Getting 36v from a 12v battery

Used or surplus UPS batteries, 12 volt 8 amp hour are cheap and readily available. New they are only about $19 each in the USA. (30ish here in Canada - wouldn't you know) 20 amp hour are also readily available on the surplus market - AGM too, no watering to worry about.

Awright, have a look at

I did a SPICE simulation (circuit simulation program) of this circuit useing guessed-at parameters for your solar cell. It works as expected. If one solar cell doesn't have quite enough moxie to keep yer battery charged you could always add another one in parallel later

-- parallel because the LMC555 doesn't want to see more than 18 volts and there is no reason at all not to parallel solar panels.

Efficiency of this circuit could easily exceed 90% if it's set up right.

The inductor could be a little wedding-ring sized toroid of about 330 uH. AxMan Surplus had a bunch of 'em on hand last I looked and I have half a dozen of 'em here. They'll handle a couple of amps without saturating. The power MOSFET is a garden-variety part. If I were to mail you an inductor I'd also mail you a MOSFET -- I have hundreds of them in my goodiebox. You can get everything else at Radio Shack and skip the minimums and shipping from guys like Digi-Key, Mouser and Jameco. The diode to the battery should be a fast diode like 1N4933 to 1N4936. I have a bunch of 1N4936 (1 amp 600 volts, fast) on hand too, more than I'll ever use.

Hi Clare,

Thanks for the detailed information, I appreciate the effort you went to. I will copy and print it and see if I can follow what you are getting at.

Hi Don,

I am not questioning the electronics knowledge of either of you as the discussion between you and Clare is now beyond my understanding.

I appreciate the effort you have gone to here, but if Clare is correct in that I can series the 3 HF panels and put them straight across the

3 batteries, why are we messing with a charge controller? I wasn't sure you could series them, but I guess all the little sub-panels within the big panel are series though.

Thanks for taking the time to do this.

Get them Taxes done man. I wimped out and got an extension, but have promised the Accountant I will have stuff done by May first. That date is now looking pretty shaky though. LOL

OK guys now I am getting confused here.

I assume you mean the output of the 3 panels will be 36v and just connect acorss the series batteries?

Why am I (we, you Don) messing with 36v out of series batteries and trying to use 12v individual charging?

There are dozens of inexpensive single-voltage models, but 36v isn't popular. Some medium-priced controllers like the Tristar are multi-voltage, but I see now that it's limited to 12, 24, and 48. The Outback MX60 (MPPT and $500) can do five voltages including 36. Compared to that, this 36V-only model might seem like a bargain at perhaps $100

These are all multi-stage chargers though, and their features would be wasted using the small supply you seem to have in mind.

Perhaps there's something on this page

but it's a bit hard to tell. ;-)

You could wire 3 of the small HF $10 (on sale) 12V panels in series. Their output is so little that you could do without a controller.

12V garden tractor batteries were $10 each (plus an old core) at Walmart last I looked. Or, under $40 delivered for 3 of these Wayne

Because initially you spoke of using three 5-watt HF panels, so it would be dirt simple to just put one panel on each battery and call it done.

That's the ones she wants. With a simple overvoltage clamp (zener and resister) for a controller, just to keep the battery from complaining about high voltage - as a precaution. You may be right - might not need it, but simple and relatively cheap.

Wow.

A zener doesn't clamp to ground when it starts conducting. It just draws current. In conjuction with series resistance, it can limit voltage. It acts very much like a ...uh ... battery in series with a diode.

The solar panels themselves can't deliver more than safe trickle charge current, so a zener not necessary here and serves no useful purpose.

How to charge a battery with a 5-watt solar panel having sufficient voltage for the battery in question:

1. Connect solar panel to battery + to + and - to -.
2. Declare victory

Solar panels are themselves diodes, though a series diode probably would reduce leakage so it isn't a bad idea -- although it does cost a bit in efficiency.

