-
Where do I start with solar?
Posted by
paulca
on 12 Dec, 2017 18:57
-
I've been thinking about having a go at solar, just initially as a play thing, but I have to start somewhere.
I have been tempted to just go for a cheap chinese panel, but I've done that road before and it often ends up with a repurchase of a more proper item.
I don't really have anywhere outdoors to put a panel. I have a sky light and in calm weather I could put it out that for a while, but most likely any small panel I get will end up in my man cave room window. The curtains are 99% of the time closed, so there is definately room for a 1m square panel behind the curtains. It's south facing, and I expect I won't get very high efficiency behind double glazing. But it is what it is.
It seems that most of the "kits" out there immediately assume a lead acid 12V battery. Do I really need to go down that route primarily? Or can I start with something else? Maybe 18650s or something less arcaic, expensive, heavy and unreliable than an LA.
I'm not looking to spend a fortune, I'd think £50 for a first end-to-end set up and is it too optimistic to expect 20-50W from that money?
Just a first foot on the ladder type thing. If I can charge my USB devices and battery devices etc. from it that would be enough to play with.
I hope to be moving house next autumn to somewhere with a garden, probably a garage and maybe a shed, but right now I'm in an appartment with no outdoor space (except my car parking space).
-
#1 Reply
Posted by
Belrmar
on 12 Dec, 2017 22:46
-
i would recommend you to use a lead acid battery , yes is not as energy dense as a li-ion battery but they are cheap AF and eviromentally speaking they are much better (basically almost 100% recyclable). also using a lead acid battery grants you the opportunity of only having to use a cheap pwm charger from ebay instead of a complex battery balancer.
Talking about the energy you wold get....expect the worse as the windows of your house are no low iron and the ammount of light comming from a window is greatly reduced but for testing is good enough
-
-
I got a 100W Chinese panel for $96 USD shipped and it works great!
I also bought one of those Li-ION solar charge controllers (you'll find the black or white plastic with blue labeling) and tried it with a 96WHr Li-ION laptop battery pack and it sucked. Maybe I could power a few USB devices, but I really did not have much else that would work well enough off the nominal voltage. I also found that the SCC would not dump the full power from the panel into my load.
I bought the exact same controller for LA battery and it works great! It will dump the full juice into my load and I currently use it with a 12V - 120V inverter to run my 42V Li-ION ebike charger and to charge other USB devices.
Adam Welch did a review of a MPPT SCC that had a programable setting and he said it worked great! I am considering that one so I can charge my ebike batteries direct from the SCC.
-
#3 Reply
Posted by
paulca
on 14 Dec, 2017 07:59
-
-
-
the 20W system has the same kind of controller that I am using. Both of these are for lead acid battery technology. Polycrystalline panels are the second tier common panel with decent efficiency. Monocrystalline panels are superior. I have the former from Newpowa.
You should look into your region solar rating and figure out how much energy to expect, and then work out which battery will work with your system and needs.
-
#5 Reply
Posted by
paulca
on 14 Dec, 2017 20:11
-
the 20W system has the same kind of controller that I am using. Both of these are for lead acid battery technology. Polycrystalline panels are the second tier common panel with decent efficiency. Monocrystalline panels are superior. I have the former from Newpowa.
You should look into your region solar rating and figure out how much energy to expect, and then work out which battery will work with your system and needs.
I was going to ask about the battery size. Lead Acids are not in any way cheap here. £60 gets me a 40Ah car battery.
If the battery was too big, does it not just mean it won't get fully charged during the day or is there more to it?
-
#6 Reply
Posted by
Seekonk
on 14 Dec, 2017 20:47
-
C/10 is generally suggested. Since this is an electronics board I would think you would want to build something. I have 2KW of panels and I control them with an arduino. Total elecrtronics cost in my setup is only about $20. I usually tell people who want to use a 100W panel to just give up and wait becuse there is not much you can do with that much power. It sounds like 20W is really your limit space wise and I would use a couple old lap top batteries with a BMS board. Grab a couple 2596 regulators for your USB and Li pack. These can be made into power point controllers with just the addition of a few parts. Lead acid batteries suck. If you use one, just buy a new one for your car and use the old one for solar.
