Hello,
I purchased a second hand “pure sine” inverter. It was cheap (15 €) and the seller said it was “like new” so I assumed it would be functional. Its robust look and the “made in Switzerland” statement piqued my interest.
link to the actual productIn fact, it didn’t work at all, but fortunately fault-finding was easy. It seems to have encountered an overvoltage situation, and a 15 V Zener and its series resistor that were across the 12 V input were destroyed. Temporarily replacing the resistor with a jumper and removing the Zener showed that the rest of the circuit was still OK.
The inverter seems to work, and I’ve been able to power a corded Dremel with it (that was the main goal of this purchase, as I’ve a little bit of grinding to do in a remote area and I wasn’t sure I wanted to buy a cordless Dremel just for that).
But the inverter emits a rather unpleasant buzzing noise. Not super loud, but definitely noticeable. The unit claims to be “for stand-alone applications”, and I’m not sure a customer who just paid 250 € / 300 CHF for a brand new unit would be happy to hear it 24/7 in his RV or boat. Maybe the noise is perfectly normal; I just don’t have any way to compare with a new unit!
The noise doesn’t change with load. But what I find a bit strange is that this is not only a constant hum, there is also a “KRRCHHRRRKKK… KRRRCHH….. KCHCHRRRCHCRRKK” noise on top of it, which appears to be random.
Here is a picture of the board :
As a learning exercise I made a schematic:
PDF versionOn the secondary side of the transformer there is only a 0.47uF MKT capacitor across the winding + two 0.01uF ones to chassis ground, nothing more.
It’s the first time I study an inverter but I think the waveforms I saw when probing around made sense.
However I noticed there is a bit of jitter on the PWM signals sent by pins 10/11 of the PIC to the pins 3/4 of the FET driver, as seen here :
That happens at the beginning of each new half cycle of the sine wave, in both polarities.
Do you think it’s normal? As the PIC is probably going through a lookup table to determine the pulse length, I can’t imagine anything introducing randomness there, but I may be wrong of course. May the PIC be making adjustments to compensate for a division problem between its own clock and the wanted 50 Hz output?
The PIC also monitors the input voltage and also there is an overcurrent protection (I guess it does so by monitoring the voltage on the primary), maybe there are timing issues with the code that takes care of that…
May this be related to the ceramic resonator? It’s marked “ 160U. “ but I’ve not been able to probe it, with a scope or even with the 10M input impedance of my DMM, it stops oscillating when I probe on PIC pin 15.
If you have any info on this, I’d be glad to hear.
Thanks!