The Light Fantastic: New Ideas for Trail Friendly Batteries

By Joseph Street, VE3VXO
Special to The ARS Sojourner
Watch the grams and the kilos will take care of themselves. It's an old backpacker's adage, and it is certainly true.

I once came across a Frenchman on the trail, a botanist. He was a huge strapping lad and he was carrying an 80 liter pack stuffed with about 35 kilos of gear and food and he had no hip belt.

Wincing with astonishment, I inquired as to his madness, and he told me that while on a month long assignment in northern Ontario it had broken and he got so used to living without it by necessity, that he never bothered to replace it!

And then there are guys like me, whose knees threaten mutiny when asked to carry much more than 20 kilos for an extended hike. I often spend upwards of a week on the trail at a time and in the cold of Canada's October month it can be a challenge keeping the weight down with that amount of food and clothing. Sometimes I look at my QRP gear with longing and my gel cell with loathing.

So when I heard about the new lithium battery packs my employer is developing for spacecraft power systems I knew I had to investigate. My colleague plopped a AA sized cell in my hand and I was amazed by its lightness.

Further research proved frustrating though. Lithium Ion cells have a linear discharge curve with quite a difference between the maximum and minimum cell voltages. When several cells are connected in series the problem is much worse.

Using enough cells to make sure that there is enough voltage at the near discharge end of the curve means that there is way too much voltage at the start of discharge and use of a linear regulator means a big penalty in efficiency.

One alternative is to parallel the cells and use a boost switching regulator which I considered to be a messy possibility with potential for RFI problems and unnecessary complexity. Then, in the darkest hour, a miracle happened.

I heard about the TADIRAN INCHARGE cell. It is a different chemistry. Li-MnO2, or lithium manganese. These cells hold 800 mAH, weigh a mere 17 g per cell, and have a flat discharge curve like a NiCad. What's more, they have no memory effect, and retain 85 percent of their charge after sitting for a year on the shelf. They can be used between minus 30 deg. C and 85 deg. C. The cell voltage is nominally 3 volts, so 4 connected in series make a nice 800 mAH, 12 volt pack that weighs a mere 68 grams.

Goodbye gel cell, hello lithium, boy do I feel skinny!

These cells are not cheap. And they are very sensitive to charge and discharge regimes. Most people would not be able to justify the expense, but then again most people do not attempt what we adventure radioactivists do routinely.

I had to build a supervisory circuit to curtail charging and discharging at the proper cell voltages. I planned to use my solar panel for a charge source which is limited to 60 mA or so, but if I charge from another source I have to use a current regulator to limit the charge current to the recommended C/10 rate.

Referring to the schematic, the circuit can be broken into two sections. One side is involved with the discharge and the other is for charging. On the discharge side there is a MOSFET in series with the negative battery lead which is turned on by one half of U1, which is a micropower op amp configured as a dual comparator, whenever the battery voltage is above the discharge threshold set by resistor divider pot and the zener reference established by D1, D2, and the 200 K resistor.

I opted to replace the 200k adjust pots with fixed values in the interest of reliability and size reduction. The silicon diode at D1 improves the stability of the reference over temperature. The recommended discharge limit for this type of cell is 2.0 volts, so the resistor divider is set to flip the comparator when the pack voltage falls to 8.0 volts.

The MOSFET then turns off disabling further discharge of the pack.

On the charge side the other half of U1 is similarly configured but this time it is used to energize a small relay when charging is complete which is specified as 3.45 volts per cell or a pack voltage of 13.8 volts. The normally closed contacts of the relay are used for charging so that only when charging is complete, is the relay energized and the normally open contact is then used to run a LED from the charger current to indicate that the charge cycle is ended.

An extra transistor, Q3, was added to act as a latch to prevent the relay from cycling as the battery voltage falls below 13.8 volts when the relay opens. I tried to think of a way of not using the anti-discharge diode at D3 which robs 0.7 volt potential from my solar panel but couldn't see an easy way to avoid it.

Also a capacitor C2 was needed, to eliminate a glitch which showed up when the charge source is initially connected due to the inductive kick effect of the relay coil. Decoupling capacitors C1 and C3 keep stray RF from entering the circuit and interfering with threshold sensing.

The entire circuit was built dead bug style on a small scrap of double clad board and mounted to the panel with two female RCA jacks and a slide switch. Two jacks simplify the circuit and also allow for discharge while the solar panel is connected.

The current draw from this circuit is less than half a milliamp but since these batteries have such an excellent shelf life it was decided to put in a pack enable switch to preserve it. The charge complete LED is set up to turn on immediately if the charger is connected with the pack disabled, as a warning that the pack is not being charged.

Because of the latching circuit you must remove the charge plug and make sure the pack is turned on before re-connecting the charger. There is no current regulation built in, but one can be made with a three terminal regulator such as a 7805 and a 56 ohm resistor between the output terminal and the terminal which normally would go to ground.

The current regulated output is then taken from the ungrounded terminal, which would have been grounded in a voltage regulator configuration.

This circuit could be incorporated into the pack, but I wanted to get the most from my solar panel so I have an external regulator for charging from mains.

I am very impressed with this little power pack and have used it for powering my 30 meter NN1G rig as well as my 24 and 10 Ghz hill topping gear. It could do wonders for your sprint score, as well as my knees!

Anyone interested in duplicating this project can contact me for more info via email at ve3vxo@rac.ca. The batteries themselves are available from performance RC under the trade name Duralite and here is their site http://performanceprod.home.duesouth.net/
Summer is coming, and I soon will be floating down the trail in lithium bliss. See you on 30 meters!

Photograph 1 of the circuit board

Photograph 2 of the circuit board

Photograph of the front panel


* * * * * * * * * *

Joesph Street, VE3VXO, is an avid builder, QRPer and outdoorsman living in Waterloo, Ontario, Canada.