Tube Heater Circuits Question

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

alexc

Well-known member
Joined
Sep 21, 2004
Messages
2,571
Location
Hobart
Hi - I have a question about tube heater circuits ..

If I have dc filtered heaters, is there any way to 'elevate the voltage' similar to doing so with ac heaters and a tap off the hv ?

I've tried floating heaters  ..  with no connection to common ground  .... very noisy  for me  ....  I'm not sure if it's possible to 'reference' the dc  heater to anything but a common ground (which of course, unfloats the heater circuit and references to 0V )

I ask because I like cathode follower type circuits  ...  which should have some heater elevation, such that the heater-cath max voltage spec isn't hopeflessly violated.

.... but I like dc filtered heaters on small signal tubes as well

Is there a solution ..  to have a heater with one side at 70V and the other at 76V  ?

or is it simply to run both schemes    ...  dc-filtered heaters for general stuff  as well  as  ac-elevated heaters for any cf/cascaded small signal stuff ?

At present I just do dc filtered,  but arrange things so as not to get too far from the spec .. and just 'don't worry about it'.  :)

It does limit things a bit, tho'.

Am I not just misinformed ?

Thanks
Alex C

 
> If I have dc filtered heaters, is there any way to 'elevate the voltage' similar to doing so with ac heaters and a tap off the hv ?

Why not?
 
Not sure, really  ...  I'm thinking that I can 'reference '  (ie. connect  the 'local' ground of) 

my  'floating'  dc filtered heater circuit  (which has it's own traffo sec winding -> bridge -> R-C filter)    ... 

to a voltage,  say +70V or so  from a divider circuit off the B+.

Would I make a 'simulated transformer winding CT'  with a pair of high value resistors across the +, - wires of the heater circuit and connect the centre node to the +70V line ?  Or not  ?  It couldn't be as simple as direct connecting it, right?

I know I should do it on paper with kirchoff's law and so on and work it out for myself, but you know, I thought someone else might have done similar.  I can't recall seeing something like  in my lib of circuits.

Thanks

 
Here is a thread with discussion of an elevated dc supply..... I've never actually seen it done..... I've been happy so far about ac elevation  in my limited journey and never thought about it further.....

Interesting...

https://www.diyaudio.com/forums/tubes-valves/235888-help-elevating-heater-supply.html

I attached the schem referred to in  the thread discussion for quick reference .....



btw....what do you mean by

alexc said:
Is there a solution ..  to have a heater with one side at 70V and the other at 76V  ?

Are you speaking of different elevations for different tubes???

 
Thanks Scott for the reference  ...  just checking it out now.

Sorry no  - I mean I have 'floating' heater with 2 wires  ..  one at +6V with respect to the other wire, which is the local 0V rail for the dc heater supply circuit.

There are no connections on the heater circuit to the common ground of the system from which the audio circuits get their 0V rail.  Just a parallel string of heaters .. 

What I want is  :eek:  to have those two heater wires sitting at +70V and +76V respectively with respect to the common system ground which is what the cathodes are eventually returning current thru.

I'll try and draw something up tomorrow .. I've done my bench time for today :)  Like I said, I may well be completely wrong!

I always had trouble with floating supplies  ..  sometimes quite dramatic.
 
Ah, John Broskie type thread ...  the circuit attachment is awol.

I guess I will need to make a dc-dc amplifier and drive the heaters in a star balanced configuration with servo control.

Sorry, my attempt at humour - I very much respect John and his work for many years!

 
Yep - thats what I'm talking about ...  let me go thru some more!

Looks like referencing one side of the floating heater to a dc voltage  by connecting directly

with

isolating that now elevated rail from any other heater circuits  via the centre node of a pair of low value resistors

and

the diode strings are not related, I think - I've read about this method of standing off 'system ground' from 'house ground'

SO - this looks like it. I'll give it a try tomorrow.  Hopefully it will not be noisy.

Cheers

 

Attachments

  • Heater biasing.pdf
    301.8 KB · Views: 12
alexc said:
Yep - thats what I'm talking about ...  let me go thru some more!

Looks like referencing one side of the floating heater to a dc voltage  by connecting directly

with

isolating that now elevated rail from any other heater circuits  via the centre node of a pair of low value resistors
You don't need the two 100r resistors anymore. Just connect the output of the voltage divider to the negative of the 12V circuit. You may want to increase the value of the cap (0.047uF) across the 47k; the heater circuit likes a low impedance connection to ground.
 
Thanks gents, that's just what I was looking for.

Now I can go to town with the higher cathode voltages  ... but in a remote-psu dc-heater context.
 
