Author Topic: Pure sinewave inverter motherboard control board/fet issue? (Read 5162 times)

fabieville · « **on:** November 01, 2020, 01:07:12 PM »

I have a pure sinewave inverter motherboard that i bought off aliexpress. It is rated for 24vdc input at 110vac 60hz 2000watt output.
I had it connected up to my transformer and operating for couple months now without any issue and out of the blue some of the fets just burst on the board and it is kind of strange as to how it happen as there was not a surge or short circuit in any form whatsoever.

Here is the link for it:
https://www.aliexpress.com/item/32976783366.html?spm=a2g0o.productlist.0.0.f2922695xNCc6s&algo_pvid=e4cb69b4-045a-4378-b196-5416785ed200&algo_expid=e4cb69b4-045a-4378-b196-5416785ed200-17&btsid=0bb0623916042512669291852e9c99&ws_ab_test=searchweb0_0,searchweb201602_,searchweb201603_

When i check the board out of the (

fets. (6) of them got damage so i went ahead and change the 6 of them then hook it up back and when i switch it on i get the standby red led lite, the driver board red lite, the blue motherboard lite and the relay behaving eratic, multiple flashes and clicks from the lites and relay before the standby lite on the inverter turn from red to green and then the blue lite on the motherboard comes on to shows operating ok but yet still i am not getting any output. Sometimes the red lite that is on the control driver board comes on and flashes 3 times and off for 3 secs and then the process repeats and sometimes it does not comes on, playing with the on and off switch for the inverter causes the control driver board red lite to stay on constantly sometimes but still no output. Sometimes even the thermal fan comes on constantly as you turn on the inverter which should not have happen as the heat has not reach the set point as yet.

The control driver board is EGS002 + IR2110 as you can see in the link below.

https://www.aliexpress.com/item/4001015245464.html?spm=a2g0o.productlist.0.0.f2922695xNCc6s&algo_pvid=e4cb69b4-045a-4378-b196-5416785ed200&algo_expid=e4cb69b4-045a-4378-b196-5416785ed200-47&btsid=0bb0623916042512669291852e9c99&ws_ab_test=searchweb0_0,searchweb201602_,searchweb201603_

I am wondering if the control driver board is the issue or do i need to also replace the other 2 fets too so in essence all fets should be replace once any amount got damage? Or what else could be the problem ?

clockmanFRA · « **Reply #1 on:** November 01, 2020, 04:58:19 PM »

Hi Fabieville,

I have replied to you enquiry on the other forum.

So here is my reply. ....

I do not know these particular boards.

However our OzInverter is using the 8010 PWM chip, and 3 flashes and a gap of 2 seconds means that the control board 8010 chip is seeing overvoltage either from a faulty transformer or with its own PSU + or - 5v for the control board.

We found that the EGS002 board to be a poor design, especially around the op amp arrangements that use the 8010 pin 6 start and stop method.

So we redesigned, and we just used the 8010 PWM chip only, and built a more substantial and better quality control board for our OzInverter.

We also found that the EGS002 board would really flake out above 2.5kw with a toroid transformer.

The 8010 chip has a good soft start process that suits our more efficient toroid transformer that we use in the OzInverter, but even so some 8010's chips we have examined have certain operating details missing in their original programming..

I also understand that Quality control with individual components is also a big issue with these Chinese boards , especially around the Caps, IR2110 Driver chips and poor quality FET's.

Those EGS002 boards are very cheap to buy at only a few dollars, so try a swap out.

I trust this helps.

fabieville · « **Reply #2 on:** November 02, 2020, 08:23:51 PM »

The controller board that I have is the red one and it has a jumper wire running from it to the main board. I realize that I am.not seeing anymore of the red boards on the.net selling. Is just the blue boards I am.seeing and seeing them connected to.the mainboard there is no jumper wire connected to them. So.I am wondering if I replace my red board with the blue board would u suggest that I place the jumper wire on it also? I am.thinking probably the blue board is a upgrade from the red board so therefore the jumper wire is not needed.

What do u suggest I do?

clockmanFRA · « **Reply #3 on:** November 03, 2020, 12:57:41 PM »

Sorry, i can not help you any more with your boards.

