Too much gain?

[Steve Conner]

"Doctor, it hurts when I do this."
"Well then, don't do this."

You optimised the amp heavily to sound its best at the volume you normally use it: it's no surprise that it sounds bad when you push it outside of that envelope. Where's the problem?

Having said that, is the crossline MV before or after the grid stopper resistors?

[Chuck H]

Here's something you can try. If it's an oscillation buggin' on peaks this may help. First, as noted, I think you may benefit from a higher value input grid stopper. There should be no difference in tone upping that value to 47k. Now, at the input triode, solder a 4.7pf 1000V cap from the plate pin to the grid pin. This is just barely audible and is especially good at stopping HF crosstalk and oscillation due to input sensitivity. If you find the tone loses too much hair try removing the 100pf cap across the V1A plate load. This should be easy enough to try.

EDIT: It's aslo worth noting that PI distortion generally doesn't sound that good when it's not filtered through power tubes that are already clipping. Most guys that use the post PI MV only use it for small amounts of attenuation and the cross line type is the least popular for tone. It IS good for testing purposes though. But how does the amp dound with the master above 10:00 if you DON'T use the post PI MV???

[daz]

Having said that, is the crossline MV before or after the grid stopper resistors?

Before.

Here's something you can try. If it's an oscillation buggin' on peaks this may help. First, as noted, I think you may benefit from a higher value input grid stopper. There should be no difference in tone upping that value to 47k. Now, at the input triode, solder a 4.7pf 1000V cap from the plate pin to the grid pin. This is just barely audible and is especially good at stopping HF crosstalk and oscillation due to input sensitivity. If you find the tone loses too much hair try removing the 100pf cap across the V1A plate load. This should be easy enough to try.

EDIT: It's aslo worth noting that PI distortion generally doesn't sound that good when it's not filtered through power tubes that are already clipping. Most guys that use the post PI MV only use it for small amounts of attenuation and the cross line type is the least popular for tone. It IS good for testing purposes though. But how does the amp dound with the master above 10:00 if you DON'T use the post PI MV???

I would like to try that, but i have no 100v caps in any range let alone pf range. I may be able to find a few i can parallel for the right value that may add up to 100v. I'll look tonite. As to hows it sound w/o the crossline, i have wondered that myself. I can't possibly turn it that loud. It would be insanely loud in my small studio home and neighbors would go off the deep end instantly. And just to address one thing conner said about how it doesn't matter if i never turn it up that high. The way i see it is whatever the issue that causes it is, fixing it might contribute to better tone when at the volumes i DO play at. Never know. I just don't like not knowing what else it may be causing.

[Chuck H]

I would like to try that, but i have no 100v caps in any range let alone pf range. I may be able to find a few i can parallel for the right value that may add up to 100v. I'll look tonite.

I just want to clarify that you need a 1000V cap for this. You REALLY don't want to put a 100V cap there. 1000V small pf caps arent hard to get. Patience is a virtue.

And I understand the desire to have everything be as it should if for no other reason than the amp should function properly in any capacity that befits amps of it's type. It's easy for outsiders to disregard that desire since at some level it is unnecessary. But I understand and sympathize.

[daz]

That was a typo...i meant 1000v.

[olddawg]

You guys know a lot more than me.....but I would start by putting it on a load bank, driving the input with a calibrated 1k sine wave, then turn up the volume and look at each gain stage with a scope until I saw the problem. You should be able to see it and measure the threshold. Until you know what is going on it seems to me you are just throwing shit at the wall and seeing what sticks. Better to know what the problem is.

[Chuck H]

I don't think daz has a scope.

[daz]

I would, but i don't know how to use, nor do i own a scope.

by the way, so i don't have to start a new thread, if i went to a tube rectifier in place of the diodes (had 5AR4 at first but soon went SS) how much lower would my 440v at the PT plates go? (yes, i know i'd have to lower any filters to around 32uf or lower)

[daz]

Good timing Chuck !

[Chuck H]

A 5ar4 would drop about 30V to 35V. You can get a similar (very similar) effect using a 120R 25W resistor between the rectifier and the standby switch. Much easier and cheaper to try and undo and no need to lower your filter values. The lower voltage will be less dynamic. As will the attack envelope. When you ass resistance to the power supply (with a rectifier tube OR a resistor) you get a higher ratio of voltage drop with more current. This is often percieved as a sort of compression effect. Which is ditinctly NON dynamic. In short, your whole tone will change and the amp would likely need to be re tweaked to your tastes. I say, if you want a tube restified amp, build one. It's my understanding that you like that sort of thing. Then you could leave the Marshall more Marshall-ish. The project doesn't need to be expensive if you're creative with parts aquisition.

[daz]

A 5ar4 would drop about 30V to 35V. You can get a similar (very similar) effect using a 120R 25W resistor between the rectifier and the standby switch. Much easier and cheaper to try and undo and no need to lower your filter values. The lower voltage will be less dynamic. As will the attack envelope. When you ass resistance to the power supply (with a rectifier tube OR a resistor) you get a higher ratio of voltage drop with more current. This is often percieved as a sort of compression effect. Which is ditinctly NON dynamic. In short, your whole tone will change and the amp would likely need to be re tweaked to your tastes. I say, if you want a tube restified amp, build one. It's my understanding that you like that sort of thing. Then you could leave the Marshall more Marshall-ish. The project doesn't need to be expensive if you're creative with parts aquisition.

