5000W Inverter Help Needed

Thread Starter

joczr2

Joined Feb 12, 2013
27
Hi everyone I am new to this site seeking help with an inverter I bought online from a Chinese company. After I received the inverter I hooked up a 12V battery and something flashed inside and the thing went dead. I emailed the company and they said send it back, well it would have costs 500 in shipping so I just told them to send me replacement parts. I do understand a little about circuits and electricity but I really am stumped here. I am attaching 3 pictures of the inverter. After opening it, it seems that only one of the mosfets...just learning what that is....This mosfet is RU6099R and under that is 492 AT121. The company sent me mosfet RU6099R, under that is 778 AT128. IS this a problem? What do these letter and numbers mean? There are also very very small resistors next to each mosfet. Using a magnifying glass I was able to see a "100" on the top. Does this mean they are 100μF? If not, what does it mean cause other small ones like it have different numbers. Also when I tested several of the resistors they seemed to give different Ω values so that is confusing. I would take it to a professional to get fixed but one, I haven't been able to locate a repair shop or a school that teaches electronics plus I think it would be fun and a good learning experience for me as I am trying to use them in solar farming and I'm sure I'll have to work on them in the future. My plan is to try and understand every component, check it for its proper unit of measurement with the multimeter and just walk the whole thing down and replace those that are bad and it should work. Sounds easy to me:) If anyone could help I would surely appreciate it.

John
 

Attachments

praondevou

Joined Jul 9, 2011
2,942
This mosfet is RU6099R and under that is 492 AT121. The company sent me mosfet RU6099R, under that is 778 AT128. IS this a problem? What do these letter and numbers mean?
No, no problem. These are probably lot and manufacturing date numbers. Doesn't matter.

There are also very very small resistors next to each mosfet. Using a magnifying glass I was able to see a "100" on the top. Does this mean they are 100μF? If not, what does it mean cause other small ones like it have different numbers. Also when I tested several of the resistors they seemed to give different Ω values so that is confusing.
Post a picture of the resistor. A SMD resistor "100" would be 10 Ohms.
Testing resistors "in-circuit" doesn't mean much, unless you know the circuit.

My plan is to try and understand every component, check it for its proper unit of measurement with the multimeter and just walk the whole thing down and replace those that are bad and it should work.

John
Good luck with that. It's not exactly a small PCB. Some testing has to be done with the proper equipment, like oscilloscope, a benchtop power supply with current limiting etc.

Look also for burnt traces, check if there are solder residues between MOSFET pins etc.
 

Thread Starter

joczr2

Joined Feb 12, 2013
27
I understand thanks so much for the help. I have attached a couple of photos of the resistor. How does one replace R8 for I am sure it is fried? There are capacitors that size as well so testing them would be fun. So would you have any suggestion on maybe why this thing failed on me? I ordered it like I said from China. Once I received it I hooked it up to a 12V car battery and poof.
 

Attachments

praondevou

Joined Jul 9, 2011
2,942
I understand thanks so much for the help. I have attached a couple of photos of the resistor. How does one replace R8 for I am sure it is fried? ......
Replace R8 with a 10Ohm resistor. That's the gate resistor.
The gate signal comes from a drive circuit. Follow the trace to see where it comes from. It may have damaged the driver too.

So would you have any suggestion on maybe why this thing failed on me? I ordered it like I said from China.
You answered your own question. :D
Seriously, I have no idea. Could be just bad soldering. Sometimes small soldering tin balls remain on the board, more or less fixed between component pins or glued to the board by flux residue. If the short two pins of the MOSFET then BOOM.
 

Thread Starter

joczr2

Joined Feb 12, 2013
27
So I can find me any 8 ohm resister and replace that little one. I've just never seen resistors like that. I will do it though. You make an excellent point about bad soldering. I took the board out of its aluminum housing so I could examine the soldering. I may be wrong but some things just looked wrong....like certain part were touching and the exact same lay out in another part of of the inverter would not touch. I have attached some pictures let me know if you cant read my writing on them. I was trying to show what I though was wrong.

Look forward to your thoughts.
 

Attachments

praondevou

Joined Jul 9, 2011
2,942
So I can find me any 8 ohm resister and replace that little one. I've just never seen resistors like that. I will do it though. You make an excellent point about bad soldering. I took the board out of its aluminum housing so I could examine the soldering. I may be wrong but some things just looked wrong....like certain part were touching and the exact same lay out in another part of of the inverter would not touch. I have attached some pictures let me know if you cant read my writing on them. I was trying to show what I though was wrong.
You mean 10 Ohm, not 8.

