JFET Preamp - Is this oscillation?

[statzern]

I just finished up an Orange Peel preamp, and testing it without an enclosure and pretty long leads (~4 inches), I get a pretty good tone from what I can hear over the noise! The noise is a combination of buzz and the typical sound of unshielded leads with a sort of whooshing sound and a low frequency hum somewhere around 500k-1k hz that changes pitch with turning pretty much any of the knobs.

I am using 2n5458 jfets, biased to about 4.5 v.

When I read about oscillation, it's usually a high pitched howling or screeching sound. This is much more of static-y droning sound. I am hoping that when I shorten up the leads, use shielded cable for the ins/outs and box everything up it will fix things, but does this sound like anything you have heard before?

Thanks all!

[GibsonGM]

Oscillation can be any frequency...sometimes it's so low in freq you really don't hear it; that's called "breathing"!  Yup, common on unshielded hi-gain builds not in the box yet.

I don't think you can make real judgement on what's causing it til it's in a grounded enclosure...tho maybe the hum could indicate some sort of ground problem.   Clean it all up, see what it does!   Does it oscillate at low gain settings?  If it does, that's an indicator of something wrong; try moving wires around if it does, see if you can locate a trouble spot...but again, probably moot until it's in its enclosure.

[temol]

Oscillation can be any frequency...sometimes it's so low in freq you really don't hear it; that's called "breathing"!

How to distinguish hi-gain circuit noise from very high frequency oscillation?

T.

[GibsonGM]

You could use an oscilloscope.  You should be able to see it with a scope!

[amptramp]

If you have a high-frequency oscillation, keep in mind that the oscillation may not continue all the way through the audio waveform.  You may get oscillation, for example, only when the audio signal is going down but is above the nominal zero point of the waveform.  I will second the need for a scope.  Used ones are cheap enough.

[statzern]

Sorry for the delay, but I have new news re: this problem. I finally got the circuit all boxed up neatly and am having continued issues with massive levels of noise. The noise I am having is basically white noise/static with a kind of crackling that is sporadic. This is what I have tried so far:

-- I used shielded cable for the input from the jack to the 3PDT and to the input on the PCB as well as the output from the PCB to the master volume pot and to an IC buffer that is in the output
-- I removed the buffer from the circuit and with the output straight from the master volume pot the noise is still there
-- I used an audio probe to probe each of the 5 JFETs at their gates - the first two are very quiet, the third has some low level static and the 4th and 5th are crazy loud static.
-- I started with 2N5458s in all 5 positions, biased to around 4.5 volts. I switched out JFETs 3, 4 and 5 with J201s with no change.

One thing I am thinking about right now is that the board may have been a bit oxidized when I soldered the components - I put it together about 2 months after etching and though I had used liquid tin with the board after etching, it was not very easy to solder to and did require more time per pad than is usually necessary. Because of this, I am thinking that there could be a cold joint(s) or other poor connection causing this. Reflowing all joints may help but will be a huge PITA now that it's pretty socked in with all these off board wires!

Any insight would be amazing at this point. If I do reflow all the joints and am unsuccessful, it would seem a whole new board may be needed to salvage this project...

[anotherjim]

If you do reflow the solder, it might be a good idea to apply some non-corrosive paste flux. It will help soldering and cool things down, since pad lifting is a big danger when you do any such rework.

The noises you describe can be...

Bad potentiometers. Give them a squirt of contact cleaner and exercise them. I'm finding grease contamination in a lot of new pots lately.

Bad capacitors. Electro caps usually. Either installed reverse polarity or they are really old stock and have lost polarization "forming". In the latter case, just leaving it powered up overnight can cure them as the circuits DC will re-form the cap. Tantalum electros can just be crap anyway - I never use them in signal paths.

Tarnished leads on old components (or were stored in damp conditions). Your solder joint can look fine, with a bright meniscus of solder looking just like it should, but there's an unreliable hi-resistance contact hiding in there.

