LM317 slow turn-on

.

Yes.

Some have used it for regulated tube heaters. There's the notion that slow heater turn on will prolong tube life and the regulator itself due to the very large start up current peak - at least on paper. This cure is disputable. Very possibly no one ever measured this and it would be too difficult to do it reliably anyway. But the parts are cheap so why not.

Others further commented that the use of DC heaters - let alone regulated! -  is pointless when heater wiring is done right but that's a whole another discussion.

I use it.  Works fine.

[silent:arts] said:

I use it.  Works fine.

Click to expand...

I guess the question was: which application?

I'd certainly like to know a few real world cases.

> notion that slow heater turn on will prolong tube life... disputable. Very possibly no one ever measured this

The ENIAC folks did.

ENIAC was a 13,000 tube computer. Every power-up, they lost a few tubes. As a side-project they built a rack with a lot of tubes and a slow-start power unit, cycled it every minute for weeks. They did not find a large difference in tube death rate.

OTOH, series-string TV sets did kill heaters. On a 110V string with a 40V sweep tube and a 4V tuner tube, the tuner tube could see 8 or 10V until the 40V tube got hot. Or vice versa. The "A" in many tubes indicates that there is some attempt to equalize warm-up times for better reliability in very mixed series heater strings. That was not a total solution-- the solution was transistor TV sets. But mixed heater strings are not seen today. (Running a few similar 12V tubes on 24V is not near as bad as the old mixed-types TV strings.)

Kingston said:

I guess the question was: which application?

I'd certainly like to know a few real world cases.

Click to expand...

I use it in a VPR module, generating my "virtual" 24V from the provided +/-16V.
Time will tell if my way doing this will avoid some reported problems other modules have with certain cheap 500 series Racks.
(start up problems, noise from uneven loaded rails, etc).

[silent:arts] said:

I use it in a VPR module, generating my "virtual" 24V from the provided +/-16V.
Time will tell if my way doing this will avoid some reported problems other modules have with certain cheap 500 series Racks.
(start up problems, noise from uneven loaded rails, etc).

Click to expand...

I would think that, in order to solve start-up problems with electronic circuits, the ramp-up provided by using a capacitor across the voltage-setting resistor would be enough. With 47uF, you get about 1 second ramp-up, I don't think you may need more.

abbey road d enfer said:

I would think that, in order to solve start-up problems with electronic circuits, the ramp-up provided by using a capacitor across the voltage-setting resistor would be enough. With 47uF, you get about 1 second ramp-up, I don't think you may need more.

Click to expand...

How would this affect the output voltage over time (when the lytic starts degrading)?

warpie said:

How would this affect the output voltage over time (when the lytic starts degrading)?

Click to expand...

The voltage would not change, but as the cap ages it loses capacitance so the timing may be shorter. This is the normal scenario.
Not so normal would be cap becoming leaky, which would introduce a resistance in parallels with teh voltage adjust resistor, so voltage would decrease. Another would be the cap shorting out, with associated almost total loss of voltage.

Makes sense, thanks .

PRR said:

They did not find a large difference in tube death rate.

Click to expand...

A trove of information.