SOLVED Found issue with Crunch Captain Deluxe

Finally got around to this one after having the PCB since last July. I try to breadboard everything first to see if there's any mods I'd like to do before involving any solder. I couldn't get it to work properly, dealing with terrible misbiasing issues. Right away I could see where I was supposed to be getting 9v I was getting closer to 8.5v. Still workable, but -9v was closer to -7.5v, and 18v wasn't even breaking 15v. I assumed it was something on my breadboard layout and trusted the board was correct. I finished soldering and was about to box it up today when I figured I'd just do a quick search and see if the schematic was available anywhere else to compare. Well, I found another forum that referenced this schematic and they pointed out how two of the polarized caps in the voltage regulator section are backwards - namely C102 that connects pins 2 & 4 of the TC1044, and C103 that's on the -9v rail pin 5 to ground. I put that together on my breadboard just now and viola, all voltages are actually correct.

The PCB needs to be updated, or at least a note in the build docs to flip these two caps around. Going to box it up today after I make these corrections and give it a spin.
 
Solution
Interesting. I have mine all wired up but haven't given it the sound test yet. Just checked voltages with C102 & 103 the "right" way and sure enough I'm getting -7v and 14.6v. Does your at least sound good? My breadboard build was all splatty and had horrible oscillation with the gain at noon and above. When I was getting the intended voltages on the breadboard it was only the charge pump section I'd put together - it wasn't connected to any other part of the circuit. I wonder what's draining it like that.
Interesting, because the schematic in the build docs looks correct, but I'll check with my own build.

Tbf, if you're using the protection diode, 39R current limiting resistor and 1N5817 as charge pump diodes, I would kind of almost expect the sort of voltage drop you're getting. I'd reduce it to 10R and go for BAT41/42/85 or similar (keep the protection diode 1N5817 though) and you'll get much closer to actual 18V.
 
Confirming that those 2 caps are the wrong way around!

BUT, flipping them around didn't change anything. Noting here that my build worked before with the caps being the wrong way around too. I really don't understand why it would work, as the datasheet says otherwise, maybe someone smarter than myself can chime in.

I was getting +15/-7.5V before and I'm still getting the exact same. No misbiasing issues either way, but if I was to guess why your breadboard build had some (apart from potential connection issues), it's potentially overlooking the fact that IC1 (the one used for the gain stage) is running at +/-9V, not +18/-9V.
 
  • Like
Reactions: fig
Interesting. I have mine all wired up but haven't given it the sound test yet. Just checked voltages with C102 & 103 the "right" way and sure enough I'm getting -7v and 14.6v. Does your at least sound good? My breadboard build was all splatty and had horrible oscillation with the gain at noon and above. When I was getting the intended voltages on the breadboard it was only the charge pump section I'd put together - it wasn't connected to any other part of the circuit. I wonder what's draining it like that.
 
Solution
Mine certainly sounds good. Make sure all your power connections on the breadboard are correct, because as I said, the 2 ICs go to different voltages and have difference Vref points. One of those being wrongly connectes is probably why yours sounded splatty and oscillated.

From experience, the "issue" is the protection diode and the 39R resistor after the 9V input. If you remove or decrease the resistor value you get closer to 9V, small value resistor without diode will get you 9V input easily and thus higher voltages from the charge pump. Also as I mentioned, BAT42 or similar schottkies drop less voltage and boost up to actual 18V, while 1N5817 and 1N4001 for example will usually be closer to 16-17V.
 
The build docs dated 4/17/22 are correct.

1N5817 is a Schottky and has a very low voltage drop. It is rated for the high surge currents present in the power section. The BATxx devices are small-signal diodes with higher resistance and will not perform as well as 1N5817. More importantly, they can be overstressed by the surge currents. The primary cause of low voltage on the rails is R100. I reduced it to 15Ω on mine. HOWEVER, there is more than enough headroom with the rails at -7V and +14.6V.
 
The build docs dated 4/17/22 are correct.

1N5817 is a Schottky and has a very low voltage drop. It is rated for the high surge currents present in the power section. The BATxx devices are small-signal diodes with higher resistance and will not perform as well as 1N5817. More importantly, they can be overstressed by the surge currents. The primary cause of low voltage on the rails is R100. I reduced it to 15Ω on mine. HOWEVER, there is more than enough headroom with the rails at -7V and +14.6V.
Build docs are correct but the silkscreen on the PCB isn't.
 
Thread derail, but sort of still related:

@Chuck D. Bones — I've run out of 1N5817 and have been subbing in 1N5818, closest thing I've got to the '17s. I looked at the data sheets, but can't fathom what difference in the real-world circuit it will make. I've also got a number of 1N5819...

PARAMETERSYMBOL1N58171N58181N5819UNIT
Maximum repetitive peak reverse voltageVRRM203040V
Maximum RMS voltageVRMS142128V
Maximum DC blocking voltageVDC203040V
Maximum non-repetitive peak reverse voltageVRSM243648V
 
Bottom line, if the applied voltage is <20V, they are interchangeable.

The 1N5817, 1N5818 and 1N5819 are are fabbed on the same equipment, at the same time, with the same recipe. The wafers are chopped up into die, then die are tested and sorted. The ones with a breakdown voltage >48V (plus some margin) end up in the 1N5819 bin. Of the remainder, the ones with a breakdown voltage >36V end up in the 1N5818 bin. Of the remainder, the ones with a breakdown voltage >24V end up in the 1N5817 bin. Anything with breakdown voltage <24V end up in the scrap heap.

This is common practice in the semiconductor manufacturing business. Bipolar transistors get sorted by HFE or breakdown voltage. BC109A, B & C are examples of HFE grading. JFETs get sorted by Idss & Vp. Microprocessors get sorted by speed. Yadda yadda yadda.
 
Just to follow up, I boxed mine up and I think it sounds excellent. A little noisier than my Russian-spec muff clone at higher gain settings, but not unusable. The EQ control on this is amazing. There are so many sound options I don't think I'll ever need another OD/distortion.
 
Back
Top