OS debug... was working a minute ago...

[Brossman]

Hey guys,

Firstly, a schemo for reference... http://www.generalguitargadgets.com/pdf/ggg_osq_sc.pdf?phpMyAdmin=4a28f86a515b7883e7bc35a68d4e7b6d&phpMyAdmin=78482479fd7e7fc3768044a841b3e85a

Got the OS working on the board using RG's rule (something is better than nothing to get it going).  Managed to get SOME semblance of this subtle effect, but really like the idea of using it as a gritty boost... whatever, that what knobs and trimmers are for! (multi purpose, just not at the same time.  But that give me an excuse to build more!!)

In dealing with making a PS from scratch, I worked it out, getting a total of 14VDC (by the math), after the Bridge.  The RC4558D has a MAX supply of +/-18V, so we're good there.  I dont have 2n5457, so I've substituted ECG456 for Q1 and NTE466 for Q2 (the MAX V+ of those are 35V and 40V, respectively).

These may also help...
ECG456 - http://www.datasheetcatalog.org/datasheet/nte/NTE456.pdf
NTE466 - http://www.datasheetcatalog.org/datasheet/nte/NTE466.pdf
NTE468 - http://www.datasheetcatalog.org/datasheet/nte/NTE468.pdf

When BB'd, I was reading +2.44V from Q1's Gate to Q2's Source...or maybe it was Q1's Source and Q2's Gate, I can't recall now...  I read on another thread about this that I should read between +1.5<->+1.7V here and should adjust a trimmer (I used a 5k trim for R5 - 2k4). But maybe they meant the 10k trimmer. (the 10k trimmer I had in was bad, so removed it entirely. Didnt hear a difference with a 10k resistor, so I pulled the whole thing, haha)  In any case, When I turned down R5 all the way, it got my lowest V reading at +2.44V, while raising R5, raised the voltage up to...I think ~+8V.

In getting a sound I thought was workable (and boxable), I decided to lay it out on Perf.  But now, my BB is not running.  Any guesses as to why?  I swapped the original Caps I had for C3,5,6,7 (4.7, 4.7, 2.2, & 6.8uF; tantalum) for something closer (4x 4.7uF, standard Al-foil electrolytic).  Then I swapped the 3 resistors I was using to make a 390kohm for the bias on 4558, pin 3 for a real 390kohm resistor.  Did the same for the 82k.  And have used a 1000pF instead of a 2200pF for C2...dunno if that makes much of a difference outside bandwidth control (or freq, or w/e).

Now I get nothing.  HOWEVER, when I lift the 390kohm resistor from the +9V, the resistor acts like an antenna (especially when my skin touches the leads), and I'm getting very loud RADIO coming through my amp. In addition, lightly strumming the guitar puts soft signal through (lots of noise), then when strumming hard, I get NASTY, crackly, loud, distorted sound...

also, and very strange to me, I NOW am reading +9V on every lead of Q1.  I check pinout, it's good.  I even swapped the ecg456 for an NTE468 - same thing...

Any thoughts guys?

[PRR]

> I NOW am reading +9V on every lead of Q1.

This alone might explain "no output".

Trace from Q1 through R5 through R7 to ground, find the break.

> when I lift the 390kohm resistor from the +9V, the resistor acts like an antenna (especially when my skin touches the leads), and I'm getting very loud RADIO

Working normally for this AB-normal connection. 390K+470K sets the input of IC1 about halfway up the 9V power supply. Disconnect the 390K, the input falls to zero volts DC and the 4558 (most opamps) can't work properly that way. However if any excitement brings the input up a volt or so, that is amplified by 220K/10K or about 23 to make a very large distorted output. Hang a 6 foot 135 pound chunk of salt-water on there, standing in a typical urban radio area, you are very likely to pick-up and badly-demodulate radio stations.

Fix the wiring mistakes.

> I was reading +2.44V ...  I read ...I should read between +1.5<->+1.7V

This depends on the Vgs of the specific JFETs. Using "right" part this can vary 2:1 from part to part. Using NTE "equivalents" it may vary much more than that.

I'm not sure why you used two different parts for Q1 and Q2.... the plan suggests identical parts.

[Brossman]

I'm not sure why you used two different parts for Q1 and Q2.... the plan suggests identical parts.

Q1 and Q2 are performing two very separate functions in this schematic, as is evident by the external parts.  My understanding is that Q1 is acting as a constant voltage to bias Q2, but please correct me if I'm wrong.  Again, if my understanding is correct, based on this logic, Q1/Q2 do not need to be matched.  Identical parts are easier for a manufacturer to buy in bulk 

As for wiring issues, I'll trace out the board sometime today and get back to the board tonight with my results.  Thankyou Paul!

[PRR]

I did not say "matched". Same-type parts.

> Q1 and Q2 are performing two very separate functions
>  Q1 is acting as a constant voltage

Right; and the voltage needed, the voltage supplied by Q1, is a function of Q2's parameters.

Q2 should idle full-off, but just barely so. On signal it should turn full-on.

This full-off voltage will vary with part-number, specific part, and temperature.

If we had matched parts we could probably derive an elegant ratio of R5/R7 which would always work. We don't have matched parts, but we can have same-type parts which will track, so the R5/R7 ratio is adjusted.

Personally I think it is a wee bit clever. Q1-R5 "could" be replaced with a fixed resistor, trimmed to pull Q2 to the barely-full-off point. True, this trim might have to be re-set for New Years and 4th July outdoor shows; using a same-type JFET may offer temerature compensation.

But that's getting ahead of the game. It worked, it didn't work, it has unhappy voltages around Q1, Q1 is fairly tough, I'd bet on wiring/soldering issues.

[Brossman]

I'd bet on wiring/soldering issues.

I'd do the same, if it were off the breadboard...   but it's not so...

I've been through with a meter checking and rechecking connections being correct on the BB, but it seems I cant make heads or tails out of what I've got on it. Maybe I got it working with certain parts on the board and when I took them out it still worked because it was already "on"...?

Might I just lay it out on perf with sockets for the IC and FETs? maybe actually getting it to a board with the right parts (no subs) and a more workable layout would be better than struggling with wether my BB is contacting. (...which I suspect to be at fault for many of my issues here)

[PRR]

> struggling with wether my BB is contacting

Hmmmmm.