The area you circled is ok. It is one large trace.
 

JDT

Joined Feb 12, 2009
657
Oh dear. Another 5000W 12V inverter.

Just do some simple maths: 5000/12 = 416.666 That's 416A !! Assuming it's 100% efficient. Have you got copper bars or just wires connecting to your battery?

It's not going to work - give up now!
 

Thread Starter

joczr2

Joined Feb 12, 2013
27
Ha...I know and I am soon giving up. I never wanted all of the 5000, just enough to run a 1.5 hp water well pump.
 

DC_Kid

Joined Feb 25, 2008
1,072
i would ask for $$ back w/ return of the unit and vendor pays shipping.

5000w inverters should be 48Vdc to minimize the input currents.

the pictures shown look like it was made on poor processes, hence the flash.

a flash should yield distinctive problem area visually and by smell, however, if the board is made bad then replacing parts is an unfruitful activity, etc.
 

praondevou

Joined Jul 9, 2011
2,942
So the three lines of solder are ok to be merged like that?
It is one trace, there can be solder or not, it doesn't make a difference.

They probably thought this could increase current carrying capability of the trace.

We built a 24V input/ 120V output DCAC converter, 2500W output, almost 3000W input. Lowest acceptable voltage was 19V input. So that's 157A input current, although peak current is higher.

It had two copper bars 12mm x 3mm from input to the H-bridge. 5x TO247 MOSFETs on each leg. With a two fan forced ventilation and a ventilation duct it still got quite hot.

In short, if this is a hard switched converter then 5000W is maybe the absolut peak power for a few ms that it can deliver.
It looks like at most 1000W continuous power. Or maybe they meant 500W and someone made a mistake and added a zero..:)
 

Thread Starter

joczr2

Joined Feb 12, 2013
27
i would ask for $$ back w/ return of the unit and vendor pays shipping.

5000w inverters should be 48Vdc to minimize the input currents.

the pictures shown look like it was made on poor processes, hence the flash.

a flash should yield distinctive problem area visually and by smell, however, if the board is made bad then replacing parts is an unfruitful activity, etc.
Well unfortunately I can't get my money back. After they, btw I bought 2, blew up they said send them back but fed ex would charge $700 for shipping so I said forget that....that's more than I paid for them. I am not able to get water to irrigate a blueberry field that I put in a few years ago. We used a 10KW generator now, electric company wanted 10K then, now 12K to put a power pole up so I thought solar was the best way. I bought panels and several marine batteries. I paralleled probably 6 batteries in order to keep the 12VDC. I guess the amps is what blew it but I don't understand why since I didn't have a load on the inverter.
 

Thread Starter

joczr2

Joined Feb 12, 2013
27
i would ask for $$ back w/ return of the unit and vendor pays shipping.

5000w inverters should be 48Vdc to minimize the input currents.

the pictures shown look like it was made on poor processes, hence the flash.

a flash should yield distinctive problem area visually and by smell, however, if the board is made bad then replacing parts is an unfruitful activity, etc.
It is one trace, there can be solder or not, it doesn't make a difference.

They probably thought this could increase current carrying capability of the trace.

We built a 24V input/ 120V output DCAC converter, 2500W output, almost 3000W input. Lowest acceptable voltage was 19V input. So that's 157A input current, although peak current is higher.

It had two copper bars 12mm x 3mm from input to the H-bridge. 5x TO247 MOSFETs on each leg. With a two fan forced ventilation and a ventilation duct it still got quite hot.

In short, if this is a hard switched converter then 5000W is maybe the absolut peak power for a few ms that it can deliver.
It looks like at most 1000W continuous power. Or maybe they meant 500W and someone made a mistake and added a zero..:)

Well I certainly appreciate your guidance. I have a few other questions and I will leave you guys alone. Being that I have 2 inverters(you know the specs), is all hope lost? Is there a way to maybe tone it down from 220 VAC to 120 VAC that would be more friendly to the DC input? Is there a way to modify the 12VDC input to a 24VDC and possibly keep the 220VAC, btw I am only trying to power up 2-1.5HP well pumps, each pump rated at 230 VAC & 9.2 Amps. I realize I cannot use these inverters for this irrigation system even if they worked as I have learned I need pure sine. I just hate to see them go to waste when I could definitely use them in other places. If you guys think they can be fixed, one way or another please let me know. Sounds like you are in the inverter business??? Would you recommend a place I could send them to to repair them...like yall said 12V to 220V isn't what they should be doing...but maybe 110...If you would want to fix them I would gladly pay through paypal but I do enjoy working with electronics and am learning so much just need someone to guide me a little.
 

praondevou

Joined Jul 9, 2011
2,942
Yes, 9Amps is to much of a load.

