DIY 2204 negative DC voltage on grid of preamp tube

[jasonvilla]

Hi there, I have a home brew 2204 that Ive been working on because it sounded a bit fizzy, otherwise is seems to function just fine (I had a post on a different forum about the fizz). On my hunt for the fizz I noticed I got -21vdc after second gain stage coupling cap with the volume all the way up AND a 1khz sine wave running through it. The negative voltage goes down as I lower the preamp voltage and it read 0 if I turn off the sine wave generator or turn the pre amp volume off. Can someone please explain what's going on and if/how I need to fix this. Thank you!

[shooter]

if/how

I'd go for a better meter and double check

[sluckey]

What's the voltage level of that 1Khz sine wave?

[jasonvilla]

This is the sine wave measured at the input jack, 260mV

[sluckey]

You're way overdriving the amp!

That's 260mV_RMS which is 735mV_PP. Lower the input level to about 100mV_RMS. Any better?

[jasonvilla]

Seems like it does the same thing, only now Im reading -11.1 on my meter.

[sluckey]

Well, don't do that.   

[jasonvilla]

lol what!?

[sluckey]

By the time your test signal reaches the second triode grid it is so large that it causes the second grid to draw current during the positive half cycle. The grid current creates a negative voltage to appear on the grid.

[jasonvilla]

Thank you Sluckey. So seeing a negative voltage, referenced to ground( on the side of the coupling cap I wasn't expecting to see dc voltage), is fine?

[sluckey]

Thank you Sluckey. So seeing a negative voltage, referenced to ground( on the side of the coupling cap I wasn't expecting to see dc voltage), is fine?

No, it's not fine. That grid has zero volts on it when it is being operated as expected.

[SILVERGUN]

Does your amp EXACTLY match that schematic?

[jasonvilla]

Pretty close, a few changes based on what I had. The power tubes are JJ6v6s. I added that 500ohm resistor to lower the screens.  Other than that, I have 2 resistors and a cap after the first dropping resistor to elevate the heaters to 60v, but it's currently bypassed.

[SILVERGUN]

If there are no other preamp tweaks to note you should post your voltages...

[SILVERGUN]

Take another look at your schematic.
Either the 500 ohm resistor in the power rail is mis-drawn or it is out of position.

[jasonvilla]

Voltages taken with no signal and volumes turned to 0, TMB at 50%.


I forgot to erase the line that connects the input jack to the first .22 cap on the schematic. I only have one input.

[jasonvilla]

I believe thats where the 500ohm resistor is in the amp.

[SILVERGUN]

Voltages

Nice job.
So, compared to the original schematic your preamp voltages are a little low.
On V2a you have 1V on the cathode. So, with no signal the cathode is 1V positive in relation to the grid.
When the grid voltage (approaches and then) exceeds the cathode voltage you get grid current.
Drive it harder with a big signal, more grid current....which you saw in your original experiment.
The chance of this happening is increased by the lower (than spec'd) plate voltage because the anode is less capable of drawing electrons away from the grid at lower voltages.

One bandaid solution would be to add a grid stopper.

[SILVERGUN]

I believe thats where the 500ohm resistor is in the amp.

There should be an additional cap node between that 500R and 4.7K dropper

[PRR]

...seeing a negative voltage, ...., is fine?

It is fine. If you want to sound like crap. Or cut-out.

What *most* loud players do is crank-up to get "small" bias shift on the loudest notes.

Watch the meter or 'scope while you really play. At some point in the overdrive profile that grid will go negative and the sound probably changes substantially. On a small tube if you beat it to -20V of bias, the sound probably cuts-out, then fades back when the beating stops.

[jasonvilla]

Thanks guys

I moved the 500ohm resistor to in-between the choke and the second cap (where I should have put it in the first place).


I switched the sine wave for a guitar and was able to see -12 volts on the v1a grid with the preamp turned all the way up and big open E chords with humbuckers. I could see the voltage go negative with the volume pot around 1.5.

