grid blocker question

[daz]

What is the lowest value that will combat *possible* blocking distortion and not affect tone noticeably? When i say "possible" what i mean is that i'm getting no blocking or oscillation that i can hear, but would prefer the safety of having them in place in case at some point it is there to a degree thats not obvious. But i don't want to change the tone. I had 100K's in place before but that muddied the tone some and i lost that jangle to a degree. Would 10k's do anything beneficial, or if not how low can you go? (no, thats not a morals question )

[ThermionicScott]

What is the lowest value that will combat *possible* blocking distortion and not affect tone noticeably? When i say "possible" what i mean is that i'm getting no blocking or oscillation that i can hear, but would prefer the safety of having them in place in case at some point it is there to a degree thats not obvious. But i don't want to change the tone. I had 100K's in place before but that muddied the tone some and i lost that jangle to a degree. Would 10k's do anything beneficial, or if not how low can you go? (no, thats not a morals question )

Depends on the input capacitance of the output tubes in question and the RC time constant of your coupling caps and grid resistors. And it's not an on-off thing, it's a matter of degree.

I would probably start by paring down the coupling caps and grid resistors so it's as "tight" as possible without losing bass response. On most Fender designs, you have some room to do that.

- Scott

[R.G.]

What is the lowest value that will combat *possible* blocking distortion and not affect tone noticeably? When i say "possible" what i mean is that i'm getting no blocking or oscillation that i can hear, but would prefer the safety of having them in place in case at some point it is there to a degree thats not obvious. But i don't want to change the tone. I had 100K's in place before but that muddied the tone some and i lost that jangle to a degree. Would 10k's do anything beneficial, or if not how low can you go?

Interesting question. The answer depends on the details. The first detail is - which tube and/or amplifier section? And here's why that matters.

All tubes have a parasitic capacitance from grid to plate and grid to cathode. The size of these capacitances determine high frequency response. Of course there are external capacitances like wire capacitance to chassis and so on, but those can be controlled. The tube internal capacitance is what it is, and is a consequence of the mechanical construction.

The capacitance to the plate from the grid is a feedback path, and as a result of the voltage gain, appears to voltage gain times bigger at the grid than it really is. So if you have 10pF from plate to grid, and a gain of 20, the grid appears to have 200pF to the source as a result of the grid-plate capacitance. The grid-cathode capacitance is added to this, so in this mythical tube, if that was, say 5pF, then the grid would behave like it had 205pF to cathode on it. That's the effective capacitance that lowers the high frequency response, and is what any source impedance and grid stoppers will react with to cut highs.

Sylvania (in their tube databook) lists the g-p cap for a 12AX7 as 1.7pF, and the g-K capacitance as 1.8pF. A 12AX7 stage with a gain of 20 would look like it had a capacitor to ground (cathode is generally bypassed) of 20*1.7 + 1.8 = 35.8pF. Notice if the gain is 30 (can be, some circuits), then that would be 30*1.7 +1.8 = 52.8pF, and as a cathode follower, simply 1.7+1.8 = 3.5pF.

If the preceding stage was a 12AX7, the effective plate resistance is 100K to 62.5K, so you might see a 50K source resistance out of a preceding 12AX7 stage into the next one. All by itself, that makes a rolloff of
F = 1/(2*pi*50K*35.8pF) = 90kHz for a gain-of-20 stage.

If you lose highs above 10kHz, it's noticeably less "sparkly" and gets progressively duller below that. You can easily hear the difference between 10kHz limited and 20kHz limited audio. So let's just pick 10kHz as a minimum, and you can adjust accordingly.

If you want to keep a rolloff with 35.8pF above 10kHz, then the source resistance can be no larger than R = 1/2*pi*10KHz*35.8pf, and that is 445K.
If you want to keep it above 20kHz, you need to make it no more than 225k.

Out of that you have to allow for the 100K to 50K of the preceeding plate, so call it no larger than 100K for 20kHz, no larger than 330K for 10kHz.

