0
\$\begingroup\$

My goal is to drive a beefy solenoid (1.6 Ohm / 1.5 mH / 50ms actuation time, once every 1-2 seconds) from a 36mF (3x12mF in parallel) capacitor bank at 24V, charged from an 1A PSU. This capacitor/low power PSU arrangement was chosen to limit the coil heating in the event of a failure.

The main problem is limiting the inrush / charge current. Possible solutions:

  • Current limiting resistor (10 Ohm): Produces a lot of Heat (10W), slow (2sec)
  • NTC Thermistor: Way to long cooldown time (60+s)
  • Transistor based: Voltage drop is undesirable, and they seem to produce a lot of heat
  • Inductor: might cause ringing
  • LED constant current power supply (for instance PLM-25-1050): they all feature short-circuit-protection, which would likely trigger
  • Soft start ICs: Soft start time is <10 ms, which is not enough

The only viable solution I can think of would be using a constant-current boost converter from 5V which doesn't care about "short circuits".

Do you have any other ideas? Is there an example / premade boost-converter? I cant seem to find a component matching my requirements.

\$\endgroup\$
6
  • 1
    \$\begingroup\$ the 1A current limit from the PSU will offer some help \$\endgroup\$
    – user16222
    Jan 23, 2022 at 16:33
  • \$\begingroup\$ Its not reallly a limit, more like a resetable fuse. \$\endgroup\$
    – UralokA
    Jan 23, 2022 at 16:41
  • \$\begingroup\$ Have you calculated the average current you need to charge 48 mF to 24 V in 1 s? I think you should do that before you go any further. \$\endgroup\$ Jan 23, 2022 at 16:48
  • \$\begingroup\$ @ElliotAlderson Thank you, it was more than the PSU could handle, I have reduced the capacity, now its .86A average. \$\endgroup\$
    – UralokA
    Jan 23, 2022 at 17:10
  • 1
    \$\begingroup\$ Choose a PSU that goes into current limiting (aka CC/CV) instead of foldback or shutdown. That should be all you need. \$\endgroup\$
    – user16324
    Jan 23, 2022 at 19:48

3 Answers 3

0
\$\begingroup\$

Resistor / PTC doesn't really work as it gets even slower as caps come closer to the desired voltage.

I would just put a shunt resistor and a series transistor (will dissipate heat though). You can use a small shunt if you add enough amplification: Active current limiter

Some power supplies can handle trying to draw more than rated current quite well, but especially for larger currents that is not the way to go, unless it's clearly specified in PSU datasheet.

To avoid dissipating heat, you could use a PWM controller and an inductor, but I would not go that way if heat dissipation is acceptable.

\$\endgroup\$
1
  • \$\begingroup\$ Also, check the related posts. \$\endgroup\$
    – Ralph
    Jan 24, 2022 at 6:42
1
\$\begingroup\$

It takes a bit for the core to become magnetized so the mutual coupling between primary and secondary is established and current flow is thus limited.

A simple fix for some transformers is to simply place a small magnet on the core. This pre-magnetization establishes the mutual coupling between primary and secondary before any current is even put to the transformer, so the in-rush current is limited.

\$\endgroup\$
1
  • \$\begingroup\$ Don’t think I agree. Current in a coil makes a magnetic field with no time delay. A transformer core can be left magnetized meaning it’s close to saturating in one direction and if you start with voltage in that direction you can get high currents for a couple cycles until core hysteresis re-centers things. That’s the mechanism where you might see high inrush. But this is for a transformer. Not sure how it applies to this solenoid question. \$\endgroup\$
    – asdf30
    Nov 19, 2023 at 4:06
0
\$\begingroup\$

Best to get a power supply that can handle the charging. If you find the right supply it shouldn’t add any cost. Find supplies that are happy to operate in current limit mode or find one with a controllable ramp rate.

Any linear method of charging capacitors dissipates half as much energy as the Cs hold. This is true regardless if it’s a resistor or transistor or if it’s fast or slow.

For a big cap bank you’ll need big parts. Transistor circuits can work but need to be carefully designed. PTC is a good choice if you have a switch bypass it after it’s done.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.