As I have already said, I do not know these boards that you are using.

noneyabussiness · « **Reply #4 on:** November 03, 2020, 04:22:14 PM »

http: // www. thebackshed.com/forum/ViewTopic_mobile.php?FID=4&TID=11169
( take unnecessarily spaces out)

Have a read of the above.. those boards can be bulletproof with a couple of minor mods... i have had the 28 mosfet version running my house for years now without fault... follow Poidas advice....

noneyabussiness · « **Reply #5 on:** November 03, 2020, 04:55:16 PM »

Picture of the first mod ( and most likely why it died in first place) the egs002 have a " fast " lm358 overcurrent sensor which is murder for mosfets if you are using a toroidal transformer or any " tight " transformer, it will introduce spurious pulse trains when it triggers . The frequency MUST remain set.

https:
//www.
anotherpower.com/board/index.php?topic=1116.0
( take unnecessarily spaces out)

Another good read on this board, explaining the above.

noneyabussiness · « **Reply #6 on:** November 03, 2020, 05:01:38 PM »

Attached picture

fabieville · « **Reply #7 on:** November 04, 2020, 12:03:34 AM »

So.basically all I need to.do.is follow this instruction that the guy wrote on the next forum to bullet proof my board?

"Here is the first part of the modifications I would suggest you do".

Background info:
There are 2 chips that drive the MOSFET's gates. These are IR2110 devices
and their outputs are controlled by a shutdown pin.
Shutdown IS in effect when this pin is pulled up to 5V
Shutdown is NOT in effect when it is pulled down to ground or some low voltage.

there are 3 tasks.

1 - remove the LM393 opamp. With fine enough wire cutters you can just cut the leads
but I used a hot air solder tool and it lifts off easy. Use a bit of solder flux to help things end up looking good.

2 - remove R4. This resistor pulls the IR2110 shutdown pins to 5V. The opamp drives this low, to enable the output while the current sense input signals an acceptably low current.
I want nothing to do with this and so remove R4.

3 - remove C19 if fitted, then bridge the terminals of C19. This now connects
the IR2210 shutdown pins to ground.

My main board as some fast zener diode 1n4148 already attach to.the gate of the mosfets so therefore part 2 of.the modification in which the guy did this that I am.quoting now below I would.not.have to.apply this modification.because I already have some fast zener diode come.preinstall already. Am I correct??

" part 2 of modification

Obtain some 15V TVS devices. I think of these as good, strong, fast acting
zener diodes. They permit up to 18V to pass in one way, and very quickly clamping any voltage over this. And when voltage is flowing the other way, they are fast diodes, conducting the current, but with a small voltage drop of about 0.4V at the working currents.

RS part no. 815-2578
Vishay ICTE15-E3/54, Uni-Directional TVS Diode, 1500W, 2-Pin 1.5KE
AUD $0.92 each in a pack of 10

We want to protect the gate drive outputs of the IR2110 chips.
The manufacturer states all outputs must never go below about -0.3V compared with
ground or the high side "ground" which is the high side MOSFET's source pin.
Additionally, the specs say to never exceed 20V.
These 15V TVS diodes are about right for the job. Read the specs.

We only need 4, and we connect them near the MOSFETs, since that is where the potentially damaging voltages are generated. We connect each gate drive output to the corresponding MOSFET source pin.
In the case of my board shown below, I have 3 MOSFETS in parallel for each of the 4 legs of the full bridge. So I only need to do each leg of the full bridge, not all 12 MOSFETS"

clockmanFRA · « **Reply #8 on:** November 04, 2020, 02:58:25 AM »

Nice one 'noneyabussiness'.

Good old 'Poida'.

Give it a go 'fabieville'. 'noneyabussiness' knows what he is talking about and achieved good results with your type of boards, but remember he has the 24 FET version.

The Power Board Mosfets.......

“In theory, a single 4110 can carry 150amps@100v... so a single one can actually do 15kw on it's own.

Practice is different, those skinny legs can't handle 100 amps continuous, apparently the silicon can, as it has a resistance of only .003 ohms.... so at 100 amps it will drop E=IR or 100x .003 or .3v losses@100amps.. so the dissipation will be W=ExI or 100 x .3... or only 30 watts.. so that leaves over 340 watts of dissipation spare.

