Tag Archives: proto5

Prototype 5

Prototype 5 breadboard
Proto 5 was built on a perfboard that I got for 50 cents at a flea market. I’m using huge wire nuts to tie together groups of three #10 wires in places. No heavy current actually flows through the board – it all goes through the relay contacts.

I’ve had to do a lot of other things lately, so progress on the solar bike has been slow.  I’m realizing now that it would be beneficial to just start putting wheels on the panel and shake out the mechanical design.

But I would also want to connect the panel electrically.  I wouldn’t want to simply connect the panel across the battery, as that could force 8 amps of current into the battery when the motor is not running.  I wanted to try just connecting the panel with no battery whatsoever, but when I did that, the motor just barely gave a twitch, then drew no more power.  I’ll have to investigate the cause of that, later.

My plan was originally to build a small analog circuit with a comparator to detect when the battery voltage dropped below a certain level.  That drop would indicate that the motor was running, so the panel could then be connected across the battery by a relay.  When the voltage rose again, the relay would cut out.  But while sketching out the circuit, I realized that I could just use an 8-pin PICAXE 08M2 instead of an op-amp, with no greater complexity on the board.  Plus, more complicated algorithms would be possible with the PICAXE, and some of the learnings might be applicable to the final Power Transfer Controller (PTC).  Thus came into being the Relay PTC, aka Proto-5.

Click on the schematic to see it full size.
(Click on the schematic to see it full size.)

The schematic is simple. A Zener diode is used to regulate the PICAXE supply – one of the few cases where this topology actually makes sense, since the PICAXE draws only about 2 mA worst case. Two of the A/D inputs on the PICAXE are used to monitor the solar panel voltage and battery voltage. A 150 ohm resistor is in series with the relay coil, because it’s a 12-volt relay, which I was able to pick up at 2/$1 at the electronics flea market. If this was more than just a quick prototype, I might have sprung for the 24-volt version of this relay, but that costs $8 at the surplus store.

ironically rainy day
We’re in the worst drought in recorded history for California, with a dry December and January. Ironically, on the exact day that I finished bench tests on Proto5 and was ready to test it outside, it rained. I’m not complaining, we need all the water we can get.
better day for testing
The next day was better, but still hazy with scattered clouds. Here is the test setup. Power connections are made by various means, including MC4 connectors on the solar panel, 1/4″ spade lugs, and wire nuts.
full view of proto5 test
This is a more comprehensive view of the test. My Hantek 6022BE oscilloscope is on the white chair, and the laptop driving it is on the mosaic chair.  If you look carefully, you can see that the bike is on a stand, with the rear (drive) wheel off the ground.
Trace without solar boost
Here is the battery voltage.  The motor is engaged for approximately 2 seconds.  When the motor initially starts, the voltage can plunge from about 26 volts down to 20 volts.  This surge doesn’t last long, just a little over 400 mS.  After that, the battery voltage returns to the pre-energized level, suggesting that the motor is drawing next to nothing in power.  There was no solar panel connected for this trace.
Trace with solar boost
This trace has the same conditions as the previous one, except that if a drop in battery voltage is detected, within 100 mS, the relay turns on for a full 9 seconds, connecting the solar panel across the battery.  Note that the acceleration time is shorter with the solar panel assisting – only about 250 mS.  However, the motor still appears to be drawing negligible  power after the initial acceleration.  This is not surprising, as the rear wheel of the bike is off the ground, and it takes very little power to keep it spinning. There is no change in battery voltage between before and after the initial surge, even though after the surge, the solar panel should be contributing.  My theory is that either the solar panel is very weak (unlikely, since I measured almost 8 amps of short circuit current), or the battery is very strong (likely).