Anotherpower.com Forum
Renewable Energy Questions/Discussion => Automation, Controls, Inverters, MPPT, etc => Topic started by: Lachlan on October 09, 2015, 03:11:08 am
-
I build my own house and went off grid - solar only. I am not so clever on the electrical front but decided to be brave and have a go. Bought my own panels 16 * 250 w @ 24 v - batteries lead acid with about 600 ah rating and the power jack. Actually had three inverter's now. One to build the house (not pure sine wave), one that blew up within 5 minutes (was a dud) and the current 8 KW model which has been a champ......so far
I have been running the inverter for about 9 months (started in middle of summer) - sweet as - runs the house well - vacuum cleaners, kettles, washing machines. Everything works well and power during the winter was not a problem. Had to ration power on the really cloudy days (i.e. no electric kettle) but no hassles. So now the summer is here and lots of sun but.........
Last week for the first time my inverter tripped off. I had just topped up the batteries (lead acid 600 AH @ 24V) - nothing unusual. So it tripped and the warning it gave was over voltage - that is about 1 minutes of beeps followed by shut down and continuous beep but green light on.
Checked voltage and was sitting at 27.5 as it had been. However decided to turn the solar controllers down to 27 to be safe. Turned back on - ran fine for about an hour then same thing again. By now it is getting to later in the day. Called a mate and we had a discussion on it all and decided to drop it to 26.9 and try again. Worked fine and ran all through the night. Yay!!
About midday the next day same thing. Turned it off and gave it a "rest" and put it back on again in the afternoon. Again it worked all through the night and into the next day till about mid day. Shut down time again. This has been the cycle for a couple of days now.
One thing I have noticed. When we run a load the big fan is coming on as it always has, but all the time I am in the shed recently I dont hear the small fan cutting in like I used to. Would have though it would have shut down with over heating warning but no. used to hear it coming on and of all the time.
Has anyone got any ideas on what is going on? And what I might try and do to remedy the situation?
Love being of grid but you feel so vulnerable when the inverter goes.
Lachlan
-
Lachlan,
You announce yourself as a newbie but could you please tell us all which country you live in? This would be most helpful so that readers could put their own situations in perspective to yours. It all helps people to understand your achievement.
David in Hong Kong
-
you have more sun now than in winter, and your batteries are getting to float much easier. Is is at this time the spike problems occur with pwm voltage control of your solar system.
Discussed many times before on here...... so, a 1m resistor across the R14 10m resistor will probably stop it happening, I use a 5v6 zenner across the 10m resistor cathode to the main chip.... this will stop it too.
...............oztules
-
Hi and many thanks for the reply David and oztules
David HK - I am in Australia. West and about as far south as you can go before getting wet feet. Out of town and no power no water (or at least not without incredible cost). A mate who has been doing wind and solar for about 6 years (same situation) said "we can do this" and so the journey started for me. Mate has been doing own repairs on wind turbines and built a few panels from scratch is pretty handy with the solder iron but the inverter is a little beyond us both although we understand the basics (him more than me). There is a few others like us about the place here - that is off grid without going through "the system" - no government rebates etc. We share between us what we can. Important to note that everything from the inverter onwards is licensed work - that is everything at 240 V is by licensed electrician. Solar panels are fairly straight forward as is the batteries. The inverter is our weak spot and nothing local to help us out. I should mention my father was a TV/Microwave technician and still pulls computers/electronic equipment/washing machines apart (generally from throw out your rubbish days in the city) and reassembles them with various "brains" as he calls them. Also have access to mates who can fix electrical music equipment/sound systems etc. I guess what I am saying is the skills are about but not the knowledge.
oztules - I have been reading your notes for about a year now and saw the recent one on the 6KW inverter and it had me thinking it may be the same issue. When I fist found this site I thought "we are not alone". I have read what you have said but is there any particular posts from before with more detail or a diagram/picture? I am the one who will collect the information - not do the work - so if you think I am a clutz - your right! but incredibly grateful - and learning all the way.
