Lipo Charger/Cutoff Tests
04/24/07 19:31 Filed in: Lipo
I've been working cautiously on learning how to deal
with Lipo batteries.
I bought a Lipo 3.7V charger circuit from SparkFun, a couple of 3.7V lipo batteries (one flat - about 2200mAh and one multi-round cell From Tenergy that is 4400 mAh). That circuit should handle the overcharging issues with Lipo.
This is the Lipo Charger from SparkFun
3.7V, 4400mAh from Tenergy
This battery has a built-in protection circuit which probably seems redundant to what this project is about except because this is my first work with Lipos, I thought it made a lot of sense to have a backup. Just as soon not set my house on fire right out of the gates.
I needed a cutoff mechanism because if you overdischarge a Lipo that can be lethal. So, I bought a cutoff circuit from All-Battery.com - it's tiny but effective.
Lipo cutoff circuit from All-Battery. This thing is small enough to easily be shrink wrapped with a battery. See pictures of this later. I don't know for sure, but this may be the circuit used in Tenergy battery - I haven't broken the shrink wrap on the Lipo pack to see what it's using.
I first wanted to test the cutoff circuit, so I wired up my bench power supply to the battery terminals of the cutoff board. I planned to simulate a discharging battery by gradually lowering the voltage on the bench supply and monitoring the load/charger terminals of the cutoff board to see if/when it cutoff juice to the load. As spec'd, the board cutoff at 2.5V.
Once I was satisfied that I understood the cutoff circuit and tested it, I wired up the Tenergy battery to the battery terminals of the cutoff board. I then fed the load/charger terminals to the SparkFun Bat/GND terminals.
I used a Pyrex container for the Lipo:
I just used jumpers to feed the battery leads to the cutoff board:
I also wired up a regulated 5V powersupply to feed the SparkFun charger. They recommend a regulated 5V though supposedly the charger can take up to 7V. I fried one charger when I was first messing with it, so I took the input seriously this time and gave it exactly 5V regulated.
To do that, I used a Dimension Electronics 5V switching regulator that takes 12V DC wall wart power and regulates it to a clean 5V. I rigged up a little DC jack cable to go from the Dimension 5V breakout board into the SparkFun charger.
Then I tested that setup with no load. I wanted to make sure I could plug in the charger with the battery and cutoff board all wired. After firing it up, I checked voltages and everything seemed stable. I was seeing 3.8 on the battery terminals and 4.6V out the Sys/GND load side of the charger. I disconnected the charger power to make sure I could still see power going across Sys/GND of the charger which would indicate the load would still be getting power, but via the battery. It was and I coudl see 3.83V going out Sys/GND of the charger which is the same voltage I read directly off the battery leads.
Finally, I connected a small load, an AppBee breakout with a single LED and 3.3V regulator on it (no XBee.) It was all working fine and could connect and disconnect the charger at will without disruption to the load.
I needed to create a more realistic and larger load and drain the Lipo down and verify that it's cutting off correctly - as I mentioned before, I tested the cutoff at 2.5V with a bench power supply. After draining it down, I need to test the SparkFun charger to make sure it's charging up the battery again.
To add a little more load to the system, I have it powering the XBee GPS I made - is probably only about 100 mAh draw but it's a similar load to what I would expect to put on a system like this. The XBee/GPS device has a 3.3V standard regulator on it so it's not the most efficient either.
[Update 4/26/07 a.m. - I let the configuration above run to see how long it would go before the GPS/XBee stopped working. Because the 3.3V regulator on the breadboard was being driven by the lipo at 3.7-3.8V, the drop on the other side of the regulator was about 1V, so the GPS and XBee were running at about 2.8 so there wasn't much margin for them to go down before cutoff. Anyway, it ran for over 14 hours before the GPS or XBee stopped sending. Not sure the exact battery level where the XBee GPS stopped functioning because it was in the middle of the night sometime. I knew the drain rate wouldn't put it in a danger zone (in case the cutoff didn't work) before I went to bed. When I came in the next day, the lipo was at about 3.6V. I plugged in the SparkFun charger at that point and disconnected the load (was done with that test) and the charger took the lipo back to 3.83V and has kept it there - charge light is still on, but the voltage on the battery has stayed at 3.83.V]
I bought a Lipo 3.7V charger circuit from SparkFun, a couple of 3.7V lipo batteries (one flat - about 2200mAh and one multi-round cell From Tenergy that is 4400 mAh). That circuit should handle the overcharging issues with Lipo.
