Custom 6x20 watts solar setup with MPPT

Whenarewethere · Guru

RE: Custom 6x20 watts solar setup with MPPT

oldtrack123 · Guru

dorian wrote:
T1 Terry wrote:
It appears you are now in conflict with another member of the MPPT choir
Quote:
For the benefit of those with an open mind
https://www.morningstarcorp.com/parallel-charging-using-multiple-controllers-separate-pv-arrays/
As a further explanation , a true mppt controller has two entirely separate operations
[a] the MPPT section, which operates independent of battery voltage ,tracking the panels peak power point continuously..It does NOT sample battery voltage when doing that. It simply is not concerned with battery SOC voltage
It then uses a 2nd stage to convert that MPPT DC power into a suitable DC power source for charging the battery [in simple words, that section becomes a PWM DC charger]
That second stage,PWM ,the battery charging stage, is independent of the first stage[MPPT]
Yes, that puzzles me, too. AFAICT, an MPPT controller does not have separate hardware for MPPT and PWM modes. Instead it uses the same hardware for both modes, but switches to a different routine in the firmware. In MPPT mode the duty cycle regulates the panel's power output, whereas in "PWM mode" the duty cycle regulates the battery voltage.
It's a kit, not a commercial product, but the software is instructive:
http://web.archive.org/web/20140626064134/http://www.timnolan.com/index.php?page=arduino-ppt-solar-charger (project page)
http://web.archive.org/web/20140626064134/http://www.timnolan.com/uploads/Arduino%20Solar/ppt.pde (software)

Hi Dorian

That is just a simpler /Cheaper Version of mppt reg similar to most on the market using just the DC /Dc stage.

It would be interesting to know the gains between that type & the much dearer one

But whichever one is used ,the output is basically a DC/DC charger

Noelpolar · Veteran Member

Good idea with the aluminium bit to take some heat out of the butter area.... will knock something up myself.. thanks for the idea.

I wouldn't worry too much about the wafflers..... most of us just skip on to the posts that interest us. Always nice to see approaches to situations that have a well defined scope.

Whenarewethere · Guru

Noelpolar wrote:
Good idea with the aluminium bit to take some heat out of the butter area.... will knock something up myself.. thanks for the idea.

I wouldn't worry too much about the wafflers..... most of us just skip on to the posts that interest us. Always nice to see approaches to situations that have a well defined scope.

This is the Schneider aluminium baking tray 4028574815277 bar code: https://www.nisbets.com.au/schneider-aluminium-baking-tray/gt145

Really happy with it & you do not need to put a fan inside the fridge, the passive cooling works a treat. This area was useless before. If cutting it yourself put some masking tape on the surface to minimize scratches & round edges of aluminium sheet. Put some tape on the back to avoid marking the evaporator.

Aussie1 · Guru

Sorry top interrupt here fella's but just had to say how much I love all those pictures and graph's. Not to sure about all the "verbal diarrhea" though !

Whenarewethere · Guru

I replaced the 75/10 controller with a 100/20 controller to see if I could squeeze a bit more out of my setup. I also added a pile of 14 x 14 x 10mm adhesive heatsinks to keep the controller as cool as possible. They have pretty much doubled the surface area, & I have a handful more I could stick on if needed. So when the battery voltage is a bit low now getting up to 11.3 amps under ideal conditions. The controller is also bolted onto a sheet of aluminium so gets a little more benefit here.

T1 Terry · Guru

Often wondered about how efficient a stick on heat sink would be. Surely the stick on bit would act as an insulator stopping the heat from soaking into the added heatsink and actually blocking the surface area that was originally available. Wouldn't a small computer fan mounted in a way that would move air across the heat sink fins an controlled by a thermostat be far more efficient?

T1 Terry

Whenarewethere · Guru

I bought a Noctua 40mm x 10mm thick fan for that purpose. They are quiet, long life & really efficient at 0.05 amps. It moves just enough air for the task. Haven't installed it yet as I've been busy with another project installing acoustic insulation. Also it could be adding a fan in my little setup there is no real gain as I am using up another 0.05 amps to run the fan & I don't gain that back keeping the controller any cooler. I might wire it in so if it is really hot weather I can turn it on as a back up option, but in general I am trying to do everything as passively as possible. This new controller is running cooler than the previous one so that is a good start. Some heatsinks stuck on the front face of the cover might be of more benefit as this side is warmer than the back. But I am getting down to the level of splitting hairs!