Q - why has the pwm (that is post wip modulation?) of the solar control changed from last summer? It did not have this issue before in mid summer so a bit confused why this has changed?
Again many thanks
-
you have more sun now than in winter, and your batteries are getting to float much easier. Is is at this time the spike problems occur with pwm voltage control of your solar system.
Discussed many times before on here...... so, a 1m resistor across the R14 10m resistor will probably stop it happening, I use a 5v6 zenner across the 10m resistor cathode to the main chip.... this will stop it too.
...............oztules
See my message reply on this - cheers lachlan
-
Hi Lachlan,
I have been constructing Oz's 6Kw Inverter.
Total cost for all the bits I needed to buy was about £500 GBP.
[attachimg=1]
Oztules notes and jottings are excellent and in my opinion sufficient to produce a good machine. My 'OzInverter' LIVES, and it 'cocks a snoot' at my expensive SMA Sunny Island 6, which never did do what what it says it is supposed to.
My 'OzInverter' is in at present, as I fit all my Meters and finish.
[attachimg=2]
The Transformer and PCB's are the heart, and Oztules way of manufacture is actually straight forward. Commercially winding them at that size is probably just not cost effective.
I will, sometime over the coming winter put my 'OzInverter' construction/manufacture story, into a Booklet form with lots of colour Photographs etc, which may help others.
This is my Not for profit 'Make your own cost effective 2kW PV Solar Tracker' booklet. And I will do something similar with the my 'OzInverter'.
http://www.echorenovate.com/new-book---make-a-solar-tracker.php (http://www.echorenovate.com/new-book---make-a-solar-tracker.php)
(Sorry Mods if this link is inappropriate please remove if you think necessary, we are Not for Profit.)
My French sheep checking my installation of No3 Tracker?
[attachimg=3]
-
try this http://www.anotherpower.com/board/index.php/topic,902.msg9296.html#msg9296
did you add panels this year?..... using less power this year... or we have a different problem. Try the zenner or the 1m resistor and see where we go from there.
The fan programming is a total mystery to me... and them I think... it changes characteristics all the to me... I don't use theirs.
It is possible that your fan is sending no tacho information too, and if it is running weird, then this is a possibility too.
................oztules
-
oztules - many thanks for your consideration and the link to info - Will try 5v6 zenner as you suggest - help at hand this end - all knew exactly what you referred to
On the issue of changes in the last 10 months - none - no new panels - usage has been constant. This is what is throwing me a bit.
The small fan (I call it the small one as apposed to the big one which is working as before) is the thing I think has changed - as well as the tripping each day about midday but never under a good load. Always trips while nothing much is happening. The small fan does not come on as it did. In fact over the last couple days I have not heard it on once.
So I have to ask what is "no tacho information" you referred to about the fan not sending means.
As we put in the zenner I was planning to replace/look at the small fan anyway as part of ticking of boxes.
Any thing else to consider and check while the lid is off as part of "while your at it" type stuff?
Lachlan
ClockmanFrance - when I see what your doing I believe more and more it can be done and schools need to teach more about all this stuff. Without taking the fun out of it it should be mainstream info
-
So - here is what happened so far............
Last couple of days were a bit cloudy and we have not tripped out once but had solar controllers set at 26v to reduce the tripping.
Got a spare inverter (same type etc. = PJ) and swapped over. Turned up the Solar Controllers to 27v and waiting for any issues - none. Over to working on the old one.
As soon as we got inside we noticed the fan in between the coils (what I call the small fan) was stuffed - and wobbly. However having a spare (with broken blade) about we swapped over good blade to good motor and tested. Knew fan worked (albeit a bit old) so put it in and set about a test. As soon as we turned the old inverter on (without any load at all = 0) the first thing that happened was the small fan started. It ran for about a minute then stopped. Have old inverter running with spare with controllers set at 27v and no issues so far.