[PRR]

> ... noise I am having is basically white noise/static with a kind of crackling that is sporadic.
> One thing I am thinking about right now is that the board may have been a bit oxidized

Sure sounds like a bad contact. If bright light and strong magnifier do not reveal a not-quite joint---

Take ALL the solder off (heat it up and bang it). Scrape/scrub every pad AND lead. Use GOOD flux, a hot iron, good technique, and re-solder everything.

[statzern]

Thanks guys, I am working on reflowing every joint as we speak. Will report back with results!

[statzern]

I reheated every solder joint in the pedal, carefully inspected for bridges, gaps and cold joints and got everything put back together this afternoon with similar results. What I am finding now is that the trimpot in the second stage is extremely noisy and adjusting it makes the crackling way worse. I am pretty sure I have found the culprit. Will replace and see how it goes. How frequently is anyone else finding bad (new) trimpots? This one is a Cermet 25 turn 250k trimpot from Tayda.

Hoping this will fix!

[statzern]

Quick update: I replaced C2 and C3 (on the board layout, which are actually C22 and 23 on the schematic), all trim pots and all capacitors in the tone circuit (to save on space since they were previously mylar caps, now polyester film box caps). Unfortunately, I am still having a massive amount of white noise coming from this circuit, despite shielded cables, neat layout etc etc.

One thing I am having a hard time understanding is the purpose of C22, 23 and 24 on the schematic below. Putting a cap between 9v and the drains of Q1-3 seems like it would let any power supply noise right through, unless drained by the 100uf C21. These caps are not present on the Citrus Graphic Mk2 (http://tagboardeffects.blogspot.com/2013/03/citrus-graphic-mkii.html), or any other JFET preamp that I know of for that matter. Maybe removing them would help?

Before all this, I did try using a battery for power with similar results which makes me think simply removing C22, 23, 24 might not be the answer to all my problems.

Anyone with experience with the Orange Peel or any insight here would be greatly appreciated.

[anotherjim]

It would help to know the value of those 3 caps.

If small pF value, they are probably meant to go Drain to Gate to reduce HF noise using negative feedback.

Because class A stages have poor noise rejection, sometimes you have local supply bypass caps at the load resistor + end to ground, but they would be uF value and be more effective with a series resistor (100R to 1k range) in the + supply feed.

[statzern]

It would help to know the value of those 3 caps.

If small pF value, they are probably meant to go Drain to Gate to reduce HF noise using negative feedback.

Because class A stages have poor noise rejection, sometimes you have local supply bypass caps at the load resistor + end to ground, but they would be uF value and be more effective with a series resistor (100R to 1k range) in the + supply feed.

Thanks for the reply - C22, 23 and 24 are 1nf, so kinda intermediate value but still 100,000 times less than C21 (100uf)

[PRR]

> It would help to know the value of those 3 caps.

AND the drain resistors.

It's ALL relative.

We now know the caps are 1nFd. Not waiting for revelation on drain resistors, I assumed a popular 50K trimmer turned say mid-way, 25K. This gives around 6KHz, the top of the guitar band.

Probably to take the high-end harshness off the distortion.

Supply should not have much 6KHz noise so that is not an issue. However they will work just as good drain to ground.

Note that the high-cut may vary radically if the "50K trim" has to be set to one extreme or the other to get good DC conditions. Maybe 3KHz to 30KHz. IMHO the choice of DC conditions should not interact with preference for high-cut, but there it is.

EDIT-- I slid my monitor and saw the note for three 100K trims. So the high-cut could start even lower, or still be way out.

[statzern]

Hey guys - sorry for the lack of info -- I have attached a board layout that might make more sense. I am not sure I am completely comprehending what you are saying - that the caps should go drain to gate makes some sense, but this is an OLC board -- unless all his boards are wrong, I don't know if changing the circuit is what needs to be done. That said, the same cap can be found on the third gain stage of the real deal Orange amp, as seen here: http://www.kbapps.com/audio/schematics/tubeamps/orange/graphicmkII.php

I am thinking about taking the first two out and will just see how that goes. I am getting this at the gates if Q2 and Q3 too, checking with an audio probe so, it is present early on, just not as loud.

[statzern]

Took out all 3 of those caps with no change — think it's time for a new PCB, I don't know what else to do at this point otherwise.