Jump to content

My PV System

Recommended Posts

So last week I got a 3600W invertor for such a good price, I had to buy it.

Granted it is a modified sine wave output and DC input is rated at 24V.


I am still struggling with the monkeypality on when we can have NET metering installed.

Now I find out they replacing all the prepaid meters as the existing ones is EOL.

If only they waited untill both options was available and the user could choose :(


The plan is now to power my pool pump and my irrigation system pumps from the solar system.

I am also not planning on installing a huge battery bank, as I do not really care to run the pumps if the sun is not shining.


For now, here is a pickture of what I am planning.

Does it look ok?




What I am still trying to figure out is how to disable the pumps running if there is not enough solar energy available.

Was perhaps thinking about a small 12V PV panel driving a relay. This relay can then enable/disable the pumps.

Any idea's is welcome in this regard.



Link to comment
Share on other sites

Ok, so I can confirm the following....

PV panels is Tenesol 235W x 6

Invertor is a EPS3624 modified sine wave, I hope it would work ok with my motors?

Solar Controller is a 60Amp Microcare with programmable relay. The relay will be used in day/night function to switch off the loads if the sun is not shining.

2 x 100Ah battery. This is only there to supply enough oomph to start the motors.

Link to comment
Share on other sites

Right, so collected the solar panels Thursday afternoon.

Let me tell you, they feel much heavier and is much bigger in real life than what it looks like on paper!

Luckly the pallet fitted in the rear of my bakkie.


So Friday afternoon was spent fixing the rails onto my roof.

I used P2000 from Cabstrut with their spring nuts. It worked great.

For fixing the outside corners of the panels I made my own brackets.


This week I will start looking at the external and internal wiring and mounting some of my geat.


P.S. I am a bit worried about the invertor driving a motor.

Spoke to a supplier of these drives and he said it is NOT modified wave, but rather square wave :(


My mounting system installed.



All 6 panels mounted.


Link to comment
Share on other sites

As stated above, I did NOT use solar mountings, as they were too expencive.

I used P2000 from Cabstrut. It is more often used in the Electrical field, sometimes known as Unistrut.

It is used as supports for cable racking systems.

It is fully hot dip galvinised, so should last a long time.


The straps I used is from the building industry.

It is normally used to fix roof trusses to concrete walls. It is about 25mm wide and 1mm thick, with small fixing holes already drilled.

I got a 10m roll from Builders Warehouse.


Just had to lift 2 tiles up to get to a roof beam, screw the strap down with 2 screws and put the tiles back.

Then I put the Unistrut down and used some self-tappers to fix the two together.


I hope to have some time over the week-end to start with the wiring and mounting the remaining kit.

Link to comment
Share on other sites

  • 2 weeks later...

Ok, so everything is about 90% up and running.


Here is an update to my wiring diagram. Changed the 24V DC wiring so that the invertor runs off the fuses and the Solar controller goes directly to the battries.

Seems like neither the solar controller or the invertor likes to run directly from eachother.

Also installed a contactor on the 230V input to the invertor, which is controlled via the solar controller.

As the solar controller have way more programmable functions than the invertor, it seemed the logical choice.

So now my pumps will at least keep running if the sun is low or the battries starts getting low.

Also the charger on the invertor is switched off, so when the battries is low and there is no solar power, the loads will then run directly from eskom power.




Talking about battries. A good friend of mine managed to score me 8 of these beauties....


They heavy as hell and still manage to keep a charge.


now the bad :(

This flipping EPS invertor is supposed to be modivied wave, but in actual fact it is a square wave output.

This is now causing my pumps to start overheating after about 2 hours of running.

Even my UPS is complaining so I can not even run my computers from the solar either.


For now just running the irrigation pumps. At least these is controlled by a PLC so I can now play around with the running times too.


Link to comment
Share on other sites

  • 1 month later...

Just a report back.

So far I have done 6kw max in one day.

Yesterday when it was a bit cloudy I ran up to 5.5kw for the day.


I have the aux relay set on the charge controller as Load Shed and Day/Night switch.

Basically it means if the solar panel voltage is lower than the battries, the relay will switch.

Also if the battery is below a set voltage, the relay will switch.

Only problem I currently have is that the relay is switched repeatedly in a short time, while the charge controller seems unsure of what to do.

I will ask Micro Care if they can include a type of timer as it is hell on the load to switch repeatedly.

Link to comment
Share on other sites

  • 1 month later...

RIght, just some feedback.

System is running pretty well to date.

Avg around 5.5kwh per day on the 2 pumps.

Where my water bill used to be over R2k per month, it is now around R200 :D

Also had no effect on electricity, as both pumps is running off solar.

As it is working so well, I have now decided to include my pool pump into the system.


To do this I have added another 2 x 235W PV panels to the system.

