PI grid leaks

[tubeswell]

heres a related question. When my PI distorts, as i said it gets very nasty sounding. yet the amp from the cathode follower thru the stack and the PI is all exact marshall specs. What could cause that? using either a crossline of dual pot at the power tune grids gives me that nastiness, so we know it can't be the PA since it's after that. It only happens when the PI is hit with a big signal. the pre is cascaded so theres a lot of signal. But it's not by any means super hi gain and even at lower gain settings it will happen with a post PI master. The pre PI master i normally use is fine because i never turn up loud enough to get to that real nastiness. Voltages are around typivcal for all stages. What could cause this?

Blocking distortion maybe? _In which case, try either 1k5 - 4k7 grid stoppers on the output tube grid pins, or 10k-68k on the PI input grid pin - or lowering the value(s) of the coupling cap(s) before those grid pins.

[iknowjohnny]

the el34 grids have 5.6k already. It doesn't sound like blocking distortion tho. that has an effect of the signal sorta cutting out or being squashed. This just sounds like you put the distortion thru a EQ and did reverse V with the center at 1k and are using a fuzz.

[Andy Le Blanc]

Try balancing the plate voltages? Have you swapped out the tube?

What in front of the PI?

[tubeswell]

heres a related question. When my PI distorts, as i said it gets very nasty sounding. yet the amp from the cathode follower thru the stack and the PI is all exact marshall specs. What could cause that? using either a crossline of dual pot at the power tune grids gives me that nastiness, so we know it can't be the PA since it's after that. It only happens when the PI is hit with a big signal. the pre is cascaded so theres a lot of signal. But it's not by any means super hi gain and even at lower gain settings it will happen with a post PI master. The pre PI master i normally use is fine because i never turn up loud enough to get to that real nastiness. Voltages are around typivcal for all stages. What could cause this?

Blocking distortion maybe? Try running 10k-68k grid stopper right at the grid pin of the input stage on the LTP, and/or lowering the coupling cap value feeding that stage. Blocking distortion is caused by excessive charge building up in the coupling cap when big bassy chords are powered into the signal chain esp at higher vols, and the cap can't discharge quick enough, so it drives the following grid into +ve bias territory in the interim, which results in a nasty blatty sounds on the signal peaks. The solution is to either lower the cap value (so it can't pass the bassier frequencies, or use grid stoppers to stop the excess current from getting into the grid.
It only takes 30 seconds to solder in a grid stopper to see.

You can also dump some bass from earlier on in the signal chain, either by roll-off with smaller cathode bypass cap(s) in V1 etc, and/or bass reduction with smaller coupling caps etc nearer the start of the signal chain

[iknowjohnny]

Same result. I have a cut control, so by bypassing the cap it becomes a post PI master. I tried that and added a 47k grid stopper at the PI input and changed the coupling cap to .01 from .022. Then i turned the post tone stack master up and used the post PI master and got the same nasty tone.

[Jana]

post an accurate schematic of the whole amp.

[iknowjohnny]

http://i24.photobucket.com/albums/c39/dazco/el34amp.jpg

[Jana]

my observations based on the schematic:

All three gain stages in the pre-amp are biased for maximum gain. V1A and V2B are bypassed with a .68 which further boosts the gain on the high end. The .022 and 68K coupling those two stages lets highs through but attenuates the lower frequencies. Same goes with that 68K on the 1Meg gain pot. There is nothing in the circuit to tame any of high end, in fact, the higher in frequency the signal, the more gain. A prime candidate for parasitic oscillations.

Also, where are the EL 34's biased?

The small value resistor (2.2K) in the B+ feeding the PI troubles me. The voltage level going to the PI combined with the cathode biasing of the EL34's could very well lead to a situation of driving the grids of the EL34's positive. When that happens, with this LTP PI, some nasty things happen. I have written in detail on this in another post a while back. Search for it.

In short, lots of high end, no attenuation on any of it, possible EL34 grids going positive, possible parasitic oscillations--Killer Bees with diarrhea.

I know you have spent a lot of time on this and it is almost absolutely perfect except for the one thing. But, it's always the "one thing."

[xk49w]

I'm wondering about the cut control inside the FB loop.

What happens WRT nasty distortion when the cut is opened and FB enabled?
How about if feedback is disabled?

[iknowjohnny]

All three gain stages in the pre-amp are biased for maximum gain. V1A and V2B are bypassed with a .68 which further boosts the gain on the high end.

the first stages are at max gain but till a week or 2 ago they were very different. First stages had cathode resistors at 1.5k and 2.7k. I finally tried every combo you can imagine and this just sounds best.

The .022 and 68K coupling those two stages lets highs through but attenuates the lower frequencies.

The 68k's aren't coupling the stages, they are grid leaks i use instead of voltage dividers to lower gain, and i use them because they worked a lot better.

The small value resistor (2.2K) in the B+ feeding the PI troubles me. The voltage level going to the PI combined with the cathode biasing of the EL34's could very well lead to a situation of driving the grids of the EL34's positive. When that happens, with this LTP PI, some nasty things happen. I have written in detail on this in another post a while back. Search for it.

Do you mean because it's letting to much voltage thru? Because the voltage at the PI plates is pretty typical i think, about 260 on each side. Or do you mean something else?

Also, where are the EL 34's biased?

You mean where as in current? I went thru that a long time ago and i can't recall. But i do remember i have them under max just enough so they aren't running too hot at all. I spent a lot of time getting that right so i pretty sure about it. Seveal gurus at another forum helped me get that right so i think it's quite good.