I do this all the time in my EZTubeMixer designs because the output SRPP stage has one cathode of a double triode at half the supply volts and the other a few volts above ground. Because of this I usually elevate the heaters to about one quarter of the HT supply. To do this you can use a simple decoupled resistive divider. However, many tubes have a specified minimum heater to cathode resistance, usually a few tens of Kohms. So I make the bottom leg of my divider 22K. This means the top leg has to be 66K to get a 4 the 4 to 1 divider I want. For power dissipation reasons I make this from two 33K resistor in series. The 22K is decoupled with a 10uuF electrolytics.

Here is a link to my HT250 PCB schematic that shows the technique:

https://drive.google.com/open?id=0B_n67A1hN3qtVlAxdXNlWEJwUUU

Just connect one side of the floating heaters to the 75V output.

Cheers

Ian
 
ruffrecords said:
However, many tubes have a specified minimum heater to cathode resistance, usually a few tens of Kohms.
Do you know why? I understand that the heater-to-ground impedance must be low, for want of forbidding whatever noise to pollute the cathode circuit, but I don't understand the need for galvanic continuity between heater and cath... :eek:
 
ruffrecords said:
Here is a link to my HT250 PCB schematic that shows the technique:

Ian

Ahhh....I've actually used the PDF as a reference! I was so caught up in doing AC heaters and trying to learn your crc concept in general that I never realized/remembered yours had the  example of using DC and elevation......  I was looking all around last night for an example and had it on my desktop all along.... :-[
 
abbey road d enfer said:
Do you know why? I understand that the heater-to-ground impedance must be low, for want of forbidding whatever noise to pollute the cathode circuit, but I don't understand the need for galvanic continuity between heater and cath... :eek:

I read in Merlin Blencowe's Hi-Fi book that the minimum heater to cathode resistance spec (pretty much never stated in datasheets) is a noise consideration as well. Essentially, there will inevitably be small current from heater to cathode because they are so close, and greater resistance between them will allow for this current to generate a larger noise voltage.

Maybe I am misremembering, but I think that's right!
 
ComodoComplex said:
I read in Merlin Blencowe's Hi-Fi book that the minimum heater to cathode resistance spec (pretty much never stated in datasheets) is a noise consideration as well.
  I would substitute impedance for resistance. Noise is AC.

  Essentially, there will inevitably be small current from heater to cathode because they are so close,
  This is very debatable... There's several possible causes of current circulating from the heater to the cathode, both DC and AC.
DC, caused by a) the back of  cathode acting as an anode for the heater, b) leakage of the filament's insulating material, is not an issue in normal operation. The order of magnitude is so low that it can hardly upset the operating point.
AC, caused by same, plus capacitive coupling, can be optimized by making impedances low enough and adjusting the DC voltage between heater and cathode, to a point.
Proximity is hardly a primary cause.

  and greater resistance between them will allow for this current to generate a larger noise voltage.
  That's why I recommend to thoroughly decouple the heater rail.
 
alexc said:
my  'floating'  dc filtered heater circuit  (which has it's own traffo sec winding -> bridge -> R-C filter)    ... 

to a voltage,  say +70V or so  from a divider circuit off the B+.

Would I make a 'simulated transformer winding CT'  with a pair of high value resistors across the +, - wires of the heater circuit and connect the centre node to the +70V line ?  Or not  ?  It couldn't be as simple as direct connecting it, right?

you're overthinking this. here's an example, http://www.michaelkingston.fi/files/Drive-1_schematic_rev2.1.pdf
 
You are right of course, Kingston.

I *thought* I was very familiar with your preamp circuit and thread about it  ..  you did that one a lot of years back and I thought it was very influential and spent a lot of time thinking over it. That's when I first started to try the old Soviet tubes - thanks.

But I never noticed that hv bias connection to the dc heater local ground  ..  exactly what I am *now* doing.

Same thing with Ian's hv pcb  - I know I went through it at the time ..  no doubt  ..  and surely your preamp work which everyone  has spent time with  ...  still  didn't pickup/register  on the  'dc heater with the elevation'  thing.

I registered the 'elevation' and I registered the 'dc heaters'  but that last bit of interconnect  .. nada.

I too was ac-heater-everywhere guy. Did go to town with 'elevation' in an ac heater context  ....    I think like Scott, who  also  mentioned something on those lines :) 

Now I'm a 'combo-of-several-different-heater- methods'  guy, as is justified.    It's been a good couple of weeks of diy.

...

I'm also working my way thru your description Abbey ...  thanks kindly ..  there's usually a lot of profound detail there and so it does take me a while to ponder, even more so for the many contributions you've always made here (this place)

I think it (dc heaters and bias from any available handy source)  all started (for me) with that crazy Japanese Encel el84 integrated hifi amp from early-mid sixties that I overhauled    ..  based somewhat on the classic us greats, as PRR confirmed for me 

  . ..  seriously loopy ,  in ways mysterious (speaker feedback coils and variable damping controls and more) 

though well engineered (55 years now and going strong, now with mods a  bit for more 'focus' on reality) 

with really great sound.


...

Regards and hNY to all :)
 
Back
Top