Author Topic: testing the egs002 inverter board  (Read 153686 times)

0 Members and 8 Guests are viewing this topic.

Offline Phred

  • Newbie
  • *
  • Posts: 18
  • Karma: +0/-1
Re: testing the egs002 inverter board
« Reply #180 on: April 16, 2016, 05:29:19 am »
This site is easy to work with   http://www.yoycart.com/Product/528093395667/

They are kits at these cheaper prices...  think the alibaba ones were built.

...........oztules
The one on Taobao is built too. The yoycart calculation comes in @ AUD 118$ assuming shipping is the same to Oz with exchange rate of 0.73 US to AUD $. The figures  might be a little higher depending what the bank decides to charge.
The Aliexpress page has some interesting apps and photo's
http://www.aliexpress.com/item/1pcs-1800AV-3000AV-5000AV-DC12V-24V-36V-48V-to-AC-220V-pure-sine-inverter-board-frequency/32637730780.html?spm=2114.10010108.1000023.15.0I5Nke

Offline RFburns

  • Full Member
  • ***
  • Posts: 124
  • Karma: +10/-0
Re: testing the egs002 inverter board
« Reply #181 on: April 16, 2016, 04:50:59 pm »
Hey Clockman looking good!  :) keep us posted on the progress. RF
Get With It ,Get Over It , Get On With It ...Or Leave

Cheap and reliable wont be fast.
Cheap and fast wont be reliable.
Reliable and fast wont be cheap.

Offline oztules

  • Moderator
  • Hero Member
  • *****
  • Posts: 1177
  • Karma: +105/-8
  • Village idiot
Re: testing the egs002 inverter board
« Reply #182 on: April 22, 2016, 07:13:26 pm »
Building and testing a new simple fet board... only a 4kw cont, but 12kw start  capability.

It  is just a simple single thickness $5 double sided 6x4 board.

It now needs full tinning and placement of copper buss bars on it to connect the heavy leads to and bring the power into and out of the fets.
The beauty is the very short fet legs, and the lack of decent heat sinks..... bolts to anything, using the board as the clamp

anyway here it is after it's first low power test, before the copper goes on, and the caps are mounted.

5543-0

and the other side... I know this is not pretty..... but it should work so that the inspire box is empty when this is bolted to the heat sink, and the 8010 board is used to drive it..... lots of room for the transformer.

5544-1


Yeah, looks crap but will solve a host of building problems easily.


...............oztules
Flinders Island...... Australia

Offline Antman

  • Full Member
  • ***
  • Posts: 51
  • Karma: +0/-0
Re: testing the egs002 inverter board
« Reply #183 on: April 23, 2016, 01:43:36 am »
Looks good Oz. Appearance isn't everything.

Let us know your results on the high power test. I'm waiting for my boards to come still. Hopefully they come on Tuesday and I can find the time to put them together :)

Offline oztules

  • Moderator
  • Hero Member
  • *****
  • Posts: 1177
  • Karma: +105/-8
  • Village idiot
Re: testing the egs002 inverter board
« Reply #184 on: April 23, 2016, 02:34:20 am »
Yes Antman, if I can get  decent buss bar soldered onto the 4 copper power planes, then it will be a winner..... gotta locate some copper somewhere on the island to shape and sweat onto the planes.

Keen to see how well it works. I expect 3-4kw continuous without a problem, and silly surge figures.... these things can start anything.

Keen to see yours running too... I like not using the 002 board, just the 8010


...........oztules
Flinders Island...... Australia

Offline Antman

  • Full Member
  • ***
  • Posts: 51
  • Karma: +0/-0
Re: testing the egs002 inverter board
« Reply #185 on: April 23, 2016, 08:30:18 pm »
Hey Oz

I'm sure you'll be able to get over 3 kw through that with adequate heat sinking.

With mine i think i'll just just use the 002 board plugged in via female headers so it can be changed out in no time at all. To buy all the parts and assemble would cost heaps of money and time plus i like the idea of saving space on the power board by going vertical.

