Heater Supply Circuits-AC and DC

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
The hi-fi Encel would be set up for class A, so bypassing is likely to have no major effect, and depends on how close the valves are pushed to grid conduction, and the valve's characteristic.

Amps set up for class AB operation need a bypass capacitor to try and maintain a fixed bias on the output stage during higher signals levels where one PP side stops conducting current (and hence a common cathode voltage would vary significantly if it was just resistance to ground).
 
Thanks for the clarification ....  that makes sense.

And PRR, that Fisher is similar to the Encel with the exception that it appears to be a full cathode bias, with bypass cap.

The thing I wonder, is what practical difference is made  with the 'combo' bias, with a fixed grid voltage providing about half of the cathode-grid dc voltage, and the cathode circuit resistance providing the rest.

My understanding is that

- 'fixed bias', sometimes also described as being 'Class AB' in the data sheets example setups, provides a less programme dependent bias (more stable)  and is generally of a higher efficiency and therefore power.

- 'self bias' or 'cathode bias' does vary the tube's operating conditions more with  the higher rms programme and usually results in a lower available power.

Fixed bias sometimes  is described as be more 'hifi' and cathode bias more 'compressed', leading some to favour one for hifi application and the other for instrument amplification.

Surely a simplification, but is 'combo' bias the 'best of both worlds'  ?

I generally use finals with fixed bias so I can closely achieve my desired operating conditions; cathode bias and combination of both is new to me. 

The only other time I've used 'combo' bias  it is with my vari-mu limiter builds, especially in the fc660 and it's local derivative, the poorman 660, where one is trying to set operation on a specific part of the remote cuttoff transconductance curve, where it has significant 'bendiness'.

----

Just found another popular 6BQ5/EL84 integrated amp using 'fixed + cathode' or 'combo' bias - the Scott 222 series.

- it takes the lowest HV rail, neon bulb regulates it to +48V and uses that as fixed bias to the grids of the finals, which also have a +60V cathode bias circuit; giving a +12Vdc cathode-grid net bias voltage.

- that same +48Vdc neon regulated rail is also used for 4x series heaters of 12V each for the input stages (4x 12AX7)

(the remaining heaters use a 6.3Vac winding with one end connected to the ground line - I don't think I've seen that before; an ac winding seemingly 'ground referenced')

The Scott 222 is a most interesting circuit for anyone interested in an integrated amp design with AX7s and EL84s.  There is a write up in Vacuum Tube Valley  Issue 13  :)
 
A few have found combo bias of benefit for their particular application  Gingertube has had benefit with hifi use in reducing distortion.
 
Hi guys,

Trying to resurrect this topic for selfish reasons :) I know this has been discussed million times (AC vs DC heaters) but I have PCBs that expect DC voltages and I cannot change that. So to me this is not about the effectiveness / noise / efficiency debate.

I tried to "implement" some of the designs posted here in PSU designer but usually failed. The numbers just don't add up :) I'm very inexperienced with this obviously and I miss something to get the desired result - at least approximately.

The circuit visible on the attachment is the B+ from this post https://groupdiy.com/index.php?topic=57166.msg728679#msg728679

Thank you for being generous with your time!
 

Attachments

  • Screenshot 2020-06-19 at 15.15.50.png
    Screenshot 2020-06-19 at 15.15.50.png
    110.5 KB · Views: 60
The schematic shows a 24Vac winding being rectified and filtered to give approx 25Vdc rail. Four 6.3V heaters (V1,V8,V4,V3) in series are powered from the 25Vdc supply. Two 6.3V heaters and one 12.6V heater in series are also powered from that 25Vdc supply (V6 presumably has a 12.6V heater).

Note that the PSUD2 simulation janos.hardi shows has an error in that the transformer secondary is shown as 240V - it should be 24V.
 
I suppose one 'advantage of 'combination biassing is that over the projected lifetime of the gear (well expected life of the valves anyway) it would reduce the risk of serious misbiassing so preventing damage to the valves but also allowing some 'self balancing' (auto correction for the inevitable drift in characteristics) so it stays working in at least an acceptable manner for longer as original customers would be unlikely to actually check or investigate performance until it was really far out.
Of course broadcast and recording gear had 'bias check' metering facilities and presumably someone had the job of checking every now and again.
 
Hello

I have been looking around but I can find a good solution to get a 6.3V DC heater supply from 9VAC.. I need 3 to 4A, for 8 tubes. I guess I can do with 6VDC. People argue online with one regulator, two regulators back to back... and I am a bit confused so I thought I'd ask here :)

Thanks in advance,

Thomas
 
Thanks for your input, I agree with the room heater haha aspect !
Most of the buck regulator boards I find deliver 2A, so I guess I'll just feed my 9VAC to two of these units and each buck regulator board will feed 4 tubes, and done.
Edit: there are regulator with bigger output... so one may be sufficient. Thanks !
 
In my PSU's I use the slow start circuit in the LM317 datasheet and it prevents them from going into overload when you first power it on. Just a nice gentle start up without thrashing your regulators or needing unnecessarily large heatsinks. For projects with more tubes, I just throw an LM338 in place of the LM317, and you get 5A of overhead there. Never had an issue with regulators since.
 

Attachments

  • gib60.png
    gib60.png
    20.7 KB · Views: 19

Latest posts

Back
Top