Automatic Battery Discharger Design
6 Jan 01 - New design added.
This was built to replace my original discharger after I tried to discharge a 12 cell pack and it melted (literally - the case melted). It was designed to discharge packs of between 4 and 12 cells, up to 5A discharge and to be easy to use. The general principles are identical to the original discharger, so I won't repeat them here. The below images are thumbnails, click on them for the full size.
As I will be able to discharge a 12 cell pack at 5A, creating 60W of heat, I decided that a cooling fan was prudent, so I fitted a 40mm (1.6") computer fan at one end of the case. The only lamp fitted was a 14V 50mA LES type, just to indicate that discharging is in process. This lamp is visible in subdued light to below 4V. The whole unit was mounted in an aluminium box acting as a heatsink, and to which all the resistors are bolted.
In my unit I have the four sets of load resistors shown in the schematic, which give effective resistances of 10R, 4.7R, 3.7R and 3.2R (looking left to right). These are only needed if you want to discharge the lower cell count packs. You could just use the 4.7R and 3.2R string if you are only interested in 7 to 12 cells. It is important that all the load resistors are of a sufficiently high rating to take the power to be discharged through them, or else a fire might ensue. All the load resistors I used were 50W aluminium cased types, with the exception of the 0.47R and 1R which were the same series but 25W rating.
I had measured (during tests) that to achieve a 12V cut-off, I needed a resistance of about 600R. As 600R variable resistor are not standard, the easiest way was to adjust the effective value by fitting a fixed resistor in parallel. I could have used a 1kR variable resistor, but that would have meant only using a little over 1/2 the travel. For improved accuracy of cut-off, I really wanted to have the setting across the full travel of the variable resistor, not just at one end. I elected to fit a 2kR resistor in parallel as that gives an effective resistance of 667R at full travel (allowing for a little margin of error in measuring the first time).
I used an Excel spreadsheet to determine the values for the load and cut-off resistors. It was also used to produce the calibration data to produce the labels fitted. I have converted it to Excel 95 format and you can download the spreadsheet here.
The only factor limiting the unit I built was the 5A meter used to monitor the discharge current. If a higher rated meter where used, or a shunt fitted, currents up to the relay rating can be discharged. Alternatively, you can leave out the meter and use an Astro Whattmeter or other indicating system. Needless to say, the internal load wiring must use a cable of suitable rating.
You will notice from the photographs that both battery connections are hard-mounted to the case. Both are insulated from the case just in case a loose lead hits the case whilst being connected. Alternatively, flyleads can be used for the connections allowing more room in the case. I had to use a shoe-horn to get everything in.
If you've any questions (having checked on the original discharger design) please me.
Return to Plans & Designs Page