Page 4 of 9

Re: MikeC Sixty-One SRT

Posted: Sat Mar 18, 2023 6:30 pm
by MikeC
Update - I've been chasing an oscillation problem when Reverb set to zero, an acceptable (at idle) plate dissipation on the EL84s and excessive crossover distortion when the amp is dimed. Farking hours of research, mods and "learnings" (hate that word)! Plate V is 360V (a bit higher than EL84s like) but screen V is lower at 340V. So with a 180R EL84 cathode resistor the plates dissipate at 97% max dissipation at idle which is acceptable and with lower screen V the EL84s should last. I've sorted the reverb oscillation - was mostly lead dress. Tomorrow (when I get one), I'm putting a 15V Zener across the EL84 cathode resistor which will hopefully eliminate the crossover distortion. What a mission this is turning out to be but worth the effort and increase in knowledge! PS, the "prototype" is starting to look alot like Frankenstein's monster :)
Frankenstein II.jpg
Frankenstein II.jpg (75.97 KiB) Viewed 34540 times

Re: MikeC Sixty-One SRT

Posted: Sat Mar 18, 2023 6:48 pm
by MikeC
robthemac wrote: Sat Mar 04, 2023 2:05 pm What are you going to do with the prototype?
Once I've got the prototype sorted and sounding great, I'll recycle all the parts to build V1.

Re: MikeC Sixty-One SRT

Posted: Sun Mar 19, 2023 7:24 am
by sizzlingbadger
97% at idle and you have cross over distortion, plate voltage too high ? Do you have plate resistors like the original Vox AC10 ?

Re: MikeC Sixty-One SRT

Posted: Sun Mar 19, 2023 8:36 am
by MikeC
sizzlingbadger wrote: Sun Mar 19, 2023 7:24 am 97% at idle and you have cross over distortion, plate voltage too high ? Do you have plate resistors like the original Vox AC10 ?
See pic below - the blue trace is the PI output & the yellow trace is the EL84s output across an 8-ohm (resistive) dummy load.
Notice that she's doing nearly 17W RMS at this setting...

The PI is beginning to distort (notice squared off top) and the crossover distortion on the EL84s is quite obvious.
Increasing the EL84s bias point by reducing their shared cathode resistor (from 180R to 150R) doesn't reduce the crossover distortion. And 150R has the EL84s idling at 110% mpd so I don't want to be there anyway.
Reducing the plate voltage didn't help either - I went as low as 290V. I am not using plate resistors.
With much googling I found info about this phenomenon in Marshall 18 watters - check this YouTube video https://www.youtube.com/watch?v=URrVYnpaE5c

I'll grab a 15V zener this morning, install it and post an "after" shot - hopefully crossover distortion will be all but gone.
FYI, at the point where there is approx. 15V across the cathode resistor the crossover distortion begins to show. Hence I'm going for a 15V zener.
Crossover distortion.jpg
Crossover distortion.jpg (1.83 MiB) Viewed 34464 times

Re: MikeC Sixty-One SRT

Posted: Sun Mar 19, 2023 8:42 am
by sizzlingbadger
The bias is shifting too much, large cathode bypass caps can make this worse too.

The zener should clamp that nicely.

Re: MikeC Sixty-One SRT

Posted: Sun Mar 19, 2023 9:00 am
by sizzlingbadger
Having said that... it may be better to just reduce the gain somewhere so you don't hit the issue in the first place. Also you should check what the crossover distortion actually sounds like, because having a bit doesn't usually matter and you may be trying to fix something that isn't "sonically" a problem.

That's a nice scope you have there :thumbup:

Re: MikeC Sixty-One SRT

Posted: Sun Mar 19, 2023 10:49 am
by RectifiedAmps
You could try larger grid resistors on the EL84s too. Never heard of the cathode zener trick - interesting idea.

