Repairing a heating blanket controller

[aheid]

It's January and quite cold here in Norway these days, so my girlfriend has been relying on her heating blanket to stay alive 

Sadly it stopped working, which caused a mild panic. I quickly found it was due to a blown fuse. Assuming the fuse had not committed suicide I poked some more and found some diodes connected in parallel with the heating wire was shorted out.

The heater seems to be driven by a SCR (marked S6004 DGS1), which seems to be ok (registered as open circuit both directions). The the shorted diode(s) are D3 and D4. I haven't desoldered them, my guess is only one of them is shorted.

From the topside image, H1 and H2 goes to the the heater element, measured ~140 Ohm. S1 and S2 is some thermal protection, measured ~980 Ohm at 22C room temp, a thermistor perhaps? For the curious ones the microcontroller is a Holtek HT66F30.

My questions are as follows:

What is the function of the diodes D3 and D4? My guess would be flyback/freewheeling diodes, due to the heating element being a long conductor hence having high inductance?

What's the point of putting two diodes in parallel? The first one that starts conducting will carry the majority of the load anyway, no? Is it OK to replace them with a single one of similar rating?

[BreakingOhmsLaw]

I think you are right.
This looks like a simple phase fired regulator. If it was a TVS (in case someone plugs it into the wrong voltage) you would not put it near the load but directly behind the fuse.

I'd replace load with a 1/2Watt 100k resistor, remove the diodes and measure the output on a scope. If you just see a positive half-wave of mains voltage, it's probably a freewheeling diode.


On another note, I will never understand why people want to wrap themselves in mains-powered wires.
But then, I don't live in Norway

[techneut]

It could be a safety issue to put in 2 diodes. If the diode fails and goes open there could be dangerously high voltage spikes on the heating element.

[aheid]

It could be a safety issue to put in 2 diodes. If the diode fails and goes open there could be dangerously high voltage spikes on the heating element.

That is a very good point, and one I should have been able to figure out myself, given that the other repair job I have going is a switch-mode converter that died because the freewheeling diode failed open... 

I desoldered one of the diodes and it happened to be the one that failed close. The other measured ~0.56V forward voltage, same as the other similar diodes on the board. With that gone the unit seems to be working fine.

Will be replacing both diodes just to be sure though.

[aheid]

If you just see a positive half-wave of mains voltage, it's probably a freewheeling diode.

I forgot to take a picture but it looks indeed like half-wave.

On another note, I will never understand why people want to wrap themselves in mains-powered wires.

The funny thing is she needs to have the tip of her feet outside the blanket, so that she doesn't overheat... 
I've given up understanding how that all works.

[bob91343]

I don't believe that at these frequencies, inductance of connecting wires is important.  As for your girlfriend using an electric blanket, it appears that you are neglecting an important thing with her.

I have used electric blankets for years and they are very nice, especially since I have no girlfriend.  My cat chooses to sleep on the other side of the bed.

[aheid]

I don't believe that at these frequencies, inductance of connecting wires is important.

So what are they there for then? I'm genuinely curious, as flyback/freewheeling diode is the only thing I can think of.

As for your girlfriend using an electric blanket, it appears that you are neglecting an important thing with her.

It's all very much tongue-in-cheek, and something we frequently joke about. Easy to forget it doesn't come across as easily via text.

[bob91343]

Sometimes a diode is used to reduce power, connecting the load only on one half cycle of the mains.

Look on a reactance chart and see how low reactance ordinary wire is at power line frequency.  The inductance of these wires is in the microhenry range, if that.

[aheid]

Sometimes a diode is used to reduce power, connecting the load only on one half cycle of the mains.

The diodes are in parallel with the load though, so can't be that in this case?

Anyway thanks for the help, the blanket has been working fine for some time now. Happy wife happy life etc 

[fzabkar]

Are the two diodes connected in anti-parallel?

ISTM that the power supply for the Holtek uC is a zener regulated transformerless linear supply using the impedance of the large capacitor (C1) as a voltage reducer. This would in turn suggest that the source is full wave AC.

[aheid]

Are the two diodes connected in anti-parallel?

ISTM that the power supply for the Holtek uC is a zener regulated transformerless linear supply using the impedance of the large capacitor (C1) as a voltage reducer. This would in turn suggest that the source is full wave AC.

I forgot to specify, the input to the board is 240V AC. The diodes are connected in the same direction, cathode above anode in the picture.

