testing the egs002 inverter board

[oztules]

I had to remove a number of turns on the 24v 5kw PJ I converted for another person over here. It had an intermittent fault I could not bother to find so just stuck a 002 in there. The head room is limited in them compared to the pj controller.

It has been running for a number of months now without problems as a main inverter on a modest cottage.

Wondering how your going to mount the heat sinks now you have covered the access holes with the buss.


Looking good.

.......oztules

[Antman]

Yep I think I'll have to remove a couple of turns or just rewind the primary with some 25mm flex that I have.

The heatsink will be fixed by screwing nylon insulating screws through the bus bar and the holes in the PCB into the heatsink. I use sticky insulating but thermally conductive tape too and once it sticks it'll never come off.

Thanks for the compliment

[Dr_Zogg]

Howdey all,
Oz would there be any chance of getting a pdf of your fan controller and gti cutout boards. like the egs board in this thread?
i am ready to get going again now, i have a bit of spare time so i will update my thread as i go

Cheers,
            Zogg

[oztules]

Zogg, it is on here somewhere... but I can't find it easily.
I will find the original on one of my computers and post it/them.

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

[Dr_Zogg]

thanks mate you are a legend. i have been looking around on here for two days trying to find them lol.
all i could find were screenshots from your computer, the pdf's will make it much easier thank you

Cheers,
            Zogg

[frwainscott]

Hey Dude...  This should be for the temp board

http://www.anotherpower.com/board/index.php?action=dlattach;topic=1044.0;attach=4322;image

  Frank

[Dr_Zogg]

thanks for that mate but i found that one. i'm chasing the pdf's so that i can use them to develop some kinsten board
thanks for looking though,

Cheers,
            Zogg

[peter]

Further info on the inverter board I bought
(see post http://www.anotherpower.com/board/index.php/topic,1116.msg12138.html#msg12138)
The EGS002 board is configured to NOT soft start (JP6 is shorted) and the dead time is set to 500ns with JP3 and JP8 shorted as supplied
by the inverter board manufacturer.

I already had two EGS002 boards purchased from ebay 6 or 12 months ago. These 2 boards are identical in function.
I immediately swapped one of the ebay boards for that which came with the inverter board to see if I have 2 spares or not and it worked perfectly.
The ebay boards came with soft start enabled and some other dead time setting than 500ns.

The "standby switch" as shown on the aliexpress product page is in fact the inverter "on/off" switch. This is handy to have.
By the way, no instructions or documentation came with the inverter board. So I saved the aliexpress page to pdf.

I captured a couple of screens from the DSO when first switching on the inverter output with the soft start disabled.
(again using the 50VA desklamp load on the secondary)
These may help us understand why a soft start is to be preferred when driving a large toriod.
Look at the high voltage change rate - it must place large strains on the FETs when you attach a large load on the secondary and do a hard start.

Blue is secondary voltage (approx 240V RMS, using a high voltage differential probe)
Yellow is DC supply current, AC coupled
Purple is a gate drive for the PWM low leg (it needs to be referenced to the DC supply's 0 volts - only got one diff. probe alas)
Light Blue is primary current via a current transformer.

first shows a short period where I enabled the inverter output. Already the DC bus caps on the board were fully charged.
I think the DC input draws a large ripple content. Maybe in the finished project I might find some big-ass low ESR bulk caps to be placed just before the board's DC supply terminals. I recall reading how lead acid batteries do not like loads with large ripple content.

next is a close up of the PWM when switching off. It seems to me that there is some soft stop effect going on.

I now show the hard start up. I think this may be where we can easily blow these nice big FETs to tiny pieces if we are not careful.

Last is the entire event showing hard start, run under constant load and stop.

I intend to do a higher power test this weekend. I think I will change it to use a soft start from now on.
I hope this is interesting for you here. If not, tell me and I will just stick to the build of the inverter system.

[oztules]

Thanks Pete.
I think I will stay with the soft start, as I know big torroids don't hard start. My big ones blow any fuse thrown at them.

