Welcome to Solid State Guitar Amp Forum | DIY Guitar Amplifiers. Please login or sign up.

March 29, 2024, 01:02:02 AM

Login with username, password and session length

Recent Posts

 

Boogie MKIV Lead channel JFET version, Oscillating like mad

Started by stormbringer, September 16, 2012, 07:20:53 PM

Previous topic - Next topic

stormbringer

Hi, in my preamp building adventure i'm building and trying out alot of distortion pedals right now to gain some inspiration and knowledge finding that "perfect" (for me) dist.

The latest build is

Problem is that i can't bias Q3 to anything near 5v, about 7 it stops oscillating, but then i got this bluesy crunch instead of the wanted raw distortion. Removed all the pots + wiring and resoldered them, using shielded cable for in/out atm. Also inspected tracks with magnifying glass, reheated just about every solder joint.

To get the oscillation to stop i can go about 2 ways.
1. Turn bias on Q3 to 7v or more
2. turn drive down VERY low.

Probing around i noticed that signal is fine At R5, and on lower side of R11. C9 is oscillating terribly, so is Q3 Gate, but i'm really having a hard time finding the culprit here.. Some of these Caps are Ceramic, as i did not find film capacitors of these values. Could these be the problem? (i read about microphonic effects etc while troubleshooting). i Feel kinda stuck here.

Checklist of what's done:
1, Short wires to pots, shielded and grounded at 1 point if crossing other cable, or board.
2. Shielded input and output Wires, not remotely near eachother.
3. Q3 is the only bias that makes the circuit oscillate, can bias the other FETS any value i want more or less without problems.

Roly

Can you post an actual circuit/schematic of this layout please?
If you say theory and practice don't agree you haven't applied enough theory.

stormbringer

#2
Absolutely, he has a schematic on the site aswell. Although Component Numbers do not match his vero layout. :/



Did some more testing today, built a standard buffer and wired up before this circuit (read that it helped some people with DR Boogey oscillation issues, and this circuit not being miles away i figured i'd try the same thing, but no difference. I'm absolutely positive that the first 2 stages works perfectly, but somewhere between 2 and 3 or at stage 3 something goes terribly wrong.

Loudthud

I would first try removing C6 (layout numbers). If the sound is too buzzy, try installing C6 from gate to ground across R11.

sowhat

Hi, I'm new to this forum but not new to FET preamps. From what you have said I suspect that Q3 is flakey or C17 is too large or a combination of both. If C19 shows strong oscillation (I'm using the schematic designations because the veloboard layout don't tell me much) then the problem indicates something amiss with the Q3 stage. These are cheap FET's and quality control may not be what you would expect with them even from reputable sources. C17 can really cause havoc if not sympatico with the FET. I have had to modify the cap values (usually making them lower) on similar stage before and this seems to stamp out oscillation problems... then again, who really knows until you really dig into things and do some "what if's".

sowhat

Oh yeah, I almost forgot. Changing the value of C17 will effect the biasing range as well and may bring you closer to the 5 volts you are looking for... all this stuff is interactive. If I got things confused with these reference designators, that's the same capacitor loudthud is talking about... he knows his stuff real good too and has helped me out of a tight spot on more than one occasion.

Loudthud

I don't really like Veroboards, I've never used one. That track that goes to Drive 2 goes half way across the board through the output section making it a good path for stray pickup and oscillation. If it's not too much trouble, cut that track around column 15 and attach the Drive 2 wire there.

Roly

Um ... wat?

It isn't that surprising that you have oscillation problems since the circuit shows a positive feedback path via R12/C15.

First thing I'd try is disconnecting these components, i.e. breaking the +ve feedback path, and see it this stabilises the amplifier.

If so you need to consider the loop gain.  If the gain around the loop can be greater than the losses then the loop will take off at that frequency.  You can reduce the gain of the loop and/or alter the phase behaviour around the loop, but applying +ve feedback is inviting instability.

Concentrate on getting your DC conditions right first before moving on to your AC conditions.
If you say theory and practice don't agree you haven't applied enough theory.