Seems we may have found the problem on mine.
Any ideas anyone on what type of fan to get to replace the old one? Like would a bearing fan be better? Would really appreciate some help on getting a decent fan to replace these ones with.
My friend who had a similar issue found his fan was OK and strong. So we have placed in a 1m resister. So far so good. Always good after these things that nothing blows up!!! Now we wait and see........
-
I use beecroft fans off ebay... they are easy to convert to tacho units to use as the PJ primary fan, or just to use as normal with your own fan circuits.
The PJ needs to see at least one fan with tacho.
Really does not matter what you buy, they can all be made to give a tacho signal... but try to get ball bearing style if possible.
..........oztules
-
Update....
As I said before we bought a new inverter with the thought of having a spare between my mate and me. The old inverter apparently had a problem with the fan (the 120mm x 25mm that sits between the two coils) so we ordered a replacement with a view that this could be a spare.
The new inverter I bought has been running fine although tripped out one sunny day when nothing was happening (but only the once). My mate who has wind and solar placed the 1m resistor at R14 in his and it has been running well and no tripping out from over voltage. Thanks for the assist with that issue.
Finally got the fan we needed after initially being sent the wrong size and having to wait yet another month. In the meantime got fed up waiting and also bought a bearing fan (Beecroft). As it turns out we ended up with two fans (nothing like a spare I thought). Today we finally both got together for fixing up what we hoped would be our spare inverter. Replaced the fan (with the bearing fan) and also put in a 1m resistor at R14 (just to be sure) and hooked it up for the big test........
As soon as we started it we noted that the fans did not trip in but understandably so as there was no load and it would have been cold from not working. So I set off and started up a kettle to give it some stick. And we watched for about 30 seconds..............still no fans kicking in and when the smoke started coming out of the power outlet end we decided it was perhaps the best time to shut down. We then (quite quickly) unhooked everything..........back to the workbench and had a look. Signs of smoke near the the fan we just replaced. We took this out and tested it and sure enough this was cause of the smoke. So we got the other fan (tested it to make sure it was working (all good). Put this fan in and hooked it all up again but this time with the top off so we could see what was going on. Same process, still no fan kicking in but this time soon after we turned it on we both saw the arc flash near the mosfets and again decided this was the best time to shut down. We looked but nothing obvious at this stage. Tested the fan again and it was still working. Again we unhooked it all but this time replacing it with the inverter that was working but now with a 1m resistor to hopefully stop that over voltage issue. All good and working as before.
So now a fair bit of head scratching on the old one and wondering if we have to strip it all down but we are now out of our depth. Like we can take it apart - perform checks - identify issues - replace parts but we would need guidance on what to do. In our discussions we agreed that it was a pity that we had no one we could take it to in Western Australia (that we are aware of?) but if we can get a guide we can have a go. Even if we could send it to some one in Australia it "might" be worth it. Alternatively in our discussions we agreed it might be just easier to buy another for our spare but it seems a cop out and a waste.........and we don't get to learn anything (which is always good ....even for a couple of over 50s die hards who believe that this is the way to go)
All the above issues might be completely unrelated - like a spider could have go in while we were waiting for a replacement fan or what happened might be the underlying cause of it tripping out in the first place - no idea.
So.....has anyone got any thoughts/views/ideas on how we might proceed?
Tested but still loving it
Lachlan and his mate
-
I have only seen one inverter with the "mid fan" in the transformer area... and that was a 48v fan... wondering if yours was 12v or 48v... that would explain the smoke part.
The failure.... no idea, unless it had something to do with the 12v fan coming from a 48v point on the board, and interferring with the power suply part....
Things you can do. If you pull out the power board. ...........Replace the fets and the resistors that will be cooked. This will give you a new power board, there is really nothing else on it.. just fets and resistors few caps too.... but no tricky stuff.... hint cut the legs out before unsoldering... so you only lift one wire out at a time, this will help NOT destroy the solder pads.