Can you see that the dissipation figures are not the power handling figures, the 370w is the figure it can safely dissipate in losses..... not throughput. It's losses is the Rds on in ohms (.003-.006 for 4110) x the current flowing through it... these losses are dissipated in the mosfet and heat sink.

Switching losses are added to the IR losses, ( and in anything I build usually outweigh the resistive losses hands down). Getting the switching losses down is a science unto itself... black science almost, add in parasitic oscillations and lots of other gremlins, and we have our work cut out to keep the losses less than the 370 watts. If we fail, it goes up in smoke.

For the most part, if we keep the switching frequency low, stick to the design rules regarding layout etc etc etc... we do pretty darn well.

With 6 fets per leg, we get roughly 150X6=900 amps of capacity X the voltage we switch... 48X900=43000w or more.. my bank is usually 56v so 56X900=50kw. ( so it should easily survive transients of this level)

Now these figures are all theoretical, and have no real place in the real world... except for fractions of a second, and there they count.

We switch at high speed........ over 20 thousand times per second... so we don't need to carry these currents for long, and our duty cycle is less than 100% (each leg is turned on only at a max of 40% of the time I suspect).

So in fractional times, 60kw is not out of the question... and surviving it is possible too. Remember of your 24 fets, you can only calculate for 6, as each leg only has 6fets..... not 24.

I use the 15kw boards because the fet spacing is greater, and so cooling is better..... mostly not of interest, but more surface is always better than less.

...........oztules

noneyabussiness · « **Reply #9 on:** November 04, 2020, 03:41:47 AM »

Thank you clockman, as always, hows your system going ?? Its amazing what can be done with these setups....

Have a look at the 2 pics i posted.. " U4 " to be pricise... that will achieve the same thing as " poida " recommended... i have on 2 occasions had a direct short on output and both times tripped breaker but no damaged mosfets... the hy4008's are awsome...

fabieville · « **Reply #10 on:** November 04, 2020, 07:28:05 AM »

What i want to.know is.if it's the same modification should applied to my control board seeing that my main board is the 8 mosfets version instead of 24 and do I need to do part 2 of the modification in which I Include the diodes on the gate of the fets even thou mines already have 1n4148 on each mosfet gate?

noneyabussiness · « **Reply #11 on:** November 04, 2020, 03:35:41 PM »

Irrelevant how many mosfets you have, its a frequency thing as oztules explained in his post. Having those spurious pulse trains will kill your mosfets, whether you have 5 or 50...

The anti-parallel diodes that are already in place are different from the ones poida has suggested, they are for " fast closing " the mosfet gates. where as the tvs diodes are for protecting the gates from any overvoltage issues ( very sensitive).

fabieville · « **Reply #12 on:** November 04, 2020, 09:42:56 PM »

So therefore would it be best to leave the default ones install or replace with the others that he suggested?

Seeing that my inverter is not working now because most likely the control board is at fault. Chances are that doing the suggested mods on it now then putting it in back it most likely will fix.my.no output issue?

noneyabussiness · « **Reply #13 on:** November 04, 2020, 11:49:53 PM »

Yes leave the anti-parallel diodes in place AND add the TVS diodes .. they both do completely different jobs.

If there is something damaged, the only way to fix it would be to test each component to find the fault... these mods are done to a already working unit to strengthen it to handle almost all situations..

fabieville · « **Reply #14 on:** November 05, 2020, 11:50:53 AM »

So given in mind that my main board has 8 fets. I can add 8 of the TVS Diode on each of the fet gates?

Also my control board maybe faulty now with the issue that I mention in the first post. If I do the mod to.it instead of waiting for.the new control.board. Chances are it might fix the no.output issue I am getting from the inverter?

noneyabussiness · « **Reply #15 on:** November 05, 2020, 02:09:10 PM »

Re-read poida's post, you only need 1 tvs diode per BANK of mosfets... unless damaged, leave theanti-parallel diodes alone...

