Electric Porsche – Battery issues and DC/DC modifications

It’s been awhile since I’ve made an update but I have been plugging away on small and time consuming projects.

I’ve found that Murphy’s Law trumps the Law of Averages.  In the two 16S3P packs (96 cells) that I built I managed to have a single group with 3 of the smallest capacity cells that I’ve found so far.  This means that my total pack was limited in capacity to the total of those 3 small cells.  These A123 pouches are already smaller on average than their rated capacity so I’m not really getting 20Ah per cell, it’s more like an average of 17.8 – 18Ah on the cells that I have tested.  I’m counting on 53+Ah total to give me the range I’m expecting.  The solution?  Pull apart the existing packs and measure the capacity of every cell, along with testing the capacity of all the remaining cells so that I can throw them in a spreadsheet and find the average capacity and then try and group them in sets of three that total measured average capacity x3.  This will mean grouping the larger cells with the smaller to make sure that all of my cell groups are within a very tight range of average capacity.  It takes about an hour to test each cell giving it a complete charge/discharge cycle while measuring the capacity.  Both the charge and discharge are automated so although each cell takes an hour I only have to be there to load fresh cells in the charger and discharger and record the results.



Another small project I put together was a modification to give me a two stage charging voltage on my dc/dc converter.  When the car is “on” I want between 14.4 and 15v to run the lights and wipers and other stock automotive systems.  Since I will have a 12v lithium battery I still want to be able to charge it to an appropriate peak voltage, but I don’t want to hold the voltage there indefinitely and overcharge the cells so when the car is “off” I want the dc/dc output voltage to be roughly equal to the no load open circuit voltage of the cells.  The dc/dc converter will remain on to keep any parasitic loads on the 12v system from draining that battery as well as avoid the requirement to switch pack voltage to the dc/dc converter which would require another EV200 contactor.

The starting point is a Mean Well SP-320-15 power supply.

I completely disassembled the power supply to look at the board traces and figure out how the trim voltage works and how I can tie into it from the top side.

There are two wires soldered to the power supply PCB, I used orange because I had some and polarity isn’t important for the resistor. One wire is soldered to the LED lead “K” on the PCB and the other wire is soldered to the lead of R73 closest to the terminal strip. These two points happen to be parallel to the trim pot, so adding any resistance here puts it in parallel with the pot.

The relay is a simple SPDT 12v relay I had laying around, I picked this one because I wasn’t sure if I would need the NC or NO terminal to adjust the voltage in the correct direction when “off”. It turns out I used the NO (normally open) contact so any standard SPST 12v NO relay will work just fine. There happens to be an open area just a hair larger than the relay on the PCB so I stuck it in place with a small dab of epoxy.

A 2k resistor (I used a pot for testing, but settled on 2k) is connected to the NO terminal of the relay and one of the orange wires. The other orange wire connects to the com terminal of the relay.

I don’t plan on using all 3 V+ or V- terminals so I cut one of the V- terminals (beside the AC ground terminal) so that I could use it as the input signal to trigger the relay, I will simply connect this to the switched 12v from the ignition. I cut the terminal strip lead close to the PCB and bent it up leaving plenty of room to solder on my relay wire. I also need a ground connection but that just happens to be right next to it as one of the remaining V- terminals since they will be connected to the cars chassis ground anyways.

When 12v is applied to the new terminal (I labeled mine “T” for trim) the relay closes and the 2k resistor is now in parallel with the pot. This raises the voltage from 13.6v to 14.6xv.



Leave a Reply

Your email address will not be published.