I have repaired a Sign WiFi Smart Plug with model number: SNSM-WASINGLE sold by Spares Nordic AB in Sweden.
The fault condition was that the device gave no signs of life: pressing the push button did not cycle the internal relay and the blue LED did not come on and it was not seen on the WiFi network.
The fault was very simple, capacitor C3 on the PCB with writing "KSJ" and "LOWESR" had literally blown its top off, after replacing it the device works fine again!
As I find linear high voltage regulator designs interesting I drew up the attached schematic of it.
Unfortunately I have been unable to find the manufacturer or model of the high voltage linear regulator U1, the only writing on it is: ZRDLD
To open the device up was easy: it is ultrasonic welded along the plastic seam but the weld can be broken by gently clamping the top part in a vice so the weld snaps.
To get the PCB out you need to desolder the live pin (right next to the MOV and fuse) and neutral pin (right next to the inductors) after that the PCB just lifts out.
However a warning is in place here: this is a non-isolated high voltage power supply, and live and neutral can be reversed since it uses the Schuko plug.
This means that any part of the PCB, even the "low voltage" areas including WiFi module are at mains potential wrt earth ground.
Do not work on a device like this unless you have the proper understanding of non-isolated power supplies!
Case in point: the below screenshot shows the voltage between earth ground and the negative or 0v side of the 10v capacitor C3, i.e. what you would feel if you touched it:
The WiFi module in this one was a Tuya WB2S. There is a Swedish website that shows disassembly steps, his uses an older revision with a different WiFi module:
https://www.telder.com/hur-flasha-om...fi-smart-plug/
The original cap is 6.3mm wide and 13mm high, I would recommend to stay with the same diameter since the Schuko earth connector sits right beside the capacitor.
(Remember that a capacitors casing is not considered to be isolated and since this whole device is not isolated there will be a potential difference wrt ground, refer to picture above).
A search on Mouser revealed Rubycon 10YXJ470M6.3X11 with a 2.5mm leg distance that would be a good low ESR substitute, also Würth 860160273020 should be good.
For testing I actually installed a used 470µF Nichicon VR 85°C capacitor and it worked fine but I would not recommend it because it is not low ESR and the device runs pretty hot.
I measured a voltage of 4.74v on the C3 capacitor, this seems a bit low since the relay is rated for 5v and there will be some additional drop in the transistor Q1 but it seems to work fine.
For another test I used a proper low ESR Panasonic FR 470µF capacitor (albeit physically too large to fit well) and it appears that the ripple is almost cut in half wrt the Nichicon VR:
As for the 2x main bulk filtering capacitors they are also of the same brand but they measured fine in circuit so I left them alone.
Remember they only see 50Hz ripple so they should last longer, I was curious if the inductors would make the caps charge to different voltage levels but I saw 334VDC on both.
That should mean they are also both working fine since this uses a single diode (D3) for rectification and not a full wave bridge rectifier so they need to support the load for a longer time.
The fault condition was that the device gave no signs of life: pressing the push button did not cycle the internal relay and the blue LED did not come on and it was not seen on the WiFi network.
The fault was very simple, capacitor C3 on the PCB with writing "KSJ" and "LOWESR" had literally blown its top off, after replacing it the device works fine again!
As I find linear high voltage regulator designs interesting I drew up the attached schematic of it.
Unfortunately I have been unable to find the manufacturer or model of the high voltage linear regulator U1, the only writing on it is: ZRDLD
To open the device up was easy: it is ultrasonic welded along the plastic seam but the weld can be broken by gently clamping the top part in a vice so the weld snaps.
To get the PCB out you need to desolder the live pin (right next to the MOV and fuse) and neutral pin (right next to the inductors) after that the PCB just lifts out.
However a warning is in place here: this is a non-isolated high voltage power supply, and live and neutral can be reversed since it uses the Schuko plug.
This means that any part of the PCB, even the "low voltage" areas including WiFi module are at mains potential wrt earth ground.
Do not work on a device like this unless you have the proper understanding of non-isolated power supplies!
Case in point: the below screenshot shows the voltage between earth ground and the negative or 0v side of the 10v capacitor C3, i.e. what you would feel if you touched it:
The WiFi module in this one was a Tuya WB2S. There is a Swedish website that shows disassembly steps, his uses an older revision with a different WiFi module:
https://www.telder.com/hur-flasha-om...fi-smart-plug/
The original cap is 6.3mm wide and 13mm high, I would recommend to stay with the same diameter since the Schuko earth connector sits right beside the capacitor.
(Remember that a capacitors casing is not considered to be isolated and since this whole device is not isolated there will be a potential difference wrt ground, refer to picture above).
A search on Mouser revealed Rubycon 10YXJ470M6.3X11 with a 2.5mm leg distance that would be a good low ESR substitute, also Würth 860160273020 should be good.
For testing I actually installed a used 470µF Nichicon VR 85°C capacitor and it worked fine but I would not recommend it because it is not low ESR and the device runs pretty hot.
I measured a voltage of 4.74v on the C3 capacitor, this seems a bit low since the relay is rated for 5v and there will be some additional drop in the transistor Q1 but it seems to work fine.
For another test I used a proper low ESR Panasonic FR 470µF capacitor (albeit physically too large to fit well) and it appears that the ripple is almost cut in half wrt the Nichicon VR:
As for the 2x main bulk filtering capacitors they are also of the same brand but they measured fine in circuit so I left them alone.
Remember they only see 50Hz ripple so they should last longer, I was curious if the inductors would make the caps charge to different voltage levels but I saw 334VDC on both.
That should mean they are also both working fine since this uses a single diode (D3) for rectification and not a full wave bridge rectifier so they need to support the load for a longer time.
Attached Files
Comment