BlackFire Debugging HELP!

[Samuel]

Hey there all. I got a working build of the blackfire. Love the sound, but it has a problem: When doing heavy palm-muting or really hitting the strings hard, the distortion seems to "back off" a bit, particular on the low E string. It's kind of subtle but definitely there.

I did a transistor bias check, and got the following:

Q1:
E: .467V
B: .901V
C: 8.64V

Q2:
E: 350 mV (yes, mV)
B: .890V
C: 5.16V

Q3:
E: .912V
B: 1.36V
C: 5.16V

Q4:
E: .780V
B: 1.345V
C: .956V

Q5:
E: .812V
B: 1.396V
C: 4.8V

So I'm guessing the problem is with Q4? Is the collector voltage supposed to be that low? I seem to remember that one of the trannies in this circuit was doing non-standard duty as phase reversal or something....

[petemoore]

voltages look too close on some of the Q's, base/emitter should have at least .7 V difference...as I recall ...HTH

[Joe Davisson]

My voltages:

Q1:
E: 1.0 v
B: 1.4v
C: 8.7v (9v rail)

Q2:
E: 0.4v
B: 1.0v
C: 4.3v

I suspect the problem is in the biasing for Q1/Q3. Try adjusting the 4.7M resistor. The collectors of Q2/Q4 should be roughly 1/2 supply.

I can't see how your Q3 value can be 5.16v if attached to the 9v rail? (mistype?)

The last inverter stage looks close enough.

If it still doesn't work, try these more conservative biasing values:
Q1/Q3: 3.3M/1M
Q2/Q4: Raise 1k resistor to 2.2k
Check Q2/Q4 collectors to be at 4.5v.
(The gain may be reduced slightly in this arrangement.)

I've tested old and new 2N5089's in this circuit, and even if the gains meter out exactly the same, the biasing can be off. I'm not sure why this is.

Sorry your having problems.. I'll be around tonight if you need more help...

-Joe

[Kleber AG]

Hi Joe!

Thanks for all your help...
Does the 2N5089 have big tolerance specs?
Or... Is it easy to damage them?

Because I've found some problems with them too, I've even thrashed some, I don't know if I've made some weird thing on them (lots of tweacks)... and got it damaged...

but it could be that if you do "something" to them, they could goes bad but still working??? Maybe not biasing right. If so I wonder how easy is to make them goes bad...  :?

I'm just wondering, because I really liked circuits that uses 2N5089...

Peace
Kleber AG

[Samuel]

Hey, thank you all for your help.

Joe - I did indeed mistype the collector value for Q3. I'm using a PCB layout I whipped up and I thought maybe I had put something (who knows what it would be!) between Q3 and the 9V rail.

So the collector of Q4 being less than the base is also a huge problem I'm assuming. I have no idea what is going on with that, there's just a 10K resistor between the 9V rail and Q4's collector.

Won't have time tonight to try out the suggestions, I'll let you know tomorrow night, though.

Also, here's the PCB I made, maybe this is the problem? I have little/no experience at this:


Thanks so much again,
Sam

[Kleber AG]

Hi,
Your PCB looks to me to have a problem...

You're taking the signal "to" Q2 and Q4 "from" the wrong pins at Q1 and Q3...

On Q1 the signal should enter at it's base (midle pin) and goes out from it's EMISSOR (just above 100K resistor) to reach the base of Q2.

But at your layout you wrong conected Q1's base direct to the Q2's base by the 10K resistor.

Hope it helps, someone else could take a better look at this to confirm?
Regards
Kleber AG

[Joe Davisson]

That's the problem. You can probably cut out small sections of the copper traces, and use short pieces of bare wire to fix it. The rest looks okay to me...

-Joe

[Samuel]

Whooooooooooooooops! :oops:

Man, am I glad I posted that. Serves me right for not triple-checking my layout! I'll make the fixes tonight and keep you posted.

Once again, this message board is an amazing amazing thing! Thank you so much you guys!

-Sam