If there is something damaged, the only way to fix it would be to test each component to find the fault... these mods are done to a already working unit to strengthen it to handle almost all situations..

fabieville · « **Reply #16 on:** November 05, 2020, 05:24:17 PM »

my original board is red colour one so with the new replacement one which is the blue board chances are that it is an upgrade to the red one hence it doesn't require any mods and the resistor that runs from the mainboard to the original red board wouldn't be valid for the blue board so I don't have to solder it to the blue board with or without any modification being done to the board.

noneyabussiness · « **Reply #17 on:** November 05, 2020, 11:33:05 PM »

Your welcome to not listen to the advice given, however if it has the egs002 daughter board it will have the " fast " overcurrent sensor which WILL destroy your mosfets eventually.. Irrelevant of board colour.. its a simple physics thing, that oztules described very well in above posts.

This is a very tried and tested combination. But its your money, do as you wish...

fabieville · « **Reply #18 on:** November 09, 2020, 08:41:24 AM »

Ok thanks for the advice.

I recently got the new control board and install it and when i turn on the inverter i am getting the same relay clicking sound and multiply flashes from the led lights on both the main and control board and then afterwards all of that comes to a stop and then i am getting a 4 flash sequence from the led lite on the control board with a 2 sec pause between the 4 flashes and also there is a momentary humming from the transformer which i assume is because some form of dead shorts is happening on the board but when i check the mosfets legs there are no continuity between each legs and my input and output fuse have not been blown.
When you just turn on the inverter there is slow rise of the ac voltage to around 38v because the board was set to soft start by default. and then then afterwards the tester starts to beep constantly showing continuity between the output L and N.
I am now wondering if i should replace all 8 fets because when the fets got damage, only 6 burst so i replaced 6 instead of 8.
Or is it the case that the transformer got burned hence that it why i am getting the humming sound from it every now and then, even thou i don't see or smell any form of burn evidence.
What do you suggest i do?

noneyabussiness · « **Reply #19 on:** November 09, 2020, 03:46:43 PM »

Ok, the first check would be the mosfets on the 23khz BANK ( both high and low) ive found both times during testing that it actually damaged mosfets it was one or all in that bank which makes sense as it switches a heck of a lot more times per second than the 50hz bank, remove them and check with meter. Then check the drive npn/pnp transistors, lastly check each 10ohm/1n4148 combo on each mosfet .. all have to work for efficiency...

AU $22.17 5%OFF | LCR-TC1 3.5inch Colorful Display Multifunctional TFT Backlight Transistor Tester
https: // a.aliexpress.com/_mrJkBCd
( again , remove unnecessary spaces)
These meters make testing much simpler...

fabieville · « **Reply #20 on:** November 14, 2020, 08:17:41 AM »

Part 3 of the modification states:

"The choke:
Find a suitable and large ferrite E core.
I used one of these
https://au.rs-online.com/web/p/products/8407648/
stock no. 840-7648
It has a 0.5mm gap. Wiseguy correctly pointed out in another post of yours
my incorrect calculations.

Other people here have made their own, in different ways, using different cores.

We need something like 45uH, and the inductor (or choke) needs to still
work with currents well over 100A if you are planning to put 5000W through a 48V inverter"

In regards to the above statement. I have already bought a inductor that lowers the idle current that the website where i bought the mainboard recommended to use with the board and i connected it to the mainboard and everything was working quite fine until the mosfets got damage. Do I still need to get the choke as he suggested in part 3 of the modification or i can just use the existing inductor that i have and skip out part 3 of the modification?

noneyabussiness · « **Reply #21 on:** November 14, 2020, 04:29:05 PM »

If it is recommended by the manufacturer, it will be fine.... for the currents you are using it at.... the above was tested as it states at 5 + kw at 48v so things become more critical at those power levels...

fabieville · « **Reply #22 on:** November 25, 2020, 10:40:13 AM »

for part 2 of the modification can someone show me where to connect the 15v tvs diodes to the bottom of my circuit board. I am kinda confuse. i bought 10 of them online as i dont quite sure where each of them should go and how much to use. my board uses 8 fets. i uploaded a photo below with the bottom of the board. i would love to get a diagram showing exactly where each diode should go on my board.
please explain in details how to do it.

noneyabussiness · « **Reply #23 on:** November 25, 2020, 04:54:00 PM »

Are the tvs diodes you purchased bi-directional or uni-directional ?? ( are they all black or do they have a white line on one side?) If they have a white line, it must point to the mosfet gate , if they all black, it doesn't matter... ive attached a picture of a mosfet, that shows gate, drain , source. The tvs diodes must go between the gate and source. On your picture, you will need to include the top of board as well , I can guess for you but lets just keep it right. You only need 1 per bank.

