Too much gain?

[daz]

This is a issue, or i THINK it's an issue i have had forever with this amp i built several years ago. It has seen nearly every imaginable tweak and configuration imaginable from the PSU to the preamp stages, everything has been configured in nearly every way imaginable. (See link below for schematic) But one "issue" has always been present and i would like to know what might be causing it or if it's normal. I never have the master past about 9:00 except on the VERY rare gig, where it's not much past that anyways. It's VERY loud. I think 10:00 would be as loud as any gig i ever played when i was gigging regularly. I used 800's and 900's and a silver jube back then, and i never had them past about 10:00 on the master either tho.

That said, the issue is this....if it goes over around 10:00 the tone goes completely horrible. I only know this because i have put a crossline master in it numerous times as a temporary control just to test it. If i turn the regular master up past about 10 oclock and turn the crossline down the PI apparently distorts and i mean HORRIBLY so ! Before that it's one of the nicest marshall tones i've ever had, but past that point it goes to hell. I have many times thought about an excess of gain, but would that even do this? I mean, if i turn it down to the point where it's just starting to sound good again, at that point it's no where near as loud as a marshall. So i can't see how it's a gain issue. Besides, look at the preamp.....does it look like hyper gain? It's OD is about as hot as a peavey classic 30 or a blues Jr. I even put another voltage divider right in front of the master to drop it down 50% and still it distort badly past a certain point.

So looking at the schematic, what do you think? Can you see any reason for this? Heres the schematic... http://i24.photobucket.com/albums/c39/dazco/amp.jpg

[nickb]

I wonder if this might be blocking distortion on the grid of the PI. Try adding a resistor (say 220K) in series with 4.7nF cap to see if it improves things. Otherwise I think you'll need to scope it out.

[Palladin]

Too much gain?

I think your plate supply voltages are too high.
I imagine you are getting something like 420-ish vdc on the PT plates.
Try 10K after the choke then 10K and finally 4.7K for dropping resistors.
That will provide voltages closer to the Marshall model which is critical to best perfomance in those amps.
Regards,
Steve

[R.G.]

Don't know the particulars of this amp, but when an amplifier suddenly does something different when the gain is turned up, I always worry about whether it's got some kind of gain-related breaking into internal oscillation on peaks; the possibility of blocking distortion exists as well, but is much easier to cure.

[pdf64]

Yes oscillation seems a likely suspect.
At high frequencies those 3 gain pre-amp stages are pretty much cascaded wide open, with little interstage attenuation.
I think adding / upping the grid stoppers would help a lot, without affecting the tone / response too much.
Say ~10k-22k on V1B and V2A.
Another possibility is that the oscillation loop is around the power amp, and maybe last gain stage; sometimes that 100pF cap across the LTP plates can cause a pole / ringing, and removing or doubling it can help.
Pete

[Enzo]

My first thought was parasitics - parasitic oscillations - and those usually inhabit peaks, and it would be consistent they only show up after you reach certain volume levels.

[daz]

Yes oscillation seems a likely suspect.
At high frequencies those 3 gain pre-amp stages are pretty much cascaded wide open, with little interstage attenuation.
I think adding / upping the grid stoppers would help a lot, without affecting the tone / response too much.
Say ~10k-22k on V1B and V2A.
Another possibility is that the oscillation loop is around the power amp, and maybe last gain stage; sometimes that 100pF cap across the LTP plates can cause a pole / ringing, and removing or doubling it can help.
Pete

Thanks all. I had 220k grid stoppers on V1B and V2A at one time ! And it did it then too....it always had as i said. The cal across the PI is only a recent thing, so thats not it. Plate voltages are much lower than that Palladin. I had 10k droppers till a couple weeks ago as i posted here about in another thread. the PI voltages now are only 200 something i believe. They've never been over 300 that i know of, and usually were around 230v most of the amp's life. Even with the 2.2k droppers the voltages at all points are lower than a marshall from the voltage charts i've seen. I will try a grid stopper on the PI tho. (tho i think i tried that eons ago)

[pdf64]

'I had 220k grid stoppers on V1B and V2A at one time ! And it did it then too'
It may be oscillation around the power amp then; does the presence / negative feedback controls affect it?
Also try chop sticking the lead dress around the LTP.
A long shot is a bad power supply cap; excessive ESR could prevent adequate decoupling and result in a positive feedback loop. Try bypassing them all in turn with a known good one.
Pete

[daz]

Not sure, but i don't think so.I have the NFB as you can see with a 250k pot and 33k R so i can adjust it almost out of circuit to a large degree. I know it does it with almost 300k as the NFB resistance off the 16 ohm tap, so i doubt it. PSU caps...i've had may different ones of different values and it's always done it with any of them. I'll have to insert the crossline master again and see how NFB or lack of it and chopsticking affects it. I'll try and get to that tonite along with a grid stopper on the PI.

problem with trying to reduce it in the preamp is while the signal level may be high, the amount of OD is not and i'm not sure how i can eliminate signal level w/o killing the amount of OD. It's already fairly low, on the order of classic rock style gain.

[daz]

I think your plate supply voltages are too high.
I imagine you are getting something like 420-ish vdc on the PT plates.
Try 10K after the choke then 10K and finally 4.7K for dropping resistors.
That will provide voltages closer to the Marshall model which is critical to best perfomance in those amps.
Regards,
Steve

I just re-read all the posts and i noticed i mis-read yours. Yes, thats about what my plate voltage is on the el34's. But with a 10k after the choke how will that lower plate voltages? The supply to the plates is before the choke. If i put a 10k there it would cause the screens to be way low and mush the tone out wouldn't it? I don't understand, but i'm probably missing something, not being a tech.
Anyways, i just went back to all 10k droppers. The higher voltages are great in a way but it's too hard to tame the bright cutting tone. Thats for a different thread tho.

