Clarification of floating heater supply

[duderduderini]

Hi Everyone
When you have a scenario where the heater cathode voltage exceeds the limit in one half of a twin triode yet the other half doesnt as in the case of the aikido and srpp I note the advice is to elevate the heater supply which i understand.
What confuses me is that no reference is made to the cathode heater half which isnt over the limit.
It seems there is no advice on say using 2 triodes per channel, one dual triode sharing the "Upper valve" duties (and having the elevated heater supply) and the other sharing the lower half duty.
Can I assume this is the convention or can you just float at say 50V or some arbitrary figure so that both are elevated but within limits (if one triode is being used as upper and lower halves).
What is the correct method?
Thanks

 

[Tom Bavis]

You've got it - set the heater to a voltage that's within H-K limits for both sections. Maybe better yet is to split the two triodes across channels - upper halves in one bottle, lower in another. But you'll need separate heater windings to bias them optimally.

 

[Akita]

I still can't get the idea. Can anyone here explain more detail on how to float the heater?

 

[Poindexter]

Here's a schem:

The reference voltage is taken from B+ with a resistive divider.  Total resistance of the divider should be about as many kilohms as the volts of your supply; you want to draw about 1mA with the divider, to get a stable voltage.  If you have a center tap on the heater transformer/winding, you attach it to the junction of the divider.  If not, a 'fork' of resistors, 330Ω or so, goes to the two sides of the winding.

Questions?

Poinz

 

[Stixx]

Floating AC heaters I have done...
but how do you float a regulated DC heater supply...?
Also by applying the fork of two equal resistors, or by connecting the divider to the return side of the supply?

Thanks for any input!

 

[Tom Bavis]

I've always preferred the term "biasing" to "floating"... the latter seems to imply that it isn't connected to anything. With DC heaters, no need for any resistors - connect the heater bias to either side of the DC supply.

 

[gingertube]

I likewise have often "biased" the AC heater supply to a positive DC voltage.

In some schematics, I have also seen the heater left truely "floating" with respect to DC and just a capacitor used from one side to tie the heater to 0V AC wise (at higher frequencies anyway).
When might you use this rather than bias the heater to a fixed DC voltage? Is there any benefit or is the opposite true (not as good)?

Cheers,
Ian

 

[aardvarkash10]

thanks poinz!

I could never get that whole part of the circuit clear in my head... I only just realised what the resistor fork is doing!

 

[Akita]

Hi Poinz,

Thank for your explaination. So in your schematic, the heater have been floated to 47v.. right?

 

[Poindexter]

It is not floated.  Floating means that it is not referenced to the main circuit (the heaters have nothing directly to do with the audio amp; they just keep the cathodes hot) at all.  You are correct, it is referenced to a DC voltage 47Vdc above audio circuit ground - roughly halfway between the two cathodes.

You can't float your heater supply.  It'll reference itself all over the place, and you will have weird fizzy humming issues.

How do I know this, hm?

Aloha

Poinz

 

[Stixx]

Thanks Tom!

I should have used the term elevating instead of floating...