I've done that Chuck, but theres too much sag. I never got that much sag with the tube recto that i recall. My thought was this....change the entire way the amp feels and breathes by getting certain aspects of the tone from opposite places. For example, i find that a low voltage PA and a small value cap can for the plates and screens gives a very springy kind of sag. I get that now with the PA at high voltage with a huge cap values because the sag comes from the low voltage pre. (i'm back to 10k droppers) But i always find that for every task there are a number of ways to accomplish it, and sometimes you find doing things differently to get the same basic results can surprise you. So instead of getting the sag i like with low voltages and filtering in the pre while the PA has big filtering and voltage, do the opposite and give the pre high voltages and maybe more filtering but cut the PA's voltage and filtering. I can accomplish this by using those small droppers for the PI and V2 nodes and taking V1's supply from right after the choke as i was doing a few days back. This way i can get the dynamics and all that i liked about the higher voltage pre but run it into a softer saggier less loud and strident PA. I've done trades like this in the past and they often pan out. So i figured it's worth a shot. I'm just going to hook a octal up and just let it sit in the chassis making sure it's safely insulated and if it sounds good i'll drill a hole for it. But i may try the resistor you mentioned first because while i never liked that when i tried it, i should probably give it a try with the PSU reconfigured as i described. the cut and punch of the high voltage pre make balance it out the same way i figure the tube recto would. I may give that a shot in a few tonite and post back.

[Chuck H]

That you didn't like the resistor when you tried it, but did like a tube rectifier in another amp isn't really a good way to decide if the resistor is a suitable solution. The power amp doesn't care WHAT is resisting the flow of electrons. If you really want to get accurate then drop the filter values to where they would be with a tube rectifier. I'd double blind test that against anyone.

The issue with PA/preamp balance, tradeoff thing is that the preamp voltages will also sag relative the the PA current. So you won't get a stiff preamp and a spongy PA.

[Gingertube]

I had an identical problem on a Trainwreck Clone - in that case sound turning to "custard" once the volume (in the same plce as your Master) was turned passed 12 o'clock.
I finally tracked it to being inadeqate power supply decoupling between the PI and the final gain stage of the preamp. Make sure that 20uF at pin 6 of V2B is properly connected BOTH ends and that it is not open circuit.
Cheers,
Ian

[daz]

That you didn't like the resistor when you tried it, but did like a tube rectifier in another amp isn't really a good way to decide if the resistor is a suitable solution. The power amp doesn't care WHAT is resisting the flow of electrons. If you really want to get accurate then drop the filter values to where they would be with a tube rectifier. I'd double blind test that against anyone.

The issue with PA/preamp balance, tradeoff thing is that the preamp voltages will also sag relative the the PA current. So you won't get a stiff preamp and a spongy PA.

Actually i was saying this amp had a tube recto before, and i've tried a "sag resistor" in it too both with and w/o the tube in it. I just tried a resistor tonite and played around with the PSU. Naaa....back to the initial circuit. The thing that i'm wondering about now is the NFB. I think turning it down to normal levels of NFB solves a lot of things, and i think helps with the loud/harsh issue. (tho certainly not solving it) But it also becomes rather soft and loses the punch and cut it has when i dial it out to a large degree. Maybe thats where i should be looking. Theres a certain magic when you can get two things to happen simultaneously that usually seem mutually exclusive. For example, like getting a bright cutting tone while remaining a fat and thick quality too. Those seem like two attributes that can't exist in the same tone but they can. I'd like to have a cutting powerful punchy feel and tone yet remain smooth and fat and spongy feeling. I've had that before to a degree. NFB tweaks may be worth investigating deeper. I have, but not nearly as deeply as PSU and preamp tweaks.

[daz]

I had an identical problem on a Trainwreck Clone - in that case sound turning to "custard" once the volume (in the same plce as your Master) was turned passed 12 o'clock.
I finally tracked it to being inadeqate power supply decoupling between the PI and the final gain stage of the preamp. Make sure that 20uF at pin 6 of V2B is properly connected BOTH ends and that it is not open circuit.
Cheers,
Ian

Thanks, but it's fine. It's not going to be a problem due to something like a bad component or solder joint or lead dressing, nothing like that. It's going to be a design issue because i have built 3 of these amps and every one has been thru endless design changes and had all different parts at different times, even transformers. Yet thru this all it's always done this. Thats why i said i wonder if maybe this is normal for this design. I had many marshalls but i never turned them past about 100, maybe 11:0 at most because they were sol frigging loud you couldn't unless you played coliseums ! So maybe it's just normal, i dunno. But whatever it is, it's not something burned/broken/bad connection/etc. It's in the design somewhere. Gotta be if process of elimination means anything.