Author Topic: testing the egs002 inverter board  (Read 62007 times)

0 Members and 2 Guests are viewing this topic.

Offline oztules

  • Moderator
  • Hero Member
  • *****
  • Posts: 1177
  • Karma: +104/-7
  • Village idiot
Re: testing the egs002 inverter board
« Reply #90 on: March 29, 2016, 11:47:09 pm »
"I was considering scrapping the EGS002 board and just putting everything onto my board but I honestly cant be bothered."

When the 2110's turn up it will be tested fully, but the wave forms at the inputs to the 2110 are looking very tidy at the moment..... that could all change when they are driving the 2110 though... hopefully they turn up soon..

5366-0


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

Offline Antman

  • Full Member
  • ***
  • Posts: 51
  • Karma: +0/-0
Re: testing the egs002 inverter board
« Reply #91 on: March 30, 2016, 12:03:31 am »
Excellent job mate! Cant wait to see the results.

I might be able to see the surface mount components but ill never have the patience you have designing everything with through hole.

If I get a couple of spare hours tonight I might throw the Arduino into the design. I'm also going to put a current amplifier in there as well and link it to an external 300A shunt. With the arduino ill be able to set any current limit and will limit power dissipation on the shunt too.

Offline RFburns

  • Full Member
  • ***
  • Posts: 106
  • Karma: +9/-0
Re: testing the egs002 inverter board
« Reply #92 on: March 30, 2016, 02:17:28 am »
I think it will drive o.k but under load thats the acid test  ;) ! .If its close to the output board and its under heavy load and still working; job done (extra lead length of discrete's is the only real problem I can see as we know the fundemental design is sound). 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: +104/-7
  • Village idiot
Re: testing the egs002 inverter board
« Reply #93 on: March 30, 2016, 03:29:20 am »
Yes RF, the extra lead length and groundplane issues is what concerns me.... but I could spend days trying to calculate it all.... and in the end, the laws of the universe will prevail... so suck it and see will be the determinate factor.
I am hopeful, as virtually all the 12mhz stuff is in the chip and out to the crystal, the rest is only 20khz or so, and the lead lengths are fairly modest for this frequency. ( is that 1500 meters?... just over the top of the audio range...)
Time will tell, but I remain hopeful.

The ribbon length seems to make no noticeable difference as testing has proved... now the next step.

Antmans version would lend itself to this better I think, but I'm not sure the difference will be noticeable. Fet heat will tell it all in the end.

But , gee if it works as planned, then what a simple solution to what I thought was a complex problem...... then  a pic chip... and we are free.


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

Offline RFburns

  • Full Member
  • ***
  • Posts: 106
  • Karma: +9/-0
Re: testing the egs002 inverter board
« Reply #94 on: March 30, 2016, 08:47:38 am »
I am pretty sure there wont be any major suprises and if there are any issues I am sure a small rejig of the layout would correct anything . For sure this is a much better try it than we will see; the maths is just to much. 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 ClockmanFrance

  • Full Member
  • ***
  • Posts: 164
  • Karma: +11/-0
Re: testing the egs002 inverter board
« Reply #95 on: March 30, 2016, 10:21:50 am »
Just my schottky diodes now.......

Best get busy!

Offline frwainscott

  • Newbie
  • *
  • Posts: 21
  • Karma: +0/-0
Re: testing the egs002 inverter board
« Reply #96 on: March 30, 2016, 08:55:46 pm »
It would be nice if there was a parts list for 12v,24v,36v,and 48v so people would be able to scale up as they can without breaking the bank. I know 12 volts is almost ridiculous due to the high amperage but a 12 volt/2kw unit could be the difference between some people getting to delve into such a project or not being able to. Whether it being for their home or experimenting.....

Frank.

Offline Antman

  • Full Member
  • ***
  • Posts: 51
  • Karma: +0/-0
Re: testing the egs002 inverter board
« Reply #97 on: March 31, 2016, 04:52:30 am »
I've done a bit of extra work on the board.

I've decided to spread the input power over 3 x 10mm squared cables similar to how the PJ boards distribute their negative. I've got heaps of these screw terminals that you can see in the center of the board which are a nice solid connection. The connection points are no more than 15mm away from the mosfets at any point so with 2 oz copper on the PCB voltage drop should not be an issue.