[jasonvilla]

BTW I was going off of Ceriatone's voltage chart when swapping out my dropping resistors. They seem lower than mine!

[SILVERGUN]

BTW I was going off of Ceriatone's voltage chart when swapping out my dropping resistors. They seem lower than mine!

Yeah, your voltages arent "the problem".
I just chose to point out that they were lower than what was listed on the original schematic posted.

[jasonvilla]

Haha oh. Any idea what is causing my problem? Should I up my voltages?  Thank you

[SILVERGUN]

Haha oh. Any idea what is causing my problem? Should I up my voltages?  Thank you

Try adding a 100K grid stopper resistor right at the grid pin on the socket.

[pdf64]

Hi there, I have a home brew 2204 that Ive been working on because it sounded a bit fizzy, otherwise is seems to function just fine (I had a post on a different forum about the fizz). On my hunt for the fizz I noticed I got -21vdc after second gain stage coupling cap with the volume all the way up AND a 1khz sine wave running through it. The negative voltage goes down as I lower the preamp voltage and it read 0 if I turn off the sine wave generator or turn the pre amp volume off. Can someone please explain what's going on and if/how I need to fix this. Thank you!

That’s perfectly normal, overdrive any single ended stage and the same thing will happen.
It’s due to RC coupling and the g-k diode. Overdrive the grid and the diode will chop off the top of the wave. All that’s left is the bottom of the wave, that’s the negative voltage you’re measuring.

[jasonvilla]

:/ haha alright now Im confused. You say it's normal, Sluckey says its not fine. PRR say it's fine, if I want it to sound like crap.

Did I screw up somewhere, or is this just what that stage of a 2204 does?




Also, you're suggesting I put a 100k gird stopper on V1a, correct? Not V2a

Thanks for the help you guys :)

[SILVERGUN]

Also, you're suggesting I put a 100k gird stopper on V1a, correct? Not V2a

No, it would go onto V2a and help to reduce grid current right where it's happening.

[jasonvilla]

Also, you're suggesting I put a 100k gird stopper on V1a, correct? Not V2a

No, it would go onto V2a and help to reduce grid current right where it's happening.

Oops, sorry, I wrote that backwards. Thank you

[SILVERGUN]

As far as the rest of the confusion goes, there are plenty of guys playing amps with grid current. And there are plenty of fizzy sounding amps.
We are here to try to identify your issue and inform you to the best of our ability.  Grid current doesn't sound good.
Does it happen, yes. Should you avoid it, yes, if you don't like the way it sounds. If you are trying to tweak your amp to function optimally then we can help you do that.
If you are a purest who believes that a masterpiece of an amp should not be tampered with then you are not alone.
One solution would be to turn down the preamp volume pot. I'm pretty sure that's what sluckey meant by "don't do that".
There are other solutions that would require modifying your amp. It all depends on what you want out of it and how far you are willing to bend the original design.
My grid stopper idea is a tiny little bend that may or may not mitigate fizz.


Before going any further I would confirm that all of my resistor values are correct. Make sure that you haven't made a mistake and then you can go forward.

[jasonvilla]

THANK YOU SILVERGUN! That is exactly what I wanted to hear.

I originally posted (on a different forum) about this new amp sounding a bit fizzy. It worked fine, but I didn't quite like the distortion. I have a few other 2204 ish amps i've built over the years, but not an actual Marshall 2204 to compare it to. Right around this time I got my first scope. Always wanted one. I've never even touched one before! I've been having lots of fun playing with it. Seeing is way easier for me. Im definitely a visual person!  Reading voltages and looking at them on the scope WITH a signal applied led me here - asking about the negative dc on the grid.

Im completely open to bend the original design! I just wasn't sure if I have done something to cause this issue. All resistors appear to be correct... I guess I could pull a leg of the 10k cathode resistor and measure it.

[jasonvilla]

I'm thinking I will lower my second dropping resistor from 10k to 4.7k and add a 100k grid stopper to V2a.