If you do insist on a zener, you can make any voltage you want with an LM336 (96 cents) and a couple of 3-cent resistors. Power zener: LM336, a jellybean PNP transistor (2N4403, 13 cents) and a TIP3055 ($1.76) that can dissipate up to 90 watts.

Danged if I know! If you have 3 panels, I'd just connect 'em to the batteries and call it done. Series or parallel, pick yer pony and take yer ride, either should work. I'd panel each battery individually if the batteries aren't the same model and age -- but we've been over that a couple of times already.

Correct.

I can only work on taxes for a little while, then I *need* coffee and a break. It's just typing, TurboTax asks questions and I type in the answers. No, I didn't have any distributions from diamond mines in Zanzibar. Yes, I did have some o' this and a little o' that, lessee, enter amount in box 3, roger. And so on.

I'll probably build the little inverter anyway just for fun. I have the few necessary parts at hand and I'm curious to see it work. Meanwhile, you could have your gate swingin' by sundown.

Yup - you can do it that way too. Just a bit more complex to describe and do. Like I've said before, I'm no engineer, and I often design / build things a bit over robust - but they generally work. My notation that the solar panels are not a "hard" supply, and therefore less current capacity would be required for both the Zener and the resistor than the calculations indicated. How much less? I couldn't say, but likely significantly. My design woudn't burn out anyway!!. I've had a small solar panel charge an 8 ah battery at significantly more than the specified float voltage, and it would shorten the battery life somewhat. I've never used a blocking, or isolation diode on an unregulated solar panel, but using the zener regulator it is a good idea. As for "clamping to ground" that is ESSENTIALY what it does. At over the rated current it turns on, and on a "hard" supply, if there is no resistance in the line, it shorts rather spectacularly. Without the resistor on a small solar panel, it would just cause the panel to heat - which generally is not a good thing. Adding a series resistance for regulation helps prevent the "magic smoke" from getting out. As you say, perhaps not essential - but I'd do it "just to be safe"

A photovoltaic array can't deliver much more current to a dead short than it can at about 80% of open-circuit voltage. If it is clamped by a zener (or zener-like circuit) that can handle the panel's power output then nothing will burn out.

That surprises me, but I don't doubt your word that you saw it happen. Given that, use of a zener or equivalent would certainly be prudent. Automotive and 80 or 105 ah deepcycle batteries can easily handle an amp of trickle charge, but an 8 ah battery might not like much over

100 mA. 5 watts at 13.6 volts (or 15 watts at 40.8 volts) is 0.368 amps which certainly wouldn't harm automotive-sized batteries but may well be too much for "float" on little 8 ah batteries. 13.6 volts is a safe float voltage for lead-acid batteries regardless of size. For 3 batteries in series that'd be 40.8 volts. 1N5259 is a 500mW 39-volt zener for 36 cents. That and the Vbe drop of an NPN power darlington transistor would make a "power zener" quite close to 40.8 volts. The transistor might be a TIP120 (5 amps, 60 volts,65 watts) , 74 cents. The base-ground resistor might be 180 ohms 1/8 watt or whatever's handy. Mount the transistor on a chunk of sheetmetal to serve as a heatsink. It wouldn't take much heatsink to make this a 15-watt zener equivalent -- no more than an actual 15-watt zener would require. The actual zener here is just a voltage reference, the transistor does the heavy lifting.

-------- | | Z | | |/C |------| \ |\E / | \ | / | ---------

The regulation of this circuit isn't quite as "sharp" as an actual power zener, but it is way lotsa plenty good enough for this application. At 40.6 volts it "leaks" about 11 mA; at 40.8 volts it shunts about 94 mA, and if shunting the entire 368mA output of the three series-connected panels it clamps to 41.0 volts (13.67 volts per battery). If ya wanted to raise the voltage a touch you could stick a 1N4004 diode (5 cents) in series with the emitter. That'd raise the voltages for given shunt currents about 0.75 volts or about 0.25 volts per battery.

