Neovibe: Depth Mod - Q11, R40

[chi_boy]

Well, I built a Neovibe and it works great and sounds fantastic.  Of course, being a compulsive tweaker, I just had to screw it up.  Well, sort of anyway.

I read several posts by the experts that both said that an N-channel JFET could be subbed for Q11.  Specifically 2N5484, and 2N5485 were identified as good subs.  So I ordered some to try.

I put in the 2N5484 and left R40 at 2M2 and on power up the bulb went bright, as if on full depth and speed, and then went dim.  After that the LFO was dead.

I replaced Q11 and Q12 with 2N5088's and the LFO was fine and worked as it did before.

I then put the 2N5484 back in, and changed R40 to 22M.  Again, bright then non functional.


So, it's not broke, I'm just doing something wrong.  But what?

I did take a chance and did assume that this was a pin for pin drop in.  So I used C -> G; B -> S; and E -> D.  This was the configuration that dropped right into the space on the board.  I'm now assuming that was pretty much wrong.

The Fairchild 2N5484's I have are GSD as viewed from the bottom, flat side down.  Is it a pretty good bet that I need to twist the leads around and match C -> S; B -> G; and E -> D? 
Or would it be C -> D; B -> G; and E -> S


    ->   

Thanks for any help and insights.

Cheers,
George

[R.G.]

I did take a chance and did assume that this was a pin for pin drop in.  So I used C -> G; B -> S; and E -> D.  This was the configuration that dropped right into the space on the board.  I'm now assuming that was pretty much wrong.

C = D
B = G
E = S

This is one place a J201 is really good for, because its Vgs is quite low. You want a low Vgs device. 2N5485 is good, a few others. 2N5484 is a much higher Vgs and MPF102 can be really bad - Vgs can be as big as 9V.

[chi_boy]

R.G.

The J201 did the trick.  I bent the pins and turned the body in the opposite direction to line everything up. 

It does oscillate, but it is still a little funny.  For now I'm just looking at the light, but with the 22M for R40, the LFO stops as the speed is turned up and then recovers and restarts 30-45 seconds after the speed is tuned down.  With a 2M2 for R40 It will reach top speed without stopping.  I'm going to play with R40 and see how high I can go without stopping the LFO.

The cycle also seems to have more dark time than before with the 22M in place.  With 2N5088's and a 2M2, the cycle would only go to dull orange on the dark swing.  I liked the way it sounded like that so I will play with the lamp driver to see what  can do to fine tune it to the modified LFO.  Right now I have a TIP122 for Q13, and R50 is empty.

Thank you again for the swift and sure help!!

-George

[R.G.]

For now I'm just looking at the light, but with the 22M for R40, the LFO stops as the speed is turned up and then recovers and restarts 30-45 seconds after the speed is tuned down.  With a 2M2 for R40 It will reach top speed without stopping.  I'm going to play with R40 and see how high I can go without stopping the LFO.

Good call. I'd swap in 4.7M and 10M and see what happens.

[chi_boy]

This is just a follow up for anyone that may come this way in a search.

I did go back and try the 2N5484 and 2N5485's pinned as R.G. specified above.  In my unit, the LFO would not start with either.  With the J201 it was fine.  The best I could do with R40 is 5.1M.  Anything higher and the LFO would peter out at high speeds.  With a 7.5M in place the LFO would actually run and then slow to a stop after about 60 seconds or so.  The 5.1M seems much more stable.

I also swapped R47 for a 47k + a 100k multi turn trimmer to get more adjustment range on the bulb.  Soldered the leads St. Louis Arch style.

So now the LFO is in good shape but now I am chasing a slight distortion in the signal path.  But that's a subject for a different post and only after more research.

R.G., as always, thanks for your help and support.

Cheers,
George

[chi_boy]

So now the LFO is in good shape but now I am chasing a slight distortion in the signal path.  But that's a subject for a different post and only after more research.

No real need to post this, but the distortion is fixed based on this post:

http://www.diystompboxes.com/smfforum/index.php?topic=70757.0

I pulled Q1, Q2, and Q3 and replaced them with 2N5088's from a different brand and the distortion was gone.  I did do the "Fat Finger" test and the distortion decreased considerably when I touched those three.  I can't tell you why it fixed it, but it did.

Just leaving breadcrumbs for others.

-G

[R.G.]

If you're really into messing with this, remove C4 and paste in a 30pF cap from the base of Q2 to its collector instead.

[chi_boy]

R.G.

I'm into it this far, I may as well try that too. Can you give me an idea what to expect? 

Please note I'm not a techie, but will relay what info I can.

-G

[R.G.]

What you'll probably get will be - no difference at all. 

The only reason I mention it is that you had "distortion" with some transistors and not others. That description could possibly be because one set oscillated at RF and caused a bias shift that made the audio distort, even if you couldn't hear the RF. In doing some circuit work and simulation, I found that C4 does not improve stability as I thought it would. It actually makes stability worse - it puts a big peak in the response out at 1MHz to 5MHz, depending on the transistor. Removing it removed the peak, and using a 30pF feedback cap on Q2 made no peak at all and gave much more phase margin. C4 seems to have no effect in the audio range at all.

Changing it to 30pF feedback has to be classed as experimental, as the hundreds and thousands of univibes and clones generally have C4 = 330pF in there.

Just a thought, while you're experimenting.

[chi_boy]

OK, I don't know if I'm crazy, wishful thinking, or just plain tired of futzing with this thing, but I think it may sound a tad bit better with the 33pf in place of the 330pf.

Last night when I was chasing the really bad distortion and I was listenting to it after swapping the trannies, I thought there might just be a hint of distortion still coming through.  But I was tired and ready for bed and I thought it might just be the strings rattling or something from the guitar.  After this change that little bit of "character" seems to be gone.  Don't get me wrong, it sounded fine with the 330, but the 33 seems a little better.

I can still make it distort if I really try, but I have to really strum hard to do it.  Whatever the case may be, the swap certainly did nothing to hurt the sound.  It would be nice to A/B the change, but I went ahead and soldered the cap in place.  Right now I see no reason to take it back out.

I've seen your suggestion about this cap in other posts.  Has anyone else actually tried it?

-G

[R.G.]

I've seen your suggestion about this cap in other posts.  Has anyone else actually tried it?

Only me that I know of.

[chi_boy]

I've seen your suggestion about this cap in other posts.  Has anyone else actually tried it?

Only me that I know of.

And ME!!    


You know, the Vibe is insanely challenging.  There are sooo many things to go wrong.

I thought it sounded good the night I got it working and swore I wouldn't mess with it.  But it is almost as if the circuit needs to be massaged to bring it to it's best and to compensate for technology, or maybe it's the lack of.  I really don't know why that is, but thanks to the breadcrumbs left by others it's possible to really fine tune this thing.

Thanks again for an awesome and challenging project.  And of course your generous support.

-George