Author Topic: Taking apart a Microsine/Ok4U gridtie inverter. (Read 7715 times)

Perry · « **on:** August 13, 2007, 08:23:14 AM »

I've had an interest in the internal workings of the Microsine Grid-Tie inverter for some time now.

One day I googled Microsine schematics and came up with an old thread on Fieldlines site. "Wanted: inverter plans or schematics "

As suggested in there by TomW I've decided to try and document my work/findings in this diary.

I had given up looking for a broken Microsine to take apart and jumped in feet first by taking one of my units apart.

Presently I have most of the potting compound off and most of the major components removed from the cicuit board.

I have begun to trace out the board and draw a schematic of the device. I have started with the 24 Volt DC input section. I am using Kicad (freeware) to draw the schematic so others will have access to this software to view the schematic. I've posted most of my files into my photo uploads. Lists of parts, pictures, and schematics, and datasheets.

I could use some help identifying some of the board components....mostly surface mount items, or even help in organizing the data I post to make it more usefull.

Comments on the circuit would be interesting as well.

Cheers, Perry

Perry · « **Reply #1 on:** August 13, 2007, 02:41:21 AM »

This is what it looked like a few weeks ago.

Higher quality pictures can be seen in my photo section.

Perry · « **Reply #2 on:** August 13, 2007, 02:55:15 AM »

And this is what it looks like today.

Notice the dotted lines on the board? The drilled thru holes still full of potting compound. They split the board up into sections. The top left section is the area I'm working on right now. It's the 24VDC input section.

The sections are isolated from each other by transformers. L7 is the input into the 24VDC section from the control section. The 24VDC section outputs into L4.

Nando · « **Reply #3 on:** August 13, 2007, 09:55:24 AM »

Perry:

That is a major effort and I can give you technical assistance if You desire it.

Please send to me a message with your email address, mine is in the heading of this message, You need to change it to remove the anti-spam wording.

I do not have Kidcad and I would have to ask a few questions about it, to be able to install it to display what you have done.

Nando

ve7khz · « **Reply #4 on:** August 13, 2007, 11:15:38 AM »

Hi Perry,

I was a manufacturing mgr in a previous life (SMT plant) and could help with component IDs. Let me know what you can't ID and I will do my best.

cheers

Paul

Perry · « **Reply #5 on:** August 14, 2007, 03:28:12 AM »

Thanks Paul, I might have to take you up on this offer down the line.

Cheers, Perry (VE6PDF.....I use to be VE3TUX a few years back.)

boB · « **Reply #6 on:** August 14, 2007, 12:33:21 PM »

Good work, Perry.

Looks like a pain but I know the results will be good.

I ~think~ I was the first in the US to hook one of these up

in the nineties when Bill Ropenecker, the president of Trace Engineering

at the time, brought one back from Europe I think before they sold the

product in the US.

Will download the software and see what I can help with as well.

boB

boB · « **Reply #7 on:** August 14, 2007, 12:41:40 PM »

C40? 1U0 K 63V PH

C41 1U0 K 63V PH

C42 1U0 K 63V PH

C43 1U0 K 63V PH

Those are most likely 1.0 Microfarads/63V caps. Just insert a decimal point

for the "u" and call it microfarads. Same kinda thing with

certain resistor designations. 1k0 would be 1.0 K Ohms

1R0 would be 1.0 Ohms etc...

Does R49 measure like a resistor even though it looks like a capacitor ??

Maybe it measures like a capacitor ??

73 (Q73?) IRF644 ( Just a cheap 250V MOSFET )

boB

Perry · « **Reply #8 on:** August 19, 2007, 08:35:53 PM »

Thanks Bob. Comments noted.

Perry · « **Reply #9 on:** August 19, 2007, 08:45:47 PM »

Ok, This is an attempt at posting an image of my schematic to date.

As you can see it's useless.

Kicad will only allow me to plot to the clipboard which I tried to paste to Photoshop and to Paint with the above result.