To your point, Clare:

"Shunt Regulator - Shunt regulators are common in photovoltaic (PV) systems since they are relatively cheap to build and simple to design. The charging current is controlled by a switch or transistor connected in parallel with the photovoltaic panel and the storage battery. Overcharging of the battery is prevented by shorting (shunting) the PV output through the transistor when the voltage reaches a predetermined limit. If the battery voltage exceeds the PV supply voltage the shunt will also protect the PV panel from damage due to reverse voltage by discharging the battery through the shunt. Series regulators usually have better control and charge characteristics."

Source:

It clamps to zener voltage (not ground) but it would indeed soon fail if driven much above zener voltage by a stiff source. My point is that a solar panel is far from being a stiff source.

The panel wouldn't self-heat any more than when it's delivering rated power to a load. The panel is generating electric power, not dissipating it. The zener (or equivalent) would dissipate available power in excess of the battery's ability to accept it at float voltage. This would be at most 15 watts with three 5-watt panels, considerably less with a battery connected and accepting charge.

Perhaps the confusion comes from the power rating. If you draw more than rated power from a bench supply or transformer, or deliver more than rated to a transistor or resistor (or zener) it'll overheat because the power rating is based on its permissible temperature rise. A solar panel's power rating is what it can deliver to an optimally-matched load in bright sunlight. The only way to get it to deliver more is to turn up the sun.

1N5260 is a 500 mW zener, which is .011 amp at 43 volts -- certainly not 3 amps!

(snip lots of stuff assuming 42 volt source)

Jenny: "But a 10-buck 12-36 charger sounds good and I do already ahve the 5W panel".

Clare: "Dead simple. Decide what your charging voltage should be. I'd say 42 volts." (nevermind what Jenny said....)

This circuit is not a charge controller, it is a 12 to (whatever voltage the battery stack requires to charge) converter. It's essentially a current supply with indeterminate output voltage. The series battery stack (and perhaps a zener clamp) will determine the stack voltage under charge.

Read back thru the thread. I did too, and saw that I didn't make it clear that this would run off your 5W panel and up-convert.

Jenny: "But a 10-buck 12-36 charger sounds good and I do already have the 5W panel."

Roger that, here ya go. I undershot the 10 bux by a bit but you could always buy spares.

Not to say that this would be better than or even as good as buying more panels and stack them in series, not at all. That'll work too, and it's simpler. Simple is best when it works, and it should here. Pick yer pony. If you use small batteries then I agree with Clare that you probably should zener-clamp whatever charger you use.

You should definitely zener-clamp this little converter circuit because it could easily produce 60 volts if a battery came unstuck. It draws energy from the solar panel at whatever voltage the panel supplies for a little while (some microseconds or milliseconds) and then delivers that lot of energy to the load at whatever voltage it needs to accept it. Then it does it again, and so on. Bit like an impact wrench, flywheel effect sort of thing.

BTW, I got my taxes nearly done YAY! I've told myself I won't go pick up a pistol that's waiting for me until I have my taxes done, and I

*want* to try it out ASAP, oh my yes!

You have several choices before you now. Pick one and go do it!

Thanks for clarifying and shwing the "active zener" implementation. I was just ASSuming she would not be using huge automotive style flooded acid batteries when the little AGM critters would do the job and are so cheap in the USA.

Just whent by how they were described. Obviously wrong.

Hi Don,

I am thinking that's what I will do. Less things to make and go wrong.

Thanks for it all, you and Clare left me behind about 2 days back. LOL

Thanks again Don and Clare,

I am going for the most simple for me which will be 3 of the HF 1.5W panels individually hooked to 3 batteries. I have some gell cells stripped from old UPS' I was given. I hadn't thought of using anything that small but will give it a shot. I already have one of the HF 1.5W panels so will order another two.

Thanks to everyone who jumped in to help. Much appreciated.

I am making new geometry mounts today and the gate should swing later this week. Got other stuff to do as well, including MY taxes before May. Haaaa! Fat chance.

Phew I am glad THAT thread is over. LOL