-
-
This is just my thought process on a small solar system.
You would rather pick a sealed lead acid battery unless you put them outside (and you will need a deep cycle battery). The controllers you linked will not work with Lithium cells.
Generally start with what power demands will be on the solar system. If just wanting USB power to charge a phone, for example, that is probably just about 10Wh per day (assuming 2.5Ah cell battery @ 3.7 nominal volts).
Then, figure about 5 days storage capacity and introduce efficiency factors/system losses, and a battery margin, so about 80% efficient and ~20% margin. I'm getting about 78Wh/day. A 12V 7Ah SLA battery would work (and so would a quality laptop battery, with the right solar charge controller).
Now determine the insolation rating to ensure enough panel to charge the battery in a day, including charging efficiency. Let's just pick a nice round number like 2. Take 78Wh/2, divide by 85% charging efficiency, so a 50W panel should work.
I'm using a 15AH lead acid motorcycle battery with my 100W panel, because that is what I had. I should be able to use a 40Ah SLA at some point (my average insolation is 5.5), or I'll get a boost MPPT solar charger for my e-bike batteries.
-
#8 Reply
Posted by
paulca
on 15 Dec, 2017 08:18
-
C/10 is generally suggested. Since this is an electronics board I would think you would want to build something. I have 2KW of panels and I control them with an arduino. Total elecrtronics cost in my setup is only about $20. I usually tell people who want to use a 100W panel to just give up and wait becuse there is not much you can do with that much power. It sounds like 20W is really your limit space wise and I would use a couple old lap top batteries with a BMS board. Grab a couple 2596 regulators for your USB and Li pack. These can be made into power point controllers with just the addition of a few parts. Lead acid batteries suck. If you use one, just buy a new one for your car and use the old one for solar.
Thanks. I like this approach. I only just discovered that 18650 cells are expensive, but brand new laptop batteries are not and contain 6 18650s. They may not be top of the range regards peak current etc, but I'm not sure I care yet. £8 for a 6 cell Toshiba battery. So I have an experimental 3s2p setup including balance lead, in the post.
For charging it, I could just use the Accucel 6 directly connected to the solar, it will take 11-18V on input, then buck or boost for the battery. The downside is it requires attention, in that it won't auto start charging without you pressing buttons. I have seen people using them on Youtube. I'm not too sure how suitable they are regarding MPPT and what not, but it seems comfortable with a whole range of voltages regardless of what it's charging. I charged two 3s lipo batteries in series (25.2V) from an 11.6V supply for example.
It might give me "Iteration 1" though. I'm starting to apply methodologies we use in software in work to my hobby projects to prevent scope creep meaning things never get finished and there is always something else to order. "Iteration 1" would be successful if I can charge my phone from a solar panel charged battery, for example. Iteration 2 would advance based on feedback from Iteration 1 as to what is most important to do next. Most likely an automatic interaction free charger.
I have an off the shelf regulator with a USB outlet and while it gets hot charging an eCigarette at 1.6Amps it works. (I think it's s dicky little surface mount regulator, I'm probably torturing it, but hey, it was about £2). Runs off a standard XT60 LiPo plug and while they are not everyone's cup of tea they are great up to 60A and what I will be fitting to my 18650 pack.
On space. It's a BIG window. Probably about 6 foot square. So even a 1m square or 1x1.5m panel. There are cupboards and fridges across the front of it anyway, so blocking the lower half of it with a solar panel is perfectly acceptable and the panel can lean back against those cupboards.
