I am playing with a heater control project, controlled by an Uno.
I am facing 1 big issue. That is the SSR used, failed closed circuit. This means that even with the arduino powered down, the heating element is still powered.
luckily I was watching the temp and realised something was wrong.
Going forward, I need some kind of fail safe, to operate in the event that the uno output is anything other than HIGH.
Ideally I think a 2 tier solution would be good. eg tier 2 would be resettable and tier 1 would be non resettable, say a thermo-fuse or similar.
Don't use the cheap Fotek SSRs
Use an SSR with a current rating at least 20% higher than needed.
Use two SSR in series
Use a good heatsink on the SSR.
Use a thermal switch to cut power if the temperature exceeds a set value.
One solution would be to use a thermal fuse connected thermally to the heat source. This needs to be totally independent of the control loop and its associated components. The second would be to replace it with an approved unit. I do not know your hardware or the environment it is in. It is possible you were lucky you caught it and prevented a fire. Remember the Arduino was not intended as a final application but for experimental and learning purposes.
SSR was mounted on a heatsink.
SSR now in bin, post dissection. ( too much resin bonded to the single MOSFET or what ever it was, to identify and research)
admittedly, it was a FOTEK (or fake tek, as it should probably be called). I will never use that brand again.
Jim-P...it was of course a FOTEK. lesson learned. but now it's "once bitten, twice shy".
I did open it up, but resin was well adhered to the transistor. no chance to identify.
Jim-p.....the SCR was screwed to the back plate, but defo no thermal transfer paste used, in point of fact, it had a layer of resin between the 2, effectively insulating it.
re TIER 2 fail safe...... I have now added a DS18B20 temp sensor to the heat sink, (with heat paste too).
I can read and display the temps ok.
But being still pretty new to all this arduino and C++ etc, I need to figure out how to process the reading and trigger an output to a pin, to activate the suggested N/O relay.
( meaning, if temps are low, relay Closed circuit. If temps rise above operating, signal goes low and relay opens. FAIL SAFE via temp).
I guess this should also have a manual button to reset.
I already have an ACS714 current sensor fitted to SSR load side. again it reads and displays OK, but again, I need to figure out processing. eg if SSR triger signal=LOW and current still flows, I would like to trigger an "SSR fault" on OLED display and close N/O relay.
if anyone has a contribution on this, feel free.
please note, this would be in addition to a tier 1 fail safe.
The whole point in using an SSR (a good one) is that they last a million times longer than a mechanical. Now you have introduced another device that will probably fail before the SSR does.
my thoughts were, use a N/O mech relay (or contactor) to control the AC supply to the 48V DC PSU that supplies just the load. ( a separate 12V DC PSU supplies all the control side).
the load side of SSR has an ACS712 monitoring the heater current.
So , my thinking is, IF arduino SSR Activation pin = LOW (SSR OFF) & the ACS712 is measuring current in excess of 0.1 Amp ( just for an example) after say 1 second, then arduino sends CONTACTOR activation pin LOW ( this disconnects the AC from the 48V PSU), also HALT heating cycle ( terminate heating loop code) and print " SSR fault !" on the OLED display.
( obviously, this idea will likely spiral a fair bit regarding code)
so the mech relay / contactor should only have 1 cycle in each period of use, versus the SSR which could be 100's of cycles.
its the closest thing I can think of presently to a closed loop safety control.
This would also just be the secondary safety. Primary will be a NON resettable thermo fuse.
The fundamental failure seems to be the device did NOT TURN OFF. Not that it failed closed circuit. It did not over heat. The factory did not bother to test their finished products.
If you add an additional temperature sensor, you need to add it to the heating device, not the control device.
I propose that you just replace the old SSR with a known industrial type SSR and forget all the additional sensors. My experience with industrial SSR devices has been they will operate perfectly for way longer than 20 years and possible longer, because other things failed before the SSR devices. I have a bin full of the old SSR units and use one occasionally in new projects.
No, was not thinking of those little boards. Something more industrial std.
Agreed, I need to watch coil current.
I do already have a 1A 12V transistorised output, that I can utilise to enable a contactor.
My biggest issue is getting my head around the requires coding...its still a challange😉
