300b SE

[Paul Barker]

I had to go back to the drawing board, due to the higher B+ required for the battery bias. So I calculated the max b+ I want to set with Black Gates, which is 510v. My main motive is to reserve the 1500v poly caps for a 40w gm70 amp.



But first I’m building the 300b amp, so here is the result for the 4p1l:-

So the 4p1l operating point is on this chart. Notice I am running the 75c1 at 60mA to get max reduction of rK



Those curves are for ug2 70v, mine shall land betweer 75 and 80v. Realistically 80v. But the curves at 80v may not change much. Nothing is exact. Best endeavour is what British Gas taught me to state! These are the nearest curves which are legible I have found searching the web. If you find some curves as legible as these g2 80v please send me the link.

I am also using the +15v g3 which demonstrates cleaner curves on my web search. So I am at variance from these curves hopefully for the better.

[Paul Barker]

The transformer has a 866a filament supply for a pair.



My transformer has 2.5v 10 amp so enough current to share. Please tell me the two filaments can share the same supply ? Anyone ?

My filament winding isn’t centre tapped, but I can combine them to b+ Can I? My brain is rust these days.

I realise I’ll have to delay the HT secondary, which may require a separate filament transformer. If so I’ll so do.
Yes I think thats the only way to hold back the power to the main transformer while feeding the separate transformer for a minute.

[Nick]

AFAIK Yes, to both questions.

[Paul Barker]

AFAIK Yes, to both questions.

Thank you. Its coming back to me. This is good for my mind.

[Paul Barker]

Now I see where the curves came from; Ale Moglia.

So now I see what Nick suggested in diagram. As Nick said the 15k can go to ground effectively.



The only parameter which I have to change is the B+, because I like to use Black Gates.

But the anode volts must go back to 250 to keep the load at 15k.

So is it the case if I rc drop my b+ 500v to 400v then use his gyrator.

Well I’ll give it a try and see what happens.

[Nick]

Though, I would put the 15k to b+, and so reduce the current through the current source, and make it a actual current source, remove c2 and the voltage ref before it. But that works as well and is a entire circuit to copy rather than just me making off hand suggestions

[Paul Barker]

Though, I would put the 15k to b+, and so reduce the current through the current source, and make it a actual current source, remove c2 and the voltage ref before it. But that works as well and is a entire circuit to copy rather than just me making off hand suggestions

Ah so you have made it simpler for me. Yes I’ll go with you’re suggestion. So I only need a dn2540.

[Nick]

Yes, but you will need to work out the numbers again. The sense resistor for the depletion mosfet will need altering to give whatever your target current is minus (with the 15k to gound its plus) the current through the resistor. With a 400v b+ and 15k load r and 17ma as the target current you will want another 7ma through the CCS. The circuit displayed may be simpler as the anode voltage is fixed by the mosfet and the load resistor by the 15k.

[Paul Barker]

Now I am confused. If I put the 15k where it is shown on the schematic, and I use the whole schematic by dropping the b+ of 500v for the 300b to 400v, for the 4p1l stage the R value should be 220ohm as he has used? Then at the cathode I can add 60mA from the vr screen voltage to calculate the rk bias resistor for .077 mA. Of which only 0.017 travels through the valve.

[Nick]

As that's using cathode bias by including the 5r in the heater circuit i would leave that alone and take the vr directly to ground.

[Paul Barker]

As that's using cathode bias by including the 5r in the heater circuit i would leave that alone and take the vr directly to ground.

I would have agreed to you. Especially since we don’t exactly know the current flow through the VR.

But something else has changed. I asked Rod to make the filament supply for 4.2v. Anyway I calculated 10.6ohm. But I looked at the curves which don’t t have to be exact in my view. We’re OK with less bias so the voltage on the schematic isnt an exact target. I’ve set a 10 ohm resistor and I’ll put the vr to ground through the negative of the bias supply.

[Paul Barker]

I’ve decided to keep it simple. Put the VR straight to ground as Nick said above.

But further not use the gyrator. Just a 15k load resistor. There fore I can use the 500v supply.

Meanwhile brand new pair 4p1l just arrived.

Secondly, as the chassis has R core 20w 4k3 into 8 ohm opts already mounted, I’ll use those. The resultant data ;



Small amount extra power small amount extra 2h d %. But not a deal breaker.

I’ll use the autobias and bypass connecting at 8 ohm secondary grounding across the secondary to the earth. Chosen to make rk simple 1k.

At the bias voltage sensitivity is 1.4v peak, not unreasonable for modern sources.

[Paul Barker]

A Baking Tray galvanised steel £3 at the charity shop. Seems a neat sise for a simple power amplifier.

Off board power supply, off board preamp.

Might find out where to get a galvanise primer then spray it with that rattle can I bought above.

Probable best do it after all holes made.

[Paul Barker]

This colour

[pre65]

Nice colour, looks a bit like British racing green ?