# Thread: Solid-state power switching

1. ## Solid-state power switching

Normal relays can fail over time and consume anywhere from 100-400mA for their coils. Commercially-produced solid-state relays are pretty damn expensive for high-current applications. How would I build a circuit with a high-current-capable transistor (Probably a FET?) to switch power to a distribution block? 40-60A current load would be nice, just for extra headroom.

2. How about these? http://www.discovercircuits.com/R/relay.htm

edit: looking at the circuits a bit more carefully, I notice they're designed to switch 120VAC load current. I don't imagine modifying it for 12VDC would be too difficult, but it's beyond my elementary skills.

3. I would suggest maybe getting some power transistors and then putting them in parallel. Maybe 4 or 5 rated at 20 amps. This would take the load and split it between the group of them lowering the current going between them well below their rating. But remember, no device is peerfect so these too will eventually fail. But not as fast if you use them below their rating by like 10% to 25%. If you get one rated for 20 amps, then only go and put 15 to 18 amps through it and it will last alot longer than if you put 20 amps through it. Same goes with the relays. But also remember with the power transistors comes heat so you need to get some heat sinks to put on them.

As for wiring the transistor, one side gets the + voltage and the other goes to the device and the third leg goes to a switch just like a relay to turn it on. Just check out the schematic in the data sheet.

Brian

4. Putting them in parallel could be problematic though...even the same parts won't have exactly the same characteristics, and it's likely that one will ramp up just a bit faster than the others, causing it to handle more than its rated current for a bit. Over time it'll blow up...

Also looking for circuit info, as it's not quite that simple...if I understand what I've read correctly, you want to send a higher voltage to the gate than what's on the source pin, in order to keep the internal resistance of the transistor low (higher resistance = power wasted as heat). Any EE types want to help out here? At least point me to some docs or circuit info on something similar so I could figure it out?

5. How often do you want to switch the equipment on/off?

Even though a relay is not as reliable but in most cases it is more than good enough.

If you are thinking of switching a device at 50Hz drawing 60A of course it doesnt make sense to use a relay.

As for the 100-400mA needed for the relay coil, you will still get that and alot more depending on the FET/transistor you use.

6. You may wanna check the ITPS circuit. As far as I can remember the FET is controlled by the PIC to provide the power into the LDO regulator.

7. Hmm, indeed, checking out an existing circuit would be good.

I did find this unit, http://www.vishay.com/docs/70776/70776.pdf but I'm not sure what numbers I should be looking at...

8. If efficiency is one of the motivators, a parallel transistor circuit is definately not the way to go. Reading the description for the circuit linked above, it states that the transistors dissipate 1/6 of the load current as waste heat. On a 60A load that's 10A!

I did a quick search to look for SSR circuits, and came across this very informative page on ePanorama: http://www.epanorama.net/circuits/se...or_relays.html

9. Bad sides of semiconductor relays

Output gets damaged quite easily by overvoltages
Typical failure mode is output short circuit
Output has minimum voltage and current which it will work
Output has some leakage current on off-state
More expensive than normal relays
Low to moderate volumetric effiency
Restricted to single pole, normally open (NO) configurations
On-resistance much larger than normal relays (means more waster power and voltage, heat sink oftern requred at high current models)
Large output capacitance (typically 1 pF ro normal relay, more than 20 pF for SSR)
Relay heats up noticeably when large current passes through it
More sensitive to voltage transients
Most types work only on AC current (there are also special DC relays available)
There is some leakage current even when relay if off
If the triac inside semiconductor relay is not driven properly or faulty, it can act like a rectifier and result in a pulsating DC to the load

10. Originally Posted by joshthepilot
Why couldn't the first reply to this thread just be "Shutup and use a relay u twat"?
Let him waste his time for a bit

#### Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts
•