Looking for type of part/specs

I have a Floor heating thermostat which stopped working, I pretty much can see where the problem is but but I have never seen this kind of component, it looks like a coil but the main runs through it so maybe a sensing device but there are many legs coming out of it. Here are some pictures, I am referring the the white plastic part. looking for specs so that I can test/or find a replacement.

It looks like a current transformer; possibly a custom part for that manufacturer.

From the photos, that is probably OK and it's the power conductor through the centre that has had a bad solder connection and started arcing/burning.

Try cleaning the scorched area with IPA and re-soldering the heavy conductor. If necessary, clean the solder resist off around the damaged area and add some heavy copper wire to bridge between the centre conductor and the PCB copper that it should connect to.

(It may need nothing more than cleaning and re-soldering; just make sure the solder flows properly around the heavy conductors).

I could be wrong but I think it is a relay.
On this image the under side of the white component have 8 connections from which the 4 red are not connected to anything and the 2 green is a coil and I get 16 ohms between them.
One of the live wire going through the middle of the coil has continuity to "A" and I have also continuity to the pad on the left of "A". There is continuity between "B" and the pad on its left but no continuity between "A" and "B".
I am thinking, feeding a voltage to the coil should make a contact between "A" and "B" but the question is how much voltage should I test with ? (EDIT: tried 1.5V but didn't make a difference)

If it has two separate conductors through the centre, it's most likely for measuring any earth leakage (imbalance between the two opposite-direction current paths).
(The designation for it on the PCB, starting with "L", also implies an inductor rather than a relay).

If so, the A and B points have the full supply or output voltage across them.

The power relays look to be next to that on the same sub-circuit board?

There are 2 power relays on the small board designated as LS1 and LS2
As for the part I am concerned about it is designated as LS1 (on the main board)
There is only one input for the mains at the top but then separated as A and B at the output which is where the black smoke happened. There are 2 mains black wires (220V) and 2 yellow wires I believe feeding 220V to the radiant floor.
it is kind of hard to see but it looks like each of the black relay connects each yellow wire to the floor and the white (mysterious relay) connects one of the yellow wire to the black mains. Here is another view...

It looks like a design failure to me. If the gap between CONT. points is 240V certain contamination levels would likely fail explosively due to breakdown voltage.

Rather than look for the cheapest 50A Contactor 240V relay on Alibaba, I would modify the design to use an OMRON part or one of equivalent quality.

- Minimum wiring size would be AWG#8 or 3.3 mm D

This is what I suggest.

Using lots of flux, a heatgun and solder wick carefully straighten flanges and remove PCB then remove LC1 from PCB. Drill holes to insert wires top side and solder bottom side. The Coil is likely 24V (verify).
Mount the Contactor nearby with PU adhesive ( subfloor adhesive types in tubes from H/W store takes 2 days to cure unlike UV or former high VOC types.) ( very strong/hard)

Coat all solder areas with Silicone spray and/or RTV Silicone after cleaning with IPA. Hopefully it will never happen again.

Their part looks like a $5 Chinese part after cost reduction that sells for $20 part on Alibaba rather than a well insulated rugged part for > $50 chassis mnt ~$500 on a DIN Rail. I think contact open arc gas and carbon may also escape between the solder joints but it's hard to tell from photos.

Alibaba example Quality of silver alloy contacts is unknown.



Feel free to use smaller ratings if you know.

The part marked LS1 is part of the inbuilt GCFI (Ground Fault Circuit Interrupter) circuit and not something to be messed with - replace the thermostat with a new one and get the heating elements checked for any earth leakage.

This looks to be an expensive field replaceable unit typically sold at 5x cost. Without a cause, perhaps rushing to fix is premature. I also over estimated the breaker current and wish to change that from 50A to 20A.

Synapse site "

• Universal Voltage (SE72 Series): Suitable for 120V or 240V applications High amperage - up to 16 Amps resistive load"

Any clues on history, age, possible causes. The evidence points to a fairly small breakdown. (no craters) It's hard to tell if the arc/carbon spray is an internal from the vent between the mechanical mount (CONT-CONT)

I wonder if you have any tools to perform a low current (< 10 uA) leakage or Hipot test on the heater wires to gnd. to 3kV also LCR capacitance on each lead to gnd.

Any evidence of water leaks?

I am not in a position to modify the design. This is a programable thermostat that is quite expensive. It was replaced with a non programable version for less money but still expensive. I was hoping to fix the original and return the non programable. I think at this point I will take apart and remove the mystery part in the hope I can see a part number otherwise I will let go. I would not sleep again if my interaction would cause a fire...

Thanks to all who responded.

There are other types of floor heating thermostats that can be programmed (with a schedule even) like:

I use their baseboard version of controllers (different from in-floor control). Control is via your phone app. They are rated 16A at 240V or 120V.

If you have a floor heating without sensor, then a basic baseboard controller should work.

Mike,
You can energize the relays standalone with 24Vdc

Does it look like a component failure (contacts stuck off or on)?
How much load do you have? kW or Amps? Is it within web spec?
How old was it?

I suspect the linear solenoid Contactor had a poor alloy and not Silver-Nickel-etc from a custom Chinese supplier as nobody else would put a contactor on an orthogonal PCB mount like this.

If it was only used for ground fault as suggested, then using a different controller is not a fix, if the fault is still unknown.

But if used for thermal cycling while the other relays for zone control. Cycle count in design is critical and MTBF depends on the quality of the design and part.