Those tantalum capacitors are pricey in small quantities and are probably the only weak point since they go into a dead short upon failure. I'll just give the boards a wash and seal with a varnish to protect from the environment. The current feedback pin will be shorted to ground and the arduino will manage startup and shutdown through the temperature feedback pin. This way I wont have to remove the op amp and modify in any way. (That's the plan anyway...fingers crossed)

Offline peter

  • Newbie
  • *
  • Posts: 19
  • Karma: +6/-0
Re: testing the egs002 inverter board
« Reply #186 on: April 24, 2016, 06:16:06 am »
Hi all here
I'm wanting to join the game but I don't have the PCB design and fab. skills so I chose to get one of
those pre-made boards from aliexpress

http://www.aliexpress.com/item/Free-shipping-1pcs-1800W-5500W-DC12V-24V-36V-48V-to-AC-220V-pure-sine-inverter-board/32637730780.html?spm=2114.30010308.3.44.ut8jCV&ws_ab_test=searchweb201556_9,searchweb201602_5_10017_10005_10006_10034_10021_507_10022_508_10020_10018_10019,searchweb201603_6&btsid=d7e2931a-fdd2-404a-ab4a-c692c960a062

and
I got for $50 a dead Aero-Sharp grid tie inverter from ebay.
When the driver board arrives I'll get busy, hook it up to the rewound transformer and see how it all goes.
I have a self built hybrid solar power system 400 Ah at 48V battery, Victron 3000 charger/inverter and Morningstar MPPT controller fed by 3kW panels.
If I make an inverter that can start my air compressor for about $300 which the $2,700 Victron can not, it will be a bit of a laugh, eh?

First question:
when rewinding the tranny, must I use single conductor copper wire for the 240VAC secondary?
I can afford to buy some new 15A PVC insulated wire from Middys or whatever is suggegested as an alternative.
Frankly I can't be bothered to straighten out the old wire.
I'm going to Middys to get the big current primary cable anyway.

Peter

 

Offline oztules

  • Moderator
  • Hero Member
  • *****
  • Posts: 1177
  • Karma: +105/-8
  • Village idiot
Re: testing the egs002 inverter board
« Reply #187 on: April 24, 2016, 04:30:28 pm »
Which Aero sharp.... the 1.5kw are a bit small to wind easily... the 3kw would be big enough to work with nicely..... and in some cases you will only have to unwind the outer layers, and a 240v layer is already there for at least 3kw continuous, or 10 kw surges..

If you wind your transformer... reuse the old wire.... do not use  pvc wire for the secondary, only the 30mmsq wire for the primary can be plastic.

Yes it will blow the victron away, and will start anything you have in your house..... the victron is a toy only, not a real home inverter.

The transformer is important to get right for maximum power transfer..... read the article on transformer rewind, until you are sure you have it correct.

If you have a 1.5kw, try to get a second one, so you can join two transformers, take some of the core out from the center, and make a decent hole size to work with. I just did one like that for a 4kw unit... I made the center hole 100mm in size, and boy does that make winding a breeze... but it reduced a single core from .9v/turn to .7v/turn... then x2 = 1.4v/turn, but with the big hole.... who cares, it is so easy to wind.

Pics later today if I get time.

Antman, ran it for  while last night 2.5kw for 20 mins everything stayed below 25c.. it was 20c in the shed, no cooling. Used half a heatsink only.
Impressed at how well it ran, will give it a better test today.... don't like burning up 2-3kw at night for nothing... better when the sun is out and driving the bank with 5-6kw of input, can test for hours on end with the battery not seeing it.... we'll see how it handles it.... I did notice that the fets were cold, but the 25mmx3mm x 150mm long buss bars got luke warm.....came out of a curtis power controller rated at 300 amps.... very surprising, thought that bit would remain stone cold.... but no.


...............oztules
Flinders Island...... Australia

Offline oztules

  • Moderator
  • Hero Member
  • *****
  • Posts: 1177
  • Karma: +105/-8
  • Village idiot
Re: testing the egs002 inverter board
« Reply #188 on: April 24, 2016, 06:40:05 pm »
looks like this
5549-0

and this
5550-1

and this
5552-2

and the victim of the short test

5554-3

hard testing comes next... 3kw for the next few hrs driving house and loads and hot water for a few hrs

.....oztules
Flinders Island...... Australia

Offline Antman

  • Full Member
  • ***
  • Posts: 51
  • Karma: +0/-0
Re: testing the egs002 inverter board
« Reply #189 on: April 24, 2016, 06:50:39 pm »
Nice work OZ. I still cant believe something so simple works so well. I'm surprised that the copper bus bars are heating up too. With that much power I guess luke warm is acceptable lol.

What caps are you using? Mine are fairly small but are super low esr high quality Japanese ones so they still should work fine.

Offline oztules

  • Moderator
  • Hero Member
  • *****
  • Posts: 1177
  • Karma: +105/-8
  • Village idiot
Re: testing the egs002 inverter board
« Reply #190 on: April 24, 2016, 07:17:11 pm »
Yes Antman,
I too am surprised that the warmest part is the buss bars from the curtis controller that is rated at over 300 amps@36v... and it does/did that too in my electric car.... so very surprised.