[Palladin]

Hi Daz,
Sorry for not being clear.
My assumption is that your EL34 plates are at about 420vdc. If that is true, my recommendation is to lower the plate voltage on the preamp tubes (not the power tubes). Your schematic shows some basic similarity with the typical Marshall topology we see so much of. Those amps typically have a 10K dropper after the choke feeding the PI (yours has a 2.2K); then a 10K feeding V2 (yours has a 2.2K), then something like 4.7K feeding V1 (yours has 2.2K). Changing to the values I suggest should lower the plate voltages on your preamp tubes to a much more reasonable level allowing you to get better control on the gain and perhaps the oscillaton other folks mention. Remedies like large grid stoppers and plate bypass caps etc etc should then not be required.
I'm also a little suspicious of the feedback circuit employed in your amp and if the lower preamp tube plate voltages don't improve things, I would go back to something very basic in that area.
I hope that explanation is a bit more clear.
Regards,
Steve

[daz]

Hi Daz,
Sorry for not being clear.
My assumption is that your EL34 plates are at about 420vdc. If that is true, my recommendation is to lower the plate voltage on the preamp tubes (not the power tubes). Your schematic shows some basic similarity with the typical Marshall topology we see so much of. Those amps typically have a 10K dropper after the choke feeding the PI (yours has a 2.2K); then a 10K feeding V2 (yours has a 2.2K), then something like 4.7K feeding V1 (yours has 2.2K). Changing to the values I suggest should lower the plate voltages on your preamp tubes to a much more reasonable level allowing you to get better control on the gain and perhaps the oscillaton other folks mention. Remedies like large grid stoppers and plate bypass caps etc etc should then not be required.
I'm also a little suspicious of the feedback circuit employed in your amp and if the lower preamp tube plate voltages don't improve things, I would go back to something very basic in that area.
I hope that explanation is a bit more clear.
Regards,
Steve

As i said above, i DID have 10k droppers in the PSU, in fact just up till a week ago. But i wanted higher voltage because actually with 10k droppers mine is much lower than a marshall at the pre plates. With 10k droppers mine goes to 120v at V1's plates with the 1.5k cathodes. But as i also said above, i went back to 10k last night, but not because of the issue this thread is about since it did it before the change to 2.2k too. It always has. I went back to the 10k droppers because the tone was just too cutting and hard. With the 10k's it's very soft, and in fact thats why i have the pot in the loop. But realize the loop is just lake a marshalls NFB when the pot is fully closed. Then it's 33k off the 16 ohm tap. But unless i run the pot wide open it's too mushy and soft feeling/sounding, and again it has always had this PI issue even when i had no NFB pot. I tried a few things last nite including chopsticking around the PI, adding a grid stopper to the PI and a few other things to no avail. I think one thing i could have tried and didn't but will try tonite is a grid leak at the PI. For some reason that woks better than anything else for reducing gain in my pre at the V2a grid. Voltage dividers there screw with the tone. It just always sounds best like that. (the 2.2k as a stopper in the schematic is wrong...i removed it recently)

[kleuck]

I think that a 10K grid-stopper at the input is a too low value, as is the 100 pf on across the plate of V1B.
Perhaps not your case, but i had similar issues (in a much simpler preamp) when i began to try to get the max gain out of it, i ended with at least 33K grid stopper on each triode, and a cap across each plate resistor (higher values : 100pf across a 330K resistor, and 500pf across a 470K one)
But the first thing i think is to use a higher grid stopper at the input, and the second triode.
Actually, i would try 33K on the 3 triodes.

[Palladin]

Hi Daz,
I have the following suggestions for improving the voicing (assuming you use the higher droppers I suggested earlier):
Raise the value of the .0047 coupling cap on V1 to something like .02.
Change the 250 pf bright cap on the 1M vol pot to .001.
Remove the 100pf cap across V1B plate resistor.
Increase the 100K V2A grid resistor to something like 470K to 1M.
Change the cap across the PI plates to 47p.
Regards,
Steve

[daz]

I think that a 10K grid-stopper at the input is a too low value, as is the 100 pf on across the plate of V1B.
Perhaps not your case, but i had similar issues (in a much simpler preamp) when i began to try to get the max gain out of it, i ended with at least 33K grid stopper on each triode, and a cap across each plate resistor (higher values : 100pf across a 330K resistor, and 500pf across a 470K one)
But the first thing i think is to use a higher grid stopper at the input, and the second triode.
Actually, i would try 33K on the 3 triodes.

Thats all just the current way i have it. I've had stoppers on every grid in the past up to 220k and a 68k at the input. The caps are also not something thats always been there.All those things have at one time been gone and like i said the issue has always been present. The reasons those things are as they are now is thats the way it seems to sound best, but they aren't the reason for the issue. I've even had all different iron in it including a 5H choke and it still did it. I keep thinking of ways that might work, but before i even try them the question always is why do i need to when my amp from the CF on back is exactly like some marshalls. I think the issue is too much preamp gain, but what i don't get is if thats true how can i get the signal level down without killing the amount of OD which is already medium at most, and without killing the tone. If i just put a voltage divider (pot or fixed) before V2a the tone changes and loses a lot regardless of the ratio.