solar

[Mongrel Shark]

The controller has a fuse. (Or at least it should!!!)

How far are your panels going to be from your controller? measure the ohms of the wire and work out what you'll get from 18v.... I bet it's under 17v.. My panels made 17.78v in full sun, I would get a bit over 15v of that at the controller... Is the 18v with the diodes attached? if you have too much, just add more diodes... Higher voltage in your line from panel to controller/battery will reduce amp drop over the line. ie if I measured the amps from my old panel's into the 10m 4mm core duplex back to the socket in wall of the house, for 10a at the panel I would have 8a at the wall. (battery and controller where inside on the same wall) I would have had thicker wire, but it gets ridiculously expensive after 4mm...


You still haven't told us how many hours of sun your expecting. You should be measuring this now as it is getting away from winter solace. You really need to do this on site as tree's, fences and other houses etc can make a big difference....
Also consider if you are going to use a tracker or not. a panel only makes full power when square with the sun. The amps drop off a lot after about 25 degrees from square... if you dont use a tracker you will get about a third of the 80ah. So maybe 25-30 ah per day. Lots more in summer... I'd still get a bigger battery. Or add to the one you are looking at asap. There are cheaper ones on ebay too btw. I got a really good 200ah gel on ebay for $600 about 2 years ago... My panels where getting old then and down to 120w ish too... I had a controller almost the same as the one you are looking at. it was good. Make sure it can do gel, Gels charge higher (mine went to 15.2 or 15.6 depending on temperature) The controller looks good, all programmable, so gel should be fine. Just check it has a fuse, just in case. Also whats the plug in the top?

Assuming you use a tracker and get 8hours sun in winter. 120w @ 12v = 10ah per hour. so 80 ah a day... 38 ah seems a bit small to me. Although it will do the things you mention. Run a ton of 40ma bright white leds, or quite a few 1w super brights (I'd go for lots of the cheap 40ma one's ). and charge your drill from an inverter. Although there is probably a more efficient way to charge it from the battery..

The controller will start to limit the amps as the battery fills too, The last 5-10% of the charge can take quite a while My 200ah gel would take an amp for 2 days (i'd be using some power at night, maybe only 4-6ah though, sometime a lot more) to get from 95% to 100% and it would only pick up about 5 ah over the 2 days..... Also bear in mind you get a better charge from less amps, as well as extending the cycle life off the battery, which is why the controller trickles the battery as it gets full... Personally I like a bigger battery for both these reasons.


Also look at the data sheet for you jaycar battery. See where it says if you charge it at 22amps, it only holds 22ah. You have to charge it on 1.94a for 20 hours the get the full amp hours in... although 15hour rate would get you pretty close. Your controller would probably give that battery the full 10A for less than an hour, then start heating the heatsink....

Here are a few battery suggestions:

First few are SLA. they'll still last quite a while if you let the controller do it's job
http://www.ebay.com.au/itm/AMP-TECH...=AU_Car_Parts_Accessories&hash=item53edf0b8d4

http://www.ebay.com.au/itm/Deep-Cyc...=AU_Car_Parts_Accessories&hash=item231aa25e47

Or 3-4 of these
http://www.ebay.com.au/itm/12-VOLT-...=AU_Car_Parts_Accessories&hash=item1c283ae719


Or if you really want gel (and a lot of people like sla better! I like gel, but for the amp hours to price you may have to consider sla) You have to look a little higher in price and they start getting Reasonable.

This is about as small as you really want to go, and only a little more than the one you sugest, with a lot more capacity and a better charge rate with 3.25A listed as the best charge rate, your controller would probably charge it at decent amps for a while if it was flat, although I notice it says charge to 10-10.8v that would be worth looking in to, I don't know if that's normal? or what it could mean.
http://www.ebay.com.au/itm/POWERSON...Caravan_Parts_Accessories&hash=item3cc1be6d9a



If it was me, with the amps and controller and with thinking about upgrading in the future, I would get this or bigger.
http://www.ebay.com.au/itm/12v-100a...=AU_Car_Parts_Accessories&hash=item3f1a1486fe
From that up to 150-200 could be really nice with the panels you want to make, which should be pretty good if you get the right glass and solder them well. If you only have 100ah or less, I wouldn't worry about a tracker for now..


With over 100 ah you'll be surprised what you can run... If it was me I'd be looking for a good cheap 2kw-4kwpeak inverter preferably with pure sine wave, look around on ebay etc. You will find there is not much in your house you cant run. With the obvious exceptions of most of the kitchen, hot water, heating etc. you should be able to get away with vacuuming though. Or a lot of lights and an lcd or two for a few hours a night. the trick is to run on 12v. as much as possible. For example a pc runs on 12v, and lower, you could make a power supply that runs from the battery and save a ton of losses inverting up to 240v then going back to 12v in the pc's power supply....