Whenarewethere · Guru

Victron MPPT 100/30 controller:

Victron quote to question: 'Its a liniair derating from full power at 40 degr C to zero at 70 degr C'

Probably the same situation for similar products. At 40°C it runs at its maximum design efficiency. Above 40°C it is downhill.

I will install a fan as the car can seriously hot, & together with heat generated by the controller, even though it is not much, anything I can do to keep efficiency at it's maximum is worth the effort.

T1 Terry · Guru

Whenarewethere wrote:
Victron MPPT 100/30 controller:
Victron quote to question: 'Its a liniair derating from full power at 40 degr C to zero at 70 degr C'
Probably the same situation for similar products. At 40°C it runs at its maximum design efficiency. Above 40°C it is downhill.
I will install a fan as the car can seriously hot, & together with heat generated by the controller, even though it is not much, anything I can do to keep efficiency at it's maximum is worth the effort.

The MPPT controllers we use for the off grid installs are rated at 70*C before derating, the same as their inverters. Unforunately gear designed in cooler climates like the USA and Europe don't believe it ever reaches 40*C anywhere in the world except the middle east (that was in an actual statement made by the lead Morningstar engineer on an MPPT webinar). Unless you can get the ambient temp below 35*C the fan will not be able to keep the controller internal temp below 40*C. Australia has the best solar in the world, yet these companies don't seem to care that their product isn't really suitable for Australian conditions. It is like they simply don't recognise us as a worthwhile market.

I'm afraid, if a company treats Australians with that sort of contempt I'm not likely to support them. Their products just are not suitable for use in Australia.

T1 Terry

Whenarewethere · Guru

The middle east even gets refrigerated parking meters so they cope with the heat. In southern France a few years ago we had week of 43°C, similar in Italy on another trip.

There is not a lot we can do about ambient temperature but can put equipment in places where it gets good ventilation & or forced cooling to keep it as close to ambient as possible. Or even a wet towel nearby for evaporative cooling which works a treat in low humidity. If you have spare water! (We were frozen getting out of the pool in 47°C heat in Mildura due to the fast evaporation)

At the other end of the spectrum we have added extra insulation to our home 80 litre hot water tank which had a heat loss of 1.6kWh per 24 hours (Australian Standards) to 1.0kWh per 24 hours. (leftover offcuts from roof).

Jaahn · Guru

Hi smile

Just to put some context on the temperature statements, for other people reading who may be dazzled by the numbers. Victron says in their specs for the Smart Solar charge controllers,

"Operating temperature -30 to 60deg C( full rated output up to 40degC ).

I have mine mounted inside my MH near the battery, vertically, in a passive ventilated enclosure which would be a normal mounting. I do not expect my MH interior will go above 35deg probably worst condition, or I would move it to a cooler spot. So I am happy to use my Victron. As always, I never run my gear at the max rated amps but buy a bigger unit and I expect and get a long life out of them and good performance. Mine is an older 75/50.

Just putting a positive side to the doom and gloom.

T1 Terry · Guru

I guess that is a way to get the very expensive product to work, massively over sizing the controller and/or building all sorts of cooling devices to get a piece of equipment you paid good money for to work as expected, but isn't it something the manufacturer should have done in the very beginning. P*ss poor that you are expected to pay so much for one of these "brand name" things and then have to do a heap more work to get it to actually do what you wanted.
What about their inverters? Really in the upper end as far as price, but if you want to run the air con when you really need it, do you mount it inside so it can stay cool too .... only works if you turn it on before it actually gets warm inside the RV. The other method is to rate it at the 70* rating, not the nonsense VA rating that just misleads the average buyer, like a 3000 unit that is actually only 2500w if you can keep it cool, otherwise it is less than 2000w continuous ..... compare the cost against a Projecta IP2000 that can actually maintain that 2000w above 40*C ....

