hacking of 14S lipo pack built with Bestechpower D126 BMS

[dnmun]

i posted up how i built this pack from hobby king lipo packs over on lee's thread: http://www.endless-sphere.com/forums/viewtopic.php?f=14&t=54024&p=804676#p804676

i mentioned already on the Bestechpower BMS thread http://www.endless-sphere.com/forums/viewtopic.php?f=14&t=50207 that i had the BMS begin to overheat above about 17A and could keep it at 20A for very short period before it got hot. at 22A it will go for a short period, about a minute before it begins to go into thermal runaway and gets very hot fast so i cut back on the power and decided to hack the BMS to see if i could make it handle more current if i doubled the number of output mosfets from 3 to 6.

i also felt that the small surface mount 2512 shunt resistors were getting very hot at 20A and that heat was causing the mosfet next to them to overheat first and go into thermal runaway. so i decided to also add some more shunt resistors in parallel to help keep the shunt cool, and i also decided to bypass the shunt resistors with all the current i am taking from the additional mosfets so i put a jumper wire directly to the B- terminal and bypassed the shunt resistors.

originally there were 4 10 milliohm surface mount 2512 resistors for the shunt. 2.5mR equivalent. i added four of the 4 milliohm 2512 resistors to the top of those four and ended up with a stack with an equivalent resistance of 2.5/3.5 or .658mR for the total resistance. this is about 1/3 of the original resistance so the heat produced is about 1/3 of the original and since i am bypassing the shunt with half the current using the copper wire to the B- spot then i can get twice the current through the BMS and only make 1/3 of the original heat in the shunts.

after i added the extra three mosfets i tested the output of the D126 now and i am only able to draw 30A through my dummy load of 5 1500W electric heaters at 60V so i could not test it above 30A. maximum power was about 1700W and it decayed to about 1300W over the 8 minutes as the pack voltage dropped. not too bad for such a tiny device imo.

i discharged the 5Ah pouches at 30A initial dropping to 27A after about 8 minutes when i stopped. the mosfets had not really started getting hot by then but i could detect some warmth, but slight very slight heating, as when i was using just the 3 mosfets at 17A.

 

[agniusm]

Dnmun, this is great info on these Bestech BMS, i am glad i found them and appreciate your efforts time and work getting them tested and then folks up here informed. This is most valued info.
Regards

 

[dnmun]

we all owe a debt to you for finding them.

they are so much better quality than the ping signalab and BMS battery BMSs and they offer them in so many formats and they really are so cheap.

shipping is the biggest expense so i have tried to keep the air freight cost down by ordering a bunch of the BMSs at once and he has been giving me good prices when i order so many at once.

 

[powersupply]

How about putting those FETs beneath the PCB?
The same layout of the top FETs, but one level below.
That way their rear is exposed and can be heatsinked if needed.
You'd only need a longer screw and a few nuts.
It would be mechanically much more stable and you could even use some of the same pins (sans Rs).

 

[dnmun]

on the underside there is no space to place the mosfets because the ground plane for the source legs goes right up next to the spot where screw holes come through and there is too much chance to short it out. this was more of a hack and not really meant as a realistic option for higher current. in that case the larger BMSs would be the better choice.

a heat sink on the underside would be in the way also since i pull the drain legs down to the copper wire where it goes through the hole in the solder lug i used to make contact.

also since the end of the solder lug i use is serrated it will dig into the tabs of the mosfets to give better electrical contact, and the hole is so large that if it were on the underside it could short the drain to the source ground plane.

 

[powersupply]

Are the holes isolated to the FETs?

Here is what I meant.

Two lockers elevate the FEt so it's body is above the solder added low ohm path beneath.

 

[dnmun]

yep, i thought of that because the legs would be in order and i could just solder them to the bottom of the the through holes where the legs on top are soldered.

but i wanted to bring the drain current over to the P- spot on a separate conductor and i wanted to bypass the shunt resistors on the source leg side because of the heat from the shunts so i ended up mounting them inverted like that.

i even considered face to face at the tabs but this just worked out for me since it was all stuff i had on hand. i have a big bag of those solder lugs that i bot for the headway cylindricals. of course they did not fit.dammmmmmmmn

 

[powersupply]

You could still do that, the same trace where you connect your solid copper wire is available beneath, same spot where the many vias are.

I just wonder how your BMS will do with all the possible vibrations.

But as you mentioned, its only a hack/test anyhow.