If the contacts get burnt out then they might look like the carbon escaped out of the PCB bottom center hole. It could be normal for a 10 yr life but abnormal if less than 2 y ( just beyond 1 y warranty)

I would complain to Synapse's Marketting group and ask if this represents their brand quality?

Sorry I don't have much info, it is from my son's house and I have with me only the thermostat.
Although the circuit boards have markings as being Synapse, the front of the thermostat which looks exactly as SE72 model from your picture above, the brand name on the faceplate is Stelpro (maybe rebranded from a Canadian reseller)

After taking a closer look, I was wrong this is not a relay. it's hard to explain in words so here is another picture.

The 2 mains black wires are actually going to the 2 relays on the small board. there is one connection (in red) on this side of the board and in yellow on the other side of the board. You can see on small board and bottom of main board where they connect. The white device contains not one but 2 coils, one being 16 ohms and the other is open so I guess this is the problem. I am wondering what could have caused this... a power surge ?
The radiant floor is currently working with the other thermostat so I think no problem there.

The two coils are often used in latching contactors/relays. 24V/16 Ohms = 1.5A , 36W which it can handle for a plunge then shut off and latch. Springs hold either position and I would be shocked if 240V is across the red/yellow heavy mounting tabs with such a small gap.

( I have been assuming L1, L2 240Vac 60Hz is switched as a double pole.)

I see Stelpro acquired US Synapse and are located near you.

Call them and explain you would like to repair the part or get it replaced. It is nice to see their history and your pride in protecting their safety image from marginal failures is motivating you to contact them.

I contacted Stelpro to see if that part is available. My hopes are not very high. This looks to be from the year 2012.
I removed the part from the board and I am curious to know what is the purpose. I have never seen this type of part. I am more of computer IT than electronics but I have seen a lot. I can understand a coil around one wire sensing current but having two individual coils around the 2 wires (L1 and L2), I don't understand...
Here is an image, on the other side it is another identical coil. There is nothing between the input and output of the mains wires, I can freely move them in and out.

It is, basically, a transformer that detects if there is any imbalance in the currents flowing on the two lines Active & Neutral - the big conductors down the centre.

If the current is equal then you get nothing out of the coil (or very minimal), if the current is not equal then you will get an output from the coil which indicates a fault somewhere, usually leakage/short to earth. This triggers the protection circuit, cutting off the power.

So rjenkinsgb was spot on at post #4.

Sorry if it took me 15 posts to understand.

No need to apologise, I've been at this a looooong time and there are still times I need the help of others to understand stuff.

Yes RJ was spot on. And Augustinez was right that for you this is untouchable. This was a cost reduced discrete version of a GFCI that you can get on Alibaba for $15 . Yet the wire does not look like it can support the current to heat the floors and only enough to power the controller. An external GFCI safety would have been a better design like the ones you see in bathrooms with the pushbutton for 120V but made for twin lines relays like your two Panasonic Relays.

It looks like a fused wiring from stress or defect caused the failure. I wouldn't say it was an unsafe design but certainly not a reliable design. They tried to save a buck to protect the user from panel leakage (?) and maybe use the same 24V relay pair to regulate temperature and also panel leakage fault detection but not the floor heater. ( Just an educated guess from what I've seen)

An external GFCI for 240V/20A versions might $50 in Canada but I suspect the dual relays might be also used for power cycling to temperature setpoint for which this part is not needed.

Looks like you may be outta luck, but maybe the factory can refurbish it for you.

lilimike said:

I can understand a coil around one wire sensing current but having two individual coils around the 2 wires (L1 and L2), I don't understand...

Click to expand...

It senses the difference in current through the two conductors; if the current is exactly equal and opposite, it's the same as no current, as the magnetic fields cancel out.

The output is proportional to the current difference. I first heard of these as "Core balance" transformers, from an electrician (who worked in a mine) friend of my older brother, around 1970.

They are now what's in all basic earth leakage trips (ground fault interrupters).

More info; see around half way down the page:

(And while searching for something showing the operation, I discovered that with industrial type setups with a remotely wired transformer & separate detector & cutoff, that's still correctly called a core balance transformer, a name I'd not seen for 50 years!)

The common-mode CM choke is "common" to leakage current detectors called GFCI's in North American or RCD's anywhere else. The designs are slightly different in sensitivity by measuring unbalanced AC current. If you rotate a 1:1 transformer it is called a Balun or CM choke. Each design follows the regional safety authority specs
Digikey calls them Common-Mode chokes or 1:1 transformers. Out of more than 8000 CM chokes listed and half in stock, none match yours on the inside. Nothing is "standard". Some may be similar in function. The plastic housing is custom. Some are 240 V rated yet need to be ~3kV rated for line impulses so they use a ~385V MOV to clamp spikes. But MOV's wear out and should also be replaced. The magnet wire insulation then becomes unprotected and with sufficient voltage will short out and fuse open from a line transient.

A GFCI or RCD have a few dozen parts including a CM choke but are simple to use with IN = Line and OUT = Load .
In theory this one could be modified with an external one (given expertise and time to do so).

Something else on the main board must measure the current difference then switch off the dual 24 relays which may (?) also be used to regulate the heat. GFCI's are a little more sophisticated inside and do a better job ( better insulation, lower threshold , better high voltage insulation, resetable with a test button.)

Here's a bunch more that almost look the same but not quite. They must be certifed, and have a safety rating for grid voltage, current and inductance.