Project Daft Idea

Ok, finally fleshed it out to a point where i am happy with it. and leaves the main host processor having to only worry about reading the 4 voltages per board, and kicking out some bits to a shift register, now this can be improved upon, and if you dont care about 3.175A charge / discharge, then some parts get cheaper, most notably the mosfets, and the through hole resistor. The temp sensors are noton the schematic, but it would be a case of placing one onthe board very close to the N channel mosfets, and one against the battery,

per board, all in american $, prices from digikey, and this is single unit pricing, if your making a batch of them, that price drops a fair amount.

Total $12.01

IC's: $2.68
1x MCP6004 63c
1x MCP6001 35c
1x NX3L2467 98c
2x MCP9701 36c

Diodes: $1.16
4x CUS520,H3F 280mV Schottky diode $1.16

Mosfets: $3.28
2xVishay Siliconix SI2365EDS P Channel $1.08 (they only act as a switch, so can be low power)
2xFairchild/ON Semiconductor FDD306P N Channel $2.20(i planned this around someone trying to charge / dischargeat 3.175A into a3.2V cell, this means about 3.5W dissipated in the mosfet)

Resistors 1% 0805 $1.89
2x 10R = 28c
14x 10K =36.4c
21x 33K =54.6c
3x 68K =42c
2x 1M =28c

Resisitor 1% Through hole 5W $3.06
2x 0.2R = $3.064

image_cuefat.jpg
 
Interesting. I'm not an expert at reading electrical schematics, but I am getting better. But I am still a lil over-loaded with it all.

Now, is this just the slave for each cell, or does this include the master as well? I'm thinking just the slave portion. $10/slave is kinda steep :s

What's with all the resistors on the Q1-Q7? How come you have 2 PMOS in the upper right?Just reversed to make the connection.
 
Ok, so yes it is per slave, if i drop the charge / discharge to 1A, things get a fair bit cheaper,

No master, really you can use any micro you want to control it, all you need is ADC's on the master to read, and 4 pins for setting the mode and dac.

The resistor ladder makes a R2R DAC, it comes out cheaper than any other kind of dac i can find, and is less effected by component tolerance, as all resistors would be from the same batch it would likely be a fair bit better than 1%,

The 2 pmos up the top are acting as a switch, if your discharging a cell, you don't want it still connected, otherwise issues will arrive, and the back to back configuration means the body diodes will not conduct if you put a cell in backwards.

dropped to 1A, i would expect about $4 cheaper, go up to 20 slaves and i would expect another 15% cheaper., equally using 0402 resistors you drop a full dollar.

you can do it cheaper by pwm'ing a resistor, i was just thinking along the lines of very little software overhead, the master can be absent 99% of the time, and no problems should occur,
 
You must log in or register to reply here.
Back
Top