What I need is a FREE ( so everyone can access/use it)schematic drawing program that will allow me to take a clean image shot of it so I can also post the image here. Any ideas?

I presently uploaded the file Microsine.zip which contains the 5 files that are in the project directory of Kicad so others can view it. ( Not much to view yet.)

I've also upload the individual files.

Perry · « **Reply #10 on:** August 19, 2007, 08:55:36 PM »

I had a little time to day to move forward.

Questions of the day.

I've got a Surface mount semiconductor with four leads. Marked 8G P124 BV (You can see it just to the left of L1 and there are three more in a row at the top right of the circuit board) Any ideas?

Transistor marked 1F p. My guess NPN from measurments.

Transistor marked SA p. Have not measured yet.

Cheers, Perry.

reiney · « **Reply #11 on:** August 30, 2007, 12:21:02 AM »

Hi Perry,

I also have an OK4 that I've started to remove the

potting from. I'm nowhere near as far along as you are

but would like to help in any way that I can. I'm using

a no/low cost schematic and CAD program called EAGLE.

(www.cadsoftusa.com). I would also like to build something

like the OK4; maybe around 250W, that could be stacked for

higher power.

I've built a couple of inverters so far that I use in my

van to run small power tools. You're welcome to the schematics

if you want (and I can find them; was a few years ago).

Thanks for all the info ...

Reiney

reiney · « **Reply #12 on:** August 30, 2007, 12:35:12 AM »

Oh, forgot; that four-legged part could be a bridge rectifier ...

Ivo · « **Reply #13 on:** August 30, 2007, 01:34:12 PM »

Hi Perry,

I also have two OK4E (Europe) made by NKF, but they are still working. Many have been broken down here in Europe due to high temperature. I have mounted mine on a big heatsink, after 10 years still no problems. Nothing to take apart for me:)

Like Reiney I also use EAGLE. May be good for you to try!

For some time I also try to build my own gridtie inverter, it's a hard thing to do. Not much information on the internet. Maybe your schematic will help.

May be I can offer you some help first.....

The part with 4 leads could be a bridge-rectifier or an opto-coupler, it's Easy to measure the 4 diodes inside or LED and transistor.

For the SMD transistors please look at:

http://www.repairfaq.org/REPAIR/F_SMD_trans.html

for Case-marking Cross-Reference.

1F BC847B

SA BSS123

Hope this helps!!

I am looking forward to you next posted Schematic!!

Ivo

The Netherlands

Ivo · « **Reply #14 on:** August 31, 2007, 08:56:44 AM »

R49 must be a Varistor for 130 Volt. It's a protection against overvoltage....

Ivo

electronrancher · « **Reply #15 on:** September 07, 2007, 06:09:34 PM »

Hey great work depotting that beast! Just happened across this thread and I'd be glad to help where I can. I'm an EE in power management, designing DC-DC and AC-DC power supply IC's. Haven't done any DC->AC yet, which brings me here.

So.. Is there any way you can measure those transformers? LCR meter perhaps, or just drive one loop with a 1v sine to get the turns ratio? That seems the most interesting thing, to me at least.

electronrancher · « **Reply #16 on:** September 07, 2007, 07:43:27 PM »

PS - The 8-pin LT part is an LT1257, 902 is their datecode. It's a DAC, most likely to make a sine wave reference from data latched in by the micro.

Soooo.. Wherever it's little sine wave is going will show us the PWM circuit. Can you try to trace pin 7? For reference p8 is Vcc and p5 is GND.

LT1257 Datasheet

http://www.linear.com/pc/productDetail.jsp?navId=H0,C1,C1155,C1005,C1156,P1425

howdy goudey · « **Reply #17 on:** September 15, 2007, 03:48:26 PM »

Perry, I'm impressed by your patience to reveal the silicon Pompeii underneath the potting goop. I have two non-functional OK4U inverters (ok, I fried them by connecting 24V-240W of PV). I wrongly convinced myself that it would protect itself by throttling the power to the inverter's max output. The higher voltage from the over-sized PV at the low current the inverter wanted, might have done the damage. It did actually work for a while before it died. Anyway, I would like to try to fix these inverters. Any suggestion about which components are most likely to suffer from this abuse, or some simple first tests to try, would be very much appreciated.