I dont think it's a good idea to add new batterys to old ones, pretty sure it messes up the new ones, although it can add some life to the old ones. So if cash is short now but you plan to upgrade in the next 5 years or less, I wouldn't even worry about getting pure deep cycle. Just get 120ah+ or big car/truck 4wd deep cycle maybe. check around the local auto places too. I had a great set of unsealed lead acid deep cycle batterys from super cheap once. Put inox conditioner in them, all up 2 65ah + treatment was $180 and they lasted well for about 5 years. I ran them a bit dry a few times too, and flattened them hard other times, they kind of copped it. excellent value for money, if you want to up grade in the near future.


If cash is not an issue, think 2v gels. But you'd need more panel for that.

 

[duke37]

"The other way you could go a further distance is if you change the DC from the panel to AC, send it to the controller, then turn it back to dc before it enters the controller. There would be some losses in this though, so you'd have to look at how long you charge line in carefully. On that note, can anyone tell me how pulsed dc travels over distance compared to AC and regular flat dc?"

DC is the best way to send power with the least loss, it is used for the interconnectors connecting the UK to the continent.
The resistive loss in a wire is proportional to the current squared, so if you use pulsed DC, say sending at 50% of the time, then the current will need to be doubled and the loss will be four times when the curent is passing. Since the current only passes for half the time, the overall loss will be double.

AC is used since it can be transformed to a lower current (higher voltage) and the resistive losses are lower. There will of course be losses in the transformers and any devices to change DC to AC and vice versa.

 

[donkey]

shark I might just get you to come and install once I have panels and charger ready lol.
the big use is lighting, so will be putting in a fuse on them.

 

[Mongrel Shark]

So say I used a joule thief type arrangement, maybe with a flyback or mot. I could make very high voltages, Assuming its a magic 95% efficient transformer. I would want high voltage and low current, then change it back to a lower voltage and get more current at the end of the line? and pulsed dc would be less efficient? and then ac last? What if it's only a little bit negative and a high duty cycle?



Sorry to drift off topic.

 

[Mongrel Shark]

donkey I'd be happy to give you a hand. just give me plenty of notice.


As for light's, order up the cheapest white, under car strip you can find. I got 2 600mm long for about $10 on ebay. they give off a lot of light and dont draw a lot of amps... Then there are the bulk packs of single leds. You can get a few hundered for $20

I've got a few of these one and there pretty good. 50 or more would light up a room pretty nice... they are much the same as a lot of the led torches getting around.

 

[(*steve*)]

Assuming its a magic 95% efficient transformer

If it were only 95% efficient, I'd take it back and get one that wasn't broken.

"Larger transformers are generally more efficient, and those rated for electricity distribution usually perform better than 98%.http://en.wikipedia.org/wiki/Transformer#cite_note-48"

 

[Mongrel Shark]

"Larger transformers are generally more efficient, and those rated for electricity distribution usually perform better than 98%."

Whats it cost for one for a 300w solar setup?

I was thinking of something I cold make out of scrap parts...

 

[(*steve*)]

Solar is DC.

The best way to reduce resistive losses is to put your panels in series. Mine run at between 350 and 450 volts.

However, high voltage DC is not something you want to play with.

In general, you should aim to have to lowest low voltage run that you can manage. this might mean heavy cables to your charger and your batteries and the inverter placed close-by.

For a given power, you lose 400 times more energy to resistive losses at 12V than you do at 240V, however DC-DC conversions are not as efficient as AC-AC, so you have to carefully weigh up the costs and benefits.

 

[donkey]

I know this thread is getting old now but I want all info in one place for me.

I got a battery 7.5ah for testing. now I thought I would find a cheap charge controller, the cheapest was about a 20A, but there are a few types, one is PWM dunno what the others are yet as they aren't listed as a specific type just charge controller so I assume its constant flow while the panels have light. but that begs the question what are the benefits of each?

 

[Mongrel Shark]

Not sure I can help much with that one.

I had a cheap 20a (which I blew up charging the house with the car when the sun came out) which would pulse the battery a few times a sec in smaller and smaller pulses. I could see it on the analog amps gauge. From full charge at around 12A the needle would flick all the way to 0 a few times over 5 secs or so. Then it would speed up as it dropped more amps, so at 9 amps it would hit 9 then drop back to 2-4 then up to 9 again. maybe 4 times a sec. As it reduced the amps it got faster. At 2 amps the needle would just vibrate really fast, with the odd big drop.

I replaced that one with a bigger flasher one, and it did much the same...


Both seemed to look after the batterys pretty well. I would say the cheap 20a one would be ideal for your test setup. After you have played with something basic for a bit, you will have a better idea what features would suit you better...

I still have my blown up 20a controller. It just needs a new transistor. All the features work except charge limiting. The faulty transistor is obvious from the large hole in the corner and the soot all around it. Part number is visible, I just haven't got around to looking it up. If you want it, message me and it's yours.

 

[donkey]

thanks shark for the input again. got some time off coming up hopefully able to knock of a few things then. hopefully I will be able to get on here and finish of the elctronic side.