I'm not tech, but i do have several decades of live playing experience and have gigged a lot of marshalls, just to assure you i know good tone. And you'll just have to trust me when i say this amp sounds like anything but bees or parasitic oscillations. All i am doing to it is playing with it for the sole purpose of perfecting it, mostly out of boredom than anything else. But i have never had a marshall i feel sounded better, and most didn't sound nearly as good. I compared it to an early JCM 800 not long ago and the 800 sounded almost pathetic by comparison. I'm just telling you this not to blow my own horn because i'm an idiot as electronics go. But to assure you the thing doesn't sound anything like what you see in the schematic.

What happens WRT nasty distortion when the cut is opened and FB enabled?
How about if feedback is disabled?

It just cuts highs in either case. Works just fine. tho i don't use it often. I only put it there in case i ever encounter a seriously trebly room.

[Jana]

The .022 and the 68k form a high pass filter--do the math on it.

The typical value for the 2.2K resistor is 20K and that is with a fixed bias amp. You are cathode biased which reduces the peak to peak voltage swing that an EL 34 (or any tube) can tolerate before going into saturation or cutoff. Cutoff isn't all that bad but saturation is not good (tone wise).

What is the bias voltage on the EL 34's and what is the max peak to peak signal coming out of your PI?

I'm just pointing out what I see, based on decades of experimenting and building amps--initially with marshalls and then evolving into my own. Not tooting my own horn either but I lost count of the number of amps I have built about 20 years ago.

And I'm not trying to rain on your parade, it's good that you are experimenting and not content to clone. But be aware that in the end stages of any amp design it is more than just haphazardly swapping parts, it takes an understanding of why. Do yourself a favor and get some test tools and start to learn the why. A signal generator, scope and good multimeter would do a lot to helping you answer your questions. As it is, you really don't know where your current issue is coming from--you can only guess based on knob settings. Maybe the guess is right but maybe not. You have been working on this amp for what, a couple of years now? Really, you will never be content with this amp, lol. There will always be something. Trust me, I know. Get some tools, some books, a couple of notebooks, let go of the notion that soon it will be perfect and you can quit. Give up, embrace the force, become one with the process instead of chasing the product.

[iknowjohnny]

The .022 and the 68k form a high pass filter--do the math on it.

I'm just saying it works well. True, i don't know why. But the amp isn't high endy at all. In fact, it's rather dark if i had to call it one way or the other.

The typical value for the 2.2K resistor is 20K and that is with a fixed bias amp. You are cathode biased which reduces the peak to peak voltage swing that an EL 34 (or any tube) can tolerate before going into saturation or cutoff. Cutoff isn't all that bad but saturation is not good (tone wise).

I came to this as with all the values in the amp by way of my ears. I have just enough knowledge to know what absolutly not to do, but within a margin of safety i experiment w/o rules. It's fun and it seems to work considering the results. Anyways, i tried what you said and it's very close to the same, tho it may sound a bit better. I will have to give it a while. But i've tried most every imaginable dropping resistor scenario.

What is the bias voltage on the EL 34's and what is the max peak to peak signal coming out of your PI?

I can only tell you the plate V is around 400.

You don't have to tell me you know a lot about this. theres no doubt of course. But because the amp is here with me and i've tried so many things, only i know what it sounds like with various value changes. So i hate to tell someone like yourself that what you describe didn't happen, but thats happened many times with guys on the various forums. they all know worlds more than I, but you can only tell so much from a schematic w/o hearing the amp. So don't take anything i say as meaning i'm doubting you. It's just that the only way for anyone to truly know what the amp sounds like is to hear it. and i honestly love the tone of this amp. But I'll probably still be tweaking it after it's the best it can possibly sound because it's fun chasing the magic

[iknowjohnny]

See, this is why i do this. that is, mess around with the amp now and then and start asking questions. Inevitably someone will say something that i will try and it will improve things. Jana, as i said i tried putting a bigger dropping resistor before the PI and now after a hour and a 1/2 or so I can say w/o a doubt it has improved things nicely. I'm not even sure how to describe how other than to say it just gave the tone/feel that certain something that makes the strings do what you ask of them. Thats the only way i can describe it. usually i can say there are more dynamics, a better balance, etc. But i can't really say anything like that. It just improved it to where it allows me to play a bit easier w/o working as hard to get what i want out of it with my hands. Kinda subtle which is why i wasn't sure at first, but yet not so subtle in the end if you know what i mean. Anyways thanks jana. all your replies here helped me greatly, as any improvement at this point is cake and not really expected. Thank you sir !

[Jana]

"Thank you sir !" Another one.

Try a 10K in the other 2.2K spot.
Try a 68K between the Master volume and the .022 going into the PI.
Try a 120pf across the plates of the PI (pins 1 to 6).
Try a .0022uf from the junction of the 33K slope resistor and the two .022 tones caps to ground--then see what that does to your mid control.

[iknowjohnny]

Another one:roll:

Just saying thanks ! If that somehow offended you i didn't mean to.

Anyways, i will try those other suggestions, especially the mids thing because i've always wanted to change the mids in this amp a bit. But that 10k is going to make the distortion more spongy and mushy which is a good thing with a hard bright sounding amp, but this amp is already very spongy feeling. But, what the heck, i'll try it. Might be a day or so tho. (roommate home, must be quiet)