Still waiting for my 4.5kW pure sine wave invertor to arrive :(

The EPS unit is ok, but it does not work so well with flurecent lights and lighting that uses ballasts.

Also tried running my computer UPS from it. That was a bad moove, as it blew the UPS :(


As the 4.5kW invertor is 48V DC, I would have to re-configure my battries to cater for that. Luckly I have enough battries to do that.

Next problem is the new invertor does not have a built-in load relay or charger, so if the battries is low, I need to build my own change-over to mains.


I am busy sorting out all these details and will post and update as soon as I have it sorted out.

Link to comment
Share on other sites

  • 3 weeks later...

Ok, just an update for now.


Have installed the new 48V pure sinewave invertor on Sunday.

My change-over contactors is also installed and running very well.

Basically if the battries is low or the invertor trips, all systems will be powerd from Eskom automatically.

I have now added all my PC equipment, that is running from a UPS, onto the system as well.

Did 7.7kwh on Monday alone :)


Now I still need to figure out how I am going to run the pool pump and it's timer.

I need to be carefull here as I do not want to drain the battries too much or have the system switch to Eskom power every couple of minutes.

Link to comment
Share on other sites

Right, so have connected the pool pump on Friday, but it was too late in the afternoon to run it.

On Saturday and Sunday with the pool pump and irrigation running, did 11kwh on both days, so very happy for now.


Problem I now have is that if one motor is already running and the 2nd one starts up, the voltage dips a bit.

This dip is causing my contactors to fall out.

Perhaps some re-wiring of my cicruit or a timer would help? Not sure.....

Link to comment
Share on other sites

Ok, for now the contactors is sorted. Everything was mounted on some marine plywood.

The 2 main contactors is big, so if they switched, the smaller contactor would shake and loose contact.

Mounted the smaller contactor in a diffrent location, so that problem is sorted for now, I hope.


With my PV solar totalling 1880W (235W x 8) and my load over 2000W, I am getting some other problems now :(

As my battery bank is very small, only 200Ah and the load very big, I find now that my system keeps switching between Solar and Eskom around every 10min.


Options would be to schedule the motor run times better, but then I will get less time on the pool pump.

Increase battery bank capacity, but as I got the existing units free, I would not like to spend more money on the battries. Also only running load during the day.

Increase the PV capacity. As my roofspace is getting smaller, I might have to mount these new panels in a diffrent location.

Anybody know how this would affect the MPPT controller? I would hate to have to go and buy another one.

I was thinking of adding another 4 x 235W panels.

That would give me 3 strings of 4 panels each.


That would get me to 2820W total. At 48V the amps would be 58.8 which is still within the MPPT controller limit.

4 x 37V (Voc) = 148V. That is also still within the limit of the MPPT controller. Max is 150V.

Link to comment
Share on other sites

148V out of 150 VoC, feels a bit close for comfort imho. Regarding MPPT, I read an interesting opinion the other day: If you don't have a lack of space, it often makes more sense to go with a cheap charge controller, and spend the cash on more panels. That more than makes up for extra energy a mppt controller adds. Buying a throwaway PWM controller might be okay... :-)


The rest of your numbers are close to what I aim to achieve over the next few years. When Eskom asked for 25% increases, that was the "vonk in die kruitvat". Even though they only got 8% (on average), the battle lines were drawn. I set out on an ambitious plan to replace 50% of my consumption with renewable over the next 5 years, and at the time my daily consumption was a whopping 40 units a day (large house with an attached flat that's rented out), so the initial aim worked out to around 20 units a day. I'd have to put up 20 x 200W panels just to make that much on a sunny day!


I have since cut my average consumption to 32 units a day, making this goal slightly more attainable. If only the women didn't take a second shower in the mornings, I'd spent no electricity on heating water...


In any case, I also started out with a small 24V battery bank (2 x 100Ah 12V batteries). At a 50% DoD it means storage of around 1 kwh... and even just going near 1Kw would easily pull down the running voltage to below 24V, a clear indication that they are working too hard. I've come to the conclusion that you have to better match the bank to the array. From the kit systems I've seen advertised, it seems you should size the bank to hold about twice as much charge as you can generate on a sunny day, so if you make 11Kwh a day, storage of 22Kwh, at 48V is 460Ah, that is you need about double what you've got.


Now the experts among us would point out immediately that doing 1Kw with only 24V x 100Ah discharges those batteries faster than a C3 rate! Much better to aim for around C10, that is the highest load will discharge them in around 10 hours. But man... batteries are expensive! It is so hard to get this bit just right and not go bankrupt. I suppose this is why the high discharge current on AGM/Lead Crystal batteries is so attractive.


