Brennenstuhl PM 231E power meter zero reading fix (#P19)
Quick fix today – a colleague recently complained to me about his Brennenstuhl Schuko power meter only reading a static “0.0W” regardless of the load that was connected to it. Display was fine, the outgoing Schuko was fine (we checked with a 2kW space heater), but the unit would display no power usage at all.
As there’s no long tri-wing bits (bastards!) at work, I took the unit home only to discover that my bits are also too short to reach the damn screws. He said he didn’t care and was ready to buy another unit (I kindly recommended the ELV Energy Master Basic 2, which will serve as comparison unit later), so I drilled into the case to widen access a little. Turns out the amount of material that needs to be removed from the recess in order to make the wider parts of the bit fit is exactly equal to the wall thickness of the case – oops
At the time I didn’t really bother making photos as it looked like a substantial part of the circuitry has died, making this a total loss. So this is the unit after given my special treatment…
So the insides are pretty…meh, I would have hoped for a little more engineering as this is a mains power meter.
The four cylinders are the inner parts that held back the screws. Single sided PCB for the display part, double sided load for the main board. The battery holder is a terrible construction fail, but very common to those low-end units. I think the terminals may even become live during operation, hence the warning yada-yada on the back.
This is, in fact, the original PCB. Flux and other yucky stuff isn’t cleaned, there is fine strands of wires going places, THT component leads are trimmed at various heights, I’d say that’s horrible for a mains powered device. (I know the photo is shit – this is all I have, sorry)
The main chip is a Cirrus Logic CS5460 variant which I did not find on the interwebs, but the CS5460A is identical in the important bits. Some marketing BS from the datasheed: It’s using a twin Delta-Sigma ADC, does a few nifty calculations for all the various figures, and has serial interfaces to e.g. run a serial display as present on this unit. It does provide power, RMS current, RMS voltage and power factor readings for single-phase meters.
Great, so I checked the positive supplies VD+ and VA+, as well as the EOUT pin that outputs a PWM signal proportional to the power draw. EOUT is completely dead, which matches the symptoms of no (digital) data reaching the LCD – probably caused by the supply voltage being at 2.7V which is just about the brownout detection threshold for the nominal 5V ±10% supply. Gotcha.
You tell me, but I cannot find a proper power supply in there, so this likely is just a capacitive dropper circuit as described only a couple hundred times in bigclive videos. And what’s the limiting component of such a dropper? The capacitor of course. The capacitor is also the component that gets suspiciously warm when there’s absolutely no need to:
Annoyingly not captured anywhere, the original X2 Wan Hung Lo branded cap was 330pF. A quick test by the transistor tester revealed that the desoldered component probably just has half of it left, meaning lower energy delivery to the bridge rectifier, delivering a lower overall voltage at a given current. That thing is just starving of power.
One (verified!) 300pF cap from the junk drawer later, everything is working again. That was easy…but I destroyed the case in the process, how do I fix it back together?
There’s virtually nothing that a bit of hot snot cannot fix
I did put everything together and just filled the long screw recess tubes with hot snot. And because I’m a dumb idiot, I missed the inner black Schuko part in my first try, which is when I had to get it apart once again. From drilling out the hotmelt glue I can confirm that this indeed will last for a long time, even though there’s not all that much of contact area between the parts, but the 200°C glue really does stick things together exceptionally well. Everything went fine the second time, and this is how the unit was delivered back.
Some comparison shots to my ELV meter – the Brennenstuhl photos are rotated for your convenience, as this bulky thing doesn’t fit my plugboards very well.
This is just the pretty bright VFD running in the professional grade Bosch Gourmet 8400 microwave/grill unit that is about my age:
Power factor of 1 is certainly not true for a 30 year old VFD driving circuit, but the power consumption matches my unit nicely. The PF reading does get more accurate with more power used, e.g. this double LiIon charger:
does consume about twice the power when fitted with those two cells:
4.5ish Watts, power factor between 0.49 and 0.66 (there’s a 5V power supply in there, which is why I have a secondary unit that is modified to take in 5V directly. Yes that’s common now, these are about 10 years old)
And finally my 65W 42T4416/42T4417 laptop power brick (yes, under varying load), which has a shockingly poor power factor for a high-profile manufacturer such as Lenovo. I understand that PFC is only required for power supplies of the next power bracket and up, but c’mon, 0.45 at just shy of 20W load? You certainly can do better.
Both meters agree at that level of power usage, and they do not drift apart all that much when going significantly higher. PF of a space heater is pretty close to 1 as expected, the wee 20W fan doesn’t contribute that much even at a terrible PF when run alone. So the Brennenstuhl does deliver power readings down to a very low 1-2W, but power factor reading is way off until about 15W. It’s only 10 to 15 bucks in retail, so that’s fine in terms of bang-per-buck. For serious hunters of rogue power guzzlers at home, I’d still recommend the ELV meter for 30€ delivered, which in addition to more accurate figures also offers a significantly easier to read (backlit!) display.
Oh, and once the Brennenstuhl fails a second time – just bin it, please