It is now connected to the house, but the time switch on the HW service was daylight savings.... and so had already heated up the water service.
It is now reset, bu the house is only using about 400w.

Just boiled the 3kw jug, and voltage sagged by 2.1v to 237.5v from 240@3.5kw..... not sure why the differential is that big,although around .8 of  a percent is pretty acceptable in my money.

See how the day goes I guess.


Caps are rubycon 5600uf 80v 85C... so bottom of the barrel, but all I have at the moment.

Normal operation they are cold, but 3kw for hours on end heat them to 45c-50c from the previous experiment.. that used three, there are now 4 on this card.

For so simple a unit, the results are world class really... I'm impressed.


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

Flinders Island...... Australia

Offline oztules

  • Moderator
  • Hero Member
  • *****
  • Posts: 1177
  • Karma: +105/-8
  • Village idiot
Re: testing the egs002 inverter board
« Reply #191 on: April 24, 2016, 07:54:14 pm »
Some more useful shots from the other night building it.
5556-0

Underneath. The copper goes straight to the fet legs and to the outputs points

5557-1

and now bolted together.... note it takes only a few seconds to release the board from the heat sink for fet replacement... it it ever needs it....

5558-2

5560-3

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





Flinders Island...... Australia

Offline peter

  • Newbie
  • *
  • Posts: 19
  • Karma: +6/-0
Re: testing the egs002 inverter board
« Reply #192 on: April 24, 2016, 10:46:23 pm »
Oztules: Answers
1.5kW
I will keep looking for another one. I understand that making one from 2x cores is about right for most loads ever encountered for this project.
But I will be happy with a solid 1.5kW result.
I will read carefully the tranny rewind pages.

(maybe off topic but..)
When pulling the Aero Sharp apart I found inside two very nice current sensors. They retail for about $30 each at RS Components.
1 of www.lem.com/docs/products/ltsr%2015-np.pdf
and
1 of www.lem.com/docs/products/lts%206-np.pdf
these are simple to use, the LTS6-NP needs 5V supply and gives a 105mV/A output, with an accurate and linear output of +/- 19A. 4.5KV isolation, fast AC response.
Perfect for interfacing with Arduino and other uCs. With the 10 bit A/D resolution of the Arduino, it resolves about 1/20 Amp


Offline Antman

  • Full Member
  • ***
  • Posts: 51
  • Karma: +0/-0
Re: testing the egs002 inverter board
« Reply #193 on: April 26, 2016, 06:08:47 am »
Update on my design...

Got the boards today and threw one together quickly with just the bare minimum to get it running. I only had 12 mosfets in my stock so I need to order some more. These are just cheapies at $1.30 each but for switching at low speed like this they will do. I like the fact that they are genuine too.

http://au.rs-online.com/web/p/mosfet-transistors/7982962/?searchTerm=PSMN4R3-30PL&relevancy-data=636F3D3226696E3D4931384E4B6E6F776E41734D504E266C753D656E266D6D3D6D61746368616C6C7061727469616C26706D3D5E5B5C707B4C7D5C707B4E647D2D2C2F255C2E5D2B2426706F3D313326736E3D592673743D4D414E5F504152545F4E554D4245522677633D424F5448267573743D50534D4E3452332D3330504C267374613D50534D4E3452333330504C26

I'm running a 250W light globe for a test load and with a tiny heat sink the mosfets remain about 5 degrees above ambient. I ran it for a while with no heat sink and the 50 Hz mosfets did not get above room temperature. The 23 KHz ones on the other hand heated up quite quickly as expected.

Next I have to get all the arduino stuff soldered up along with the current sensing and fan control to make sure it all works. After that I'll make up the bus bars that are meant to run along the outside and screw in through the five holes along the boards edge. These will carry the high current AC to the transformer primary. I think this board will be good for about 1.5 KW max continuous in a 12V configuration. The current is huge at such low voltages. The good thing about this design is I can just extend the board longways adding mosfets.

The only problem I have now is the only transformer I have is a tiny 200VA one. I can either try and source a second hand one locally and rewind or be lazy and get one made in china. I got a quote some time ago for a 2000VA to be wound and it was about 450 AUD delivered which isn't too bad. The primary was solid flat copper too. Half of that cost was delivery too lol.

Offline billy

  • Jr. Member
  • **
  • Posts: 28
  • Karma: +0/-0
Re: testing the egs002 inverter board
« Reply #194 on: April 26, 2016, 07:01:59 am »
Nice work Antman


Billy