-
#9 Reply
Posted by
Seekonk
on 15 Dec, 2017 19:54
-
It is fairly easy to modify any buck converter into a pseudo MPPT controller with just a few added common parts. Technically it is called MPPC (constant panel voltage). Power point only changes with temperature and since inside that won't change much. Many of the MPPT chips are actually MPPC with a temperature sensor. BMS boards take care of all the charging requirements.
-
-
-
-
Technically it is called MPPC (constant panel voltage). Power point only changes with temperature and since inside that won't change much.
While this may be true enough for very small systems it is not true of large array's (several hundred watts up) because the MPP also varies with power output and shading, considerable loss of efficiency would be incured by trying to operate the array at a fixed voltage even if temperature compansated.
-
#12 Reply
Posted by
rhb
on 18 Dec, 2017 00:28
-
You might just setup a PV panel and use it to run LED lights during the day. Harbor Freight sells a 100 watt 4 panel array for $150 in the US. That's enough to run 7-8 1500 lumen LED lamps. Great for daytime lighting for detail work. That is about to become my first experiment. No battery. A battery to store charge adds about $80 and has only a 5-7 year life vs the 15-20 for a good PV array. Life of HF array and battery as yet unknown.
Solatube and others make skylights with reflective tubes, but they are expensive and also require roof penetrations. A PV array and LED lights is a more convenient version that costs less and can't cause a roof leak.
-
#13 Reply
Posted by
Someone
on 18 Dec, 2017 01:53
-
You might just setup a PV panel and use it to run LED lights during the day. Harbor Freight sells a 100 watt 4 panel array for $150 in the US. That's enough to run 7-8 1500 lumen LED lamps. Great for daytime lighting for detail work. That is about to become my first experiment. No battery. A battery to store charge adds about $80 and has only a 5-7 year life vs the 15-20 for a good PV array. Life of HF array and battery as yet unknown.
Solatube and others make skylights with reflective tubes, but they are expensive and also require roof penetrations. A PV array and LED lights is a more convenient version that costs less and can't cause a roof leak.
It may not be cost effective for everyone but running LED lamps during the day where lighting is needed can result in lower overall energy use than letting light through the windows (due to heat loss/gain through the windows).
-
#14 Reply
Posted by
kathy45
on 26 Dec, 2017 07:05
-
This looks to be more complicated than I thought. Interdot, the fridge does work when reaching next to the inverter. I have no other items joined to the inverter. Pouching, I made another cable to check if the lead was indeed the case.
-
#15 Reply
Posted by
paulca
on 30 Dec, 2017 22:56
-
BMS boards take care of all the charging requirements.
If only this was true. I've been searching around BMS boards and to be honest I haven't found a single review of one that actually works.
I know the single cells 1S BMS boards work fine, but it's when you move to 2S to 7S that things come apart. I've seen reviews of boards that just don't work period, boards which are sold as charging BMSs, but are just protection boards, boards which will not cut out to over 4.3v+ per cell, boards that let the voltage drop to 2v per cell and the better ones still appear to use a resistive drain on the high cells to drain them down to the low cells for balancing and can't keep up with a sensible charge rate.
I am still to find a multicell BMS which supports:
Proper cc/cv charge
Proper active cell balancing, or at least cross-cell balance charging.
Proper cut out voltages
Supports discharge AND charge
I can find a dozen Lithium chargers which support all of the above, but they don't typically like being in the circuit with the load and don't like being charge/discharge controllers, so they would need relay switching between charge and discharge. All of which makes them fairly useless with solar.
I watched lots of Adam Welsh's videos and Julian Lett videos and I don't think between them they have found a reliable multi-cell BMS. The best they seem to have come up with are Frankenstein devices with BMS to resistive balance the cells, a 12.6V cc/cv charge module to feed it and a boost converter to bump the 12V supply up. I'm not sure this could be connected to a solar charge controller, I'm pretty sure it couldn't as it would be trying to sense a battery not a charger board.