That's for a stage gain of 20. If the stage gain is 30, the corresponding numbers are 200K and 50K to 75K.

So yeah, using 100Ks in series with a 12AX7 stage with high gain will dull the top end a bit. Big surprise - you knew that.

The numbers say that to be sure, you need to keep it over, say, 47K. To be even safer on not losing anything in the audio range as a result of higher gain, call the maximum stopper 33K. It's quite important for these calculations that you put the stopper *right at the tube socket lug*, body of the resistor as close to the lug as you can solder it.

Now - will that correct any possible blocking or oscillation? ...er... possibly...

It's hard to say what possibly would happen.

One good thing to do is to make it not produce a bias shift if the incoming signal is bigger than would make Vgk positive. You can do that by making the gate conduct in the cutoff direction if the positive peaks make the gate conduct in the positive direction. It's a neat trick. You put some device that conducts to cathode when the grid is driven more negative than the grid/cathode bias voltage. For most 12AX7s, a string of diodes should work, two or three silicon diode junctions to get to 1.4V to 2.1V. Maybe even four. What this does is let the diodes conduct the excess charge that's put on a grid by grid conduction in a balanced way, so the capacitor leading to the grid does not build up the unbalanced charge that leads to bias shifts and blocking distortion. It's worth a try - or not, if you're not getting any blocking distortion now.

[daz]

Oh %$#@ ! This will take some reading ! I'll get back to it later when i have time and hopefully i will understand more. But for the record i'm talking about 3 stages of 12ax7 in a high gain circuit with 820R/.68uf cathode on both V1A and B with a 220k (split load, 120k/100k, 120k on the plate side) plate on A and 100k on B. 3rd stage and CF are the same as a jcm 800. Voltages are around 150-200 at V1 plates roughly. .022uf caps and one 470k/470pf filter going into V1B. I just found the 1M gain pot i've had in there forever is a 200K Labeled 1M tho !!! Sp i'm in for some mega gain reduction via a second gain pot till i find the right combo and install a set VD w/cap.

[Merlinb]

If grid stoppers really bother then you might prefer to use catching diodes instead, ala Paul Ruby's article:
18-Watter Buzz Info
You can do the same on 12AX7s, with a 4.5V zener and 1N4148 diode.

[R.G.]

If grid stoppers really bother then you might prefer to use catching diodes instead, ala Paul Ruby's article:
18-Watter Buzz Info
You can do the same on 12AX7s, with a 4.5V zener and 1N4148 diode.

I said:

One good thing to do is to make it not produce a bias shift if the incoming signal is bigger than would make Vgk positive. You can do that by making the gate conduct in the cutoff direction if the positive peaks make the gate conduct in the positive direction. It's a neat trick. You put some device that conducts to cathode when the grid is driven more negative than the grid/cathode bias voltage. For most 12AX7s, a string of diodes should work, two or three silicon diode junctions to get to 1.4V to 2.1V. Maybe even four. What this does is let the diodes conduct the excess charge that's put on a grid by grid conduction in a balanced way, so the capacitor leading to the grid does not build up the unbalanced charge that leads to bias shifts and blocking distortion.

That being a long winded way of saying the same thing. I may have gotten the diode voltages wrong.

[R.G.]

Oh %$#@ ! This will take some reading ! I'll get back to it later when i have time and hopefully i will understand more. But for the record i'm talking about 3 stages of 12ax7 in a high gain circuit with 820R/.68uf cathode on both V1A and B with a 220k (split load, 120k/100k, 120k on the plate side) plate on A and 100k on B. 3rd stage and CF are the same as a jcm 800. Voltages are around 150-200 at V1 plates roughly. .022uf caps and one 470k/470pf filter going into V1B. I just found the 1M gain pot i've had in there forever is a 200K Labeled 1M tho !!! Sp i'm in for some mega gain reduction via a second gain pot till i find the right combo and install a set VD w/cap.

