Help with high-current DC switching

Question

I need to remotely switch a 48 V battery that is supplying a 15 A load (which is slightly inductive), as part of a low voltage protection circuit for a PV solar system.

Am familiar with the challenges around breaking DC current, have used contactors for 24 V designs, and know what a DC1 rating means. But looking online at various contactor’s data sheets, 48 V seems to incur a much higher price.

Can anyone suggest a low cost contactor that is up to the job? One with DC rated contacts, not just a 400 V AC one that may (or may not) be OK at 48 V DC. There must be ones typically used in 48 V solar designs, but I haven’t found any.

Answer 1

I’ve decided to answer my own question, in the hope it may help others in my situation:

I started off asking about a low-cost relay or contactor that could switch 15 A at 48 V DC, and later qualified my budget at around £25. I was looking for something that wouldn't impose too much of a drain on the battery.

There are actually very few relays or contactors specified for 15 A at 48 V DC – and consequently they are $$$. Why is this? Because most LV applications only go up to 24 V DC (resulting in a wide range of product options – traditional power relays from Omron, TE, etc., and DIN-rail mounted contactors from Schneider, Finder, etc.). Above that, the next popular DC voltage range is around 60 - 100 V.

Also, as the DC contact power (and more so the DC switching voltage) increases, the coil power for larger contacts increases to be able to move them fast enough to avoid DC arcing. So typically coils for contacts rated 15 A @ 24 V DC need around 2 W, while those for 15 A (usually you have to go up to at least 20 A due to the scarcity of 15 A parts) @ 48 V DC need 4 W.

You can of course buy a higher spec’d part, like the Albright SW60 mentioned in comments – but this has a 6 W coil power requirement.

This is all fine for many Solar PV high power applications, where a continuous 4 W or 6 W drain is negligible. However in my application the 48 V battery has only an 8 Ah capacity as the load is applied infrequently for short durations. A constant 4 W drain could quickly cause a low-voltage condition by itself.

Plus it’s worth noting that a relay/contactor consumes its coil power whether the load is drawing current or not.

Looking at other options:

• Traditional solid state power relays using thyristors will have a forward voltage drop, typically 1 - 1.6 V. This results in a high on-state power loss, which requires substantial heat-sinking. Not really a preferred option. However, unlike with a relay, that power loss only occurs when the load is drawing current.

• More recent MOSFET-based SSRs look very interesting. Some in the right power range have a max on-state resistance of 0.01 Ohms. That's around 2 W at 15 A, which requires minimal heatsinking. But they are pricey.

As this is for a LVP circuit for a Solar PV system, another option is to use an NC power relay which consumes no coil power in normal operation. When at the low voltage threshold I'd use the remaining power in the battery to turn it on breaking the supply of power - which is a viable approach as it's likely the sun will recharge the battery before the relay drains it completely. This of course only protects the battery from the load taking too much power, and not the case where the solar charging system has failed...

So to summarise, my extensive online searching hasn't identified a relay/contactor that fully matches my original criteria. I’d probably use a MOSFET-based SSR like a Crydom CMX series part - above budget, but there’s very little choice - knowing I’m only incurring that 2 W loss when the load is drawing power. Or for budgetary reasons I might use a NC power relay, which only protects the battery from discharge by the load and not from failure of the solar charger.

Answer 2

You could use an isolated photovoltaic MOSFET driver to build your own solid state relay. This is just the driver part of a MOSFET solid state relay, you can pick MOSFETs that suit your specifications separately, so if you don't care about switching speed, you can get lower RdsON.

If current will only flow one way and the goal is only to protect against over-discharge of the battery, perhaps you can use only one MOSFET instead of two in series.

If you don't need isolation and you need only one FET, you could use a high-side NMOS driver with integrated charge pump which makes it able to stay ON continuously, unlike drivers with a bootstrapped supply. Here is an example. This one only goes to 36V so you'll have to do some shopping.