I don't even think they showed this set up under proper test, regards the BMS balancing cells or the proper cut off voltage of 12.6V (for 3S).
The Adam Welsh videos on Lithium solar charge controllers didn't seem to go well either. I don't think he found a reliable lithium charger which didn't float charge and ended up using a "Gell lead acid" mode and a 7 series pack to match the voltage of a 24V gell LA.
-
#16 Reply
Posted by
paulca
on 30 Dec, 2017 23:52
-
As always once you say you can't find something on a forum, the very next web search turns up 3 of them.
This one here, reviewed by Adam Welch supports Lithium and can have their float voltage configured well below peak charge voltage meaning the cells should not remain trickle charging all day but will get a break. I'm not sure how they would fair if there was a constant but small load as it will discharge to float voltage and then the controller will start trying to maintain that voltage on the cells "floating them" but at least it won't try and float it at 4.2V per cell!
https://www.ebay.co.uk/itm/391950642450?var=660913082749Here's a bit of a basic question. The solar charge controllers I have looked at don't seem to support fixed current charging. They seem rated for a maximum charge current and if there is solar power available will pump that current into the battery. So if I buy a 10A SCC and give it enough panel to provide 10A, it will charge the battery at 10A.
This would suggest that I need to match panel, scc and battery in ascending order of rating. For example:
8A panel -> 10A SCC -> 12A max charge current battery.
If I'm doing initial tests with small batteries I would assume I can get away with a small panel.
2A panel -> 10A SCC -> 3A max charge battery.
Does that work? If I up the battery capacity I could then up the panel to a maximum of 10A on that SCC.
Given my 18650 pack I made is 3S2P, 2600mAh cells making it 5.2Ah, that has an absolutely maximum charge current of 5.2A and I'd prefer to run it at 2.6A. Converted to watts that's 2.6A * 12.6V = 32.7W - efficiency and margin.
That would make my max panel size 32W or something like this:
https://www.ebay.co.uk/itm/222373840601Alternatively as I will be running it indoors and it isn't 100% efficeient anyway, I could aim for the 50W panel and be lucky to get more than 2.6A but know I'll never get more than 5.2A.
Does any of that make sense?
-
#17 Reply
Posted by
rhb
on 31 Dec, 2017 01:43
-
For the level of sophistication you are seeking I suggest that you set up a sample panel, dump the output into a resistive load and measure voltage and current in situ. Then set about designing the charger and battery system. Of course, you may want to consider a tracking mount for the panel.
In my case, I'm going for dead simple. A couple of 35 Ah gel cell lead-acid batteries, a 100 W Harbor Freight set of panels and charger and a small Harbor Freight gen set. All mounted on a hand truck fitted with BMX wheels for rough terrain and sporting a telescoping mast for a multiband inverted V dipole for HF and a vertical for 144 and 432 MHz bands. But before I actually build it, I'm going to characterize the PV panels myself. However, most of my effort will be on building the package and mounting the panels as a foldable array.
-
#18 Reply
Posted by
Seekonk
on 31 Dec, 2017 07:38
-
I can find a dozen Lithium chargers which support all of the above, but they don't typically like being in the circuit with the load and don't like being charge/discharge controllers, so they would need relay switching between charge and discharge. All of which makes them fairly useless with solar.
This is a universal problem in solar with even flooded cells. Any active solar system never gets much out of bulk charge. For that reason I pull as much bulk power as I can directly from the panels instead of the battery. Still, everyone muddles through with how they are set up. Those multi stage charge routines only work when you are on vacation. Good enough for this application which appears to be pointless.
-
#19 Reply
Posted by
paulca
on 02 Jan, 2018 18:29
-
So I took the plunge and bought:
EPEVER 10A Tracer MPPT charge controller
50W Mono-crystalline panel
Yuasa 26Ah Deep Cycle 12V SLA
The panel comes with 5m cable, I have plenty of 14 AWG wire for the battery. The battery max charge current is 6.5A the panel short output is 3A, charge controller is 10A, so all in spec and could later take a 100W panel with the same battery.
