Burning Relays 2

[crimp]

I have a couple of questions about SSRs(Solid State Relays). In the factory that I currently work in, I have an oven that I use to cure painted parts. This is an electric oven with heating elements driven by solid state relays. On 2 occasions now over the last 6 months, I've had one of the relays fail causing a small fire. Both times, the relay ended up with a hole about the size of a dime on the power side right between the 2 switched terminals. The power for these is 480 volts. The first time this happened, the 480 volt side shorted directly to the control side which is 24 volts. This caused in excess of $10,000 worth of damage to a plc, a robot and a power supply. The most recent event wasn't as bad, blowing several fuses and damaging the thermal control module on the PLC. Has anyone else seen failures like these in SSRs? I don't think it's heat related as it seems there was an arc of some sort that shorted the relay internally the first time. What would be a good solution to protect the rest of the equipment from future short circuits? I appreciate any help or advice you could give me.
Thanks,
Greg Thomas
Prime Finish

 

[itsmoked]

Wow! Sounds like a lot of fun.. NOT!

Need much more info. You've talked about what occurred not "what you have".

Name names...

Single phase or 3ph?

What model SSR?

What current?

What type of heating elements?

Do you know the resistance of the element?

How EXACTLY are the SSRs mounted?

Where exactly are the SSRs mounted?

What protection do you have on the power side?

Do you have a temperature gun? (If not, why on earth not?!)

Is there any chance you can shoot us a picture of the panel/installation?

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[MagnusElec]

The use of MOV's and semi-conductor fuses, in that order, downstream of the SSRs helps.

 

[crimp]

Lots and lots of questions!!! I'll shoot you some pictures Tuesday, right now I'm getting ready for a long weekend... it's been a long week!!!

 

[crimp]

OK...

3 phase 480 Volts

Dayton 5Z983 480VAC, 75Amp, 3-32VDC Control Side Voltage

2KW Heating Elements

Not 100% sure on resistance but I think 12ohms

Each SSR is mounted on its own heat-sink inside a panel box and the panel box has a cooling fan mounted in the side

Power side is protected by JLLS60 fuses

I have a temperature gun, but I'm waiting on the Omron temp control module to arrive this morning to get the oven back into operation, so I'll have to get back to you on the operating temperature of the SSRs.

Pictures and schematic to follow.

Thanks for the advice!!!

 

[crimp]

Here are the photos:

Panel:

http://i243.photobucket.com/albums/ff73/crimp111/DSC01275.jpg

SSR with Hole:

http://i243.photobucket.com/albums/ff73/crimp111/DSC01281.jpg

SSR Cover:

http://i243.photobucket.com/albums/ff73/crimp111/DSC01282.jpg

Panel Door:

http://i243.photobucket.com/albums/ff73/crimp111/DSC01283.jpg

BTW... the equipment is back up and running. I'll get the SSR temps after it runs for a while. Also, current through each SSR is as follows: (Left to right, see pictures)

SSR #1: 27 amps
SSR #2: 41 amps; this is the one that blew
SSR #3: 22 amps
SSR #4: 38 amps

Thanks again!

 

[itsmoked]

Hi Crimp.

Groan.. Let me ask. Do you go to a fine Chinese restaurant and order a hamburger? That's what you're doing buying poorly specified (nay - pathetically specified!) SSRs from Grainger. If you must use Grainger supplied relays you should buy a name-brand one that has useful specs you can look at elsewhere.

Example:
What's the derating of your Dayton?
What's the maximum i²t fuse rating for your dayton?
What's the heat dissipation?
What's the maximum dvdt allowed?

These are all important aspects of SSR applications.

Anyway. It looks from your pictures, (thanks, by the way), that likey the SCR's cooked, failed, then arced their way thru everything. Nasty!