The control card.. this can be pretty disappointing if the problem goes past the totem drivers and the opto isolators.... may be best to just buy a control card.. they are fairly cheap...... else replace the totem driver transistors, and the opto isolators.. and cross your fingers.. big time...
No harm in trying. Fets from aliexpress are of less than perfect specifications, but for the momney ( usually can get 50c/piece) generally around the .006-.009ohm... which is pretty good, and makes for cheap experimenting. Dinky di ones may be a few bucks each, and be very expensive to blow 24 of them up at once.
I wonder how I know this...
...........oztules
-
Oz Thanks for the reply. Really appreciated
Ours are both 24V systems (both have the fan between the coils and one at the input end) and the fans are the same V. Pretty sure we got the info off the the old fan right when we had a look at it the first time and magivered a 24V fan which worked but was not a reliable fix.
The rest of what you said, like always, we will have a chat about over a few days and try to absorb the information you have provided and see if we get our heads around it all. Your right ...you know a lot of stuff.
You really have me wondering about the fan V. The input is 24V from the batteries and the fan is connected to the input.......... is it possible that the mid fan is only 12V ....how does that work?
Lac
I fully realise how much I don't know is amazing
-
Yep you were right - just checked the bearing fan I bought. They had sent me a 12V and not 24V that I ordered. Lesson in checking what was ordered. This would explain the smoke?
Now onto the rest of it..........
Lac
-
So we tried (you have to try)
Stripped it all down. Found the arc point which was at the point where the transformer lead connects to the heat sink (the black one). Problem seemed to be a combination of loose screw and heat shrink over the terminal - poor contact. Rather black. So we made good the connection..
Decided while we at it we would go all the way and make sure nothing happened around the fets. All good and no signed of damage. Testing no short circuits and resistance was the same on three sides.
Just started it up and it works but now giving up to 270 V output. So now way way over.
Have we perhaps cooked something else? Is there a point of regulating output voltage? Or is there a point at which we can test for damage to something else?
Must say we are learning a lot (but never enough)
Lachlan
-
There is a resistor string that the computer uses to sense the ac voltage and direct traffic from there... cant remember which ones ( it is mentioned in one of my previous answers somewhere). You will see the resistors coming away from the 240v area... strings of them, one string is for syncing to the mains for charging/ups, and one for voltage reference ... I think that is how it goes.
If I get time I will try to get the numbers for you.
They are very high R, so very high impedance, and easy to change the ac by even dampness or dust..... bit scary really, but seems to work in practice, but they are in the order of megohms..... if there are any problems with any of the resistors in the staircase, then the voltage will move. A shunt across any one of these will drop the voltage, as it sees a higher reference sooner.
..............oztules
-
Cheers for the reply OZ - always something new to consider/learn and always gives me hope!
Will look as suggested - I have my 80 year old father with me at the moment (ex TV and Microwave Tech from the old days) so this is keeping us in constant theory (Ha!)
Numbers would be good if you have the time
We do not use charger/UPS function - only inverting from 24 to 240 - does that make any difference with potential issues? Eliminate any?
Note from basic me only (don't laugh...too much)
On the back of the control board there was a couple of bits of of masking tape taped across the back - I thought "what is this doing here?" and it was .......ummmm "tatty" But decided if that is what they want then so be it.......... it was grubby. Could that be enough to alter V ? Sounds ridiculous I know but...... My father tells me it is to protect the components from people wearing silk shirts! I will take a day (maybe two) to get to the bottom of that! Is this really what is needed to cover back on control board? I am really sorry to sound so thick but.............I am
Lachlan
PS I am becoming at ease with the quantity and quality of the massive amount of stuff I have no idea about.
-
Your right! They are very susceptible to dust. Removed the control board. Removed the tatty bit of masking tape from the back of the control board. Gave the whole board a really good clean - all back in and BINGO 239 to 240 V output in stage 1 testing (via jumper leads). Stage 2 testing to follow shortly (proper hook up and ask it to make 3000 W).
