SMPS for tube HT

[ruffrecords]

Ian, why did you say the HT ripple was too high - was that mains frequency related? 

Because for mixer or mic preamp use I need it to be below 1mV. I forget the frequency but it was not mains.

Did the heaters eventually power up in the earlier version? I would expect that any such switchmode heater supply would hiccup its way through initial powering of the heaters - that is a common control method and has no concern for the application - I would be amazed if it was 'catastrophic' as that just doesn't sound like any switchmode control mechanism used nowadays.

I agree commercially available SMPSU tend to use hiccup mode controller chips but who is to say which chips are used in this unit. There are plenty of other low cost chips out there with no hiccup mode. even with hiccup mode, I have yet to find a SMPSU that that will fire up a heater load equal to its rated current output. The problem is the cold heater current is at least twice the nominal value. You have to reach a point where the energy put into each hiccup is enough to keep raising the temperature of the cold heaters so that their resistance keeps increasing and will eventually allow the unit to exit hiccup mode.

Simple SMPS chips have a fixed (safe) current limit way below their max rating (foldback) so they will never power up heaters at the rated current. For the particular unit we are discussing we have no idea what current limit if any is implemented hence my comment about possible catastrophe.

Bottom line is hiccup mode SMPSUs are the only ones that can successfully power up cold heaters where the nominal heater current is a significant fraction of the SMPSU rated output current. As  rule I rate SMPDU for heaters at twice the required heater current to be sure they will fire them jup.

Cheers

Ian