Someone with a lot of time and experience could try to modify them. And someone with the right equipment like an oscilloscope.

You would need to draw the schematic, find the feedback loop that determines the output voltage and could possibly bring down the voltage to 110V without having to change magnetics or semiconductors. Increasing the input voltage would for sure be more work since the whole circuit is made for 12V or maybe up to 16V...

It's a shame to spend a few hundred dollars and then they just blow up. That's the price we pay if we buy in China. I had my experiences too (not that expensive though).
 

Thread Starter

joczr2

Joined Feb 12, 2013
27
Oh btw...A friend of mine gave me a variable freq drive, AF-300 G11. If that is something that could be used in this whole mess of a situation please let me know...thanks so much.
 

praondevou

Joined Jul 9, 2011
2,942
Oh btw...A friend of mine gave me a variable freq drive, AF-300 G11. If that is something that could be used in this whole mess of a situation please let me know...thanks so much.

They come with input voltages of 230V or 460V. So you can't use them with a 12V battery.
 

takao21203

Joined Apr 28, 2012
3,702
It is one trace, there can be solder or not, it doesn't make a difference.

They probably thought this could increase current carrying capability of the trace.

We built a 24V input/ 120V output DCAC converter, 2500W output, almost 3000W input. Lowest acceptable voltage was 19V input. So that's 157A input current, although peak current is higher.

It had two copper bars 12mm x 3mm from input to the H-bridge. 5x TO247 MOSFETs on each leg. With a two fan forced ventilation and a ventilation duct it still got quite hot.

In short, if this is a hard switched converter then 5000W is maybe the absolut peak power for a few ms that it can deliver.
It looks like at most 1000W continuous power. Or maybe they meant 500W and someone made a mistake and added a zero..:)
250W is a lot of power to conduct it, rectify it, and transform with an SMPS. I mean 250W permanently 24/7.

Everything heats up and need cooling. Even the (large) capacitors heat up. The large toroid coils heat up (2x 32mm each). The large power PNP has a VGA cooler (also a large rectangular one). Does not get hot, but also heats up.

System voltage = 50V, upto 100V, not tested.

For 5kW, well that's 20x the current. Not a good idea to use 12V. You'd need extraordinary screw terminals. Even 30A terminals are BIG. And so are the cables. 30A is really some kind of threshold current which you still can handle with soldering and regular cables.

After that, you need special tools, special terminals, and thick cables, as well good fusing.

The idea to bank many smaller inverters (photo) is not bad- the cooling system is- it is crap, not capable of 5kW.

Large VGA coolers are really good, the one I use could carry the heat for maybe 350W to 400W.

I could crank up my technology to 500W or 1kW if I wanted to- 5kW? I would not know how to do a bus properly for that power level, don't deal with such cables or terminals.

Maybe they used inferior MOSFETs- they really exist. Smaller chips, thinner bonding wire, and very brittle legs made from weird amalgam.

I mean I saw such MOSFETs.

Or they did not apply the heat conducting grease properly. Things like that. The board does not look totally bad- but 5kW from 12 volts- I do not think.
 

takao21203

Joined Apr 28, 2012
3,702
Think about this dc/dc converter. 150W class. That is largely incorrect.

In order to power 2x 20W LEDs at 60V, I need a much larger cooler for the MOSFET. And that is only 40W.

http://www.ebay.com/itm/DC-DC-Boost...Up-Adjustable-Power-Supply-150W-/110953873333

Correct the input current is the double if I double the voltage (by patching it). But also that only makes equivalent of 80W at 30V. Which the converter is not capable to deliver in unmodified form.

Passive cooling is not very effective, I have done some experiments and measurements, forced cooling increases capability by factor 5 to 10.
 

Kermit2

Joined Feb 5, 2010
4,162
See if you can sell the AC pump motor and then buy a DC motor for 24 or 36 or 48 volts.

The pump will run intermittently so figure out a good schedule to let the panels do the charging as close in time as possible to the end of a battery discharge.

For this scheme you will need a charge controller for the solar panels. I am not endorsing anyone, but I have an 'OUTBACK' brand charge controller and have found it to be very reliable. I do however keep it well protected from any outdoor weather with a steel enclosure. They come with different input and output ranges so you can choose the proper one for your solar voltage input and your load requirements.

This arrangement will avoid the conversion losses in the inverter since the batteries will power the motor directly.
 
Top