testing the egs002 inverter board

[oztules]

Fan controller

resistors
47k         X1
10k         X3
50k trimpot X2

Diodes
3v zener    X1

semi's
IRF640 ?    X1 is this suitable?.............yes
LM339       X1


Gti cut out

caps
100uf electro X1
25uf  electro X1
100n       X1 ?? what type of cap is this? ceramic or similar

diodes
1n4007        X2
20v zener     x1
40v zener     X1
5v  zener     X1
red led       X1
grn led       X1

resistors
1w 270r       X1 ??1W?  I think they were 3w from surplus store 20 years ago
1w 560r         X1 ??1W?... ditto
1k5         X3
15k         X1
12k         X1
5k6         X3
20k         X1
5k trimpot    X1


Misc

LM339        X1
CD40106        X1
IRF640        X1



CONTROL BOARD

Caps
4.7uf        X5  ?? ........can i use tantalum?? I use ceramic or electro now, don't know what tantalum will do... should be ok.
100n        X9  ?? ........what type are these? ceramic or poly .2" and .1" pin types check board spacing for each
100uf 63v    X1
10uf 35v     X6  ?? correct voltage?  I use 63v there are 7 including the one under the 8010
22pf ceramic X2
1nf          X1  ??plastic film capacitor?? ceramic or poly


Diodes
13v  zener   X1
5.6v zener   X1
1N4007        X12..... The charge pump uses2 x FR107 high speed stuff. the rest don't matter.. I only see 8 normal, and 2 fast
led          X1


Resistors
5W 120r      X1
10k          X4
100r        X2
1k        X3  only 2 of these
50k        X4
0r        X10..... think there are 12 of these including one under the 8010
2k  trimpot  X1
10k trimpot  X1

xtal 12mhz hc49/us  Is 12.288mhz ok? use 12mhz, they are only 10 cents each. 12.288 should in theory work fine, but I don't know how much that will effect dead time etc. with machine cycle time... not much i suspect.


Remember I just build things, I am not an engineer..... just the village idiot. Most times I use things because I have them on hand. Not often have i found that different caps make appreciable differences... but this time they did/do. I know their behavior is slightly different in different circumstances, but did not think it was important here... wrong again ... but I didn't appreciate fet gate capacitance much years ago either

[Dr_Zogg]

cheers Oz your the man.

[Dr_Zogg]

howdey,

just another couple of questions.
for the fan controller is it ok to substitute the 3v zener for a 3v3?
instead of irf640's can i use some of the 4110 that i have laying around?
for the high speed diodes can i use byw29e I have a few of them laying about?

Cheers,
            Sam

[oztules]

First, the coil you refer is pretty small... so no.

yes 3v3 is fine, it only establishes a reference voltage.

Any power fet will do.... we are only dabbling with 200ma at most... so 4110 is great..


Was fiddling with the board today, and hooked it to the house AFTER I had set the voltage... It ran local loads of up to 4kw perfectly and stably..... when I hooked it to the house circuit... without even turning  the contactor on... the voltage moved by 15v.. down.

Tried all sorts of things to see why.

1. replaced the 14007 with high speed fr107... small difference.... changed the  position of diodes with the 50k resistors... ie rectify first, then through the 5ok resistors.... small change again... cut off the 10k resistor in parallel with the pot... small change.... but have not solved it at all really.

If I set it to the house circuit... it tracks fine within a few volts over a 5kw range... but there is something of a high impedance fault I can't get to grips with yet....   

Small filter before the 240v goes to the voltage sense network helps a bit too.



,,,,,,,,,,,,,,,,,,,,,oztules

Always something I didn't think of apparently... it is annoying, not a game changer though.

[Dr_Zogg]

thanks for that oz.

hmm that is a funny quirk indeed, keep us in the loop it will be interesting to see what you come up with as the cause

Cheers,
           Zogg

[oztules]

It gets even weirder.... The solar panels are still hooked up at night, and I don't bother to block them... I am able to switch them out as 4 banks of approx 2.5kw per bank..

Well I switched off two of them don't know even why, but noticed the voltage fluctuated on the inverter... turned off the next two, and voltage on the AC changed by 15-20v?? how?... so tomorrow  some investigation is in order... first the house without the power actually turned on changed it by 15v, then the solar panels at night.... changed it by 15v.

Makes no sense, as it s the AC side that controls it. A 5kw load makes almost no difference at all... maybe a volt......, but unlit panels and a turned off but connected house makes a huge difference.... wondering if the MEN earth system has something to do with it... no idea at the moment... but on all the testing "local loads " no matter the size, don't effect it at all....... thinking.....nah, must be ghosts..... .

..........oztules

[oztules]

Ok, I have isolated the problem. As a stand alone inverter, it needs no changes.... but as a off grid type unit, it needs changes to the ac reading part.

Fortunately this is a very easy mod... we simply pull out the 4 x 50k resistors on the voltage divider, and replace them with a very small ac transformer any voltage so long as it is over 5vac to probably 30v or more.

It seems that because the original circuiit was designed to not place the AC  either the load or neutral to ground... as this would blow things up if the MEN earth was messed up... safer for the unit to float....... so, they used the 200k as quasi isolation... the isolation you have when you have no isolation... that kind of isolation.

If you look at how they implemented it in the eg8010 data sheet, they have not accounted for a earthed neutral... which will short out one half of their voltage divider... this is what happened to me, the 240v became 120v just by plugging in the house leads... by using all kinds of tricks I reduced this to 15-20v.... but unstable with different environments.

