Chinese PID controllers ex Ebay

[Archibald Tarquin Blenkinsopp]

Hi, I connected a REX C100 up, it went splat, blew the fuse and 32A ring main breaker.

Opened it, investigated the thing, the mains input track was intact, but the thermocouple input connections were burnt.

Tried another, my last one with the alarm relay function. I removed everything bar the mains input. This also went bang, blowing the fuse.

The same problem was apparent, the mains input tracks on the internal PCB were intact, but the thermocouple tracks leading to the unconnected terminals, (I only hooked up the mains) were burnt.

I suspect that the units are actually 24V with a protective device in the case for the PT100 in case someone connects mains to the temperature input.

I found a last one without an alarm relay unfortunately, it worked. The only difference was a yellow caution sticker by the mains input. Yet they were all ordered as 240V and it ctearly states 240V on the case.

Does this sound familiar to anyone?

I cant remember the vendor, so it is for interest and I could do with ordering replacements. Ideally working ones. I do not have the luxury of waiting for warranty replacements if it's a standard problem.

AB From the well pummelled keyboard of Archibald Tarquin Blenkinsopp Esq

[Theo]

Not exactly, but I think there are a number of variations of the REX C100 clones and they don't populate all the components. Try comparing yours with some of the teardown videos. For example this:

like there's a chunky transformer included - is that missing from yours?

Theo (used one successfully powered from mains)

[Archibald Tarquin Blenkinsopp]

Thank you, it suggests that there may be an inherent problem as mine blew on switch on too, both of them.

Mine has the transformer. The disk appears to be a transorb on the photograph though, yet my device uses an NTC thermistor.

It is electrically similar though, the thermistor on mine coming from the land labelled 1 on the photo, and the other land "2" is where the resistor connects into the bridge rectifier. It,s a minature surface mount unit on mine, not discrete diodes as in the photo.

It really is odd though, there is no sign of damage on the mains in [1 and 2], yet the amount of print that is stripped and burnt from lands

9 and 10 is unreal.

The second unit only had 1 and 2 connected to the mains and 9 and 10 were completely open, yet in both units 9 and 10 are splattered.

My theory that it was 24V was obviously wrong, in comparing the unit to your linked photo I checked the capacitor fed from the mains input in my unit it's 4.7 uf 450V, so it must be the Volts stated!

Oh well, I'll saw the outer case open in the morning there is something damned weird about this.

Reminds me of a load of PIR switches I had, I fitted a fair few when doing some DIY wiring, but for no obvious reason three blew, quite a flash & bang and they never worked again, but I could find no evidence of a short at all. All surface mount, so how something can produce noise and light yet show no damage beats me.

Clever these Chinese :-)

AB From the well pummelled keyboard of Archibald Tarquin Blenkinsopp Esq

[Fredxx]

I have purchased items that were sold as 240V but in reality were 110V. The transformer saturates taking out fuses without any visible damage on the primary side.

If you had a variac you might have been able to check function without the light and smoke.

[Paul]

Could the problem be tin whiskers ?

Maybe a second attempt to use them will find them functional.

Or, maybe a fault other than across mains will reveal itself.

Before the ROHS era, we had an experimental assembly at work which did this. Some genius changed the previous generation gold-on-ceramic design to a pure-tin-on-ceramic design. You could plug them into equipment, and roughly 365 days later, a short would develop. Once we knew what to look for (it was a low voltage assembly), we could take a microscope and a hobby knife and cut the whiskers and reinstall the module. They would of course, grow back. Cutting them, is a workaround.

What I don't know right off hand, is if tin growth needs bias. And whether tin would grow if the assembly was just stored in the box for a year.

The idea is, the Chinese production could do functional test, the unit could pass, they put it in the box, a tin whisker grows and shorts mains, you plug it in and it blows.

Try using the module again (if you dare) :-/

Paul

[Archibald Tarquin Blenkinsopp]

Thanks for the input everyone. I have stripped the cases of the offending units this morning. As with all these things it is just a shell anyway.

From the rear.

