86-Special ready with noise problem

[10thTx]

Mat,

You are a VERY gifted guitarist and a great builder with superb innovative ideas. And I appreciate you sharing your talent and creativity.

I am cheering for you and I trust you will get this to work out just fine.


Best regards and respect, 10thtx

[mat]

Thanks Jeff for the praising (I'm sure, not deserved!) words.

Tillydog, few new pics at: http://www.flickr.com/photos/fuzzface19 ... 620310996/

I removed the extra solder and retouched almost every joint on the amp.

At work right now. Will test the amp later today..

[mat]

Think I got it. I tested the amp with low volume and it seems to be ok !

I'll test it better the day after tomorrow.

The bad part is that I dont know now what was wrong with it..

Will get back.

[mat]

I tested the amp with louder volumes. No noise on clean channel at all and little (bearable) noise on od channel. The od volume jumps really quickly so I'll check the curve/condition of the pot for more controlled volume sweep.

The sound seems quite tight. Any tips for loosening/getting more compression out of the amp ? Bigger screen resistor (now a shared 1k) ?

I have to tweak the OD section for better drive sound. The sound now is too fizzy.

I'll will also change the bright cap for more crisp.

Just noticed, I have the bright cap on VOL1 on wrong side of the pot..

[mat]

Ok, the amp is lots brighter with the bright cap on right side of the pot

I'll start to tweak the od section for smoother overdrive. Think I start with a feedback loop at V1b and them focus at the EF86.

Any suggestions for smoothening the EF86 ?

Any harm trying bigger screen resistor on the EL84 grids ?

Thanks,

[10thTx]

Mat just tossing out things to consider. I have NOT used an EF86 but have used the 5879 and both of these options with good results.

1) lower cathode cap value

2) smoothing cap across plate resistor and smoothing cap plate to ground

3) add a triode CF or a mosfet CF which was the favored solution for me.

With respect, 10thtx

[Tillydog]

Glad to hear you're making progress.

Any suggestions for smoothening the EF86 ?

Don't be afraid to try a really big grid stopper - I went to 560k (where the 150k is in 10thTx's diagrams) with an EF86 as stage 2. You don't lose top end like you would with a triode, because there is very little Miller effect, so the input capacitance is very much less than a 12AX7, say.

It had the biggest effect of the fizziness of the EF86 distortion of very many things I tried with biasing, etc. (FWIW, I did have a MOSFET follower after the EF86, too, but it was fizzy/harsh until I added the huge grid stopper).

I suspect that the huge cathode bypass cap above was aimed at the same thing - smoothing the effect of driving the EF86 grid positive. That circuit looks to have the EF86 biased very warm, so I think it will be driving the grid positive almost from the get-go.

[mat]

Thanks Jeff and Tillydog for the suggestions, I'll try them asap.

How would You think upping the EL84 screen resistors. Any relief there for the stifness ?

[10thTx]

I often will use 2.2k screen resistors on power tubes. They give a more compressed tone than say a 470R to 1k, IMO.

10thtx

[mat]

I often will use 2.2k screen resistors on power tubes. They give a more compressed tone than say a 470R to 1k, IMO.

10thtx

Thanks Jeff, I'll put a pair of 2k2 in. The clean channel needs also more chime or more 'juice' to the playing feel.

13h work tomorrow, but I'll try to throw them in at coffee break

[mat]

Ok, much better but not complitely satisfied with the od. Have to test it with proper volume. The VOL2 jumps at very high level very soon. I did cut the connection between VOL2 pot and (od) tone pot.

I changed the EL84 screen resistors to 2k2's.

Changed EF86 grid resistor to 470k.

Did put 220p and 470p caps on the spots Jeff suggested.

I would like to try the mosfet option. Where is the output on that option ?

[Tillydog]

I would like to try the mosfet option. Where is the output on that option ?

On the diagram above, it's where the 120K resistor meets the MOSFET source (think cathode follower).


Side note: When I tried to use a follower like the above to drive a tone-stack from an EF86, I ran into problems because the huge voltage swing available from the EF86 means the cathode follower needs a significant current output capabilty to cope with the capacitive load. It was OK driving the load positive, but the MOSFET was cutting off completely for large negative peaks. The negative peaks were being clipped, and decaying dependent on the RC time constant of the Source (think cathode) resistor & load capacitance of the tone stack. I reduced the Source resistor to 33k(?) at which point the zero signal current through the MOSFET follower was heading towards 10 mA, things were getting very warm, my power transformer (already marginal) was struggling and I still had a slew rate problem tracking large, higher frequency signals. So I switched to a 'constant current source' instead of the Source load resistor. I used both of the following - either works OK, but I can't remember what bipolar transistor I used (may have been a BD115 and my fingers crossed - is only good for < 180V)

I switched from the bipolar NPN transistor to the delpetion MOSFET (LND 150) for a higher B+. The IRF740 is good for <400V B+ (IRF 840 will go to 500V max, as will LND150).

Hopefully you won't need to go down this route...

Any comment on which changes were the most effective?

[mat]

Any comment on which changes were the most effective?

No, I did all the changes at once, sorry. I finally changed the VOL2 pot to 500kB and now it works as supposed. I did put today a negative feedbackloop to V1b and it gave nice dumbly chime to the sound and playing feel.

I'll try the mosfet option and maybe something else but if I'm not satisfied I'll change V2 to 12AX7 with a more traditional OD.

Try to record something at friday.

Will get back..

[mat]

Hello again. Sorry to it to take so long (busy at work). I recorded something at band rehearsals but it was no good after listening to it. The OD is too fizzy still.
What I'll do next is play the amp as much as possible and after a while will try more tweaks, maybe get rid of some of the shielded cable..

Will be heading at London tomorrow for couple of days. Any Londoneers going to PLASA ?

[10thTx]

Mat,

IF the EF86 is still too fizzy, I'd consider just using a more traditional D-style OD or ........... try a 5879 with a cathode follower which I found reasonably smooth.

with respect, 10thtx