Hi
I don’t have any site where I could post this, so I decied to post this here.
I got one cheap Zigbee Gledopto DIN-rail mounted energy meter - relay device from Aliexpress some time ago just to check if it’s good/safe enough to be used. The price was some 17€ which was very compelling.
Unfortunately the design follows the way-too-common cheap-as-possible, and this does not comply to EU standards.
Power input circuit for the high voltage buck regulator:
- No X rated EMI capacitor
- ZNR 07D471K — 7mm disc, 470V varistor (Bourns equivalent) with 4.7R in series.
- M7 marked single rectifier diode
- Aishi CD11GAS — 4.7µF 400V 105°C elco (2.5mm pitch!)
- 100R series resistor
- Second Aishi elco
Picture of the top side with some components removed.
Picture of the bottom side of the DC rectifer on the PCB:
I measured approx. approx 1.3mm distance from the rectified mains VAC to the 0V/GND reference of the DC side with a Mitutoyo caliper between the elco legs.
(Paid) AI generated conclusion from a long thread
The device uses two 400V electrolytic capacitors in a half-wave rectifier supply operating from 230V mains, where the rectified peak voltage reaches 325V — leaving only 75V headroom before the capacitors’ absolute rating. The onboard 7mm MOV provides inadequate surge energy absorption and insufficient clamping speed for fast transients, meaning voltage spikes from the mains reach the capacitors before the MOV fully conducts. A fast spike can momentarily overstress the capacitor dielectric, initiating internal breakdown and heating that boils the electrolyte, ruptures the case, and ignites the flammable vapour.
The PCB clearance distances between the high-voltage supply section and the low-voltage Zigbee circuitry measure approximately 1.3mm at the electrolytic capacitor pads — well below the 3mm minimum required for basic insulation through air at OVC III per IEC 60664-1. The creepage distances along the PCB surface are correspondingly short, falling well under the 4mm required for basic insulation and far below the 8mm required for reinforced insulation for FR4 material group IIIb at this installation category. While inadequate clearance creates a direct arc risk through air under transient overvoltage, it is the creepage path that becomes the dominant hazard under real installation conditions.
In unclean or outdoor-adjacent installations — including distribution panels in garages, outbuildings, or industrial premises — condensation, dust, and contamination deposit conductive films on the PCB surface. These films dramatically reduce the effective surface resistance along the creepage path, initiating tracking at voltages far below what the dry board would withstand. Once a carbon track forms it is self-sustaining even after the contamination dries, as the carbonised path has permanently lowered resistance regardless of subsequent drying. The creepage path from mains potential to the logic side then supports a sustained arc capable of igniting the FR4 substrate — a failure mode that neither the installer nor the upstream distribution panel fuse can be relied upon to prevent.
Also to consider
I made some preliminary thermal tests with 5A constant load through the relay. It seems to heat up too much compared to it’s specifications and a high power heat pump is likely to be too much for the relay.
I have a M.Sc degree from 2003 and I have been designing industrial electronics ever since, so I have some expertice of what I’m talking about.
The PCB might be possible to modify to meet clearance and creepage distances but the time and component costs are likely to render that unpractical - if not for hobby purposes.
I hope this is seen by any who consider the device for their home automation or any other project.
Kind regards, ej
.
