I had a strange experience yesterday. I turned my Devel inverter on and it wouldn’t turn on, however the lcd display came on and said the battery is sitting at 19.4v (this is a 48v battery)

I tested the battery with a multimeter and got 50.8v. I tried a couple things like disconnecting the inverter from the battery and connecting again, I also pressed the reset button on the inverter a few times, but the battery voltage (according to the inverter) stayed at 19v to 20v. 

Then I turned it on and left it on. After about 5 minutes I returned to find the display saying the battery is at 50v, but still not coming on, only the display was on. After switching off and on again it finally came on. 
 

Any ideas out there as to why it would do this? Is there maybe a capacitor that needs to be filled with power before it switches on? 
 

Regards

Solaris

Edited January 8, 20205 yr by Solaris
Typo

5 minutes ago, Solaris said:

Any ideas out there as to why it would do this?

My guess would be one of the battery voltage divider resistors is intermittently high resistance, due to a manufacturing fault, crack, or bad soldering. If you are experienced with electronics, the various voltage divider resistors are fairly easy to spot, since there will be 3-5 resistors of the same value in series with no other parts nearby. Don't confuse with the voltage dividers for measuring bus voltage, inverter output voltage, or AC-in voltage.

Hi Coulomb 

Thank you for the quick reply.

3 hours ago, Coulomb said:

If you are experienced with electronics

I’m still an amateur. 🤦🏻‍♂️
I know of a “friend of a friend” who could perhaps help. 😁

3 hours ago, Coulomb said:

Don't confuse with the voltage dividers for measuring bus voltage, inverter output voltage, or AC-in voltage

The funny thing is that sometimes the AC-in voltage shows a reading like 57v or 120v or any random number... However there is no AC fed in at all. It’s is an off grid system. Perhaps there are a few bad resistors...

I think it will be best to take this inverter to the “friend of a friend” for assistance. 😔

on my 2KVA model the 1M smd voltage dividers are just above the fan and in high humidity condensation was building up on them, eventually they failed due to corrosion, i only noticed after inverter kept shutting down saying (i think error 09) and high battery voltage and reading was >30v even though batteries were at 26V (2kva  is 24V) 

I cleaned up the board and replaced the 1M resistors, and also put a small 1M preset instead of one of them to calibrate the reading as the divider is not that precise in my opinion then I sprayed some varnish over the area after drying it properly to avoid this condensation building up as my inverter is outside in a cabinet but in my country it could be 25C with 99% humidity and the fans make the situation worst due to venturi effect i think

2 hours ago, 9H1LO said:

in high humidity condensation was building up on them, eventually they failed due to corrosion

This could be a similar problem for me. The house where I was staying for the past 6 years was very damp. I’ve now just moved to a new place which is dry. Strangely enough the problem only started in this new place. Anyway I suppose that is just coincidence. 
I will tell whichever electrician I use what you said about the 1M resistor and 1M preset. 
By the way there is a preset on top which is numbered 0-9 what number should it point to as a standard setting? 
 

Thank you 

Edited January 8, 20205 yr by Solaris
Typo

i do not know about this preset, i don't think mine has it and i would not go around varying things unless you have a service manual explaining what they are or you know what you are doing, it could make things worse

i don't think its something for an electrician (the inverter installer type) find someone with electronics repair skills that can go down to component level if you don't have the skills to do it yourself 

I looked inside and see several resistors of different kinds. There are round ones that have different colour rings on them right next to several mosfets on the DC side. Then there are the flat kind on the main board. which ones are likely to be faulty.  
Sorry my electronic knowledge is very limited. 😢

 

 

22 minutes ago, Solaris said:

There are round ones that have different colour rings on them right next to several mosfets on the DC side.

Those are through-hole resistors. They are likely gate resistors for the MOSFETs. Not actually in a series string.

Quote

Then there are the flat kind on the main board.

Those are Surface Mound Devices (SMDs). The ones you will need to look for are the larger ones, a size called M2012 (though strangely often also by their imperial size, 0805). That's because the larger size (up to a point) have higher voltage ratings, important when you are measuring high voltage. You have identified at least 2 strings of these; well done. Here is the bad news: there are just as many parts, if not more, under the PCB 😕

Quote

which ones are likely to be faulty.  

They can all be faulty. After a big bang event, as well as the semiconductors, it's actually more likely that gate resistors are damaged.

But the M2012 SMD devices I mention above are more likely to be used in the voltage divider strings that measure battery voltage, among other things. Since the battery voltage is significantly lower than the bus voltage or the AC voltages, the battery voltage divider is likely to have fewer resistors total. So look for a shorter string of M2012 identical value resistors. You could try cleaning off any dust (perhaps with methylated spirits), and coating with a PCB lacquer to see if that fixes your problem. Of course, if you find a cracked one, it needs to be replaced.