1. Mains in 6. ALM

2.Mains In 7.ALM 3 Common 8. RT

1. SSR+ 9.TC-

5.SSR- 10. TC+

On the first I had the mains in to 1&2

1. is not used, although labelled there is no internal connection on the PCB anyway.

4&5 went to the working [proven SSR]

6 I strapped to 1, with 7 taken to a 240Vac relay

8 no connection

9 and 10 went to a K type

That was the first unit, it was connected correctly as I was using the label visible from the fixed side of the JB that I was working in to guide.

The controller might be rotated to put mains on the thermocouple inputs, but this wasn't the case as the clip securing the PCB's to the shell is obviously right or wrong.

On the first wired as above, the PCB tracks to 9 and ten were blown, the shell to PCB contact clips on 9 and 10 were burnt also.

The track that vapourised does not look anywhere near the current handling of a 32A breaker, which tripped. The 3A fuse was taken out also incidentally.

There is no damage to the PCB lands or track going to 1 and 2.

Then I tried controller 2 with only the mains connected, nothing else. Again splat, the supply fuse 5A this time as I had no more threes, blew and the same damage was evident although the PCB was burnt on the tracks leading to the shells 9 and 10 terminals, there was no damage to the terminals this time. Not too surprising as I had no Thermocouple connected now.

Much testing and head scratching later, I dropped in my last controller which unfortunately has no alarm relay.

It works!

If someone presented me with the problem, I would have decided mains had been applied up the thermocouple input and this would account for the damage, but the physical arrangement makes this an obvious thing to be picked up and even if senility is setting in, the third is in the same place, same wiring and exactly the same appearance.

One of lifes little mysteries.

I have ordered a replacement from Amazon this time. I guess that there are reasons why Eurotherm controllers dont sell for twenty quid or less :-(

AB

From the well pummelled keyboard of Archibald Tarquin Blenkinsopp Esq

[Theo]

Hang on a mo. According to a different datasheet:

Power input is 6 and 7, alarm relay contacts are 1/2. Thermocouple 8/9 or thermistor 8/9/10, output 4/5, 3 can part of alarm or triac output.

If you put the relay contacts 1/2 across the mains that might explain things...

Of course, in the Cambrian explosion of clones who knows how yours has evolved.

Did these all come from the same vendor?

Theo

[Archibald Tarquin Blenkinsopp]

It is tempting to think that I might have been in error at some point, my sanity wouldn't be at risk if it were mere incompetence :-).

If you look at the image you originally sent though, the sticky label suggests pins 1 & 2, if you then follow the image to the last PCB layout, the photo shows the mains input bits clearly going to the lands labelled 1 & 2. If you then look at the PCB layout and the case accommodating the PCB, it has to go with the lands connected to the thermistor, bridge and resistor to terminals 1 & 2.

Although in the case of your drawing, my thermistor in series with the mains input becomes a transorb, which although difficult to see the connections must be across the resistor feed to the bridge.

I do not think all three came from the same vendor. My third working one would have been a trial to see if the bargain was too good to be true. :-)

The two faulty ones would have been for long term use, one being the spare.

Hence the relay out function on the service units, as protection wasn't an issue for the trial [that never happenned, incidentally].

Anyway thanks for the suggestion, I must admit this one has me flummoxed.

AB From the well pummelled keyboard of Archibald Tarquin Blenkinsopp Esq

[Archibald Tarquin Blenkinsopp]

To those that provided an input, thank you.

My Amazon replacement arrived Monday and I will have to mount this in the JB that I am currently using to contain the objects of the enterprise, 20 Khaki Campbell duck eggs.

I only just found out that humidity controllers are quite cheap also, but most are relay out on both channels, so I am using a bath covered with an aluminum foil "throttle" to control manually.

It will take a month to find out how successful the system is, and then perhaps someone might have some tips on duck shed manufacture.

'Trouble is with DIY, the first attempt is always the point where you find out what you should have done :-(

AB

From the well pummelled keyboard of Archibald Tarquin Blenkinsopp Esq