I'll stop the rant now, it just really peeves me when a big brand name product is flaunted but no one mentions the extra work they had to do to get it to work, leaving others to get sucked in the same way they were sucked in, I thought these forums were aimed at helping others avoid the traps ....

T1 Terry

oldtrack123 · Guru

It seems that the rating methods & details required under the relevant standard are confusing to some members.

All AC power sources are rated in VA or KVA.

All AC inductive loads are rated in VA or KVA ,except for motors using "HP" but even those do not operate @a PF1.

The reason being that apart from pure resistive loads, all other AC loads have a PF less than PF1.

Knowing the PF & the rated voltage the actual Amps .drawn or available can be calculated

Since Amps are the key factor in heating of both power sources & loads "IxIxR", using Watts is irrelevant as it only applies @PF1, but the Watts equivalent can still be calculated if the actual PF is known .

Ac motors' PF varies with the load & a PF meter is needed to determine the actual PF @ that load

-- Edited by oldtrack123 on Tuesday 19th of February 2019 10:22:00 AM

Whenarewethere · Guru

Jaahn wrote:
Just putting a positive side to the doom and gloom.

Years ago I ran a handful of dual Xeon computers with 130+ watts per CPU. Apart from the room getting hot in summer, about 38°C with ventilation. The computer fans throttled up to the maximum. One computer sometimes shut down.

I replaced all the CPU heatsinks with radiator heat tube heatsinks. The computer fans stayed on the lowest speed as the new heatsinks were far more effective at removing heat from the CPUs, & I could actually work in a quiet environment, only the heat was an issue!

There was about 2400 watts of computers running flat out 3D rendering week after week via my 3.0kVA extended battery UPS.

Hence why I have been putting in a bit of effort to squeeze the most efficiency out of my solar setup!

denmonkey · Senior Member

Afternoon When,

I'll bite mate :)

Like the set up and the folding aspect of the panels.
They seem to provide a good compromise between size and power.
Nothing better than a bit of DIy to get exactly what you need.

My fridge has been suffering in this heat and I'm also considering adding an additional fan into the cabinet to assist in in cooling.
With such high ambient temps in the car, though it won't matter how much hot air I move, but it's worth a try.

I've also got some water cooling radiators from some old PC builds which I nearly through out the other night. Perhaps, I could work these into the system some how.
I had considered trying to make them worth with my 3 way fridge.

I'm also working with an mppt controller for my van panels.
The projecta dc/dc/solar in the car has not impressed me from a solar perspective.
I think the higher voltage panels and a deciated controller has been working much better.

I've also noted the siginificant heat in the panels when moving them. They were literally too hot to touch.
Only cooling them with water, there was an immediate improvement on the power output.
Interestingly the heat was very focus on the aluminum frame.
Perhaps the frames are responsible for a lot of the maintained heat in the panels rather than the glass etc.
To that end, the idea of the raised panels with a small pc fan on the back it of interest to me but its further down the list of projects and things to do.
thanks for sharing.

cheers Brett

Whenarewethere · Guru

We have wind deflectors on the car & leave the windows open a touch in really hot weather, make a big difference with a bit of air flow.

Often we don't run the fridge when the car is closed up, & cover the fridge with doona & sleeping bag. Makes a big difference in slowing the fridge warming up.

Also have cut up some windscreen sun shades for every window so we can put them in which ever side needs the sun blocked or all windows if needed, they work really well. Also great at night in cold weather.

The waeco fan I pulled out of the fridge is temperature sensitive & will run faster when the temperature is hotter. But the sensor is up wind of the heat generated by the fridge, stupid!

Also some of the fans air was recirculating around the side of the fan so I put a baffle in to stop it.

The latest fan I installed is actually reasonably quiet once it is in place, & the 28 litre fridge's compressor is a fair bit quieter compared to all the lager models as they have a larger compressor which is louder. I listened to them all in a shop.

The Noctua fans are expensive but are well made, I replaced the 3 80mm fans in my UPS 10 years ago & they run 24/365 & they are as quiet as the day I put them in.

