A123 20AHr Pouch Cell Battery Build & Info Thread

Might have got lost in translation.
There is no vaseline/grease on the mating faces only on the bolt, as i suggested stainless and it can bind sometimes if its really tight and over the finished bolted connection to inhibit corrosion.
Edit the bolt's not there to conduct juice.
 
the bolt threads into the plate which is the 'mating surface'. as the bolt screws in the grease is stripped off the threads and ends up on the plate around the hole which is the 'mating surface'. i don't see how you can keep it off the contact. has anyone actually done this with any success? the bolt will conduct if it is in contact with the lug, just not as much since it is stainless. but the threads have a much larger contact area imo.
 
I thought flexy said he was using like a plain copper bus style clamping method, copper clamps and bolting through the tabs :? :D Has that changed.

I think Stevo's is like you said and looks like Allen screw into a clamp.

The bolt doesn't want to conduct and the 5A it would in 200A wouldn't matter anyways.Trying to minimize the contact resistance of the joint by clamping down tight on the mating faces, standard practice, though yeh any grease on those faces would be a pain.
So maybe should say hold faces together, bolt goes through, tiny amount of grease goes on thread only, then torque up nut, smear grease over finished connection.

Edit should add flexy, if you've got nice bright copper try not to scratch it too much and after you drill i would countersink the burrs of the holes, not file the surface, so its still got a shiny un scratched surface, sounds like your cluey on this stuff anyways, slap me if i sound patronizing :lol:
 
FWIW, I've seen vaseline used on car battery terminals, coating the post and inside and outside of the clamps, then sliding the clamps on and clamping them down still works to start the car, so I guess as long as the clamping force or whatever is sufficient to move the vaseline out of the way of the contact surfaces, it should still work.
 
Expect what's happening is the vaseline is squashed out from around the microscpic peaks.
So they are still making contact.
Though vaseline would be a good insulator even with those additives they do to to help conduction,
the research says better without on dc stuff.
Its good for an overcoat, would say.
 
wasn't trying to diss your idea, the antiox paste used on Al wire and to coat the weather exposed wire splice crimps does get mashed outa the way by the screw torquing down on the wire in the box or the wire nut jamming the strands together as with the crimp.

would not think it needed though with a stainless bolt, but i do use a drop of light oil on the threads when i put the lug nuts on the car and truck wheels because the are exposed so i do understand and appreciate the idea.

the exposed aluminum tabs really are the weak spot and keeping moisture outa that contact is crucial imo. i still like JD's split pin twist and spot weld technique.
 
No worries dis away buddy, I could well be wrong :D
With the vaseline thing i meant instead of antiseize if it were also going to be used to coat the finished joint, yeh could use some other spraycorrosion inihibitor and all over the exposed tabs edges as well would be the best, didn't consider that, important, need to do mine, sitting in the shed, pretty damp overnight, thanks.
I did remember to buff/clean all my connections with a really soft cloth and alcohol before clamping them though :D.
Good to see all the different build methods going on.
With mine personally, i liked the idea of keeping the tabs as they came and ended up with an external pack clamping method, no drilling or soldering, so i can dismantle and revamp fairly easily.

Note Ive done 132/275kV ac, aluminium crimped to 5 tonne and Ive seen some of the finished cable fail microhm test due to the type of grease :?
 
Thanks guys,
I was mainly concerned that the flat base of the ring terminals is a little rough so I might lap them prior to fitting ,also their cross section is smaller than anything else in the conduction path.

I will threadlock all the clamping bolts (learned that lesson on a headway pack build) .

I was going to coat all tabs and bus bars with vaseline to keep moisture out and avoid any corrosion, I dread the thought of moisture wicking in and rusting it all to hell.

I was also thinking of coating all the contacts and copper bus with silicone once it's assembled and tested, the vaseline would allow it to be peeled off later, if repairs were needed.
 
If their tinned, can you hit them lightly with a blow torch get tinning flowing,
is that what you mean by lap?
Get a very clean wet cloth and wipe it quick in one stroke.
Should give a smooth flat finish then.

