Buck boost converter problem

[rcbuck]

Has anyone used this converter: http://www.ebay.com/itm/141723797418

I purchased two solar panels that put out 21V @ 270ma. I am going to wire them in parallel and feed them to the input of the converter. The converter output will be set to 13.8V and connected to the battery in my truck. The truck is only driven about once per month and the battery is usually dead when I go to start it. Ten minutes on my 10 amp charger and I can then start it. The reason the battery drains is because of the constant current draw of the alarm system.

I have found that the converter will not start working if you slowly power it up. I confirmed this with a variable power supply. So when the sun rises in the morning, the converter doesn't start. However, if I disconnect the solar panels and reconnect them, the converter starts and puts out the 13.8 volts.

I have come up with a solution that seems to work on the bench with the variable power supply. I have a 6.8 volt zener diode in series with a 10K ohm resistor connected to the power supply output. The 10K resistor drives the base of a 2N3904 transistor. The transistor collector switches a relay which applies the power supply voltage to the converter. When the variable supply exceeds roughly 7 volts, the zener breaks over and drives the transistor base causing the relay to operate. The converter starts and puts out the required 13.8 volts.

I haven't put this circuit on a bread board yet and connected it to the solar cells. I will do that sometime tomorrow. If this circuit doesn't work with the solar panels, I will try using a PIC and read the panel voltage with one of the A/D pins. When it gets above 6 volts I will turn the relay on to apply the panel output to the converter. A few lines of code will be all it takes.

I was just curious if someone had used this converter. Or is there a better solution that I can use?

[Simon]

you could use a power mosfet instead of the relay. it should work. But it might be a problem when just enough current flows in the base to activate the transistor but not enough current passes to activate the relay. You will get most of the voltage and power dissipation across the transistor. You could use a comparator - with hysterysis.

[rcbuck]

Simon,

I thought about a mosfet. I was going to play with that and see how it worked. I would have to set it up for low side switching to work properly.

Do you think it would hurt anything to connect the panels directly to the battery? At full sunlight they would only put out a little over 1/2 amp. As a test, I connected them to a 12 volt gel cell and went out in the sun. The battery voltage stayed at the 12 volt level even though in theory the panels should have been putting out 21 volts. The panels will only be in the sun about 6 hours per day.

This is the first time I have worked with solar panels.

[rcbuck]

Simon,

I have one other question. The two panels I bought off fleabay have a thin clear mylar film over the panels. Should I remove that film or leave it on? The panels will sit on the dash inside the truck. The solar cells (72 on each panel) are on .062 fiberglass material like would be used for a PCB.

I just connected both panels in parallel and connected them to the truck battery. The battery was reading 12.19 volts before connecting the panels. After I connected them and pointed the panels directly at the sun, the battery reading increased to 12.23 volts. I then checked the current draw the battery was pulling from the panels and that was 300 mA.

[Simon]

You can use a high side mosfet as well, you just invert the comparator logic or that of your transistor and pull the gate low, If you are certain the panel voltage will never go over 14.5V then yes you can make a direct connection. I wouldn't personally. Also at low voltage you won't get much charge.

If there is a film that easily peels off I'd remove it.

[rcbuck]

Simon, thanks for the info. I will run a couple more tests and decide what I am going to do. I want to test the converter under lower sunlight conditions to see how much current it will put into the battery under boost conditions. If I can get at least 100 mA from it, I will go with a mosfet circuit rather than direct connect.

[Simon]

If you are getting less than 13V in full sun then it sounds like the panels do not make enough voltage to charge the battery and so the current capability will be wasted. A fully charged battery at rest (not under charge) would be around 12.6V due to the battery internal resistance you need more than that to make the charging happen.