This is the Lipo Charger from SparkFun
3.7V, 4400mAh from Tenergy
This battery has a built-in protection circuit which probably seems redundant to what this project is about except because this is my first work with Lipos, I thought it made a lot of sense to have a backup. Just as soon not set my house on fire right out of the gates.
I needed a cutoff mechanism because if you overdischarge a Lipo that can be lethal. So, I bought a cutoff circuit from All-Battery.com - it's tiny but effective.
Lipo cutoff circuit from All-Battery. This thing is small enough to easily be shrink wrapped with a battery. See pictures of this later. I don't know for sure, but this may be the circuit used in Tenergy battery - I haven't broken the shrink wrap on the Lipo pack to see what it's using.
I first wanted to test the cutoff circuit, so I wired up my bench power supply to the battery terminals of the cutoff board. I planned to simulate a discharging battery by gradually lowering the voltage on the bench supply and monitoring the load/charger terminals of the cutoff board to see if/when it cutoff juice to the load. As spec'd, the board cutoff at 2.5V.
Once I was satisfied that I understood the cutoff circuit and tested it, I wired up the Tenergy battery to the battery terminals of the cutoff board. I then fed the load/charger terminals to the SparkFun Bat/GND terminals.
I used a Pyrex container for the Lipo:
I just used jumpers to feed the battery leads to the cutoff board:
I also wired up a regulated 5V powersupply to feed the SparkFun charger. They recommend a regulated 5V though supposedly the charger can take up to 7V. I fried one charger when I was first messing with it, so I took the input seriously this time and gave it exactly 5V regulated.
To do that, I used a Dimension Electronics 5V switching regulator that takes 12V DC wall wart power and regulates it to a clean 5V. I rigged up a little DC jack cable to go from the Dimension 5V breakout board into the SparkFun charger.
Then I tested that setup with no load. I wanted to make sure I could plug in the charger with the battery and cutoff board all wired. After firing it up, I checked voltages and everything seemed stable. I was seeing 3.8 on the battery terminals and 4.6V out the Sys/GND load side of the charger. I disconnected the charger power to make sure I could still see power going across Sys/GND of the charger which would indicate the load would still be getting power, but via the battery. It was and I coudl see 3.83V going out Sys/GND of the charger which is the same voltage I read directly off the battery leads.
Finally, I connected a small load, an AppBee breakout with a single LED and 3.3V regulator on it (no XBee.) It was all working fine and could connect and disconnect the charger at will without disruption to the load.
I needed to create a more realistic and larger load and drain the Lipo down and verify that it's cutting off correctly - as I mentioned before, I tested the cutoff at 2.5V with a bench power supply. After draining it down, I need to test the SparkFun charger to make sure it's charging up the battery again.
To add a little more load to the system, I have it powering the XBee GPS I made - is probably only about 100 mAh draw but it's a similar load to what I would expect to put on a system like this. The XBee/GPS device has a 3.3V standard regulator on it so it's not the most efficient either.
[Update 4/26/07 a.m. - I let the configuration above run to see how long it would go before the GPS/XBee stopped working. Because the 3.3V regulator on the breadboard was being driven by the lipo at 3.7-3.8V, the drop on the other side of the regulator was about 1V, so the GPS and XBee were running at about 2.8 so there wasn't much margin for them to go down before cutoff. Anyway, it ran for over 14 hours before the GPS or XBee stopped sending. Not sure the exact battery level where the XBee GPS stopped functioning because it was in the middle of the night sometime. I knew the drain rate wouldn't put it in a danger zone (in case the cutoff didn't work) before I went to bed. When I came in the next day, the lipo was at about 3.6V. I plugged in the SparkFun charger at that point and disconnected the load (was done with that test) and the charger took the lipo back to 3.83V and has kept it there - charge light is still on, but the voltage on the battery has stayed at 3.83.V]
