PL519 Heaters in series blow fuse

[wiysel]

Greetings to all! What a great forum this is!

I have a question about connecting eight PL519 heaters in series across a 320 Vac tranny rated at 450mA. This series circuit is completely independent of any other. I turn on the power switch and watch all eight tubes begin to glow, 2 seconds later the 6A slo blo burns out. What in the world, am I mistaken about the PL519 specs? Any advice?

[ilimzn]

The answer is really quite trivial - cold heaters draw a LOT more current than hot ones, even heaters with controlled heat-up characteristics like on the PL519. It is not uncommon for the initial surge to be many times the nominal current. With heaters as big as on the PL519, it takes a while until they are hot enough for their resistance to go up sufficiently, so the surge lasts a long time.

You should also check that the independant 320VAC does not violate the heater-cathode voltage spec. If it was me, i would never use such a heater circuit. Even if it is independant, heater isolation leakage will establish the potential difference between heater and cathodes, some of which (depending on where in the chain each PL519 is), may be out of spec.

Keep in mind that these were used in TV sets with series heater conenction, which operated on 220VAC (not 320!) and the order of heaters connected in the series chain was quite important due to heater-cathode maximum potential specs. Also, series heater chains are never connected directly to the mains, but through a surge dampening resistor (usually a large and partially adjustable wire-wound) and/or a NTC element to decrease the surge current on power-on.

[wiysel]

Thank you for the clear explanation. Do you think would be better to power two sets of 4 heaters with separate tranny coils?

[hemgjord]

IIRC the recommendation is to use a series resistance in series heater chains.

/Olof

[kevinkr]

Might be better to run them in parallel if you need to get another transformer anyway.. Two separate sets of 40V windings is probably a good idea because of the currents involved. Each winding would only have to supply 1.2A of operating current - no concerns with voltages not dropping evenly during warm up, and you could also use a small varistor in order to limit cold inrush current either on the primary or on each of the filament windings. You could use a single 2.5A winding as well..

To minimize cathode to filament coupling (hum, etc.) you could float these filaments on about 100V of well decoupled dc. It appears that this type has relatively good filament to cathode isolation already so this technique may not be necessary at all. (Consider it if building a 0 feedback type amplifier.) Don't leave the filaments floating if you don't reference them to dc, preferably ground them through a center tap or through a psuedo center tap created with a pair of power resistors.. (A pair of .22uF ceramic caps may work as well and will dissipate little "real" power.)

[316a]

I often use PY500A which have similar heating requirements (42V@300mA) . With these , I always use a 2 x 22V transformer and run each heater with a series resistor . The primary also has an NTC in series with each primary connection .

cheers

316a