PS I stopped at LW11.6

T1 Terry · Guru

The reflective window shields are an excellent way of keeping the inside temperature down when the vehicle is not moving. They work best on the outside of the window because they stop the glass from heating up heating the air that passes the window. Keeping them attached to the outside of the window is a task all of its own, but the ingenuity of the average handyman has come up with some great methods and it would be good if members could share their ideas here, as long as Jonathan doesn't mind of course

T1 Terry

Whenarewethere · Guru

Latest addition to my 100/20 controller to remove the last bit of heat from the controller. Have been in Tassie for a month so didn't need to connect the solar panels. Preparation for hot climate. I should space the heatsinks but they don't stick to the plastic quite as well so stuck the heatsinks next to each other for support. Next task is to think of any others ways I can squeeze more than 11.3 amps out of the controller. Maybe another stint in the Pilbara or swap all the wiring to 4awg!

But seriously, very happy with the setup & I can simply get on an enjoy my travels & offer cold beers to friends after long walks which is what it is all about at the end of the day!

T1 Terry · Guru

Under neither that plastic cover is a blue gel like substance. No idea if it is a heat transfer type material or a heat insulating material. Its major purpose is to stop inquisitive people having a look at what's inside. I think a fan added to increase the airflow across the heatsink on the back would be enough to stop it from shutting down due to overheat. All the heat generating items are attached to the heatsink. A water cooled heatsink would move 3 times as much heat energy as an air cooled heatsink, but only use distilled water, any other additive will reduce the cooling ability of the water.

T1 Terry

Whenarewethere · Guru

I have pretty much got the heatsink side down to the point that it is enough surface area. I could have used a spare heaksink tube radiator off one of my old xenon CPUs but it's not worth the effort milling down the fins on the controller & designing a new mounting system.

The was a bit of heat out the blue cover of the controller, none around the sides, so that's why I chucked on the 9 heatsinks.

Next task on the list: back into upgrading the last of my acoustic insulation!

Whenarewethere · Guru

Just got around to short circuiting the six 20 watt solar panels. They are set up as 2 in series x 3 sets with no diodes.

Max voltage was 43.06 but typically around 41.6 volts.

Maximum amps was 4.175 (16.7amps four times through clamp meter / 4), 4.1 amps one time through clamp meter. Typical it was 4.0 amps.

Done at 12.30 & the sky was probably about 95% as good as it will get. Where was a touch of haze.

I also tried without watering the panels first, then with water for cooling the panels & didn't really see much difference.

