Author Topic: Inverter showing wrong voltage output  (Read 1395 times)

0 Members and 1 Guest are viewing this topic.

fabieville

  • Full Member
  • ***
  • Posts: 135
Inverter showing wrong voltage output
« on: May 27, 2018, 02:32:56 AM »
I recently got a used 24v 3000watt 120vac 60hz reliable pure sinewave inverter. It has a 2 display on the front that shows the dc voltage input and the output voltage reading. In normal operation it always shows about 120 -122vac output on the display.

Recently i was pulling about 1100watt from it constantly and I realize that the case was getting very warm to a point that i assume the fans should have chip in by now which they did not so I assume maybe it wasn't time yet/not hot enough or probably the fans or not working. After a while I heard a poof sound and the inverter was still on but no output. I turn it off same time. I waited a while and turn it on again and it was working normally but this time the output voltage now drop from 122vac to 110vac.

What could have gone bad in terms of when I heard the poof sound? what could have cause this to happen is it because it got too hot and the fans did not chip in to cool it down. What would cause the voltage to drop so low.

I have not open it as yet to check I am just waiting to hear someone's input on the matter first so I can now what to do when i open it to repair it. It is still working now but just with a much lower voltage output and I really want to get it back to the 120- 122vac reading please.
So please help me out in this regards.
I anxiously await your responses.

Mary B

  • Administrator
  • SuperHero Member
  • *****
  • Posts: 3169
Re: Inverter showing wrong voltage output
« Reply #1 on: May 27, 2018, 06:34:58 PM »
Brand and model number? Without that we would be totally guessing... if it is a cheap chinese inverter you probably fried a fet or a trace went poof on the board...

fabieville

  • Full Member
  • ***
  • Posts: 135
Re: Inverter showing wrong voltage output
« Reply #2 on: May 28, 2018, 12:03:17 AM »
I open it today and discovered that one of the fuse was burn and one of the fets got damage.  The inverter has four small transformers and each transformer has 2 fets assign to each of them because I notice that once one of the fets got damage the fuse for that section blow and therefore  that section got cut off from the inverter so you are still getting power from the inverter but with decrease output voltage pulling it down from 120vac to 114vac.

The inverter uses (8) big fets. IRFP4368.

When the fet got damage i was pulling about 2000watt and I realize that the case was getting quite hot and the fans did not chip in and suddenly i heard a poof sound. So after i open the inverter i replace the blown fet with a slightly smaller one IRF3705 as i did not have any of the big one. I replace the blown fuse. After testing the inverter i realize that I was getting back the 120vac for the output. I was not sure if the temp sensor on it was working so i connect the fans straight so as the inverter turn on the fans came on also.

So i went ahead and did back the estimated 2000watt test and after just a short time i heard the same poof sound again and when i open it the other irfp4368 on the same section was blown and the fuse blown also but the small irf3705 that i just put in was still good. I am kinda confused because the fans were running constantly and the case was not hot so how comes the other big fet blown and I was not anywhere near the max 3000watt rating for the inverter. And how comes the big one blown and the smaller fet did not blow too?

I am wondering if is just a bad batch of irfp4368 came pre-install on the board? Would it be best to replace all 8 fets with 8 irf3705 instead of using the big fets that came in it? Using (8) irf3705 would still give me back the 3000watt rating or which other big n channel fet would you suggest i use to replace the irfp4368(s) that i have now?

I am thinking about connecting back the fans to the sensor because maybe the temp needs to reach very high before they chip in because I was thinking that it was heat cause the big fet to blow but after running the fans constantly and drawing back the same load without the case getting hot and the big fet still blow i get to realize maybe it wasn't heat cause it to blow.

So any ideas guys as to the questions that i asked?

SparWeb

  • Global Moderator
  • Super Hero Member Plus
  • *****
  • Posts: 5452
  • Country: ca
    • Wind Turbine Project Field Notes
Re: Inverter showing wrong voltage output
« Reply #3 on: May 31, 2018, 08:21:03 PM »
Hi Fabieville,

I take it you meant "8" big FET's.  But we know you are the cool one, even if your inverter overheats.  8)

I suspect that the FET's are arranged in pairs on the 4 legs of a H-bridge, making the total of 8.  When the first one failed it left the other one overloaded for a time.  It may have been overloaded for a long time, so that there is not much you can learn from the later failure of the second FET while your replacement is still working.  It's likely that the FET you replaced is now overloaded and going to take its turn to go poof soon.

In order of priorities, I figure you should get all of the FET's replaced, all from the same batch if possible, to even out relative differences between each one, since they have to work in parallel on your inverter.  The next thing to do is investigate the gate signals controlling the FET's and the temp sensors controlling the fans.  If there's just 1 sensor somewhere on the heat-sink, then one failed FET might not cause a detectable temperature rise, even if one FET is hotter than the rest.

Hopefully we can get Oztules' attention because he's he world expert on keeping these things running.
No one believes the theory except the one who developed it. Everyone believes the experiment except the one who ran it.
System spec: 135w BP multicrystalline panels, Xantrex C40, DIY 10ft (3m) diameter wind turbine, Tri-Star TS60, 800AH x 24V AGM Battery, Xantrex SW4024
www.sparweb.ca

joestue

  • Hero Member
  • *****
  • Posts: 1759
  • Country: 00
Re: Inverter showing wrong voltage output
« Reply #4 on: May 31, 2018, 09:07:24 PM »
I don't think you need to replace all of the fets but i think you need to replace or add more capacitors. if they have dried out then their esr has increased and the over voltage spikes at turn off may be exceeding the avalanche voltage of the fets, which leads to increased heat losses, and or failure.

but you need a good oscope to look at the actual drain to source voltage at turn off to determine if this is a problem. alternatively you could add an appropriate capacitor as close as you can to the dc bus at each H bridge.

the fets you installed are 55v 8mOhm, stock units are 75v 1.5-1.8mOhm. however, the 3705 has far less gate capacitance than the other units so you might be better off replacing each 4368 fet with 3x3705 fets.

given that the same h bridge failed again, you could have a failing gate driver that is not driving the gates hard enough, the 3705 is able to turn on at a lower voltage than the 4368...


btw, you don't have an H bridge, you have a push-pull topology. when one fet turns on, the other one has double the supply voltage delivered to it, through the transformer. 55v rated fets are probably not adequate for a 24v inverter.

there is also the possibility though i consider it unlikely, that you have a two switch forward converter. anyhow, need an oscope to trouble shoot why just that one section is failing.


if its a push pull topology, the +24v battery wire goes to the transformer (almost certainly through a fuse mounted on the board), two leads from the transformer go to both drains of the fets, the sources are tied to ground.

there is probably a resistor and a capacitor in series connected in series from one fet's drain to the other. check to see that the resistor or the capacitor hasn't burned up.
« Last Edit: May 31, 2018, 09:17:18 PM by joestue »
My wife says I'm not just a different colored rubik's cube, i am a rubik's knot in a cage.