Fingers crossed
Lachlan
PS why was the tape across the back??
-
"PS why was the tape across the back??"....................probably to stop just that.... except it did not work as they expected me thinks.....
.............oztules
-
Just had my latest set of 15kW boards apart........
-
A little excess solder flux splatter.......
-
One of their better boards.......
..........oztules
-
The picture of the tape on the back of the control board looked just like mine except mine was a bit more worn and torn and looked like dust had got in underneath. Should I be re-taping it? (with new stuff!).
I tested the unit fully today and the results were as follows: -
- Start up OK and making 237 to 238 V
- Got the output over 3000 W and no sounds of "snap crackle or pop" or smells that worry you or flashes that do the same and it held steady (we are nearly there me thinks)
BUT (ha ha not so fast) - No fan kicking in. So I boiled up a bit more water...and still no fan. I left the unit on for about an hour or so doing what it normally does (which is very little - about 300 W runs the house most of the time). The day was very hot here getting to about 36c in the shed. The case was starting to feel a it hot so I got worried (I am a coward and did not want a melt down) and shut it down and had a look. The heat sink was hot as was the coil - hot to touch and you would not want to leave hand on for too long. I am thinking something is not switching on the fan.
My backup inverter fan will kick in as soon as I load it up over 2000 W.
I know the fans work because I tested them. I am wondering if I have damaged the heat sensor when I put the 12 V fan in error before (the one I made smoke as mine is a 24 V unit - whoops and lesson learned there).
Is there some way I can test to see if the heat sensor is working? or the control board end? I assume it is the one on the top of the coil - wired to the control board? Other thoughts?
So close and yet......
Lachlan
PS when you learn you learn there is more
-
Hi,
My 2 cents worth;
On my inverters, the sensor on top of the coil has some bare metal pin sticking out where they attached the wires. A screwdriver shorting those pins will run the fan. A set of wires and a toggle switch instead of the screwdriver gives you a switch to turn on the fan whenever you want, irregardless of temperature. Nice to have.
None of this tests the sensor, though.
I use a non contact laser thermometer to read coil temp or whatever, but I haven't tested to see what temp turns on the fan. You need to know that value to be able to tell if the sensor is working or not. With that value in mind, run the unit til the coil temp reaches the value and go from there
Hope this is of some use to you.
dochubert
-
Another thought just came to mind... (and boy did it hurt)
The sensor on the coil nearest the power switch end of the unit (if yours has more than one coil) turns on the main fan. Also, the sensor on the side of the mainboard heatsink (it plugs in near the main fan power connector) also turns on the same main fan. When testing, you will want to be sure which sensor is turning on the fan. If the unit will run with that heatsink sensor unplugged, then you would be sure only your coil sensor signal was controlling the fan. Be sure to monitor heatsink temp while running its sensor unplugged, and not something I would do for too long a time or run unattended.
Maybe you could just run it for awhile to warm it up, then use a heat gun to cook it up a little more to test the sensor. It would tell you it works, but not if it was turning on at the correct temp, Still need that value... maybe someone else here knows what it is.
-
Thanks for that - good thoughts. Had considered the heat gun but thought there had to be another way. I will try just getting the fans to run as you suggested and then move down the line to see what is and is not working.
I always thought that there was some sort of other thing (in addition to temperature sensor) that tripped the fans on. Because when I use more than 2000 w the fans would always kick in immediately. I assumed that at a certain output it came on and the brains for it is near the main chip.
However my fans on this unit always seemed to have a mind of their own. I am sure that the fan in the middle of the coils would come on by itself (which when I look at it seems impossible!).
Your assistance and thoughts are greatly appreciated.
Lachlan
-
No problem. Hope my ramblings helped
On my 48v unit there are three coils and three fans. The additional fans are powered directly off of the dc input (no fuse or breaker) and have no connection to the control board at all. If you trace out the wires back from your 2nd fan you should see that they connect to a sensor on the 2nd coil. Same deal; shorting those sensor pins should turn it on, and you can put a switch on that fan too, if you want.