Current will still be sensed by an external shunt and amplified by a INA271 high side sensor. Its pricey but very accurate and can handle over 70V. This will be monitored by the arduino and or the EGS002 via a selectable jumper. This way I can get the arduino to start the fan if the power level goes past a certain point.

Fan is controlled from the arduino and can be PWM'd to vary the speed depending on heat or load.

Thermocouple is monitored from the arduino

Input voltage is monitored from the arduino for low/ high voltage shut down

DC to DC buck converter is included for 24, 36 & 48 volt use. A shorting jumper is used for 12V instead of the buck converter.

Pins for attaching an on/ off switch which simply disconnects the control voltage.

I added pins for a LCD from the arduino in case I want to go overboard. Its a shame the EGS002 board  doesn't have a serial connection breakout to read and set all of the parameters. I may have to remove the EGS002 board all together and place everything on my board ;)

Now the hard part is to connect all of the pins up and tidy up the layout when time permits.

Offline frackers

  • Full Member
  • ***
  • Posts: 239
  • Karma: +9/-0
  • If it moves - computerise it!
Re: testing the egs002 inverter board
« Reply #98 on: March 31, 2016, 05:03:03 am »
My boards have arrived - I ordered the LCD to go with them and it is tiny!! 28x12mm - a total waste of money!

Hunting through the junk box I came across  a 1700VA UPS but checking the number of windings on its transformer I think I'll stick to the original plan of a simple 15v transformer backwards for the test jig!!

I might even get my logic analyser on it over the weekend.

I'm also working on 20KHz low side, 50Hz high side software for an Arduino Nano clone. Looks pretty straight forward, just got to work out how to get the loop gain right for the voltage feedback to keep the output stable. Unfortunately, my study of Smith charts and stability criteria and such like in closed loop systems was over 40 years ago and that was all analog anyway  :-\
I can see some of the advantages of this approach - less current required from the charge pump driving the FET gates on the high side and reduced losses due to less switching but is that it (or am I missing something). It seems to avoid the dead-time issues of fast switching!

The all singing and dancing 'from scratch' controller will use an stm32f103 or stm32f303 and will monitor both input and output voltage and current but will probably be 20KHz both high and low side as the dead-time is handled in the hardware of the micro itself. Or is might be 50/20k as I learn more about inverters.

Robin Down Under (or are you Up Over!)

Offline frackers

  • Full Member
  • ***
  • Posts: 239
  • Karma: +9/-0
  • If it moves - computerise it!
Re: testing the egs002 inverter board
« Reply #99 on: March 31, 2016, 05:09:30 am »
DC to DC buck converter is included for 24, 36 & 48 volt use. A shorting jumper is used for 12V instead of the buck converter.

I hope the buck converter isn't like one of these http://www.aliexpress.com/item/10pcs-lot-Power-Converter-Step-Down-Module-LM2596-LM2596S-DC-DC-1-5V-35V-adjustable-step/1967756079.html because they are only rated to a maximum 45v input :(
Robin Down Under (or are you Up Over!)

Offline Antman

  • Full Member
  • ***
  • Posts: 51
  • Karma: +0/-0
Re: testing the egs002 inverter board
« Reply #100 on: March 31, 2016, 05:21:04 am »
Hey Frackers

It'd me more like this. http://www.aliexpress.com/item/48v-Step-down-Module-DC-DC-4-5-60v-LM2596HV-Adjustable-Regulator-Module/32266096532.html?spm=2114.30010308.3.1.DVzWIH&ws_ab_test=searchweb201556_9,searchweb201602_5_10036_10035_301_10034_507_10032_10020_10017_10005_10006_10021_10022_401_10018_10019,searchweb201603_9&btsid=283f427b-af11-44e4-8325-00af349b11b8

It's rated up to 60V but to be honest I wouldn't trust it to be too reliable. On a 48 volt system you would be pushing the limits and the Chinese aren't known for leaving much headroom ;)

This board is more of a test board for me. If it was a final solution I would include my own high voltage buck converter circuit in there but my component list is already growing larger then I'd like. Using this i'll be able to test control circuit current consumption too since it is way too hard to work out now.

