Math and how my hair caught fire... (grid attenuation)

[Reeltarded]

Happy New Year! My head hurts!

For the life of the chance gimp I have in Schrodinger's cat box, where do I find a formula for the corner frequency for a 12ax7 vs a grid stopper?

I am aiming for the 5-8khz area. I know I am looking for a big number. I probably shouldn't be approaching it this way, but it's how you learn.. if you live.

Muhahahahaha.. and such. I don't want to knock my amps teeth out, I just want it to grin a little less.

The chance gimp thanks you.

[ToneMerc]

Here's something to toy with.

http://www.ampbooks.com/home/amplifier- ... d-stopper/

TM

[Reeltarded]

Do you rock? Yes, yes ya do! Thanks!

Trying to tame some shassh, and make this thing sound old and mellow.. like me..

but mellow.. sorta. Uhoh.. question.. another post in a minute.

[Structo]

Is someone still a little inebriated?

The input grid stopper's main purpose is to stop radio interference from entering the amp.

You can go higher and attenuate high frequencies but it can sound bad to go too high on the resistor value.

Fender and Marshall typically use 68K.

Dumble uses 22K-33K.

[selloutrr]

If you had a specific tube in question I could put it on the Amplitrex and adjust the test specs and run curve traces until you get the desired result.

With that said you could build a test jig and use a scope to dial in the values you are looking for. Using the Tubes spec sheet so you don't exceed it's max limits then dial it to behave as you wish.

[Reeltarded]

Ahh.. yes, I need a scope. Hmm.. I wonder what's up at garage sales around here?

[martin manning]

Nah, just try some values from 150-220k. That should get you in the ballpark. 220k should get -3dB at around 9kHz for the typical 12AX7 stage.

[Reeltarded]

I am going to go melting on it now. Thanks for the input!

[martin manning]

BTW, what else is around this grid stopper? Is it an input stage with a 1M to ground before the grid stopper, or something else?

[Reeltarded]

V1b on a Marshall 4 holer i bought to solder on..

1m then 68k(n) 34k (b) input resistors. Expected normal channel from there (820/250), bright is 270k plate, 2.7/.68, mixers are 470k bypassed with 470s both channels, v2b is 820 bypassed with 10uf. All 22n couplers.

This isn't guess work, huh Martin? 220 touched it, 270 is what I am starting at volume with in the morning. Little work on the normal channel cathode arrangement should make it a great amp.

The 270 plate and the added 68k over input 1 did some nice things to the grind, but made it harsh at volume! Very useful though.

This is fun!

[rdjones]

Bear in mind that the corner frequency will change a bit with the gain of the stage/tube, and is also affected by the tube's capacitance.
(The Miller effect)
This will cause one tube to sound darker than another when a larger stopper is used.

rd

[Reeltarded]

The frequency moved up when I jacked up the plate and input (by ear)

I am sure I just gave back some of the gain I set out for, but it still promised to push back at volume. I could use a few more capacitor values. I have robbed all my donors. Aww! I want .1 on the PI I think.

[martin manning]

I invite you and everyone else to play with this: https://tubeamparchive.com/viewtopic.ph ... highlight=

On the Input Data worksheet, the grid stopper value is cell B14, and the HF roll-off corner is cell L17 (~25 kHz for the nominal 68k stopper). Have a look at the frequency response plot to see where that and the other breakpoints are in the full response.

I just uploaded a very slightly modified version (some text editing and setting default to 12AX7). Read the notes page to familiarize yourself with the assumptions and limitations.

[Structo]

Now if it only came with instructions.....

Too far above my head for sure.

[Colossal]

I invite you and everyone else to play with this: https://tubeamparchive.com/viewtopic.ph ... highlight=

On the Input Data worksheet, the grid stopper value is cell B14, and the HF roll-off corner is cell L17 (~25 kHz for the nominal 68k stopper). Have a look at the frequency response plot to see where that and the other breakpoints are in the full response.

I just uploaded a very slightly modified version (some text editing and setting default to 12AX7). Read the notes page to familiarize yourself with the assumptions and limitations.

Thanks again for this Martin. That is one badass utility. I started out making my own a few years back but upon discover of yours (v7.01), I gave up and have used yours henceforth. No sense in trying to reinvent perfection already achieved! Well done! :clapping hands: