Physical switching for a ~25VDC 5A battery powered circuit

[etaripeka]

Hello,

I'm wondering what is a good way to safely switch on my compact portable battery powered device, which:
1. is powered by a 6S battery (so ~18V - 25.2V DC)
2. usually draws an average power of ~80-100W, so the current could be up to slightly above 5A
3. and when I measured the inrush current with a BM037 clamp meter, I measured around 90A(!) - (should I be very worried about this value?)
4. must be able to handle vibration and shocks (my current system seems to sometimes restart itself when knocked around a bit - I have a feeling it's due to the rocker switch not dealing with shocks well, but it's quite hard to replicate)

My current solution simply uses a high-inrush current rocker switch between the battery and a custom PCB (Bulgin 1350 high-inrush rocker switch, supposedly rated ~24Vdc/10A), and my PCB is protected by a 5x20mm F 15A fuse (which hasn't blown yet!).

I was thinking of changing my rocker switch to a push-button switch with a lower current rating (with some shock resistance rating) (are there smarter alternatives?), with maybe some sort of solid state relay - my idea being that no moving parts should mean higher robustness to shocks / vibration. However, I've also recently learnt of ICs such as "load switches" and "eFuses" on Texas Instrument's website which can perform the switching, along with current inrush limiting, and also remove the need for my fuse holder on the PCB, which sounds brilliant in theory. I also saw that a solid state relay that can handle my circuit is quite physically large and expensive (655-SSRDC-200D12    -    $70!), so a cheaper load switch IC makes  sense.

So I'm contemplating this option:
1. Use a simple push-button switch rated for >25Vdc ~1A, which will connect the battery voltage to a 5V LDO.
2. The output of this LDO is used to set the enable pin of a TI TPS27SA08 load switch, which then connects the battery voltage to my main circuit.

Is this idea functional? Am I missing anything? Are there smarter ways to do this?

Thank you!

[mariush]

Solve your high inrush current with a varistor ( resistor that decreases in value as it heats up) and a p-channel mosfet or a solid state relay / classic relay to short out the varistor (take it out the circuit) a few seconds after startup.

If needed use a cheap 5v buck regulator to produce your stand-by voltage from the battery, even for something low power no point wasting energy in a linear regulator.

you can make it soft start with a on/off slide switch ... switch pulls gate of a mosfet either to voltage or to ground ..... a p-channel mosfet will be on by default, put voltage on the gate to turn it off... pull to ground to turn it on....  the other side of the slide switch can be connected to a time delay circuit (a one shot 555 timer for example turn on the mosfet or relay that shorts the varistor)

[etaripeka]

Is there a reason why a relay could be preferred over the load switch I was looking at (TPS27SA08)?

[Kleinstein]

Mechanical swiches are a bit less likely to fail with a short over a solid state switch. Also mechanical swiches usually have less leakage, better isolation. However there is wear with mechnical swiches. Swiching DC is also not ideal, as an arc could persist for quite some time and do damage, even at only 25 V.

Ideally one would find a way to limit the inrush current. High currents at low voltage are a common thing in automotive (trucks often use 24 V). 5 A are actually not that much - somewhat comparable to 1  halogen head light and these also have some inrush current.