Problems with heat controller...

[Davek0974]

Ok so i get a call to a machine which blows the breaker as soon as the heaters kick in after power-up.

Diagnosis - disconnect one or other of the two SSR's that control the heater bank.
Summary - One SSR is shorted to earth.
Repair - Replace SSR.

Two hours later i get another call, machine is 200c over temperature but controller is showing off on the output.

Summary - SSR shorted internally.
Repair - turn machine off and go for tea break, wait for smoke to clear.

Any ideas on what can kill SSR's like this?

I am guessing the machines were not designed with this fault in mind as there is nothing to kill the power in an overheat situation.

I have only seem them fail short-to-earth before.

???

[MrSleepy]

Any ideas on what can kill SSR's like this?

Over Current , Over Temperature , Over Voltage. Too high DVDT.

I'd bet its the latter two . I would put some snubbers and Varistors across the SCRs ,if there not fitted.

Are there any old style inductor started Flourescent lights flickering away on the same supply circuit , or derived from the same supply.

Rob

[vincemulhollon]

Assuming the obvious like heat sinking / temps / maybe there's a massive inductive load fan in parallel with the heaters, one thing that kills SSRs pretty well is voltage spikes and the like.

Assuming you've got a massive inductive load somewhere (maybe even another machine?) could it be making massive spikes blowing the SSR?

Could be something like they designed it to run 3ph input and heaters on ph1 and ph2 and a big fan (or pump, whatever) on ph3, but someone did some rewiring and now the fan / pump / whatever and a heater are on ph1 so the inductive spikes kill the SSR.

Optoisolated SSRs are almost impossible to kill by gate voltage spikes, but a plain old non-isolated MOSFET can get blown up pretty well by weird spikes and stuff on the gate.

Most of the FETs I've blown up have been in RF power amps not SSRs so commentary on VHF oscillation and the like would be pointless.

[Davek0974]

Hmm,

temp is not an issue as they do not even get warm, these are zero-crossing ssr's, don't think they are opto isolated though.

There are big inductive loads elsewhere in the factory but not in this machine, can not see any snubbers fitted any where.

I'm going to go through the wiring to see if there are any broken or loose wires etc in the heater circuits, may possibly be a shorted heater cartridge too.

[MrSleepy]

What style of SCR are they.

ST has a range of high commutation SCRs , based on their Alternistor type Triacs usually in a To-220 or 247 ..

but if they are a metal stud or hockey puck style,then chances are they should have snubbers fitted to limit the dV/dT voltage rise across the device.

And adding Varistors is always good practise.

Rob

[Davek0974]

Hi

they are / were Crydom D2425PG units.

[J Tiers]

Just about any SSR these days is opto-coupled. There may be a few that are coupled in a different manner.

dV/dt is a big issue, but it doesn't usually directly "kill" the SSR, it tends to turn it on when it should be off. Eventually the conditions that create the dV/dt may kill the SSR. That may be through becoming more extreme in voltage, or possibly by some condition of internal dissipation due to switching on in an unintended manner repeatedly.

If you had an SSR that was shorted to EARTH..... how did that happen? Was the case split open etc?

If a condition occurred that shorted it to earth, that means the internal ceramic insulation between the semiconductors and the heatsink was destroyed or bypassed. That is reasonably rare, and argues for a serious problem, possibly a serious overcurrent that melts the ceramic (arcing typically), or even a really bad voltage spike and arcing with high "follow current" that goes to earth along the ionized or carbonized path created by an initial voltage-related arc.

Then also... WHY DID A SHORT TO EARTH NOT BLOW A FUSE/BREAKER? Seems as if there are may be some design issues in the machine. (there are some valid but less common reasons why it might not open the fuse, an isolated secondary on a transformer, etc.)

So:

Does the SSR in place pass a "sanity check"? Meaning for instance that if you have a 230V heater rated at 5kW, that's 22A, is the SSR rated at a current comfortably over that current? Or, if the unit is operating on 230V, is the SSR rated for at least 600V? That sort of thing.

Is the SSR from a known quality source? Some of the low cost SSRs that purchasing agents demand to be able to buy are not really the equal of the good brands like Crydom etc.

Is the SSR of the same type? Usually it is easier and cheaper in the long run to replace with manufacturer specified parts. They should have done the needed engineering, no need to re-engineer the machine in the purchasing and maintenance departments....

Is teh SSR a true equal? Aside from voltage and current specs, some have internal snubbers, others need external snubbing. If the unit expects an internal snubber, you need to replace with the same thing, OR put in an external snubber which can do the job.

Might there be a fault in or around the heater unit? An intermittent fault, such as a partial or complete short of the heater element or a short to earth (which could be in outside wiring) might damage the SSR, but be cleared by the time you show up. Especially possible if any of the heaters etc move, or anything nearby moves and might touch the wiring, etc.

Are there any other clues? Any marks of possible arcing in the machine? Unusual noises or smells? Did something else occur at the time, controls go crazy, problem elsewhere in the plant, etc, that might point to an external cause like a voltage spike or surge, etc?

[Davek0974]

Thanks,

the heaters are only 2-3kw and the SSR's are rated 25A so should be plenty capable.

I have the machine head in bits now and think i have uncovered the fault.

The heaters are in a head which moves, only about 2" but movement still. The glass-fibre wrap on one pair of heater cartridge wire has failed and shorted intermittently to each other so a L-N short. This would explain why the breaker blew the first time around and probably expalins the short demise of the replacement SSR.

I will order up a new SSR and rebuild the head, see how it goes.

Its not built in a very robust way though as there is no kill-circuit for massive overheat scenarios caused by a short in a SSR.

I was only surmising that the first unit was a short to earth as the fault cleared when it was disconnected, of course at that point i was not aware of the wiring short further along.

[Noitoen]

Apart from solving the SSR issue, you can install a main circuit breaker with a trip coil and wire it to the controller's alarm output.

[MaxHeadRoom]

t


Its not built in a very robust way though as there is no kill-circuit for massive overheat scenarios caused by a short in a SSR.

.

Obviously the correct protection is needed, proper fusing etc this should prevent SSR's blowing.
Max.

[macona]

I bet the heating coil is partially shorted. Check current draw.

[J Tiers]

Glad you found it. Makes total sense, too..... I like it when the discovered fault and the results agree.

Might be time to consider a thermal switch that cuts the "go" signal to the nearest contactor to it if the heaters go overtemp.

(it isn't gonna do much good to tell the SSR to turn off! )

[rdfeil]

I build a control system for a machine that has a heated air supply. The heaters vary from 3KW to 100Kw depending on air volume needed. The heaters are controlled with 2 SSR's on the 3 phase feed power. For a run away safety the heater power is fed through a dedicated contactor which is controlled by a seperate temp controller. This system helps to provide redundant failsafe operation in the event of a SSR failure OR a controller failure. So far all has worked as designed even with shorted SSR's and shorted heaters causing the SSR failures.

I would look into adding an over temp safety system of some type.

[Davek0974]

All sorted.

The breaker did in fact do it's job as it turns out the ssr's were ok.

The wires were intermittently touching on one heater cartridge. I checked them all for resistance and short to earth but all were ok despite the massive over temp ! Just over-sleeved the wires with some high temp sleeping and rebuilt it all.

Working fine again now.

I will dig the plans out and look to fit some fail safe devices to all machines, we have four of these, all controllers are different though and at least one has no alarm features.