T1 Terry · Guru

21v open circuit per panel 2 panels in series is 42v, an average for a pair of 12v nominal panel connected in series. When panels are connected in series, the voltage adds up but the amps stay the same.
I hate to burst your bubble, but 4 amps short circuit current shows it is not a 20w panel. This is the specs for a 20w panel from Low Energy development www.ebay.com.au/itm/MOST-POPULAR-New-20-Watt-12v-Polycrystalline-Solar-Panel-MC4-Plugs-NEW-SIZE/322339354506:~3YAAOSwcUBYN8Ub the open circuit voltage of 21.6v is roughly the same, but the 20w panel only has a short circuit current of 1.27 amps.
This 60w panel www.ebay.com.au/itm/60W-12V-Solar-Panel-Kit-Mono-Generator-Caravan-Camping-Battery-Power-Charging/231133326316%3D1&hash=item35d09eabec:g:QsMAAOSw92Jdw6yx&enc=AQAEAAACUBPxNw%2BVj6nta7CKEs3N0qV42ZsndMD83ackMm39AByyDjrzkOal4jx9yiaGUFxtxnrZ8KaKo8A6W4Tq6Pa3DqW31OkEZIOFaXanmOsR0Kz88a0GuS5dv3pE%2BaNLBfjj9FxULXogz9e4LtMmYO88k6pdn%2FT14gCR2urpz4K0USkZoco64YKqCa6kn3oT25UFMrQsmaW6zq8nTNkhdS7yUlEqI3KPzhubzbz6%2BXky2uR3SouvtYhzfeM0OaNqidjblLZ%2Ba6vA8LZ8JLI3HW3GuF4MbWknEotse37iwSv7H%2Bu90yA1qHxb3KVWsiaq0CgAYrTpdKfIuJmyFKvNN8h%2FllOye%2F98X7BcrqRXJ3HhVy4aGO%2FylZVlmfYWs%2F4LMBqjgbjtiBIRl3hvcohnd7yT%2FPdUgzyef5%2FwoIdQl4I%2FJNNA%2BQ9sAcS%2B%2BbQp2zGdkMIu6c4miVXLqC59Tsm2vIWg8f9cYkfdW43n9Jc1DrwGbCEtM5tFNPuAFCanyHdpZtkHN9D9tn3qIh%2FtnzEYKF62zPFj%2FJYIvSmVReWrCscyW2UN0TACNkPZQwIneAy8hhKVvPhwXGJ7OgCuwoc4fk3%2FY5L7gZMba3TgRsRlzfLJI1AVN2xxqw38Y6bWHwpOjGDPIOe0B2S5rYnSGSL17cC9q%2FF3AtKQD%2FbhrnGa%2BBLd93q5NplhTQtfHt%2Frkhs3mmBBtRnlb8Rq8ocM%2B7wUvVvrLQ8TZxogyKPB3WycGvrncoMtaVMCSvDTNrA3TEG1pSRiEidbr%2FDZzwdt3TWBwgLhUw0%3D&checksum=23113332631651d2721ff2d14b88942339048a927540&enc=AQAEAAACUBPxNw%2BVj6nta7CKEs3N0qV42ZsndMD83ackMm39AByyDjrzkOal4jx9yiaGUFxtxnrZ8KaKo8A6W4Tq6Pa3DqW31OkEZIOFaXanmOsR0Kz88a0GuS5dv3pE%2BaNLBfjj9FxULXogz9e4LtMmYO88k6pdn%2FT14gCR2urpz4K0USkZoco64YKqCa6kn3oT25UFMrQsmaW6zq8nTNkhdS7yUlEqI3KPzhubzbz6%2BXky2uR3SouvtYhzfeM0OaNqidjblLZ%2Ba6vA8LZ8JLI3HW3GuF4MbWknEotse37iwSv7H%2Bu90yA1qHxb3KVWsiaq0CgAYrTpdKfIuJmyFKvNN8h%2FllOye%2F98X7BcrqRXJ3HhVy4aGO%2FylZVlmfYWs%2F4LMBqjgbjtiBIRl3hvcohnd7yT%2FPdUgzyef5%2FwoIdQl4I%2FJNNA%2BQ9sAcS%2B%2BbQp2zGdkMIu6c4miVXLqC59Tsm2vIWg8f9cYkfdW43n9Jc1DrwGbCEtM5tFNPuAFCanyHdpZtkHN9D9tn3qIh%2FtnzEYKF62zPFj%2FJYIvSmVReWrCscyW2UN0TACNkPZQwIneAy8hhKVvPhwXGJ7OgCuwoc4fk3%2FY5L7gZMba3TgRsRlzfLJI1AVN2xxqw38Y6bWHwpOjGDPIOe0B2S5rYnSGSL17cC9q%2FF3AtKQD%2FbhrnGa%2BBLd93q5NplhTQtfHt%2Frkhs3mmBBtRnlb8Rq8ocM%2B7wUvVvrLQ8TZxogyKPB3WycGvrncoMtaVMCSvDTNrA3TEG1pSRiEidbr%2FDZzwdt3TWBwgLhUw0%3D&checksum=23113332631651d2721ff2d14b88942339048a927540
shows the same open circuit voltage and only 3.63 amps short circuit or 3 times that of the 20w panel shown above. A rough back of the envelope calculation says these are around 65w panels.
If we use the output for the 60w panels they say their maximum current output is 3.39 amps @ 17.7v 3.39 x 17.7 = 60w, so that adds up about right, but in real life if you get much more than 2.5 amps out of these panels you are doing well. With 6 panels that would mean you should still be able to get 15 amps out of them, so it is quite possible they have deliberately derated them to cover their otherwise poor performance. I see that quite often with panels a mob is clearing out, they advertise them as 175w panels but their physical size is the same as a 250w panel. No one complains if they get more than the 175w they bought so everyone is happy.
I'm guessing this is the same in your case, you bought what you believed were 20w panels and they deliver a lot more than 20w each so you are happy and got more than what you paid for and the system does everything you are asking from it. You could try connecting 3 panels in series to increase the open circuit voltage to around 63vdc, well within the 100v the controller can handle but the higher voltage might reduce a few of the cable and connection losses resulting in a bit more output from the controller.