Also, any tips on the technique of potting removal would be good, too. I actually want the thing to be functional, so I'm worried about damaging the components as I remove the potting.

I have four other OK4Us that have been working reliably for a few years. I'd like to try to communicate with them on their RS485 bus. I picked up a RS485-RS232 converter, but I haven't had any luck using the OK4manager software. Any tips on the serial communication would be great, too.

Thanks,

Howdy

Perry · « **Reply #18 on:** October 19, 2007, 12:45:20 AM »

Just a note to let every know I haven't drifted off on this project. I've had a busy last few weeks repairing some damage to my house after a drunk driver tried to drive thru it.

November, will see me out of the country for a little bit, but when I get back the snow should be on the ground and I'll have a lot more time to play.

Thank you for the last few comments. I'll try out the other schematic software in the mean time and see if it'll work.

Cheers, Perry

Perry · « **Reply #19 on:** October 19, 2007, 01:13:07 AM »

Howdy,

I can't really give you any tips on repair. Yet

Do the 24V input lines read as shorted out with a meter?

As for the potting compound. Paitence and time. You can use a heat gun to heat up the potting compound. Not too hot. Just warm. Then use your favorite scraper to scrape away. It tends to crumble off when heated up. The hard part is seperating the case from the board. Slow. Very slow work. If you decide to move forward with this let me know. There is a couple of components I damaged a little. I could use information from your board to finalize my schematic.

It would be easier the second time as I know what to look out for. It's hard to know how fast to scrape away with out knowing whats underneath.

The back of the board has to prongs soldered to it that act as heat sinks I believe, that are connected to the case half. I could not figure these out with all the potting compound under the board and ended up using a hack saw blade to cut thru them

to get the board to seperate from the case.

Even if you get down to the board level you will most likely have to desolder alot of the components to get down to the troubleshooting level. Putting it all back together can be done, but.....you think I have paitence......you'll need even more.

Data read out,

I use the ATEN IC-485S. Switches selected to T-RTS/R-RTS (Center position) and DCE for the other switch. (Top position when holding unit with text readable.) R+ and T+ connected together and R- and T- connected together. I then connect the common + lines to one side of the OK4U and the common - lines to the other side of the OK4U.

Hope this helps.

Perry · « **Reply #20 on:** October 19, 2007, 01:17:10 AM »

Yes, I need to measure those.......

All in time.......I've got so many things to do on this side.

Will make this a priority.

electronrancher · « **Reply #21 on:** October 19, 2007, 12:08:01 PM »

Oh no! Well I'm sorry to hear about your house being damaged - hope it's sealed up before the snow comes!

Have a nice trip - I'll keep an eye on the thread as I'm always interested in this device. FYI - I use PSPICE schematics for all my board-level circuits. You can get a free copy (student edition 9) from many places on the net, but the number of components per page is limited.

http://www.engr.uky.edu/~cathey/pspice061301.html

Perry · « **Reply #22 on:** October 22, 2007, 01:27:44 PM »

Always a good side to every thing. You see that Solar popcan heater in my diary. They need to re-stuco my house so that unit will have to come off the house and then be re-instaled. The contractor will not touch it. The insurance company said to hire whoever installed it to remove it and reinstall it afterwards. That would be me!

We've agreed on $500.00 dollars for this task. It cost me just over $150.00 to build.

Payback period was shorter then I figured.

nander · « **Reply #23 on:** August 09, 2008, 11:01:48 PM »

Perry, what is your current progress with your tear down of the Ok4U inverter? I've been looking for one of these for a long time and understand that they are unavailable. Did Microsine ever offer a skematic? Have you been able build one of these inverters? Hope you don't mind sharing. Thank you.

News:

Author Topic: Taking apart a Microsine/Ok4U gridtie inverter. (Read 7715 times)