I once did a similar preamp. I had to use 100K stoppers on all of them just to get the thing to not oscillate. But it did sound good.

[daz]

heres the thing. i've gotten this thing to sound so rich and amazing that i did the same tweaks to the first one i made like this for a friend a week ago and he flipped out at how good it sounds. (it's different than my current schematic) But the thing is, he doesn't gig and i have done a lot of gigs, quit when i got to around 50 years old, and i'm now starting to gig a bit again. He loves the tone and thinks it's god's gift to amplifiers. I feel the same except for one thing....what sounds phenomenal can in fact sound like cr@p at a gig. And i know this well. he just hears it at home and hears the rich amazing tone. i hear that and i want it, but i also know that unfortunately it won't work in a band mix. so since i will be gigging again i can't afford to deal with inconsistent tone. So i'm throwing in the towel. I just redid part of the pre like a 800 with a bit of plexi thrown in, and tomorrow i intend to finish that, tho likely without the cold stage which i never seem to be able to get past. Maybe because my voltages are lower than a 800. I may try it again and see, but i will more than likely go with a 800 style pre with a regular stage in place of the cold stage and possibly a split load 220K plate at V1A like i have now. In any case it's already sounding like my old 800's with that tone that just simply works and fits in a mix with no ill effects.

I spent a lot of time on this thing, but now that i will be playing out regularly for the first time in 6 or 8 years i need a workhorse. JCM's always worked well for me, so i'm going that way, only this time it will me my version of an 800.

[Bruce / Mission Amps]

I also gig regularly... the one thing I've always noticed is that players here at the shop or at home use too much flat mids and woofy bass for self accompaniment and on stage it simply sounds flat, dull or stodgy with little thrill.
An amp set up to be a little more middy with a bright edge always sounds better in the mix to me.

[daz]

I couldn't agree more. The mids give it body and supply the richness while the highs that sound sorta nasty at home give a nice jangle in a mix and no longer sound harsh like they do by itself. Thats why i always play to a CD at home to test tweaks i try. It's not the perfect solution, but i does give you a idea how well it will likely fit in the mix.

[Steve Conner]

I also gig regularly... the one thing I've always noticed is that players here at the shop or at home use too much flat mids and woofy bass for self accompaniment and on stage it simply sounds flat, dull or stodgy with little thrill.
An amp set up to be a little more middy with a bright edge always sounds better in the mix to me.

It's what you call the "wanker woofy drone tone", right? :-) Been there, done that...

I think the Tubescreamer is a great example of this principle. Sounds all harsh, fizzy and buzzy on its own, but when you stomp on it at a gig, it's just right.

Maybe one day all amps will have a big red switch that you can flip from "practice" to "gig". It would swap the woofy tone for a mid and presence boost, and increase the power by a factor of 10.

[Merlinb]

I said:

That being a long winded way of saying the same thing. I may have gotten the diode voltages wrong.

Sorry, I was waiting for the abridged version to come out!

[daz]

by the way, i truly believe that this is an agenda that many large companies use when designing amps with sales figures in mind. especially amps that they know are geared more to younger kids. They go for killer tone that works great at home but doesn't cut it live because they know the tone will sell the amp in the showroom. Older players aren't as easily fooled by that because they have been there. I remember years ago buying a crate VC30 because it sounded great and when i brought it home and set it next to my marshall and A/B'd them i thought to myself what a piece of %$#@ the marshall really is ! I sold it, then a year later after having countless horrible gigs with my new "amazing" amp, i sold it and bought another marshall just like the one i sold earlier ! lesson learned....well, almost. Guess i'm not telling anoyone here anything they don't know, but it still amazes me how an amp can sound so different in a band mix.

[R.G.]

Sorry, I was waiting for the abridged version to come out!

Abridged version?? Why use one word when 27 will do?

OK, so I'm long winded...

[Old Tele man]

Abridged version?? Why use one word when 27 will do?

OK, so I'm long winded...

..."descriptive" is how my Tech Report writing professor would call it.