I am still voting for overheat as that is a very easy thing to do to SSRs. When you check the temp try to angle your gun to see the base of the SSR so we can get a peak temp reading. We want to see the very highest reading you can muster. Get in real close.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[itsmoked]

Oh another thought. That relay has an internal snubber. The capacitor for that could easily have gone blitzoid and taken everything else with it. Your SSR damage looks like the same damage I saw with a SSR checker that took down a fourteen million dollar machine when the internal snubber cap lost its cool.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[crimp]

The 2 relays that carry 22 and 27 amps respectively are around 150 degrees. The 2 that carry 38 and 41 are around 157 degrees. I also checked the temperature in the area where the hole was blown in the case and it was at least 12 degrees hotter in that area with the hottest relay being the one that keeps popping. It's temperature was 171 degrees where the hole was blown.

 

[sreid]

I'm with itsmoked. Switch to Crydom D4890's.

http://www.crydom.com/products/productFamily.aspx?id=87

 

[itsmoked]

157F = 70C
You are dumping about 42W off those hotter units. 70C is awfully hot in my experience for running SSRs reliably. Remember with the door shut, those temps are probably 10C hotter.

First tell me you have heatsink grease applied under those SSRs right? You cannot use them without it.

If you do. Then I would boost the cooling dramatically with a fan. Yes, you have a box fan but I would put forced moving air directed right on the heatsinks of the hotter units. Air applied that way triples or quadruples the heatsink effectiveness. That or you box fan should be seriously supercharged so you turn over all the air in the box many many times(20X) a minute. Think hurricane.

If those units run hot, too hot to touch, they will age very quickly. VERY quickly with an end game like you experienced. I see exotic failures like yours in many cases when they are run at high temperature for long periods like I believe you are. If you get those units down where you could touch them they will out last the rest of your equipment.

No metal dust around is there?

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[crimp]

Yes, there is heat sink grease on them! No there is no metal dust anywhere near. Metal dust and painting applications don't mix well together!!!

What would you consider to be an appropriate operating temperature for these SSR's? Also, if I up the cooling, should I still consider changing to a different brand SSR?

 

[itsmoked]

I'd probably switch brands after your experience...

As sreid suggests a Crydom D4890 is a very good unit.
It has a higher current rating 90A (and 530VAC).

https://www.alliedelec.com/Search/P...&R=682-0034&sid=46DCA000C58E17F&tab=specs#tab

I like and use Continental Industries (Cii) SVDA/6V75
Same current as yours, (75A) But I like its 660V rating. Heck of a nice margin for 480VAC applications. I know that the Cii, (last I bought them), had a thick copper base which is far better at transporting the heat away from the internal dies.

http://www.mouser.com/search/refine.aspx?Ntt=518-4575

Personally if it were my equipment I'd be looking for a maximum 'base' temperature of about 140F in a fully warmed up box.

It all comes down to your heat sink. If your box can get to say 140F very likely in a normal case.
140F = 60C

Then the maximum you can stand is 90C-60C = 30C rise due to heatsink efficiency before you start harming the SSR.

If you have a 1.5C/W heatsink system that means you can stand:
30C/1.5C/W = 20W
coming out of the SSR.

That would limit the Cii unit to 22A even though that SSR can actually handle 75A, three times that current.

This is why I suggest a fan as that will turn a piece of flat boring metal plate into a sub 0.5C/W heatsink. Of course the fan must be working or you're back in the sh#t.

Some considered sheet metal creativity could make sure serious air is being directed past the SSRs and thru their heat sinks.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[crimp]

Great information!!! Thanks itsmoked very much for your advice. I really appreciate it.

-Greg-

 

[LionelHutz]

Keith has you really well covered.

As soon as I looked at the pictures and saw one module was running at 41A I thought "not enough cooling".

I'd also consider isolators between the controls and these modules to keep from damaging anything again. Offhand, I think Weidmueller would probably have an isolation module that could work for you. Then if a SSR puts line voltage on it's input again it won't feed into your PLC.