Remember to make sure that the current shut down works without glitches, or a torroid will blow the fets away. I suspect they are using the bypass rather than the on board current shut down?

Keep posting... everything tells a story that may tie into something later... we never know what scrap of info completes a picture......eg today I rebuilt Clockmans  board again, as I used ceramic caps for the 10uf in the charge pump this time ( had plenty laying about).... but it distorts the wave form... needs to be electro.... who knew?.... so it is out there now if any one else is silly enough to use ceramic 10uf caps for that function.... took a long time to wake up that that was doing it too

ok pdf for fan... I think this is the latest.... been a long time.... it's raining and I 'm not going out to the shed to check component values tonight... but that pattern should work.

 TEMPFAN2_2.pdf (7.85 kB - downloaded 555 times.)

Ok ran the clockman units, worked perfectly once I replaced my stuff up caps...... busy, but ran at 2500w for 20 mins or so without fans or anythine else.... heat sinks stayed cold....  interesting. Voltage stayed at 237v unloaded, and 237v under 2500 watts.

Pulled on another 2000w, and the voltage dropped to about 127v. Have not had time to check why.... assume the new tranny has to many turns on the primary... will follow it up tomorrow. Cannot be anything else but run out of head room.... back to 2500w, and perfect 237v....
EDIT see later post.... was a faulty meter!
Edit 2 no it wasn't. It needed a small 1w 240v:12v transformer for sensing the AC output, otherwise it found a earth return through the power system in the house to effectively make the voltage half what it should be... see later on.

Amazed at the 25w idle.... but a single stack tranny... which I don't normally do probably explains it.

Interested in changing the tranny tomorrow, fully expect 5kw easily, 3kw cont with this tranny..... ran cooler than expected... don't know why...maybe the 17C shed.

Here are the last known things that shut off the grid tie... I don't use them now, as the gti looks after itself in the end.

 64VSTOPA.pdf (15.5 kB - downloaded 349 times.)





And looking at the gti thing.... have no idea why  I made it like that... way over complicated for a simple voltage switch.... not sure what I was thinking of turning it into....

There are some track changes but essential the same. The component values are pretty flexible, as you can see......

There will be better simpler ways to do it.


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

[Dr_Zogg]

thanks oz your the man,
quick question. would there be any harm in me doing all 3 of these boards on the one sheet and mount it on the heat sinks as per standard
or would there be a problem with high freq switching in the circuits interfering with each other?

just thought it may be a way of cutting back on a little wiring, and keeping things a bit neater.

Cheers,
           Zogg

[oztules]

I can see no problem with that.... maybe go Antmans way and use a nano for everything?

....oztules

[Dr_Zogg]

looks good but i'm more of a hardware guy than a software guy.

Cheers,
           Zogg

[oztules]

Ok cleared up the fault with the voltage running amok... turns out it was the digital watt/voltage meter....
EDIT .. not I didn't see later posts
Today tried to chase down the fault of the changing voltage.... I could not replicate it no matter how much load was applied... then the voltage meter up and died completely... so it looks like that was it.....

Some pics to conclude.
Firstly.... no load and voltage


then 1500w load.....


Then near 4kw of loads



Driving this for while, heat sinks still cold, no cooling after 20 mins.... still less than skin temp.... damn impressive.... even with the ambient temp at only 16c... still very very good.



So all seems to be fine.... spent a long time trying to fault it, but it would not replicate the single  fault I saw last night.


All seems well.


[ Specified attachment is not available ]

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

[Dr_Zogg]

another quick one.
seeing as i am putting 3 board on one,
is there anywhere on the control board that i could pull a 12v source for the fans and controller from? 
could save me having the converter in there

Cheers,
            Zogg

[Dr_Zogg]

just thought i would chuck up a copy of the 3 boards combined. probably no much use to anyone but you never know.
i turned out pretty much the same size as the standard pj control card.

Cheers,
            Zogg