bypass Module.
  • The "blobs" around the wires coming off the board are bits of hot melt glue I used for strain relief. (For some reason I forgot to apply some to the LED wiring at the bottom.) Also not new, but I think it bumps the professionalism up a notch.

Now that I have a working PCB, I'm going to build the Cornish GC-1. Per Aion's trace, the GC-1 and CC-1 are the exact same circuit, but have about a dozen component differences. The GC-1 is much higher gain.

I have extra PCBs I'm happy to give away if anyone wants one, just PM me!
 

Attachments

  • IMG_1739.jpg
    IMG_1739.jpg
    417.9 KB · Views: 63
  • IMG_1740.jpg
    IMG_1740.jpg
    480.9 KB · Views: 63
  • IMG_1741.jpg
    IMG_1741.jpg
    416.6 KB · Views: 61
I never get tired of seeing a MattG build. Your treble bass mixup is a perfect advertisement for the MichaelW School of graphics though... 😆

Awesome build. I love the over the top choices to really make this one special.
 
Says the rest of the forum, "Sheesh Matt, haven't we seen this a few times already?"

Yes, but I finally got it right! I started with the Aion Cepheus. Then I made a custom PCB version of the circuit that had all kinds of problems, and though I got it working, it was too nasty to post a build report. I did post my second try, which also had issues, but not nearly as many as the first go.

So here's the most recent version, which worked perfectly on first power up!

...almost. The circuit is perfectly fine, but when I did the enclosure artwork, I forgot that I swapped the Bass and Treble knobs in the most recent PCB revision! I was playing through it, putting it through the paces, and thought, "Where's my treble? Did I make another circuit mistake?" I kept turning up the "Treble" knob, but definitely wasn't getting more treble... then I turned up the "Bass" knob - there's the treble, d'oh! But I consider that a cosmetic error, so no need to revise the PCB.

Anyway, this build has a few features that are either new to me or somewhat different from typical builds:
  • The stock circuit, as traced by Aion is in my opinion kind of weird in that it uses a single transistor buffer, a dual-opamp and a single-opamp. That single transistor buffer is used in many (maybe all) Cornish designs, it's been discussed here and on other forums. I don't have the tech chops to weigh in on the engineering behind its design, but all the other experts seem to agree it's nothing particularly special, just a bootstrapped emitter-follower. It's cheap to implement, but a fair number of parts. So my thought was, why not use a good ol' opamp buffer instead? Then we can use a dual-opamp in place of the single for no change in required PCB space, and an opamp buffer uses fewer support components than the transistor buffer, for a net saving on components and PCB space.
  • If you look closely at the top of the PCB, you can see a large-ish transistor with a metal backing. This is the Infineon IRFU5305PBF, a power P-channel MOSFET used to implement the polarity protection described here by RG Keen: Advanced Power Switching and Polarity Protection for Effects. This scheme is indeed more complex than the classic series 1n5817 that we typically use. The downside of the series 1n5817 is the small amount of voltage drop. Using the P-MOS scheme has no fixed voltage drop like a diode, only some resistance-based voltage drop due to RDSon of the transistor. So to minimize voltage drop, we want the lowest possible RDSon. I spent some time looking at P-channel MOSFETs on Digi-Key, and getting past the TO-92 package, into the MOSFETs I assume are mostly intended for high power/high current applications is where we find the best RDSon numbers. This IRFU5305PBF has an RDSon of 65 milliohms! It's capable of handling 30 amps - the metal backing is for mating it to a heatsink. It's in a TO-251 package, which actually isn't that much bigger than TO-92. File under deliciously overkill!
  • Since I saved approximately $600 by DIY'ing this circuit, I figured I had room left in the budget to splurge on a fancy footswitch. So when Antique Electronic Supply (aka Tubes and More aka @Amplified Parts) had a sale recently, I grabbed this Lehle Long Life SPST Footswitch. This is a ludicrously overbuilt switch, but I'm a sucker for overkill!
  • This is the first time I've used "teardrops" in the PCB design (see here for KiCad docs on teardrops).
  • The engage/bypass circuit is integrated on the board. This isn't new, but you can read about it here: Buffered Electrical Bypass Module.
  • The "blobs" around the wires coming off the board are bits of hot melt glue I used for strain relief. (For some reason I forgot to apply some to the LED wiring at the bottom.) Also not new, but I think it bumps the professionalism up a notch.

Now that I have a working PCB, I'm going to build the Cornish GC-1. Per Aion's trace, the GC-1 and CC-1 are the exact same circuit, but have about a dozen component differences. The GC-1 is much higher gain.

I have extra PCBs I'm happy to give away if anyone wants one, just PM me!
I'm going to have to dig out my Cephus. It was one I built very early in my pedal building journey. I remember liking it a lot. But it's been years since I messed with it.
 
If you look closely at the top of the PCB, you can see a large-ish transistor with a metal backing. This is the Infineon IRFU5305PBF, a power P-channel MOSFET used to implement the polarity protection described here by RG Keen: Advanced Power Switching and Polarity Protection for Effects. This scheme is indeed more complex than the classic series 1n5817 that we typically use. The downside of the series 1n5817 is the small amount of voltage drop. Using the P-MOS scheme has no fixed voltage drop like a diode, only some resistance-based voltage drop due to RDSon of the transistor. So to minimize voltage drop, we want the lowest possible RDSon. I spent some time looking at P-channel MOSFETs on Digi-Key, and getting past the TO-92 package, into the MOSFETs I assume are mostly intended for high power/high current applications is where we find the best RDSon numbers. This IRFU5305PBF has an RDSon of 65 milliohms! It's capable of handling 30 amps - the metal backing is for mating it to a heatsink. It's in a TO-251 package, which actually isn't that much bigger than TO-92. File under deliciously overkill!
I'm going to have to switch to these, I was using VP3203N3-G in a TO-92-3 package and they're more than double the cost of the IRFU5305PBF.

And awesome work!
 
I'm going to have to switch to these, I was using VP3203N3-G in a TO-92-3 package and they're more than double the cost of the IRFU5305PBF.

Yup, more expensive and higher RDSon. A lot of those power mosfets are in the TO-220 package, which is huge (for a guitar pedal anyway). But I read the specs for TO-251 and saw it's barely bigger than TO-92. I even did an image search because I was initially somewhat incredulous. In terms of PCB footprint, I think the TO-251 might actually be ever-so-slightly better. Plus if you ever want to build a pedal based on an arc welder circuit, you're good to go! ;)
 
Yup, more expensive and higher RDSon. A lot of those power mosfets are in the TO-220 package, which is huge (for a guitar pedal anyway). But I read the specs for TO-251 and saw it's barely bigger than TO-92. I even did an image search because I was initially somewhat incredulous. In terms of PCB footprint, I think the TO-251 might actually be ever-so-slightly better. Plus if you ever want to build a pedal based on an arc welder circuit, you're good to go! ;)
Don't tempt me. It sounds like a Neil Young and Crazy Horse themed pedal to me: Arc Weld Fuzz.
 
I forgot about that.

Uhm, it's what I used to do... apparently I still do it. It got under my skin.
Mind, it's always easier to proofread other people's stuff than it is your own.


You've definitely progressed. I think I've regressed — but I digress...
 
Back
Top