[perieanuo]

It's January and quite cold here in Norway these days, so my girlfriend has been relying on her heating blanket to stay alive 

Sadly it stopped working, which caused a mild panic. I quickly found it was due to a blown fuse. Assuming the fuse had not committed suicide I poked some more and found some diodes connected in parallel with the heating wire was shorted out.

The heater seems to be driven by a SCR (marked S6004 DGS1), which seems to be ok (registered as open circuit both directions). The the shorted diode(s) are D3 and D4. I haven't desoldered them, my guess is only one of them is shorted.

From the topside image, H1 and H2 goes to the the heater element, measured ~140 Ohm. S1 and S2 is some thermal protection, measured ~980 Ohm at 22C room temp, a thermistor perhaps? For the curious ones the microcontroller is a Holtek HT66F30.

My questions are as follows:

What is the function of the diodes D3 and D4? My guess would be flyback/freewheeling diodes, due to the heating element being a long conductor hence having high inductance?

What's the point of putting two diodes in parallel? The first one that starts conducting will carry the majority of the load anyway, no? Is it OK to replace them with a single one of similar rating?

if D3/4 are parallel, maybe T1 is thyristor, you gotta respect the polarities when you replace them!!
you removed the little blue alien, it's some ptc/ntc I'd guess, is he good?
i don't get the thermal fuse you mentioned, the regulating block have thermal contact with the pillow???

[aheid]

if D3/4 are parallel, maybe T1 is thyristor, you gotta respect the polarities when you replace them!!
you removed the little blue alien, it's some ptc/ntc I'd guess, is he good?
i don't get the thermal fuse you mentioned, the regulating block have thermal contact with the pillow???

Yep I took great care to respect the polarity of the diodes.

I did not remove the blue alien    As far as I can see it's a TVS or similar, it's in parallel with the inputs.

As you can see the thermal fuse has thermal paste between it and the two parallel resistors. The fuse is in the return path of the heating element, while the resistors are part of the microcontroller power supply path. My guess was that it's to cut off the heating blanket in case there's excessive current drawn by the controller circuit.

[perieanuo]

if D3/4 are parallel, maybe T1 is thyristor, you gotta respect the polarities when you replace them!!
you removed the little blue alien, it's some ptc/ntc I'd guess, is he good?
i don't get the thermal fuse you mentioned, the regulating block have thermal contact with the pillow???

Yep I took great care to respect the polarity of the diodes.

I did not remove the blue alien    As far as I can see it's a TVS or similar, it's in parallel with the inputs.

As you can see the thermal fuse has thermal paste between it and the two parallel resistors. The fuse is in the return path of the heating element, while the resistors are part of the microcontroller power supply path. My guess was that it's to cut off the heating blanket in case there's excessive current drawn by the controller circuit.

in your 2'nd photo (bottom side with pcb tracks), the terminal of the ptc is unsoldered, isn't he in series with the thing marked "fuse' on bottom side?
the power path is AC>fuse F1>ptc>...
the photo says so

[aheid]

in your 2'nd photo (bottom side with pcb tracks), the terminal of the ptc is unsoldered, isn't he in series with the thing marked "fuse' on bottom side?
the power path is AC>fuse F1>ptc>...
the photo says so

I could be wrong of course. I attached a better photo of that area of the board, along with a rotated and flipped version of the topside, so components are on the same side as the bottom side when viewed side by side. I annotated the bottom side according to my interpretation.

I also added a closeup of the switching element, I think it's an SCR like these[1]?

[1]: https://www.littelfuse.com/products/power-semiconductors/discrete-thyristors/scr/sxx04xsx/s6004ds1.aspx

[perieanuo]

you're right, tvs is protection varistor, should be 'open'
so i'd guess, if the thyristor came out undamaged (no short between A and K), just put new diodes and test
sometimes gate can dye as well, test also G-K junction for the thyristor to be sure
of course if you have a scope you can test if the micro gets the thyristor ignited, if you don't just check with the diodes removed if you got the digit working, that means the micro and his power supply are alive

[GORS]

Hello,
I have the same controller and it is broken at the moment. I replaced the thermal fuse (which was activated and turned into an open) and I replaced the R12 (which I assumed is a 50 ohm resistor). The display is alive now but when I turn “on” the system, the display goes to “4” (which is normal), but then it switches immediately to “E.”, which I assume is an indication for “ERROR”.
Any ideas?