ROG 18 drifting voltage question ....

[MartyMart]

Hi All,
       Returned to an "Eighteen" build ( www.runoffgroove.com ) which has an odd problem
that I can't understand.
All works fine apart from Q1, which once set up at 4.5v drain ( tried numerous J201's ) VERY slowly
drifts down to 0v, over around 30 minutes  !!
Changed trimpot, it's a "known" good one, still does it and obviously ends up "splatty"
Q2/Q3 are fine and stable at 6.4 and 4.5v

My only change was to use an input cap of 68n and a 33pf cap from gate to ground at Q1.
Could this be the cause of the shrinking voltage ??

Wierd   and I'm stumped on this one, made worse by the fact that during first 15 minutes of use
it's sounding real good :-(

Marty.

[petemoore]

Wierd   and I'm stumped on this one, made worse by the fact that during first 15 minutes of use
it's sounding real good :-(
  Dunno except is sounds 'normal' or like what I get, not the voltage drift to 0v, I've never measured that, but Jfet Drain Bias drift...happens to me all the time! I don't know if it really happens, just about 1/2 consistantly because I only 1/2 consistantly want to admit and look for exactly that.
  But it does seem that once in a third or so Jfets, one just won't sit still, won't mind it's bias manners, and plain refuses to cooperate.
  I'm tending toward like 1 Jfet at a time or Mu Amps, I've never had a bias issue with a Mu Amp once biased. I've had lotsa so-so to plagued results stringing Jfets together...probably me, I'll start counting closer?...probably 10 Jfet strings, 0 survived.

[MartyMart]

BUMP !
Any ROG chaps have an idea ??

MM.

[petemoore]

 Wow that sounded harsh..
  But I have had trouble with Jfet builds, and need to say that I can't verify the newness of the electrolytics and trimpots used, these two questionable [because the trimpots are often plain funky/sensative...it just seems like vibration could set them to voltage 'bobbling', they 'scratch' when touched].
  And I should get Thor back up there and pick 'tone' Fets, then set bias on them, eliminating the trimpots, on this one I lifted the Electrocap, but it's waterfall continued unabated.
  Many of the Jfet circuits sounded superb, like Supreaux and Prof. Tweed, but just didn't make the PB selection process even though the light to med distortion tones were excellent, Supreaux got a noise thing like some of the other 'longer Jfet stage strings', probably something in my circuit was faulty. A shame because it sounded really great.
  Proffessor Tweed did his thing very well,  and it did tweed sound very well [light to med distortions.
  Judging from reports and soundclips, either 'lucky' or crafty Jfet S.amp builders seem to be getting super results...I'm still trying to figure this out.
  Dr. B has gotten great reviews, and soundclips...and it's a string of Jfets.   
  ...Almost enough to make me want to try a nice fresh build !
  Mu Amp seems to be 'springier', responding to various amplitude inputs with usable and variable results from harder distortions to basic Jfet Mu sounds. I've been opting for the 'place the fragments' approach using a Mu mostly as final stage to various preboosts including single Jfet stages / Mu / Opamp distorter etc.  Having a Mu in there makes a great before/after device to build a PB Dist sound around, sounds kinda ampy by itself, med gain w/boosted input, high gain with Distorter before, high gain with Distorter after. Always sounds perfectly biased, resonds to being pushed hard by still staying 'curved' but distorting harder, pushes other circuits hard well too. They always seem to turn out fairly high gain/low noise, and are a great tool for distortion tone shaping.
  This allows me to work the individual stages, tossing noisy, pesky ones.
  Sparkleboost, and some Mu type thangs borrowing voicing options from here and there hold the Jfet positions most solidly on my PB.
  So...could well be those trimpots making noise, I can't really think of anything else, and the sensativity of Jfet bias makes me think that could be much of the reason for the problems I was having.

