PDA

View Full Version : SSR failures



ags0000
12-28-2018, 03:33 AM
I just had a 25A "hockey puck" SSR fail in the middle of my show. Luckily I had a spare. I was surprised at this failure though.
It is a 25A 120/240v rated unit. I have max (all white, full brightness) of 20A through it (typically much, much less). I do not have a heat sink in use.

Are these failures common (or not usual)? I wonder if I should go to a 50A rated unit (without heat sink) or add a heat sink to a 25A version. This was not an easy fix (due to enclosure & location) and I would like to prefer a repeat.

MartinMueller2003
12-28-2018, 06:33 PM
heat kills electronics. a little heat kills slowly but still kills. heatsink is a good thing for those units.

ags0000
12-29-2018, 02:20 PM
This is now clearly not a freak occurrence but rather a design flaw. My spare SSR died yesterday during testing after pixel replacement. I now suspect the culprit is power supply inrush current. I have used these SSRs rated at 25A for a previous design which used 4 Meanwell supplies (inrush spec: 20A each). This year I have a totally different design which uses 8 Meanwell supplies (inrush spec: 60A) - and thus much higher turn-on current. I will try using a different SSR rated at 50A continuous, but even that is well below the 480A max inrush per the specs. BTW, I will be sure to use a zero-cross SSR to reduce the initial current.

Has anyone else used hockey puck SSRs to turn on banks of switching supplies, and if so, what configuration (supplies and SSR rating)?

davros
12-29-2018, 09:04 PM
What brand is the SSR? Many units with false specs out there.

Barnabybear
12-30-2018, 04:37 AM
Hi, the link below is a teardown of a 25A SSR, 4:40 looks at the triac.

https://www.youtube.com/watch?v=DxEhxjvifyY

ags0000
12-30-2018, 03:51 PM
What brand is the SSR? Many units with false specs out there.

Potter & Brumfield (apparently now part of TE). It served without issue for 4 years, but with less load. The single cycle surge spec is 250A, which reinforces my suspicion. I found an "NTE" branded SSR at Fry's with 50A on-state current and 450A single-cycle surge and installed that. I then setup a test to cycle it on/off every 5 seconds for 30 minutes, and no failure. There are other 50A SSRs available with 750 and 900A single-cycle surge rating, which I'll prefer and possibly upgrade to to be further from a failure in the future.

algerdes
12-30-2018, 05:51 PM
...The single cycle surge spec is 250A,...
Can you explain this "single cycle"? Not familiar with it. Thanks.

ags0000
12-31-2018, 10:08 AM
Hi, the link below is a teardown of a 25A SSR, 4:40 looks at the triac.

https://www.youtube.com/watch?v=DxEhxjvifyY

This was very informative - particularly the issue of fake components on eBay and in this case, the potential for a real hazard. I strongly encourage anyone working with mains voltage to be certain of the legitimacy of the parts used. In the case of this teardown, although rated at 25A on the label, the components were actually only good for approximately 10A.

ags0000
12-31-2018, 03:39 PM
Can you explain this "single cycle"? Not familiar with it. Thanks.

SSRs are spec'd for maximum continuous current and "single cycle" current. The latter is the current sustained through a single cycle (1/120th of a second for 60Hz AC). It is helpful when considering the inrush current when voltage is first applied to the connected load. For example, the Meanwell supplies I am using are rated at just under 7A input current but 60A inrush current.

algerdes
12-31-2018, 09:50 PM
Thank you.

ukewarrior
12-31-2018, 10:13 PM
As others have noted, many of the hockey puck SSRs sold out of china are the worst of fake name brand products.
I actually cruze ebay and buy Crydom SSRs which can cost upwards of $50 from Grainger and Mouser.
Many of these are pulls from industrial equipment and many or may not be used.
But, I get them for $15 or so.
I do know they were built right and will most likely outlast any chinese clone even if they are used.

This is now clearly not a freak occurrence but rather a design flaw. My spare SSR died yesterday during testing after pixel replacement. I now suspect the culprit is power supply inrush current. I have used these SSRs rated at 25A for a previous design which used 4 Meanwell supplies (inrush spec: 20A each). This year I have a totally different design which uses 8 Meanwell supplies (inrush spec: 60A) - and thus much higher turn-on current. I will try using a different SSR rated at 50A continuous, but even that is well below the 480A max inrush per the specs. BTW, I will be sure to use a zero-cross SSR to reduce the initial current.

Has anyone else used hockey puck SSRs to turn on banks of switching supplies, and if so, what configuration (supplies and SSR rating)?