AMZ Mosfet Booster trouble (lend me your eyes)

[knutolai]

OK so I put this thing together with a few modifications and....... it does not work  

Some notes on additions and changes:
-I didn't have any 2k7 resistors so I matched and put some resistors in parallell giving me 2k9 for R4 and R5.
-I put a switch on the pot as suggested in the AMZ article so I could use it as a booster and a buffer. I intend the device to always be in the chain, so no bypass switch was added.
-R2 is swapped for a 100k trimmer and "A" is set to be exactly half my supply voltage.
-D2 and C6 is your usual polarity protection and power filtering.
-R7 and R8 was added for circuit protection.
-The zener is a BZX55C15 15 volt zener. According to this (http://www.diystompboxes.com/smfforum/index.php?topic=106637.0) thread  a 15v zener should work for this circuit.

Now I've looked over this thing a billion times. Checked every lead, off-board wires, jack-cables, diode polarity, tried 2 different BS170s, my power supply... and everything seems to be a-okey, but the only sound I get is a faint high-pitch whine.

any suggestions on what the problem might be?

Here's the schematic:


and here's my PCB layout:

[GibsonGM]

Q: how are R7 and R8 for "circuit protection"?  They constitute a voltage divider, no?  I 'get' the pulldown thing, but...

I have always used the 9.1V zener, and had NO problems. This is my go-to booster, and I use it as drawn by Jack.  I've had a few of 'em, made 'em for friends, etc....never had an issue.   Figured, if it ain't broke, and he went to the trouble of setting it up for us, why tweak it?

All I can suggest is to audio probe it from the input, and also to make SURE that YOUR BS170 pinout is what you think it is.  I understand they can vary somewhat by manufacturer, so you may need to take a magnifying glass and go look it up (?)

[knutolai]

I 'get' the pulldown thing, but..

To be honest R7 is just something I've adapted from other circuits and kinda forgotten the purpose of. If I remember correctly its for static protection of the semiconductor (I might be totally off here). You can see it in all of the ValveWizards pedals.

Ill have a closer look at the BS170s, Thanks for the tip! It's wired correctly according to the datasheet provided by the seller (taydaelec),  but the model might deviate..

Figured, if it ain't broke, and he went to the trouble of setting it up for us, why tweak it?

Well I wouldn't think of the voltage divider, power filter and reverse polarity protection as substantial tweaks. Swapping D1, R4 and R5 was due to parts I don't have in my possession

Edit:
Its now working!.... at least on one of my amps. Turns out it had to do with the trimmer, which is kinda odd...
According to AMZ:

The dc voltage measured from point A on the schematic to ground should be 4.5v to 5.5v with the circuit idling and no input. The value of R2 should be tweaked until the point A voltage is correct. If two or more of these circuits are mounted in the same enclosure, only one R1/R2 network is necessary and the R3 from each stage can be connected to the Vr point.

This is what I did at first, but it required R2 to be way lower than 62k (closer to 41k), so instead I set the trimmer to exactly 62k which gave me 6.3v at the Mosfet Drain.

The sound I get with my Ibanez amp is as advertised. With my DIY Ruby amp on the other hand: with the SW closed is kinda warm soundig/lightly overdriven. With the SW open (buffer mode) the sound remains at unity level, but is crudely distorted. Almost starved sounding.  

[duck_arse]

and here's my PCB layout:

I don't follow the red jumper on your board. (I now see it's the switch connection.) is that the earth link for in/out, or is that link at the jacks?

I woulda slung a 200~250k trimm between 9V+ and ground instead of resistor and trimmer. you can try trimming by ear, see if the sound improves. R7 looks like it might be a current limit, like for an opamp output. the bs170 pinout is opposite of the 2n7000, which catches some out.

[GibsonGM]

Glad to hear it works now, Knut...yes, they can be tough to dial in, I have heard - all the ones that I've built have been easy, but others have written here with the same problem you describe. Like Duck says, maybe bias by ear?  Then it probably won't be good with your other amp, though!!
I don't know WHY that would react badly to the Ruby  (?)....as long as you have an input and output cap, it should be isolated....

[PRR]

> how are R7 and R8 for "circuit protection"?  They constitute a voltage divider, no?

Not really a voltage divider. It is only 0.999 reduction; and you could put R8 on the other side of R7 for "no" reduction.

As you say, R8 is a pulldown.

R7.... if R6 is zero, and you have a long cable on the output, and you feed big high-frequency in the input, the transistor can die instantly. Shocked the heck out of me the first time I did that. I had a dummy-load of "300 feet of cable" and fed 50KHz in to get full-swing out. And Square Wave. (Yes, this was an abuse-test.) Stopped working. The transistor will try to discharge all 10,000pFd of capacitance in a couple nano-Seconds. The peak current can be many Amperes.

While there are many ways to defend against this (and it is an unlikely abuse), often it is possible to pick a series resistor which prevents ABSURD current yet does negligible harm to signal. With 1K under 9V you can't have more than 9mA in the transistor. And 1K driving typical 50+K guitar-cord loads is a 0.98 divider, 2% loss, in this case on top of an uncertain yet adjustable gain-- insignificant.

> half my supply voltage

May not be optimum because of the large voltage across R5. At a glance, 2/3rd is closer to right. Maybe more for MOSFET drops. Up or down according to your flavor taste.

[GibsonGM]

I knew the 'divider' ratio was such to be almost inconsequential, but I did NOT know that, PRR....  :oO      Nice abuse test!!  And a good protection, then.