power supply switching

The MeanWell LRS-200 series specs a 60 amp typ inrush current. Times three power supplies.

The PFC front-end is a boost converter. The AC line initially sees a bridge, a smallish inductor, and a big cap.

--

John Larkin         Highland Technology, Inc 

Science teaches us to doubt. 

  Claude Bernard

Don't over-design this! The customer wants a power switch? Give them one that handles sufficient current to do the job, then +50 to +100% over.

Fluke 9010As had a solution for the power switch (noise, electrical safety, etc.) up front - they used a long push rod that worked the push-on/push-off switch buried in the power supply box at the back of the test gear.

Won't something like that help you with UL approval as well?

KISS

John ;-#)#

The inrush could get close to 200 amps.

These people probably don't care about UL, but we may as well try to stay within the UL/CE rules. A "soft" power switch would help some there. A pushrod might be helpful, but we still need a switch. Maybe DC switching after the power supplies, with giant mosfets, is the way to go... if it doesn't get out of hand.

--

John Larkin         Highland Technology, Inc 

Science teaches us to doubt. 

  Claude Bernard

On a sunny day (Thu, 03 Sep 2020 09:51:07 -0700) it happened snipped-for-privacy@highlandsniptechnology.com wrote in :

Would a big triac work as power switch? Drive it with some DC from a small supply that is powered by the power switch Is that legal? Over here you may need a real switch in the power line. So maybe 2 switches, one big one to cut the mains, and a push button thing or touch screen or whatever for on off. No noise no sparks. ?

Can't you do that with a single multi output supply. Might be cheaper than 3 supplies and whatever added circuit it needs to fix the surge issue ?.

Have a stack of rechargeable tools and other stuff that needs to be topped up every so often, 8 or 10 chargers, time switch for on/off. Solved it by fitting a big 25 amp ind relay from the timeswitch, Noisy, crude, but gets the job done and no blown line fuses yet...

Chris

The MeanWell LRS-200s are crazy cheap, around $30 each, and the triples cost $250 or so. Don't know why.

Some giant relay is a possibility, but I need to get the coil voltage from somewhere. AC line voltage could be 120 or 240.

You have to consider also the house wiring and house fuses.

If you use separate relays for each PSU, you could delay the turn on for some of the PSUs, limiting the peak current to a manageable level.

n the US, if you have split phase (2 x 120 V) you could reduce the current, but this requires a two pole relay.

If the issue is only relay contact rating (and not house fuses), just get a three phase relay, so each contact has to handle only 60 A peak. In the US, if 120/208 V three phase happens to be available, you could run each PSU from own phase.

Of course, in the rest of the world a 1x230 V or 230/400 V three phase could also be used with the same relay and only select between one and three phases.

This won't be residential, but it could be used anywhere in the world.

Single phase.

I don't think that 600 watts of switching power supply will trip an AC breaker on startup. I'm confident it would eventually destroy a small switch or relay.

The bog disk NTC thermistors are intellectually challenging but do seem to work.

Don't you miss big old mercury contact relays? :-) Tough critters!

torsdag den 3. september 2020 kl. 21.35.36 UTC+2 skrev John Larkin:

use the output from one of the supplies to turn on the rest?

The LSR-200 says max inrush of 60A. Find a power relay with a 12 or 24 volt coil and drive the coil via a small step down transformer. That way you have lv to the front panel switch.

--
Chisolm 
Texas-American

The cheap supplies don't have a control input.

I could power them all up and add load switches downstream, controlled by a little switch. These days, most power switches are fake.

3 supplies is 180 amps... typical.

The transformer would have to work from 120-240 VAC.

I can get a universal-input 12VDC supply for around $10.

I've used Meanwell supplies. They are not bad, but look very minimalist in terms of circuit design and the pcb's on some are, shall we say, hastily constructed. Nor is there any ovp protection. Cheap and cheerful and ok for the occasional project round the lab, but not sure I would be happy selling a product with them into a professional environment. What voltages are needed and, for example, would a standard pc psu get the job done ?. Various form factors, sizes and power outputs.

I would shop around for a single 3 output psu, but if you must use 3, why not a add a cheap wall wart style psu to drive the relay ?. Gives the capability of remote switching as well...

Chris

A mains switch & NTC is very simple, does the job. Snubber & parallelled poles help a bit. As someone said, a plastic rod could keep mains out of wherever you don't like it. Case flexure can make those fun.

240v rated relays generally run fine on 120, some on 100. Testing required. Obviously you need spare pulling capacity else contact pressure is weak. And permitted mains voltages are >240.

A PTC thermistor in series could in principle reduce the i variation, but you'd be left needing a coil voltage that's not off the shelf. Or perhaps a transistorised i limiter, eg a fet that cuts out peaks when 240v is used.

If your market excludes Japan a series cap could make use of the 50/60Hz difference. 240 then drops a bit more than 110 - but again a useless coil voltage is then needed.

Or series PTC R & parallel NTC or Z etc etc.

Or a 2nd relay that switches in a series C when it gets 240.

Or... so many options.

NT

torsdag den 3. september 2020 kl. 22.32.31 UTC+2 skrev John Larkin:

switch for first supply, the supply turns on relay to power other supplies switch only has to handle inrush of one supply, and the relay voltage is fixed maybe use a second pole on relay to bypass inrush limiter on first supply

The canonical approach is a switch feeding a wirewound resistor that is shunted by a relay.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
ElectroOptical Innovations LLC / Hobbs ElectroOptics 
Optics, Electro-optics, Photonics, Analog Electronics 
Briarcliff Manor NY 10510 

http://electrooptical.net 
http://hobbs-eo.com

So, instead of 'switch' have standby power and a logic pushbutton to ON/OFF the heavy power parts. They mass-produce this for PCs, in a variety of form factors and ratings.

Rotary switch actuator on front panel, plastic rod, magnet, reed switches. Need inrush current limiting.

Or use relay or contactor manually operated by cam or pushrod from front panel actuator, maybe use the coil for latching?

--
Cheers 
Clive