T1 Terry

Don't ya just hate it when the link adds all that other stuff you don't want evileye

-- Edited by T1 Terry on Friday 8th of November 2019 04:59:36 PM

Whenarewethere · Guru

I've wired the panels back to parallel once & back to series again. I'm not doing it another time, besides the Kyoritsu 2046R clamp meter confirms the Victron output. I use the clamp meter on holidays to do random checks & I can't complain with what is going into the batteries.

This photo from the top of the thread when the panels were all in parallel showed 7.5amps going into the MPPT which corresponds to the panel meter.

Probably 2 controllers, clamp meter & panel meter are all faulty.

Whenarewethere · Guru

656 x 296mm the $25.99 20 watt panel.

638 x 278mm the Projecta 20 watt panel.

The Projecta's area is 8.6% less.

The Projecta was $42 a panel when I bought them, but for my requirements size, especially the 278mm width was far more important than price. Not having the panels stick up higher than the spare wheel prevents the forward facing panel being covered in bugs & dirt so I didn't need to have a cover.

My 6 panels cover an area of 1.064m2 (not including hinges)

dorian · Guru

AFAICT, the efficiency of these panels is about 10%.

20 watts / (656 mm x 296 mm x 1000 watts per square meter) x 100 = 10.3 %

Were they old stock, or is this what we can expect today? AIUI, this figure is at the very bottom end of the range.

Whenarewethere · Guru

The panel was all I could find at the time that would fit the space on my roof rack. If I find some new ones which are better I'll chuck these out.

They could pack the cells closer together & brought the frame in closer which would have increased the energy density & reduced weight. The wasted space adds up pretty quickly.

dorian · Guru

https://news.energysage.com/what-are-the-most-efficient-solar-panels-on-the-market/

These panels are for domestic and commercial rooftop solar use, but ...

Most solar panels are between 15% and 20% efficient, with outliers on either side of the range. High-quality solar panels can exceed 22% efficiency in some cases (and almost reach 23%!), but the majority of photovoltaic panels available are not above 20% efficiency.

I checked a 20W panel at Jaycar, and their efficiency is about the same as the Projecta panels, but there is a substantial border due to the frame, as you say.

If I take 10mm off each side to allow for the frame, then ...

20 watts / (618 mm x 258 mm x 1000 watts per square meter) x 100 = 12.5%

A 20mm border gives ...

20 watts / (598 mm x 238 mm x 1000 watts per square meter) x 100 = 14.1%

dorian · Guru

Whenarewethere wrote:
Just got around to short circuiting the six 20 watt solar panels. They are set up as 2 in series x 3 sets with no diodes.

If you shade one pair of panels, does the combined current fall to 2/3 under normal conditions? Can you detect any reverse current into the shaded panel?

Whenarewethere · Guru

I tried covering each 1/3 of the panels & the total power dropped only by exactly 1/3 according to the Victron app, & covered 2/3 or the panels & power was down by exactly 2/3. I also did 3 panels in series in 2 sets & power dropped by half.

Covering a very small portion of a panel knocked out power of the whole set in series, so I learnt that even a small shadow is detrimental & I would loose a 1/3 of the power.

I would have gone with 3 in series x 2 sets to keep power up in very low light conditions, but decided to go with 2 in series x 3 sets so if I get a shadow I'm more likely to only loose 1/3 of the power.

These test were to understand solar panels as I had no experience to date. I needed to know if 120 watts was enough as everyone says how much less you get than what the manufacturer says. I had intended to install bypass & blocking diodes as recommended in the Projeta manual that came with the panel but decided after these initial tests to see how I would go without diodes. Maybe on larger panels diodes are more of an issue but I don't know.

I have gone completely overboard with heavy wiring as I really did not want to do that again. Now even having a 4awg cable to connect the panels to the controller, a couple of 6awg cables & 8awg if I have to reach 15 metres for the sun. I could even wire in my jumper leads for another 3 metres!