J M Fahey

Agree and add.
Didn't get into the Vero layout, but the schematic shows various errors, many of those show a beginner who copied what he *thought* he saw in the MK IV schematic.
Anybody can make mistakes, of course, but later common sense or experience make them stand out and beg to be corrected.
I have *big* doubts it's an acurate rendition of a MK IV, FET or otherwise.
1) as noted, the R12//C15 direct "clean" path makes no sense.
It stinks of the 3M3//20pF attenuator in a Twin  reverb channel, and maybe MB "borrowed" (to say it lightly) from that design, but there it has the useful purpose of mixing one channel with the other; here it's ridiculous and only offers an uncontrolled feedback path.
2) C16 and C25 are in parallel.
Whatever purpose in paralleling a 47pF cap with a 500pF one escapes me.
Also the C15 "brightness compensation" role loses meaning when said highs are grounded by huge 550 pF (combined) capacitance.
3) C2+R9 are *useless* being bypassed by R2/C1.
4) C24 is useless considering it then is grounded by much larger C25//C16.
5) C26/R22 are also useless.
6) Given the gross , still uncorrected errors shown here, I wouldn't trust Harald Sabro Bolstad instructions to make a cup of tea.
Sorry.
PS: won't even waste time checking the Vero layout.
I would only add that a very high gain preamp built in such a small board and without much consideration to grounding, layout and decoupling is a recipe for disaster.

stormbringer

Thanx for your input. :)

Guess i got hooked by the big "Verified"-stamp and thought it might actually work, but seeing the problems you point out here i think i'll actually scrap this one. There's some good things in this experience anyway, as i understand way more after your comments. :) Money/components spent on learning = good investment in my book.

I'm quite new, and not ready to design my own amps yet, but i'm aiming to get there, so i'm building lots of stuff found online and trying to put the pieces together. Again, thanx for your help. MUCH appreciated!

Regards
Fredrik

J M Fahey

Dr Boogie has been built by many, so I trust it more.
Try it ;)

stormbringer

Have actually built 3 Dr Boogeys. ^^

One with spare/scrap parts to test it, and later 2 with quality parts (1 for me, 1 for the other guitarist in my band).
Really like them, although a bit tricky tonestack, but i like trying out different circuits, you never know when you might stumble upon something awesome. :)

Roly

Quote from: J M Fahey
:grr
...
I wouldn't trust Harald Sabro Bolstad instructions to make a cup of tea.

Quite right.  I am altogether too trusting.  I did notice some strange things there but I must have been asleep, concentrating on the +ve feedback path.
:embarassed:

I've just had a look at the original valve Mk4 circuit at;

http://www.freeinfosociety.com/electronics/schemview.php?id=287

...and some of the crazyness can be explained by leaving all the switching out!  For example the apparent idiocy in the cathode/source circuit of the first stage makes a lot more sense when the three missing switch contacts are replaced.  I haven't gone through it in detail, but the FET circuit is supposed to be an emulator of the original valve circuit, except that omitting all the switching (and there is quite a bit) has turned the FET version into a nonsense.

What we seem to have is a valve circuit that was a bit odd to start with, which has been translated into a FET version by somebody who didn't really understand the valve version (and hence all the trim pots).  Avoid.
If you say theory and practice don't agree you haven't applied enough theory.

spud

Hey, try the Plexizer or Plexi-izer:

http://www.ssguitar.com/index.php?topic=2685.0

It's a Marshall Plexi (1959) emulator done in Jfets.  I built it on perf-board but I think someone's done a PCB template/mask for it.  Check DIYStompboxes, do a search for Plexizer and Mojotron (the user on that site that originally designed and built it). 

Jim

Roly

What is it with this reluctance to commit?  Another rank of trimpot drain loads (but at least this one doesn't seem to have a positive feedback loop  ::) ).

{I wonder if anybody would mind if I knocked out an atheist preamp, say for use by heavy metal and Alice Cooper cover bands?  8| }
If you say theory and practice don't agree you haven't applied enough theory.