[SILVERGUN]

I was most concerned with the 470K:470K voltage divider right before V2a. If you have correct values there then you have the design pretty much right where it was intended. (except the slightly lowered voltages)
Raising the volatges and installing the grid stopper makes sense.


Since you are willing to bend the design, and you're looking for more overdrive, there are bigger changes that you can make to get it there without having to nail the grid of V2a. Let us know how it goes.

[jasonvilla]

Thank you, I'll report back after I make those changes.

Here is the amp, and the board. The 470k voltage divider is at the bottom.

Thanks! :)

[SILVERGUN]

I'm thinking I will lower my second dropping resistor from 10k to 4.7k

Ohm's law tells me you can go even lower with that value if you'd like to get closer to the listed voltages on the original schematic.
Not suggesting you have to, just sayin'

[SILVERGUN]

I'm assuming I understand your layout and there is an extra resistor here that I don't quite understand. 

[jasonvilla]

That is the 33k tone slope resistor. I just didn't have the wire on it when I took that picture.

I think I originally settled on 2.7k, 4.7k, 10k as my dropping resistors, before changing it back to what I have now.

[jasonvilla]

Alrighty, I changed the dropping resistors from 6.8k, 10k, 10k TO 2.2k, 5.6k, 10k.  I also still have that 500ohm resistor in after the choke.

My B+ is now: 390, 384, 358, 334, and 323.  V1 is now 289 and 245. They're still a bit low, but mostly after the 2.2k resistor.

I also added the 100k grid resistor. With the sine wave playing I was getting -13v after the second coupling, and now im getting -10v. So its getting better...

[SILVERGUN]

Ok, you can set aside the B+ adjustments...you are close enough for rock and roll.

2 things you can try to reduce the onset of grid current:
-I personally would see what happens when I up'd the cathode resistor on V2a. (maybe 1.5K, 1.8K or 2.2K)
-You can also increase the value of that grid stopper (220K up to 470K) and see what happens.

Just understand that I have no allegiance to the JCM800 and I'm just trying to get you to see cause and effect.

There's one bombshell move in the holster but I want you to get past your focus on the grid current.
I was walked through a similar situation a long time ago and learned a valuable lesson along the way.

In no way shape or form am I the only person who you should listen to.
My goal is to be helpful and not push my opinions on you.
I am in treatment for tweakers anonymous.

Where is the preamp gain control set during your measurements?

[jasonvilla]

Lol Thank you Silvergun, I really appreciate the help.


I left the dropping resistors alone and changed the cathode resistor from 820R to 1.5k and changed the grid stopper from 100k to 470k.

Now the most I see is -5.3v at that spot, this is with the pre amp on 10.

[SILVERGUN]

I would say you're in a better place now.
Plug your guitar in and set the preamp volume to where it sounds best. Probably not on 10.


Report back on how the amp sounds to you now and where you would like to go from here.


If not on 10 doesn't yield you enough overdrive we can source it elsewhere.

[PRR]

....-5.3v at that spot...

Follow along to the speaker, I figure 3Vrms at that spot implies 1,000V at the speaker. Which is about 125,000 Watts. Which is a couple thousand times more than a couple EL34 can do.

The bias shift IS "normal" when you OVERload a stage. Most audio designers avoid this. Here, you *may* want it.

And when you don't, Turn It Down. There most be some reason the factory spent all those pennies for pots.

Can you *play* the amp at these settings and levels? Not just 'scope it, but play it?

[jasonvilla]

I would say you're in a better place now.
Plug your guitar in and set the preamp volume to where it sounds best. Probably not on 10.


Report back on how the amp sounds to you now and where you would like to go from here.


If not on 10 doesn't yield you enough overdrive we can source it elsewhere.

Well if definitely sounds different. I think it sounds better, but I feel like it has lost something. The distortion sounds different past about 7 on the preamp volume. Id say the amp sounds louder too.

[jasonvilla]

....-5.3v at that spot...

