I had to do some tweaks, but the little ***tard works! I'll drag a pair of6S4S in to work next week and try to light them up. I tested the supply initially with a pair of Kikusui electronic loads operating in constant resistance mode, and of course a bench supply for the input.
I was able to light up a pair of 6S4S with my switching filament supply. Well, the filaments don't really "light" up all that much as they're probably oxide-coated. Voltage across the filament is about 5.9-6V, and I assume the constant currrent regulator is doing its thing, based on its performance with an electronic load in constant resistance mode.. I'll check that to be sure. One of the other things I need to check is whether the current through the transformer is balanced from 1/2 cycle to 1/2 cycle, one of the reasons I'm using a current-fed converter.
I just put a current probe on the transformer primary of my filament driver board, and the balance is as nice as can be expected. I need to decrease the overlap between primary phases so the the board doesn't act as much like a boost converter - otherwise, mission accomplished. Schematic will follow when ultimate circuit values are worked out.
Last edited:
Member
Joined 2009
Paid Member
It's sitting in a box waiting it's turn in line. I haven't had enough of a yen for triode power to push things. Too many ideas, too little time, and often an new idea comes along that shoves older projects out of the way. The major consideration for this project is a front end circuit that will swing the required amount of voltage to drive the low-mu triodes properly. I could cascade a pair of triode diff stages, or I may try something more heretical to get the required gain in one go. A pair of high-gm pentodes in a differential pair with plenty of bias current might be one way to go. Then again, one of the dual dissimilar triodes for vertical oscillator/amplifier duty might work as well, with the high gain triode current source loaded and driving the low-mu monster as a cathode follower with a CCS on the bottom. You could tinker around in the breadboard phase forever with an amp like this...
One of the other things I need to work out are proper positioning and value for some RC delay elements to ensure some crossover/overlap between the two push-pull switches of the converter to prevent a voltage spike at switching transitions due to the presence of the inductor in the transformer center leg. Not a biggie.
With a voltage-fed push-pull you would avoid any switching overlap like the plague (we call it shoot-through),but with a current-fed converter it's necessary, though you don't want to overdo it, as it will tend to boost the output voltage.
With a voltage-fed push-pull you would avoid any switching overlap like the plague (we call it shoot-through),but with a current-fed converter it's necessary, though you don't want to overdo it, as it will tend to boost the output voltage.
- Status
- Not open for further replies.