Right now, I have a 450W array with 240Ah x 24V storage. That works for now, even if it only makes 2.5kwh a day. It's a start :-)

Link to comment
Share on other sites

Regarding different directions on the same mppt, I read somewhere that this is okay when your panels are mounted east-west, I suppose because the sun goes across them in a straight line and the array is always balanced. But I don't think it would work if you had half your panels pointing North. I'm contemplating the same thing, my house is oriented west-north-west, meaning that the side where the panels are points north-north-east. This means that in the afternoon the panels end up shading as early as 3:30pm, as the sun moves too far west. I'm considering adding some panels on the west, but I'm pretty sure it won't work to put them on the same mppt controller.

Link to comment
Share on other sites

Ok, first the MPPT.

I sent a mail to Microcare and according to them it will NOT be a problem having diffrent legs of a string face diffrent directions.


Here is my shopping list for mounting the panels to the roof.

Planning two sets of four panels, but only installing 4 panels for now.


Hardware can be ordered from any electrical supplier. Part numbers from Cabstrut.

4 x P2000 (5m sections, 1mm thick)

20 x SN110/M10 (m10 spring nuts)

20 x P1063 (Square to hold 2 panels on the sides)

10 x P1026 (90 degree angle, to hold outside panels)

20 x M10 bolts, 80mm long


This time I will try and document the installation step for step.

Link to comment
Share on other sites

I've been deep in thought most of the morning (while dropping off kids at school, etc) why it wouldn't matter if you had two different legs on the same controller, and I think I might have it figured out.


What the MPPT controller does is find the optimum power point, which it does by finding the V for which it gets the highest P=I*V, which if I recall what the current charts look like means finding the highest V before I starts dropping off. A bit like a surfer finds that optimum point on a wave.


Now if I have two legs on the same controller, the effect would be that the controller will look for the optimum point across both legs. If the legs are similarly sized, and the panels have similar power curves, then one of two things happens. Either both arrays gets similar sunlight, and the optimum point across both legs would be pretty equal to the individual optimums anyway, or one leg will get more sunlight than the other, which will cause the optimisation to essentially optimise for that leg (getting the right V will essentially cause the biggest difference for that leg, so that leg would essentially drive what V is chosen), running the other leg slightly too low for optimum performance, but because the other leg is lower anyway, the losses might be low enough that one can ignore it, especially compared to the cost of another mppt controller.


I think that for best results one should use separate controllers so each one can run at it's optimum point, but the question would be how much more you would actually get out of such a setup (I'm guessing not much) and compare that to the cost of the extra controller. I suspect that in terms of bang for buck, having the two legs on the same mppt controller would work out even better than adding a cheaper controller for the second leg.


The only issue I have with my setup, is that my 40A controller won't be able to handle more than 6 panels (my array is built from 150W panels), so I'd need a second controller for my second leg anyway :-)

Link to comment
Share on other sites

I really should put all my thoughts in one post :-)


I'm returning to the 148V VoC you mentioned. I worry that that is too high, that's like less than 2% margin. The way these controllers work, if I understand it correctly that is, is that they push the voltage "off the cliff" when they want to limit the incoming power, in other words, they essentially push the voltage up so that the panel operates at the point where current drops off dramatically. That means your controller would have to push the voltage up to 148V when it wants to stop charging, and on a semi-cloudy day when reflection off the right colour of cloud causes inbound insolation to be more than 1000W per square meter (those are the days when I see my 450W array operate at 510W), I have a feeling you may just be steering too close to the shore to avoid all the rocks.


Though asking an expert is likely the way to go. I'm just thinking out loud here. If it was me, I'd feel safer with 4 strings of 3 panels (though I can imagine that could cause arrangement issues on your roof), for a more manageable 111V Voc.


In addition, my local installer said I should try not to let the VoC be too much more than twice the "full" voltage of the batteries. 24 cells at 2.45 volts is around 58V, so you want your VoC not much higher than 116V or thereabouts.

Link to comment
Share on other sites

Ok, for now I will take the 148V under advicement :)

My initial plan was to have 2 strings of 6 panels each. Just right for a 3kw grid-tied system.

This is also the reason I am unwilling to spend more money on MPPT controllers or battries.


Currently I have 2 x 4 panel strings and adding another string of 4.

Up to now I have not had any problems, but if I do get some, I would have to go to 4 strings, each with 3 panels.


I like your explenation regarding the diffrent strings on the same MPPT, thanks.

Link to comment
Share on other sites

Ok, so I spent another week-end on the roof contenplating which way to go with my panels.

Finally decided to go with 4 strings of 3 panels each.


Basically it came down to the Jinko and Tenesol panels not matching 100% in performance if I mix them in the same string :(

So now it will be 6xTenesol and 6xJinko. Could be a nice match-up.

Jinko to do the early morning sun to afternoon, Tenesol to do late morning to late evening.


So to make things legal and easier for me to measure and test, I installed a solar combiner + fuse box on the roof.

The cable is a 4 x 6mm trailing cable, so more than enough insulation and protection.

Each 6mm core should be good enough for 40Amp, so I'm covered there.



Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

  • Create New...