The solar panels/battery and their earthing regime.... water dust etc ( is has been raining for 2 days here... another reason the problem became much easier to see) also changed the earth side of the DC and so this affected the result as well, so we can see that this system could be put off by competing earth loops.... high impedance ones as well



It is a happy coincidence that clockman had plenty of room at that end of the board, and that 4uf was generous, and 2uf was plenty.

So remove 1 x 2u2  cap, and used that space to place a small transformer like this.



The 240v leads go straight through the board to the 240v output pads.
The 14v (in this case as I got 10 for $12 years ago.. odd voltage never used them till now) now goes straight into where the 50k finished off before thwe diode bridge... and we are done.

Here is another look.



and again




and again




Now it is totally independent, and has very good stability no matter the earthing of the loads, or the battery side.


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

Ps sorry Clockman I never anticipated this... although I had a sneaking suspicion this might have been the case, but local load testing never looked like finding this out.

[ClockmanFrance]

Hi oztules,

What can I say but Wow!

Very good Excellent news on your finds and solutions.
 I went to bad last night in serous thinking mode, 'current sense ceiling reached on the 8010 and holding back voltage'.? 'PV leakages mine do'? 'Do we need that AC transformer back'?etc etc.

I wake up in the morning to come to write Brief notes the OzInverter (been asked) on its History, its name, substance and its development.
And 'oztules' has the voltage problem sorted!.

No problems on the OzControl board, mods. On my/Mrs Web site, my page is a dread full mess, I now have an update/amendment section so alterations are not desperate.
Plus I have small quantities of Print runs and PCB runs and the PCB stuff masks I print direct with my Mono Laser so things are quick to alter. (from today anyone asking for the boards would get this new one).

Oztules, Can you Link please to a suitable small transformer that will do the job, I will amend board. Many Thanks.

 

[oztules]

Something like this is designed for pcb
http://www.ebay.com.au/itm/220v-12V-1W-Output-Voltage-Isolation-Dry-Type-Transformer-Toroidal-Tube-/201347873880?hash=item2ee144bc58:g:Q2kAAOSwqu9VUETR

or a normal type would be this
http://www.ebay.com.au/itm/AC1W-Dual-LineTransformer-220V-to-12V-Power-Transformer-Power-Supply-Transformer-/191578313687?hash=item2c9af513d7:g:R8sAAOSwhwdVUZ77

Bigger is here:
http://www.ebay.com.au/itm/220V-50Hz-to-12V-3W-Ferrite-Core-Vertical-Single-Phase-EI-Power-Transformer-/391252396214?hash=item5b1875c8b6:g:ps8AAOSwJb9Wq6TV

etc etc etc.....


....oztules

[Dr_Zogg]

so that i can get one of these local would any of these be ok?
http://www.wiltronics.com.au/catalogue/465/electronic-components-and-parts/transformers/low-current-transformers-240v-ac

Cheers,
           Zogg

[oztules]

yes 9v or 15v  one would be fine....

Also to get better definition using the trim pot, I found that 500 oms instead of 10k with a 7k6 padder resistor on the  plus input leg, a 1k8 resistor between the gnd leg and ground. This gives a much tighter range to trim over... from 215 to 270v, rather than almost 50v to 390v.... and very tight to trim.... ie 1 degree was probably 10v, now less than 1v...easier to get to exactly 230v or 240v whatever....

I was a using a 14vac transformer, so your resistor value will be different I suppose.

Since swapping to the transformer for voltage control, the range during the day under any loads from the house was less than  a percent variance from the looks of it... probably closer to half a percent

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

[ClockmanFrance]

Trim pot from 10k to 500r ?

100r resistor replaced with the 7k6?

10k resistor  replaced with a 1k8?

Thanks.

[oztules]

No, 10k pot to 500r pot
1k8 added
7k6added
10k resistor deleted
transformer added
power LED pins added
proper SPWM inhibit with pins added for on/off rather than the temp pins.
secondary LED pins for fault led output as well as on board LED.

Here is the tentative pdf.
Will build one today all going well.
If no changes then you can action it.
Don't use this except for perusing at the moment.

 80109B.pdf (28.51 kB - downloaded 551 times.)

........oztules

[MarNet]

Hi all,
Antman i heard that you need a bit of help with Arduino? Well i won`t charge you   . If i can help you there is no problem
This thing is made and designed by me ... ->

one question: the 4110`s gate power is 12V right? i`m stuck with an mppt controller here... I wanna make an 60A MPPT(arduino powered) for my wind turbine with the fixed output of 30V ... The mosfet get`s verry hot at 100khz ... Ill try to put an P channel transistor to run the fet from an 12V source
 

[Antman]

Hey Mar

Nice work that design and putting it all together neatly. Yep I'm a noob with all the Arduino programming. I understand what I need but just don't know how to put it all together.

The gate would work at 12 volts but it is best driven with a dedicated mosfet driver. What does your schematic look like? I have a buck converter board designed that is a single phase Design. Good for about 15 - 20 amps and 96 - 98% efficient without blocking diodes on the output. I also have a dual phase one mostly designed which would be good for double. Here's a quick vid with some basic software showing it working.

Only thing limiting the design is software. It needs a 100khz pwm with at least a 10 bit resolution. Dual phase would need a complimentary pwm.

Where abouts are you located?