• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

peak current 300B heater with start up

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Administrator
Joined 2004
Paid Member
With regulated dc supplies I have measured inrush currents of >3A with cold 300B filaments, where 3A was the current limit of the regulator in question.

(JJ 300B with nominal Ifil of 1.5A, not the 1.2A typical of most 300B)

I now use constant current heating which results in the tube taking I estimate about twice as long to heat to operating temperature as constant voltage mode.

For voltage regulated filament supply just filament dcr may be adequate if the regulator does not drop out , in the case of unregulated supplies or regulated supplies that are in drop out you need to do as Sheldon suggested.
 
If you use three-terminal voltage regulators for heaters/filaments, it's easy to slow down the turn-on voltage ramp with a couple of extra parts. This prevents the surge of current when the filament is cold, reducing stress on all parts. I do this routinely for even indirectly heated tubes.
 

Attachments

  • slow_turn_on_reg.gif
    slow_turn_on_reg.gif
    3.4 KB · Views: 320
Brian Beck said:
If you use three-terminal voltage regulators for heaters/filaments, it's easy to slow down the turn-on voltage ramp with a couple of extra parts. This prevents the surge of current when the filament is cold,

Well' not quite. If the psu doesn't droop, then the dissipation has to go elsewhere in the regulator pass transistor. for example I notice the instant cold rush heater current spiking in one half NOS ECC83 is over 1A. This drops fast but the initial surge can still damage chips.
In the case of a common reggy 78 series or 317, these devices have thermal + internal foldback current limiting, so the device is essentially self protective on s/c's solong is heatsinked.. In the case of discrete parts, i.e simple pass transistor and zener config can be easily destroyed without additional protective features.

Amp builders using DC for the big game heaters, 6550/KT series and 6C33 stuff, carefully do the current homework. The instant on current v.s resistance represents a darned good s/c scenario.

A car headlight bulb is a cheaper simulation than a vacuum tube.

richj
 
You'd have to work overtime to blow out an LM317 or one of its brethren when used this way. The point is that the slow turn-on circuit prevents the current surge from ever happening to begin with (with properly chosen time constant). And LM317-style devices are both over-current and thermally protected. They would shut down in a fault event, but even that won't happen here. The idea is to gradually raise the filament voltage so that the current never does surge. Current will track the voltage monotonically, although not linearly. In the case of the 300B, it would gradually grow to its final value of 1.2A. BTW, that's just under the 1.5A limit on the standard LM317. It would probably work fine, but the LM350 can be used for more margin. During the short time when most of the supply voltage is dropped across the LM317, the current is not yet at full value, and even if the momentary power dissipation is high in the LM317 due to higher voltage drop, the thermal inertia will carry it through the ramp-up safely, unless you've chosen a very long delay. I've built a number of these delayed heater supplies, many driving a large number of tubes simultaneously, and I don't ever recall an LM317 failure. Likewise, I've used them to regulate 1000V with nary a glitch. You gotta love this little part.
 
thanks so far

I asked the question because I bought 2 switched PSU's very cheaply in a surplus shop. They are 5 volt/4A and are protected against overload. I wonder if it is safe to try them. Perhaps I can mount them in a tin box to prevent HF hash (?).
 
Oh well, in THAT case... You will have no easy provision for a slow start, but the current capability ought to be fine. I would double-check that the output voltage is isolated from the AC ground, since the filament voltage is your cathode voltage. It probably is isolated, but I would still check. Noise is then the question. Yes, I would mount them inside metal boxes, and perhaps consider placing these outboard of your main chassis. You may find that you will have to add LC filters after the supply to get rid of switching noise. This can vary a lot from supply to supply depending on the original application it was intended for. A ferrite clamp on the output wires might be useful.
 
That's not my circuit; it's straight out of the application notes in back of the data sheet. R2 sets the output voltage. If you keep the 240 upper resistor, then R2 would have to be about 720 ohms for 5 volts out, and about 970 ohms for 6.3 volts. A 1K or 2K pot could be used to set the desired voltage. Since this is a voltage source, the output current is determined by Ohms' Law in the load and is not an adjustable parameter within the supply (the LM317 can also be configured as a current source, but not here). The transistor is not critical. Any small signal PNP will work. Once it's done it's job of slowing the turn-on, it takes itself out of the circuit.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.