I just calculated the value for the KT88(tetrode connected) grid stopper for 20kHz roll off, and it seems very high (usually its from 5k-10k in guitar amps).
Here is the math:
http://www.drtube.com/datasheets/kt88-mov74.pdf
Gm@g2=250V = 11,5 mA/V
rp = 12k
mu = Gm*rp = 138
I use 5k primary on a SE transformer so
Av = mu*Rl/(Rl+rp) = 138*5k/(5k+12k)=v 40,6
and the Miller capacitance is Cg1(all less anode) + Cg1a*(Av+1)=
= 16pF+1.2pF*(40,6+1) =
=66pF
so the so 20kHz the R in this internal RC filter should be
R= 1 / 2*pi*f*C
= 1 / 2*pi*20kHz*66pF=
=121k
Where did I go wrong?
Thanks,
Miha
Here is the math:
http://www.drtube.com/datasheets/kt88-mov74.pdf
Gm@g2=250V = 11,5 mA/V
rp = 12k
mu = Gm*rp = 138
I use 5k primary on a SE transformer so
Av = mu*Rl/(Rl+rp) = 138*5k/(5k+12k)=v 40,6
and the Miller capacitance is Cg1(all less anode) + Cg1a*(Av+1)=
= 16pF+1.2pF*(40,6+1) =
=66pF
so the so 20kHz the R in this internal RC filter should be
R= 1 / 2*pi*f*C
= 1 / 2*pi*20kHz*66pF=
=121k
Where did I go wrong?
Thanks,
Miha
Why would you want to impose a 20kHz roll-off with a grid stopper resistor? The grid stopper is there to prevent parasitic HF/VHF oscillations and for no other reason.
I didn't look at your math but the numbers seem ok, use a 1 - 10K grid stopper and if you need to introduce a roll off for feedback stability you can do it in the driver stage.
Bandwidth in excess of 20kHz is a good thing if it doesn't come at the expense of something else like stability.
Most of my amps are designed to be flat to something considerably greater than 20kHz if possible.
I didn't look at your math but the numbers seem ok, use a 1 - 10K grid stopper and if you need to introduce a roll off for feedback stability you can do it in the driver stage.
Bandwidth in excess of 20kHz is a good thing if it doesn't come at the expense of something else like stability.
Most of my amps are designed to be flat to something considerably greater than 20kHz if possible.
Last edited:
I don't think you did go wrong- I think you've got it right.
12.1 k Ohms would put the response down 3dB at 200 kHz, which is usually plenty low enough. You're not trying to set the upper frequency open loop response with the stopper resistor- only trying to prevent oscillation at (usually) several MHz by reducing the gain above the intended passband.
Indeed, 5-10 kOhms seems about right based on your calculation.
12.1 k Ohms would put the response down 3dB at 200 kHz, which is usually plenty low enough. You're not trying to set the upper frequency open loop response with the stopper resistor- only trying to prevent oscillation at (usually) several MHz by reducing the gain above the intended passband.
Indeed, 5-10 kOhms seems about right based on your calculation.
High value grid stoppers are often used in guitar amps, partly to cope with poor layout involving long leads, and partly to limit the HF to less than the full audio range. In my view this is poor engineering. It makes them very sensitive to details of valve capacitance, and so encourages tube rolling as a crude way to adjust capacitance.
In hi-fi amps smaller grid stoppers are used simply to avoid VHF oscillation. Tone controls can be added separately if required.
In hi-fi amps smaller grid stoppers are used simply to avoid VHF oscillation. Tone controls can be added separately if required.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.