Follow along to the speaker, I figure 3Vrms at that spot implies 1,000V at the speaker. Which is about 125,000 Watts. Which is a couple thousand times more than a couple EL34 can do.

The bias shift IS "normal" when you OVERload a stage. Most audio designers avoid this. Here, you *may* want it.

And when you don't, Turn It Down. There most be some reason the factory spent all those pennies for pots.

Can you *play* the amp at these settings and levels? Not just 'scope it, but play it?

But I like playing it with the preamp on 10! :D Or at least around there... Turning down the volume on my guitar makes the amp clean up pretty well.

Im not sure I know what you mean about playing at those levels.

[pdf64]

...I ... changed the cathode resistor from 820R to 1.5k ...

I think that may not be a good idea, as it will mess up the DCCF operating point, ie the pn 1 and hence pin 8 cathode Vdc will rise.
This is a well proven circuit, a bad sound is more likely due to a fault than a design issue. So I'm not sure that mods should be a first resort.

Grid conduction is a perfectly normal mechanism of tube amp overdrive, for non master volume p-p amps it is by far the principal one, ie the 'other' p-p tube takes over before cut off. I don't see why correlation should be drawn between grid conduction per se and a bad tone?

Are you sure that the amp is stable? A fizzy, cut off sound can be caused by instability, as well as extreme bias shift.

[SILVERGUN]

Well if definitely sounds different. I think it sounds better, but I feel like it has lost something. The distortion sounds different past about 7 on the preamp volume. Id say the amp sounds louder too.

Well, that's because it is slightly different. Remember what pdf64 told you about the top half of the wave being chopped off. Now it's doing that less.
Plus, by adding the 470K grid stopper resistor you will lose some highs. It is common to then run a cap across that 470K to allow some highs to bypass it (like you see in other spots). A good value would be like 100p up to 500p depending on what sounds good to you.

Don't let your ears fool you. Sometimes different is better and sometimes it's not. None of what I've had you try is so far out of whack that it would be a huge concern. Although there are millions of JCM800 phanatics that probably think I'm leading you down a dark path. Everything is reversible.

You are developing a skill set. If you like something, keep it. If you don't, don't. If you have questions, please ask them.
I agree with pdf64's point that there are other possible sources of fizz and bad tone, but since you've mentioned having other Marshall amps I'm assuming that we really aren't troubleshooting a major issue. Once you get to a point of acceptance as to where you are now we can help you make informed decisions on how to get the amp where YOU want it. Without being there we have no way to gauge how much fizz or what you are actually hearing.

Everything affects everything and one of the best techniques you can use is to make small incremental changes one at a time and then decide whether to keep or not..

So far all we have done is reduced your 3rd stage grid current. That ain't crazy. If you'd like to run another experiment, try a different 12AX7 in there and rerun your measurements.

[jasonvilla]

Oh boy. Well...

I can always put the cathode resistor back to 820 and leave the grid leak.

What do you mean by stable? The amp seams to operate normally, I just think the distortion sounds a bit fizzy. I know thats almost meaningless over a forum. I wish there was a good way to record it.
On past builds Ive put a .68 cap across that 820R and added a few snubbing caps to mitigate the fizz, but this amp sounds different. I guess it could be the lower voltages, OR something that Ive missed or messed up. It also has smaller transformers and a smaller choke, etc.
These are all new tubes, and Ive also tried old ones. 

On a side note, wouldn't you guys expect to see more than ~390v from a Hammond 272FX 300-0-300 PT?
https://www.mouser.co.uk/datasheet/2/177/200-1390014.pdf

[sluckey]

On a side note, wouldn't you guys expect to see more than ~390v from a Hammond 272FX 300-0-300 PT?
https://www.mouser.co.uk/datasheet/2/177/200-1390014.pdf

That PT will put out 424V unloaded. The heavier you load it, the more it will sag.

[pdf64]

‘Stable’ means that (almost) whatever conditions of input, load, control settings are applied, the amplifier exhibits no tendency to oscillate.

[SILVERGUN]

Do your other amps have 6V6 power tubes?