Lux Inverter Problem

[Stinsy]

It looks as if the connection between the busbar and the PCB inside the inverter has become loose, overheated, and damaged the PCB.

I have emailed the manufacturer to see what they suggest. But the inverter is well out of warranty...

[Marcus]

I've seen this before on a 3.6kw lux ac coupled battery unit - ended up doing a diy repair to the filter pcb assy - it's hard to tell if it was a poor solder joint, a poor terminal connection/ busbar connection after the fact.

If you do have any luck sourcing replacement parts could you let me know please, as the last time I checked that inverter the other battery pole (I.e. the one I hadn't repaired) was getting hot - it's on the to-do list to go back with the big soldering iron and repair it before it goes altogether.

[Saladin]

Looks like a line filter. If so: you don't need it, it's just for noise reduction. You can bypass it if you're not too fussy about the noise and harmonics.
It's a generic stand-alone module and pretty common on lots of stuff. I alway keep a few kilos of e-waste for that sortov thing. a 16Amp one is probably a bit rare because not too many things need that power.

Is the melty thing a relay or a capacitor? What are the cables north and south?

Theoretically (from my armchair) the loose terminal cooked the cap (or relay?). If it's a cap that's a parallel component though so the problem goes deeper If it's a relay it's probably expired contacts.

Check the inductor for continuity, cut the krusty cable back to shiny metal with new terminals and clean the mounting posts. With that sotov heat you might have burned a PCB trace and open circuited the board. You can bodge in a jumper wire and replace the cap/relay if that's all tis.

[Marcus]

Yes a line filter on the battery input (I assume - that's what it was on the one i patched up), but if it's 3.6kw at 48v nominal it's a bit more than 16a - more like 60-70a depending on battery voltage, and more amps if it's >3.6kw. I suspect that's part of the problem: the pcb seems a bit weedy for those currents - if they're not generously soldered then sustained loads get them too hot, and thermal cycling does the rest.

The plastic boxes are just caps (assuming it is the same as the one I've seen) - victim of the heat in the pcb I expect. If they're still readable they're easy to sub - and values not too critical if they're not readable - as it's d.c. bigger may be better.

[Stinsy]

So, I've been emailing LuxPower who've been helpful. They have aggreed that the EMI board is toast and not disputed my assertion that it is a manufacturing defect. However I need to deal with Infinity Innovations to arrange a replacement.

So now the problems are:
- It isn't clear if I'm getting a replacement Inverter or a replacement PCB.
- Infinity Innovations don't exactly have a stellar reputation for CS.
- Infinity Innovations aren't currently answering their phone and haven't responded to email.

Hopefully II are back in the office on Thursday and we can get it sorted.

It wouldn't be difficult to remove the board and bypass it. It seems the purpose is to prevent interference from the batteries from affecting the inverter and/or interference from the inverter from affecting the batteries. How necessary is this? Surely if it wasn't required it wouldn't be there?

[Marcus]

Well good luck on getting the replacement - a whole inverter would be great if to can manage it .

In my case the installer wasn't interested in honouring the warranty for the new house owner, so the owner asked me to patch it up.

The emi (electro magnetic interference) would likely come from the inverter rather than the batteries. The filter is there so the inverter can meet the required emc standards - without the filter board the battery cables would work as antenna transmitting the noise from the inverter - you could try it and see I suppose: bypass it and run the inverter flat out, charging or discharging and see if it affects anything. An analog radio would give you some idea.

But ideally repair or replace the filter if you can - even if it doesn't affect radio signals in its vicinity, the interference could affect the a/d voltage measurements of the bms and cause problems.

[Saladin]

I don't think I've seen a line filter on a battery before. I wonder if it's my lack of observation or my preference for low-frequency topologies.

I'd elect to repair rather than bypass. By all means bypass it to get it up and working again while waiting for parts to arrive but it's a bittova long term system shock absorber if you will. Inductors resist changes in current, capacitors resist changes in voltage.

It reduces noise, therefore harmonics & ripple.
Ripple can slowly degrade batteries with unwanted heat and transient voltage instabilities.
Sometimes it will increase audible noise & RF, which can be transmitted along the cable of your installation, make transformers rasp a bit more, motors shudder, SMPS's get grumpy, wifi devices to attentuate to overwhelm the harmonics etc..
Omitting it will send the transients into places you don't want like the often feeble surge protection passives in your downstream loads and other connected hardware.

It's probably not very hard to reverse engineer if you can read the component parts. The inductor's prolly A-Ok..they're a lot tougher than capacitors.
I'd just ask Lux if you can buy a replacement, it won't cost a lot and it'll expediate the process.

Sights like that board are the reason that I always check terminal torque on devices and consumer units every time I work on them regardless of whether I am working on that part or not.
A lottov HF and PWM devices can untap terminals with resonance from mechanical harmonics and or high vibration applications...like anything with an engine.
They're a bit like wheel lug nuts. Check after 30km of driving every time you swap a rim.

[Lincs Robert]

Ive seen this before with Lux units. There was a non-functional LXP3600ACS unit on ebay some months ago. The guy had opened it up & posted pics of this exact same board - which was similarly charred. If no luck with Infinity/Lux on a replacement then you could try Southwest Renewables based in Axminster, Devon who had one of these on ebay a while back. I suspect the same board is used in a range of Lux products?

I’d be especially interested in hearing how you get on with Infinity on this.

Ive just found these pics. The last one shows what it should look like, the other two the charred item from the ebay listing. All from LXP3600ACS.

Cheers - Rob

[Saladin]

Jaysus! I take it when Lux say "manufacturing defect" they mean an underspec-ed component. The CSA is too small. I bet they get loads failing. Manufactured obsolescence much?
You could beef it up with a heap more solder on the underside.


The last pic is an upgraded board. Those extra resistors are probably addressing issues the bean-counters ignored.