Pure sinewave inverter motherboard control board/fet issue?

fabieville · « **Reply #24 on:** November 26, 2020, 05:11:35 AM »

So basically i need to install a total of 4 in all?

On the bottom of the board i notice that on each side of the board 2 mosfets source pin or connected and then on the same side of the board there are another 2 mosfets that shows that the 2 source pins are connected too.

So when installing the tvs diodes should i place the white line to connect to one set of the mosfets which means the white line would connected to both gates of the set of mosfets or just one gate of any one of them and then the black part connected to the source pin and i repeat this for the other section on the same side of the board which the other 2 mosfets source pins are connected as one?

If i am not sure how to do it, would placing one diode for each mosfet have any negative effect or cause any issues on the board?

noneyabussiness · « **Reply #25 on:** November 26, 2020, 04:40:32 PM »

Like this, line is white line on tvs diode... repeat on other side... yes 4 in total... on 1 gate will be fine as they are both connected on each bank through gate resistor and diode..

fabieville · « **Reply #26 on:** November 28, 2020, 07:01:52 AM »

I did all the steps and it's working perfectly.
Thanks for the support.

noneyabussiness · « **Reply #27 on:** November 28, 2020, 01:12:39 PM »

Good to hear , best wishes

clockmanFRA · « **Reply #28 on:** November 28, 2020, 01:37:45 PM »

Nice one 'noneyabussiness'.

fabieville · « **Reply #29 on:** December 27, 2020, 06:23:36 PM »

The board is working ok but today i just discovered an issue.

My transformer is by default a 220vac split phase one with 110vac on each hot leg but i abandon one of the hot legs and was just using one side which is just 110vac i was using up til today i decide to use the 220vac which is the 2 hot legs. When i went ahead an test the 2 hot legs i am only now getting 110vac and when i test the neutral and one of the hot i am getting basically 60vac on each leg.
So in essence the transfomer as now decrease to 110vac split phase instead of 220vac split phase.

It is strange because i have a little meter that comes with the control board and it is connected to the 7 pin slot on the control board and it is showing 220vac at 60hz. Is it because i change the frequency from 50hz to 60hz make i am only getting 110vac from the 2 hot legs?

What could have cause this to happen? is it because of the modification i did on the board and on the control board or is it because of when i was buying the board i chose the option that ask about output voltage and i chose 110vac for the output even thou my transformer is rated for 220vac output? So probably the board came in output 110vac by default even thou the transformer is 220vac split phase it just ignore that and still produce just 110vac on each hot base upon some configuration on the board.

When i test the primary stage/transformer input wires that goes to the board i am getting 8.6vac going to the board, shouldn't it be about 12-14vac going to the board? when i test the battery - with anyone of the transformer input wires i am getting 12.6vdc which is half of my battery voltage. I am confused guys.

how can i remedy this issue?

the board is working quite fine still because i am getting a well regulated 120vac on the output even under high 1KW+ wattage.
is just that i would need back the 220vac capability that the transformer originally had.

mab · « **Reply #30 on:** December 29, 2020, 01:42:52 PM »

if i understand what you've done: you had it wired with one 110v winding of the transformer connected to the feedback transformer of the inverter board, and with the other 110v winding unused (disconnected) and it works as intended as a 110v inverter?

Now you have connected the two 110v windings together (0-110_0-110) and connected the two ends to the voltage feedback of the inverter board (N_NFB_0-110_0-110_L&FB)? And you're getting 110v out. Correct?

Do you want 220v L-N only or do you want 110-0-110 L-N-L split phase?

for 220v L-N: connect feedback N to 0v end of the two transformer windings and the feedback L to the midpoint of the two transformer output windings (N_FBN_0-110_FBL_0-110_L). If you're only using 220v out then it should be OK although you would have been better getting a 220v inverter board.

for 110-0-110 L-N-L: connect feedback N to the midpoint and feedback live to the 110v end
(L2_0-110_FBN_N_0-110_L1_FBL)
The inverter board can only regulate one of the Lives (L1) so when using split phase try to keep the larger loads reasonably balanced between the two lives.

News:

Author Topic: Pure sinewave inverter motherboard control board/fet issue? (Read 5162 times)