Yorkshire, you'd see a decent amount of damp all year round, id bet.
Been there few times up to Harrogate, The other place with the crooked spire, forget :?
Best mate was from Killamarsh, if its spelt that way, his dad was a coal miner, god rest his soul.

Yeh thread locking, good idea, i was going to say fat spring washer there.

Like to know how your coating finished product, buddy?
 
megacycle said:
is that what you mean by lap?
Click to expand...
lapping is a process to create perfectly flat glass-smooth surfaces; the most common use most people might know of is for CPU heatsinks for ultraperformance systems, so that you get the absolute best thermal contact surface for heat conduction between CPU and heatsink.

Commonly can be done by securing extremely fine sandpaper to glass, then rubbing the flat contact surfaces on that. Depending on your final purpose for the item being lapped, you can use finer and finer grades, then polishing pads/powder, until you get a flat mirror finish with no ripple distortion in it. Takes a lot of patience. :)
 
Dont know if that's a recommended method for electrical contact, dont need a mirror, the surface of bright copper bar should be good to go unless its been scratched or green, then linishing Edit or buffing sounds similar to lapping, but no thermal paste :D
 
Fishmasterdan said:
When you (anyone) get the pack building figured out let me know. :D and how much you will charge me to build me a pack!!! :wink: My time is very limitied till july.
Click to expand...
I found the whole experience of building a pack using Agnius' kit very rewarding and not so difficult. I'd be willing to build you pack if you're interested at cost plus a fixed hourly rate. Let me know if you're still interested.
 
ohzee said:
Once home I sat down with the battery and tested all the cells to see where they were.. I had one cell near the middle of my pack which was noticeably lower then the others.
I am very happy with this battery , but my amp draw is quite low.
Click to expand...
I just finished replacing one of the cells in my 12S pack and am still cogitating on why the cell went kaput. It was in the middle. There's also another cell that is consistently lower in voltage than the others. Strange enough its the next one in line. So cells 6 & 7 were the bad guys. Coincidence?
 
arkmundi said:
dnmun said:
lifepo4 is fully charged at 3.65V and considered fully discharged at 2.1V.
Click to expand...
How can it be fully discharge and still retain voltage? What's the "safe" discharge level for these cells, without doing any damage? Thanks.
Click to expand...

Stated safe limits. bottom is 2.00V, top 3.65V. If you go lower or higher, you most likely will damage your cell. From what i read to proper treat cells, discharge to 2.1V and charge to 3.55V, this would prolong cell's life with loosing something like 0.5AH
 
Yea that's something that concerns me with my eva assemble BMS.

They list 2.0 low and 3.9 high. Not sure if I am misreading that wrong. I usually shut the charger off before it's done just because.

http://www.evassemble.com/index.php?main_page=product_info&cPath=6&products_id=46
 
Let's just hope the replacement cell doesnt go bad too. How many replacement cells do you have?
 
I know what you mean flexy, with cabling from the packs.
Im not going to terminate with a lug or use connectors, with say 200A draw.
all adds to give additional points for failure and voltdrop.
View attachment 1
Should be able to see the copper bus sticking out.
I want to use flexible braid connections for my pack interconnects and power takes off's.
Like these copper-flexible-expansion-connectors-19751-2607201.gif
If i use say a Kelly controller, has bolted flat connections and so basically the whole connection arrangement is bus, no circular cable.
Might even leave all that copper naked, all going to live in panniers anyways, hopefully.
Maybe use your coating method mate, beautiful.
 
ohzee said:
Yea that's something that concerns me with my eva assemble BMS.