It's plenty to be playing with. The plan is to start small.
-
-
The Adam Welsh videos on Lithium solar charge controllers didn't seem to go well either. I don't think he found a reliable lithium charger which didn't float charge and ended up using a "Gell lead acid" mode and a 7 series pack to match the voltage of a 24V gell LA.
I recently received the boost MPPT CC that he reviewed and so far I've seen it shut off the current when the target voltage has been reached. I need to do some longer testing to see when/if it comes back on (and I will have forgotten if he mentioned it). That will not actually matter much to me as I plan to use it with a 100W 18V panel to charge 36V LiION e-bike packs that will be rotated quite often.
Looks like you got a decent small system to experiment with.
-
#21 Reply
Posted by
paulca
on 03 Jan, 2018 21:02
-
I recently received the boost MPPT CC that he reviewed and so far I've seen it shut off the current when the target voltage has been reached. I need to do some longer testing to see when/if it comes back on (and I will have forgotten if he mentioned it). That will not actually matter much to me as I plan to use it with a 100W 18V panel to charge 36V LiION e-bike packs that will be rotated quite often.
Looks like you got a decent small system to experiment with.
I thought about the Lithium stuff. I decided to go with the SLA because they are generally considered standard, simple and safe. I can use it as a ballast to charge Lithium IONs or Lipo's under supervision. Summer evenings and weekends when I'm home during the day I can experiement with the Tracer and it's settings for the battery voltages via the software giving quite fine control that 'should' do lithium packs. Remains to be seen and doesn't solve the balance problem. The constant discharge balancers like Adam is using worry me somewhat, but if the cells are balanced it shouldn't in theory be drawing much current, except to power it's LCD etc.
-
#22 Reply
Posted by
paulca
on 06 Jan, 2018 21:07
-
Basic wiring questions.
I could just bare the ends of the solar panel wires and stick them into the charge controller, but I want to put a switch in there to kill the solar power without having to unwire it.
I got some nice switches, they are DPDT.
Two questions:
1. Are there standards in the solar world for connectors, does it matter and would I be shunned for soldering a male and female XT60 onto the switch pins?
2. Related to 1. I know I usually only need to switch either wire in a circuit but as the switches are dual and the XT60 sits so perfectly on the pins... is there any downside to switching both + and - of the panel, battery or load? I know it would then leave the components isolated and able to float voltage. Is keeping the ground/- connected preferred as it keeps the voltage of all components in sync? Or does it matter at all?
I also have 10A fuses for each panel, battery, load, just in case.
-
-
I'm using XT60 connectors on the output side since that is what my ebike batteries use. The solar side uses solar industry sealed connectors, but I am also considering using the XT60 connectors there as well for when I want to swap my boosting or SLA controllers. (I really want to get a second panel though)
-
#24 Reply
Posted by
paulca
on 08 Jan, 2018 20:38
-
XT60s it is.
Panel arrived. Looks lovely. Heavier than expected.
Of course I had to play with it.... at 8 o'clock at night indoors. Slightly surprised it gave 8V open circuit under dim relaxed lighting and 12V when I pointed a 20W ceiling spot light at it. I even hooked up a 3W power LED knowing if the panel managed to produce more than 3W the LED as no more. The LED lit, but only just. Probably 0.1W.
Still... indoors, 20W halogen spot light from the ceiling that isn't bad.
It will be Saturday morning before I get to put it in actual sunlight, if we actually get any.
Charge controller hasn't appeared yet and the seller has some negative feedback about tardy delivery, so I expect it's another "I'm not really relay ordering from China" idiot. Grrr. The battery hasn't arrived either and hasn't even been marked as despatched. I may have to message him with "WTF?". The battery was cheap as it was 1 year old stock and the seller says he will charge and test it before despatch. But it can't take a week to do that!