Re: MikeC Sixty-One SRT

Posted: Sun Mar 19, 2023 11:24 am
by MikeC
The 15V Zener has definitely reduced the crossover distortion (see pic) and again I'm driving the amp to about 17W RMS output. Yes true with "what it actually sounds like" and the overall gain reduction (ergo I can just turn it down on the volume knob), but I want to be able to drive the EL84s to clipping. I want to get these things right before I experiment with power scaling (only on the power section), perhaps a "cut" control and the overall voicing of the amp.

Re: MikeC Sixty-One SRT

Posted: Sun Mar 19, 2023 11:26 am
by MikeC
RectifiedAmps wrote: Sun Mar 19, 2023 10:49 am You could try larger grid resistors on the EL84s too. Never heard of the cathode zener trick - interesting idea.
The grid resistors on the EL84s are currently 4.7K. What value would you suggest I try? P.S. I presume you are referring to the control grid (not the screen grid)?

Re: MikeC Sixty-One SRT

Posted: Sun Mar 19, 2023 1:17 pm
by RectifiedAmps
MikeC wrote: Sun Mar 19, 2023 11:26 am
RectifiedAmps wrote: Sun Mar 19, 2023 10:49 am You could try larger grid resistors on the EL84s too. Never heard of the cathode zener trick - interesting idea.
The grid resistors on the EL84s are currently 4.7K. What value would you suggest I try? P.S. I presume you are referring to the control grid (not the screen grid)?
Yep, control grid resistors. You could try 10k and see if it makes a difference. You can still see a crossover notch on your scoped image so it might be good to reduce it further. The other option would be to use smaller coupling caps between the PI and EL84s, but that might kill too much bass.

Re: MikeC Sixty-One SRT

Posted: Mon Mar 20, 2023 12:03 pm
by Anvil Amps Alan
RectifiedAmps wrote: Sun Mar 19, 2023 10:49 am You could try larger grid resistors on the EL84s too. Never heard of the cathode zener trick - interesting idea.
Fender Bassbreaker 15 has Zeners which always go short circuit and cook the output bottles. I have removed them on 2 amps now.
They C*#ts of amps to work on but you can clip the diode leads without dismantling the sandwiched PCBs.

Re: MikeC Sixty-One SRT

Posted: Mon Mar 20, 2023 12:29 pm
by MikeC
Anvil Amps Alan wrote: Mon Mar 20, 2023 12:03 pm
RectifiedAmps wrote: Sun Mar 19, 2023 10:49 am You could try larger grid resistors on the EL84s too. Never heard of the cathode zener trick - interesting idea.
Fender Bassbreaker 15 has Zeners which always go short circuit and cook the output bottles. I have removed them on 2 amps now.
They C*#ts of amps to work on but you can clip the diode leads without dismantling the sandwiched PCBs.
Interesting to know, I used a big boofy 4W zener with leads like tree trunks. Hopefully it'll last!

Re: MikeC Sixty-One SRT

Posted: Sun Apr 02, 2023 9:33 pm
by MikeC
Update - I've just finished documenting the layout of v4 which now has Power Scaling on the PI & power amp. The prototype is now looking like a dogs breakfast and I also have a bag of "spare" parts which have been swapped in/out on the journey to date. It's time to build a new turret board...

Re: MikeC Sixty-One SRT

Posted: Mon Apr 03, 2023 6:07 am
by Slowy
MikeC wrote: Sun Apr 02, 2023 9:33 pm Update - I've just finished documenting the layout of v4 which now has Power Scaling on the PI & power amp. The prototype is now looking like a dogs breakfast and I also have a bag of "spare" parts which have been swapped in/out on the journey to date. It's time to build a new turret board...
looks like your typical Vox!
Great effort by the way. :thumbup:

Re: MikeC Sixty-One SRT

Posted: Sun Apr 16, 2023 10:23 am
by MikeC
A little bit of progress on the SRT... I've built a new turret board ready for the new build. This will be the extent of the progress for the next 7 weeks as I'm out of the country. Can't wait to get on with this!