845 Grid stopper value

[Salas]

What do you think is the largest grid stopper value allowable for 845 tube in SE with 10k plate load OPT and a 400k Rg? I have seen from nothing to 1k. I ask this bcs from 1k to 6.8k I saw positive results for overall noise, but I don't want to damp the highs. 845's grid to anode capacitance is 11.5pF times 5 gain for Miller. It must not restrict the audio bandwidth up to 20kHz. No FB is used so to get some bandwidth back.

[gingertube]

Salas,
I've never found it necessary to use a grid stop on an 845. I take it that you are using cathode bias and not fixed bias, otherwise that 400K Rg is too high.

6K8 of grid stop will not influence audio at all.
845 Input capacitance will be mu x Cga + Cgh + strays, assume about 70pF.

With 6K8 that gives a high frequency roll off at over 300kHz - of course to work out the true roll off you need to add in the output impedance of the driver stage.

Cheers,
Ian

[tubelab.com]

I run my 845SE in A2. The grid current hits 50 mA on peaks, so that means no grid stopper. The amp has been operational for 4 years without issue.

[Salas]

Originally posted by gingertube
Salas,
I've never found it necessary to use a grid stop on an 845. I take it that you are using cathode bias and not fixed bias, otherwise that 400K Rg is too high.

6K8 of grid stop will not influence audio at all.
845 Input capacitance will be mu x Cga + Cgh + strays, assume about 70pF.

With 6K8 that gives a high frequency roll off at over 300kHz - of course to work out the true roll off you need to add in the output impedance of the driver stage.

Cheers,
Ian

Its cathode bias, and the driver stage has 1k output impedance. There are some issues with the UHT supply power up, and the stray capacitance of the OPTs and chokes. Without grid stop they tend to put the 845 in ''hss-hss'' cyclic noise song, during startup mainly. It settles when warm, but it can sing again erratically. I run some sims and there is no audio bandwidth limiting issue, you are right. Maybe I will try 10k carbons and ferrite beads. Beads on anode wire from OPT too.

I don't think it can go A2, but with 2W resistors I assume its going to test fine. Maybe I will shroud the chokes with copper foil too, as a last resort. I already got the singing cut down a lot.

[Salas]

Hmmm grid stoppers and anode stoppers and good coax cabling dressing, cut hum and that peculiar noise very low, but it still triggers from somewhere. I will see to use NTC thermistors for a milder power up.

[Zoran]

try to find Cdyn dynamic capacitance
Cdyn=Cgk+A*Cga where A is absolute value of stage amplifinig factor.
Then calculate R=1|(2*pi*Fo*Cdyn)
You can choose Fo like -3db
If You choose 131KHz -3db then You will have loss of 0.25db at 20KHz
approx.
(For A2 class Cdyn could be found from diferent eq...)
And that is valid if You have low enough impedance, and adequate power from driver stage...

[Salas]

OPT leakage inductance is the culprit. Noisy in such UHT. I compensated it. No fireplace sounds anymore. Thanks.