directly heated valve biasing

[Nigel Goodwin]

also i was wondering about.... iv forgotten the name of it. its the resistor that's on the input to stop oscillation and pickup radio interference! well anyway what ever its called do i really need it and how do i work out the value?

It's called a 'grid stopper' - it's not something you need to calculate, just use something in the ten's of kilo-ohms.

 

[bychon]

I just checked my old tube manuals. The pentodes I'm finding have half a puff or less Cgp. A grid stopper would have to be in the multi-megs range to work at audio frequencies. You can give up on the grid stopper idea or add a capacitor intentionally if you really want a pole there, but it's not necessary.

 

[JimB]

remember i want distortion!! and 3 battery's doesn't fit in an effects pedal.

OK, use what works for you.

how did you choose the 150k resistor and 0.1µf capacitor for the future??

Inspired guess work based on experience from many years ago.

and im guessing the output isn't connected to the screen grid?

Quite right, I just could not be bothered to go around dotting the junctions, I hoped (quite correctly) that you would understand.

I disagree with the diagram in post 16.
Your third grid is in 2 seperate parts and internally connected to pins 3 and 5, the cathode (heater), not Grid 1. (That's the end of, "where do I connect grid 3"!)

You are quite correct, the suppressor grid should be connected to the cathode, not to the control grid.

JimB

 

[Nigel Goodwin]

I just checked my old tube manuals. The pentodes I'm finding have half a puff or less Cgp. A grid stopper would have to be in the multi-megs range to work at audio frequencies. You can give up on the grid stopper idea or add a capacitor intentionally if you really want a pole there, but it's not necessary.

Perhaps you're too young?, grid stopppers are to prevent RF not audio frequencies - they help to prevent stray RF entering the circuit, and help prevent RF oscillations. It's something you add as a matter of course, not really an 'option'.

 

[unclejed613]

to figure out the cathode bypass capacitor, choose a rolloff frequency (usually 20hz, but for a guitar it can be higher, maybe 100 to 300hz). then find the capacitance that has a reactance equal to your cathode resistor at that frequency. capacitive reactance is Xc=1/(2πfC) so rearranging to find the capacitance is C=1/(2πfXc). so using your 3k resistor from the above example, and choosing a rolloff frequncy of 100hz, we get 0.5uf

ok, i forgot the low E is 80hz, so make the bypass 1uf.

 

[the cracken]

OK, use what works for you.


Inspired guess work based on experience from many years ago.


Quite right, I just could not be bothered to go around dotting the junctions, I hoped (quite correctly) that you would understand.


You are quite correct, the suppressor grid should be connected to the cathode, not to the control grid.

JimB

iv built the amp on bread board the only thing i guessed and didn't calculate was the grid leak resistor!! can anyone tell me how to work out the right value?
it works though there is a load of 50hz hum i guess because its not in a metal earthed case! and there is high frequency oscillation which im not sure how to get rid of! maybe increase the resistor on the grid?(100k) (iv played with it and put the screen grid from the anode to ht as seen in the pic and it slightly better but still oscillates )
here's what iv got just to clarify!
**broken link removed**
thanks very much for all of your help

 

[Nigel Goodwin]

Well you don't have any decoupling capacitors on the HT supply, and you've ignored the advise about a grid-stopper. The grid leak resistor is determined by your input impedance requirement - 1 meg would be reasonable for a passive guitar. G2 should connetc to HT, NOT to the anode.

 

[bychon]

1) You've got grid 2 on the plate, with no resistor and no filtering. Your tube specs say grid 2 wants 15 to 22.5 volts...much like what a battery supplies.

2) Yes, I am too young to remember RF grid stoppers...I'm only 60...but the principle of Miller effect works and it's useful. If cracken puts a few puff from plate to grid one, it will cause an audio frequency roll off. The last time I did this, the answer was 18 puff, but that was a triode. Whatcha got in yer junque box, cracken?

3)Increasing the grid resistor will have the same effect as adding picofarads, but I've already told you, that resistor would have to be in the multimeg range to have any effect at audio frequencies.

4) 50 Hz hum is a sure sign it's not in a metal box.

5) A grid leak resistor is used to develop a bias voltage on grid 1. A cathode resistor is used to develop a bias voltage on grid 1. I don't know how to calculate a grid leak resistor. Good thing you have cathode bias and don't need it.

 

[bychon]

WTF?? Somebody added G2 filtering while I was typing...but G2 is still not connected to the battery.

 

[the cracken]

im sorry everyone i was up at like 4 last night lol tinkering

i did change the circuit. first i tryed the grid stopper like you said (100k) but i had the screen grid connected straight (no resistor/cap) to the anode to act as a triode. it work but not as much gain as having it connected to ht. i made a little mistake in the diagram (now i have it connected to ht with resistor/cap)
i took the screen stopper out to see if it made a difference and it didnt stop the oscillation. i had a look at the response and it was at 4v 100hz and 2v 100kh it started rolling off at about 20kh. is that going to be alright?
ill put the grid stopper back in to see if it made a difference and its still ringing.
how do i decouple the ht supply???

i double checked and i think iv got it all in there apart from the ht decoupling!
**broken link removed**

thanks for all the help its really appreciated

 

[bychon]

To "decouple" the HT supply, put capacitance from the HT line to the ground.

 

[the cracken]

To "decouple" the HT supply, put capacitance from the HT line to the ground.

what size capacitor?
thanks for the quick reply

 

[the cracken]

ps, 22.5 volt batteries exist, and with an idle current max of 100 to 240 microamps, this is a viable option. They look a lot like 9 volt rectangular batteries. In fact, I expect this tube was designed for use with that battery.

I see also grid 2 current as 25 to 60 microamps. That's how you find the value for G2 resistor. It calculates like about 470k.

A guitar puts out more than the listed 1.2 volts, so beware, you might get distortion!

do i need to take the grid 2 current into consideration when calculating the cathode current or is it small enough to ignore?? thanks

 

[unclejed613]

put a 100uf cap across the 22 (or 18) volt battery. that's how you decouple the ht supply.

 

[unclejed613]

somebody want to explain to me why several messages just happen to appear in the time it takes for me to type a single line? and some of those messages are over 1/2 hr apart?

 

[bychon]

I met that problem, too. Website hinky?

 

[the cracken]

im in a time warp

 

[unclejed613]

where's Rod Serling? i expect him to step out of a shadow any minute and tell me i'm in the Twilight Zone...

 

[skajam66]

The cathode resistor is there for biasing. If you bias a different way you can remove the cathode resistor (and capacitor). See attached... This way you don't need a floating supply. You don't have to use batteries either - any source will do
**broken link removed**
This biases the grid to -1.5V. If you want a lower (less negative) bias voltage, split the 25R resistor into two and tap the 1M resistor from the junction of the two

 

[JimB]

Did someone mention a time warp?

JimB