[MartyMart]

Pete, sorry that was NOT intended to come off like that at all !!
I just didn't get much from your reply other than " my Jfet builds all drift"
and as I dont generally have that problem, I was just trying to identify what
I could do to MAKE Q1 stable  !!
I'll read your second post now

MM

OK, got that down, I've only had perhaps three "problem builds" from 15-20 Jfet emu's
some of which are my own "babies"
I like them a LOT, OK possibly not "vintage amp" replacers, BUT the Bassman and Classic 30
emu's come VERY close, enough to convince me and the Bassman owner of their merits !
I just can't figure this one, why te heck should it "bias" and then drift down to 0 volts over
a half hour or so period !!
My other two problem ones just wouldn't bias properly, once sorted, they stayed where they
should have been.
Recently I've used better trim pots ( a buck or so each ) which seem good, so it's not that.
Perhaps I'l just rebuild the whole of Q1 again, it's only three caps and two resistors, perhaps
a cap is " bad " and I dont have a way to check 'em .

MM.

[petemoore]

All works fine apart from Q1, which once set up at 4.5v drain ( tried numerous J201's ) VERY slowly
drifts down to 0v, over around 30 minutes  !!
  That's really wierd, I assume you socketted and swapped the Jfet?
  Drain drifting to 0v over time...hard to say...and the other transistors hold bias? I assume the S bypass cap is newer? I don't see though how anything 'reasonable' could be pulling drain all the way to ground!
  I hope that my addendums to the 10/0 remark are taken with less of a grain of salt than the remark, those Jfet Sims sound great, take a look at my reviews of them, but for one reason or other, I just have a Mu amp and a Sparkleboost on my PB. The reason for that is it's probably my builds and preferences.
  Sorry to take over your thread MM, I wish I could delete some late night typings sometimes...

[MartyMart]

Thanks Pete, yeah I use a couple of MU things, one of my own, a preamp off the net and a booster 2.5
which is a recent one, sounds real good and is STABLE too !
I'll figure it out, just sometimes it seems like  a real waste of time, debuggin a little
circuit for 1/2 a day !!

MM.

[gzed]

JFet termal runaway?   

[Ardric]

What's the voltage at Q1's gate when the drain voltage drops?  There's no DC blocking input capacitor on the Eighteen; some kind of negative DC bias voltage on the input, like a wonky capacitor in the pedal before it, could do this.  You could also try putting a 10k resistor and 100n cap in series with the input, before the 1M bias resistor.

BTW, note that you need a fairly high input impedence on your meter to check the gate voltage properly.  An old analog voltmeter stands a good chance of pulling the gate to ground during the test.  Maybe you could try powering up and playing through it until it drifts down, then for a moment short the gate to ground.  When you remove the short, is the drain voltage restored?

Another reason why drain voltages may refuse to settle down is oscillation.  It may be too high pitched to hear.

[MartyMart]

My input/Q1 is this :

Jack to board, 1M to ground ( switch pop resistor ) then 68n cap to another 1M to ground at gate
820 Ohm and 47uf cap from source to ground, trimpot to drain and 10n/4n7 from drain to the
tone control .... etc.
When Drain is 4.4v , gate is 1.2v and source is 0v

MM.

[Ardric]

When Drain is 4.4v , gate is 1.2v and source is 0v

Are you sure that's right?  We're looking for 0V at the gate and almost 0V at the source.  Pinout problem?

[MartyMart]

Correction source at 122mV !! 
ROG values are D- 4.61 G- 0 S- 118mV so at first it's very close ..... then !!
Another error, my input is 1M to ground -> 68n ->33p to ground -> gate NOT an additional 1M !
..... hmm perhaps I'm on to something, as ROGs original is just 1M to ground -> gate ??

MM.

[Ardric]

Ack!  You're missing the 1M bias resistor for Q1!  There ya go.  It needs to be on the gate side of the blocking cap, or there's nothing to set the DC voltage on the gate.

[MartyMart]

Ack!  You're missing the 1M bias resistor for Q1!  There ya go.  It needs to be on the gate side of the blocking cap, or there's nothing to set the DC voltage on the gate.

Thanks ... JEEZ that's SOoooo  obvious and I SHOULD know better  !!!!!



MM

[Ardric]

You're welcome!  We can all benefit from a second set of eyes sometimes.

It would've been more straightforward if you'd told us your circuit was modified from the ROG version.  You didn't mention it earlier in the thread, just posted a link to the official schematic.  If you'd posted your own modified schematic to start it would've been easier to help.

[MartyMart]

Now "stable" @ 4.54v 

"Yeah Baby Yeah"

[Paul Perry (Frostwave)]

sometimes it seems like  a real waste of time, debuggin a little
circuit for 1/2 a day !!

Only 1/2 a day?? I should be so lucky!!