fixed bias

Paul Barker · Post by **Paul Barker** » Sat Sep 14, 2013 7:06 pm

Thinking about it the amplification of the driver won't be affected by different bias voltage as the two resistors top and bottom from grid to ground are the dominant factor in the thevenin calculation.

You'll be fine.

So you just have to get the total dc resistance of these two resistors to the amount allowed by the output valve, and live with that as your load. Size your coupling cap accordingly.

Watch you bias your output valve so that it doesn't go into A2 as there would be nasty blocking distortion, and havock to the bias voltage.

ed · Post by ed » Sat Sep 14, 2013 7:31 pm

I think it's best to leave this here........

Only Nick came close to understanding my question........

Yes!!!!! replace the cap with a wire...it's sitting in the middle of the wire, so, if you remove the cap...you are left with a wire.....which is why I couldn't understand when it was suggested remove the cap because what is left is a wire...the load path to ground....

I never envisaged that the load would be grounded through the bias network, which is what I think you guys envisaged....

my bad for not explaining where I was coming from......

so cap to ground to block the bias or resistor to ground to load the bias instead.

I will replace the reverse cap with a 10k resistor(as per drawing number 2) and see if there is a change.

Apologies for the unclear description.

Post by **Nick** » Sat Sep 14, 2013 7:50 pm

Regarding Nick 2nd use of the cap. Not required imho. You want the valve to take the AC you don't really want any shunt caps stealing some of it away.

Not required no, but its a case of the AC grid leak to ground being through a network of resistors and sliding contacts, or via a controlled signal path through the cap.

The valve will need its grid to be referenced to ground, and the max value of that will be set by the spec of the valve.

Mike H · Post by **Mike H** » Sat Sep 14, 2013 7:57 pm

Yes shirley

if you remove the cap, the result is an empty space, not a wire shorting to ground, lol

I agree with Nick the cap will give the signal an AC path to ground, otherwise it will have to go via the bias presets paraphernalia. Not a huge problem I don't think provded these are grounded close to the valves' circuit's own earthing point.

HTH

Paul Barker · Post by **Paul Barker** » Sat Sep 14, 2013 8:35 pm

Mike H wrote:Yes shirley if you remove the cap, the result is an empty space, not a wire shorting to ground, lol

I agree with Nick the cap will give the signal an AC path to ground, otherwise it will have to go via the bias presets paraphernalia. Not a huge problem I don't think provded these are grounded close to the valves' circuit's own earthing point.

HTH

When cap removed it is replaced by a fixed resistor. the resistance in the bias mechanism is parallel to this resistor and of greater resistance so the ac finds ground through the fixed resistor predominantly.

As said already the pair of series resistors have to satisfy the valve spec the choice of coupling cap and the amplification required.

Post by **Nick** » Sat Sep 14, 2013 9:25 pm

When cap removed it is replaced by a fixed resistor. the resistance in the bias mechanism is parallel to this resistor and of greater resistance so the ac finds ground through the fixed resistor predominantly.

Ok, but then you are wasting a lot of bias voltage, for the resistor to ground to be less than than the source resistance of the bias supply, the bias supply is driving into that resistance so you start to demand more current from the bias supply and need to start with a greater -ve voltage. Seems to be a lot of work to avoid a cap.

But that's just IMHO, YMM (and probably will) V.

ed · Post by ed » Sun Sep 15, 2013 8:50 am

Nick wrote: Seems to be a lot of work to avoid a cap.

But that's just IMHO, YMM (and probably will) V.

that was the original question in post #1, i.e is there a better way....

the rest of this thread revolves around the way my silly brain views the world, i.e

Yes shirley if you remove the cap, the result is an empty space, not a wire shorting to ground, lol

in my brain, when I remove the cap what remains is a wire because that was what was there before I put the cap in.....it seems I don't quite view the world in the same way as you guys...

the alternative, with a resistor, I already covered the parallel resistor values in an earlier post...which was my suggestion for a possible alternative to the reverse cap......

so, back to the original question.....I like Nick's response quoted at the start of this post.....'we always latch on to the answer we wanted to hear'

edit: another thread populated with posts caused by bad terminology and misunderstanding....sorry guys, I really must try harder to describe my questions better.

Paul Barker · Post by **Paul Barker** » Sun Sep 15, 2013 9:47 am

Nick wrote:
When cap removed it is replaced by a fixed resistor. the resistance in the bias mechanism is parallel to this resistor and of greater resistance so the ac finds ground through the fixed resistor predominantly.
Ok, but then you are wasting a lot of bias voltage, for the resistor to ground to be less than than the source resistance of the bias supply, the bias supply is driving into that resistance so you start to demand more current from the bias supply and need to start with a greater -ve voltage. Seems to be a lot of work to avoid a cap.

But that's just IMHO, YMM (and probably will) V.

Yes. I was answering from the premis Ed wanted to rid himself of the cap.

Personally I would direct couple it from a Cf floated on a +/- supply by a similar bias mechanism but also no cap.

This discussion makes auto bias look very easy in comparrison. No wander it's so popular.

Paul Barker · Post by **Paul Barker** » Sun Sep 15, 2013 9:51 am

ed wrote:
edit: another thread populated with posts caused by bad terminology and misunderstanding....sorry guys, I really must try harder to describe my questions better.

Not at all. It's good to get different people discussing these issues and we all learn from each other. We all have a tendancy to have a favourite end, and we fit our discussion to arrive at that end. Because of that fixation we may miss the blatantly obvious which others who have different ends in view can see clearly and help us to see our blind spots. Then of course we can help them find their blind spots and we all grow.

ed · Post by ed » Sun Sep 15, 2013 7:40 pm

Paul Barker wrote:If you give me the driver valve it's anode load resistor it's cathode resistor and whether it is bypassed I will work out for you the effect on gain of the possible extremes you may encounter.

I dug this out for interest sake. I can't find the notes that went with it at the mo but I do recall it was with a uni-t and a fluke, which didn't entirely agree with each other...when I get the breadboard back together I will do some more measurements with the digi scope which I know to be accurate.....
this was with the config shown in post #1 so the load is 100k with a 1uf reverse bog standard electrolytic. I do remember being disappointed at the time because the gain wasn't anywhere near that implied by Broskie.

I imagine all at 1k signal.

full story circuit and valve/gain variations :

http://www.tubecad.com/2006/05/blog0064.htm

Paul Barker · Post by **Paul Barker** » Sun Sep 15, 2013 8:32 pm

The driver is really a CF so the load of 100k is fine. I don't know why the output from the CF is taken from the bottom of the bias resistor and not from the cathode but if Broskie designed it like that I would be in the wrong to change it. But don't understand it.

All the voltage gain is in the first stage, should be enough.

ed · Post by ed » Sun Sep 15, 2013 8:55 pm

Paul Barker wrote:
All the voltage gain is in the first stage, should be enough.

it's enough for me.......I said disappointed because in my config I was expecting a gain of 39 so 30 was a bit shy.

audio-talk

fixed bias

#16

#17

#18

#19

#20

#21

#22

#23

#24

#25

#26

#27