They list 2.0 low and 3.9 high. Not sure if I am misreading that wrong. I usually shut the charger off before it's done just because.

http://www.evassemble.com/index.php?main_page=product_info&cPath=6&products_id=46
Click to expand...

yes, you misunderstand what the numbers are. the 3.9V level is the HVC that causes the charging mosfet to turn off.

if any cell reaches 3.9V then the BMS stops charging until that cell has drained some of the excess charge off, usually down to 3.8V before it allows the charging to start up again.

this is different from the shunt transistors turning on in order to divert current around the cell when the cell has reached full charge. since the charger puts out more current than the shunt resistor can absorb, the voltage will continue to rise on the full cell until the HVC is reached.

eventually all the cells catch up and that one cell that caused the shut down is not longer so high in voltage and the pack will charge at a reduced rate determined by the sum of the shunt resistor current flowing around the already full cells. this is when the balancing is actually taking place because this is where the low cells finally reach the same level of charge as the high cells. that is called balancing.

after the pack has been balanced by the BMS the first time and for all the following cycles, the cells will all charge up together and there is not a problem with one cell causing the charging to shut down for HVC. then the balanced pack will just charge up with all the cells climbing in voltage the same amount so there is no cell that sticks out of the group and causes the HVC to trip. so HVC hits are an indication of an unbalanced pack and why the new packs take so long to balance the first time, and when they have sat around uncharged for long time so they self discharge to different levels with age and abuse.

overcharging and over discharging make it harder for the cells to balance to each other because the abuse causes the internal resistance and capacity of the cell to change.

if you always turn off the charger before the pack balances then it always has undercharged cells that can cause the pack to shut down for LVC before the other cells have been fully discharged, so you lose a lot of the available capacity of the pack.

you are then essentially taking a 15Ah pack and converting it to a 12Ah pack. this is why none of the guys who refuse to balance their packs with a BMS ever post up the results of the capacity tests on their packs. because they have lost so much capacity running it without a BMS.

it does not damage the cell for the voltage to climb up to 3.9V and then drain down. if you allow the charge to continue and push the cell voltage over 4.2V then there is some unquantifiable damage. when it gets over 4.4V the damage can be noticeable as reduced capacity and increased self discharge rates and increases in internal resistance which make the cells much harder to balance because the cells now do not have the same resistance when connected in series to the bulk charger so the current from the charger flows preferentially to the lower resistance cells. above 4.6V makes you wanna cry.
 
megacycle said:
I know what you mean flexy, with cabling from the packs.
Im not going to terminate with a lug or use connectors, with say 200A draw.
all adds to give additional points for failure and voltdrop.
View attachment 1
Should be able to see the copper bus sticking out.
I want to use flexible braid connections for my pack interconnects and power takes off's.
Like these
If i use say a Kelly controller, has bolted flat connections and so basically the whole connection arrangement is bus, no circular cable.
Might even leave all that copper naked, all going to live in panniers anyways, hopefully.
Maybe use your coating method mate, beautiful.
Click to expand...

Ah, that's a neat build, those braids look the business. Unfortunately my pack has to be installed in separate modules so cabling and connectors are a necessary evil.
 
Was looking at your set up again mate, 200/6 =approx 33A per cell max, not much load really.
Do you want to contact all of the tab area.
As i was hinting at before, if you even just snag 10mm of tab under your clamp, though it would have to well clamped, there should be a similar current flow in the tabs that have 25mm clamped, so you dont need to strain the tabs to get the 6S under your clamp, sounds counter intuitive, streamline effect.

Was thinking with mine, similar to others using 1S for bikes, if the individual cell's are carrying high load say 50A+ then a full tab connection to the bus or clamp might be recommended, so they would help sink the heat away from the tab some.
 
Found this excellent review/ reverse engineering job of the evassembles cap bms.
Hope its helpfull, if you Havnt seen it before, mine are still in the box teasing me :lol:
http://www.avdweb.nl/solar-bike/electronics/bms-ae-lmd17-rev-a1.html
 
Yea m8 do whatever makes you happy. Like I mentioned before I have a buncha lipo and I balance monitor bulk charge etc..

I just like not having to do that with my 20ah cells.. plug it in and forget it for the most part.. I have a 20s 20ah with a ping BMS
and a 24s 20ah with a eva BMS.. I like the ping bms because of size and lights , but the eva one can put out a lot more amps.

I have a couple 12s high power BMS from cell_man still that id let go cheap. No lights , but they are rated high power. I can get a
link if you want.. if your not needing high power id check with ping and get one of his.. it's signalab..

still BMS's are like opinions.. or is it a$$h0l3s ? hah dunno but everyone has their own ideas.
 
You must log in or register to reply here.
Back
Top