Offline oztules

  • Moderator
  • Hero Member
  • *****
  • Posts: 1177
  • Karma: +104/-7
  • Village idiot
Re: testing the egs002 inverter board
« Reply #101 on: March 31, 2016, 06:24:35 am »
The 2110's turned up and I plugged them in......


Well tonight's news is that the unit at the top of the page works... not sure yet how well, as I only took it up to 2kw resistive element, and then decided to be brave and start the compressor next to me.... all worked and not a dead fet in sight.... decided to go to bed with success in the mind.....bit wimpy I suppose.

Tomorrow I will get a bit more game and push it to 6kw... it would have seen peaks near there starting the compressor I would imagine.

Now the good news..... the fet banks stayed dead cold after 15 mins at 800 watts for the first test..... the current control works smoothly, the temp control works smoothly, the voltage was incredibly stable ... even switching on the compressor... the scope did not see it correct at all..... nifty indeed.

Now the odd news.... this chip has excellent voltage control, but unlike the 002 boards, it has no idea at all of under AC voltage cut out...... no idea why.
The wave form is very good for the top 3/4  ( and bottom) of the wave.... but there is some wobble at cross over.... may have to vary the dead time a bit... but apart from that, seems very solid... tomorrow will tell as I dump all kinds of loads on it.. then finally, the big welder.....we'll see.

So a dead simple unit is very much possible.

Frackers...... a nano?... I would not have guessed it could do it... but if anyone can, I suspect you can.

Antman, I am hoping this success will goad you into ditching the 002 and do a better job of what I have done... this is getting very interesting now.


Camera has flat batteries so hopefully tomorrow some pics.


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

Offline RFburns

  • Full Member
  • ***
  • Posts: 106
  • Karma: +9/-0
Re: testing the egs002 inverter board
« Reply #102 on: March 31, 2016, 04:56:51 pm »
Things are getting more interesting....

From the tests sound's like it will work well enoungh with a bit of tweaking

If you gound pin 13 on the 8010 directly under voltage should kick in after 3 sec's - I would test this

What is the wave like to/from the 2110's? (If this is ok -which i am thinking it will be- is there noise on this under load?)

Frackers yes the LCD's are tiny and really only good for test's

Antman I am sure there are probably a few people who will be following your build who want a beefy 12v inverter for the 4x4/van etc so keep us up to date. 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: +104/-7
  • Village idiot
Re: testing the egs002 inverter board
« Reply #103 on: March 31, 2016, 05:28:41 pm »
No pin 13 on this one will not turn off the pwm on low voltage. +5 will turn it off, ground won't. 5v means over voltage to hell.... so it has no under voltage protection at this time.. big wow really.

By that I mean to go low acv... we will be drawing well over 10kw from the looks.... this thing has grunt.... and runs cold under 2kw.

I ran it at 2400w for 10 mins this morning, it took 5 mins before i could feel  difference in the heat sink .. after 10 mins it had risen to 27C... no air movement.... that must be very very efficient......., better then the 002 board... start temp was 17c..... thats pretty startling really.

Just like the 002, starts anything I have tried so far. it is all looking too simple... not understanding why i didn't do this years ago... save a lot of time and effort getting the pj up to scratch.

This will run the house loads today as a good test of all kinds of loads, mixed and solo, high and low from 177w at the low end, to 3kw or more for a few hours, then back to normal 200w-6000w cycles.

I expect it to do it easily from what I have seen.... stupid simple seems to work this time....egad.

Should be not much tweaking RF... there is nothing much on the board. Not sure what is causing the apparent distortion, but it is not showing up as a heating source thats fairly certain... it runs too cold under high loads, thats where I would have expected bad switching to show up.



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

Offline oztules

  • Moderator
  • Hero Member
  • *****
  • Posts: 1177
  • Karma: +104/-7
  • Village idiot
Re: testing the egs002 inverter board
« Reply #104 on: March 31, 2016, 05:52:58 pm »
This is the scrappy little devil....
running the house, hot water on so about 3kw... certainly does not look like much... but gee does it power along.

5382-0

Not much more I can say... it just works....


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