As far as I know, the fans are only temperature controlled.
My 8kw 24v unit only has one coil and one fan. Not sure if thats good or bad, but it never had any trouble handling loads. The fan did seem to run alot on that one. I suppose the single coil gets hotter sooner, so the fan runs more.
Hope you solve your problems
dochubert
-
I just had a bit of trouble with my middle fan, it would not go off,so i pulled the sensor off to have a look at it, it is held on by some black goo.easy to take off ,checked it out it was shorted all the time. so i put it in the freezer to see if it would close it, that worked.put it back in and its ok for now so i ordered some more because its liable to do it again.The middle fan sensor is a 45 c degree one(KSD-01F 45C NO Temperature Control Switch Thermostat),as this one comes on first i would say the one for the big fan is probably a higher degree sensor, they sell them in 50c and 70c as the nearest so it could be one of those.
john
-
The big fan is dynamically controlled via the temp sense and the computer.
I have never nailed down how and why it comes on, but it has tacho incorporated too.
The center fan is a dumb fan via therm switch as you discovered, but it usually also has a sensor that the computer can see, and will shut down the unit if too hot.
...............oztules
-
The more I learn about the fan controls and their apparent inconsistent (read; unreliable) performance, the more I am inclined to follow Oztules lead and put in place independent fan control. I remember reading that Oz built his own, but I'll probably buy a temp control card off ebay (dirt cheap) and just parallel its switch with the existing inverter's sensor. That way, whichever control wants the fan on first wins, and my inverter stays cool either way. Maybe even two cards; one for heatsink temp and one for coil temp. Two temp control cards with sensors for less than $10US total is too cheap NOT to use.
I have noted in the past that some days the inverter seems to run hotter than other days, but without actual temp readings, I just thought it was my imagination. Not so sure now... The cards I'm looking at will also give a constant temp reading, which is a nice plus.
I guess that just leaves us with the question of where to put the setpoints. On at 40-45 maybe and off at maybe 35? Oztules, what setpoints did you use and did you sense at the heatsink or on a coil, or both?
dochubert
-
bought 3 of the 45 degree sensors the same ones that are on there, and only 1 of them worked.gee no wonder they have so much trouble with the fans.
-
I bought a five pack of those too for spares. Guess I'd better test mine also.
Thanks for the heads up
-
I got those temperature control cards and finally got around to hooking them up. Got 3 but only have the 2 in use so far.
[attachimg=1]
The top unit is hooked up to the heatsink. Just tucked the sensor into the fins.
The middle unit sensor is on the first coil near the stock thermal switch. Loosened the center bolt on the coil and tucked it between plate and rubber
So top temp reading is heatsink and middle is first coil with third unit just reading ambient. (might later have it sense my third coil temp as no switch on that one)
The larger middle reading on each unit is sensor temperature, with the left number the high setpoint or kickin point and the right number is the kickout. Buttons below each for adjusting settings. Easy
Good thing its easy, as the directions were a bit cryptic; ::)
[attachimg=2]
Anyway, just paralleled the two control unit contacts together and then paralleled to the pins of the thermal switch on the first coil or to the toggle switch if you have one installed, whatever is easier. That way all the stock fan controls still operate (hopefully) but now these adjustable controls ensure main fan operation when its needed, without having to change or modify the fans themselves. You could, of course have one card operate the main fan and another run the second fan but I went with simple and easy
I won't be sure of setpoints until I get some hot summer days for real testing but so far so good
Power for the cards is 12 or 24vdc. Since this is all for testing, I just have wires hanging out of the case. When its time for more permanent hookup, easy enough to route wires thru the endplate or notch the clamshell
Sorry about picture quality. Had trouble getting the lighted numbers to show up. As a photographer, I make a good electrician