APL's Pressure Contact Battery box

Small battery bank for phone back up. Had one piece with flash light, could not change the battery, mother is using the last one from 3yrs ago. This is what the cells were made for not welding.
Pressure contact

Like that pic. Like Lexan to see thru, pretty. Could have the Lexan closer together where the holes would fit around the cell ends with removeable caps on the end. Cap would connect the cell to the bus bar. The Lexan would replace the cell holder. The cap could just be a dimpled copper strap that screwed thru the copper into the Lexan. This way you take off a cap/end and remove the cell with out taking it all apart. Yes, lots of work! I like it.
 
Just a couple of reference charts for whatever reason. I think gauges differ slightly for different metals, this one is for silver. Conductivity, resistivity,.. silver is best of course, but only slightly better than copper. Doesn't oxidize as fast though.
(For approximation, can't guarantee accuracy of this info)

GAUGE
Thickness conversion chart for silver.png

RESISTIVITY
Approximate-electrical-conductivity-of-selected-materials-at-25C.png

MELTING POINTS
Melting points of metals.png

Thought about getting a roll of silver jewelry bezel, and punching 'dots' to solder onto the dimple tops. :lol: Fuzzy logic.
A lot tougher than plating, but still cost a lot. Keep on keep'n on... more updates soon.
 
The contacts on authentic Anderson Power Pole connectors are silver-plated copper-core. The "wiping" action of insertion is enough to clean the tarnish.

The fakes are aluminum core with zinc-plating.
 
Trampa boards used a PCB with a small disk with a resettable fuse inside for current carrying. They didn't give any details about that proprietary part.

Does anyone know of an available part that could be surface mounted on a PCB to provide electrical contact to a perpendicular cell?
 
drdrs said:
Trampa boards used a PCB with a small disk with a resettable fuse inside for current carrying. They didn't give any details about that proprietary part.

Does anyone know of an available part that could be surface mounted on a PCB to provide electrical contact to a perpendicular cell?
That sounds like a PTC. They can be used like a fuse or circuit breaker, but self reset when they cool off. You could use a small one for each cell like cell level fuses.
 
Exactly right. They were using a proprietary part with a PTC placed inside the disks on the PCB. That part is probably unobtainium.

I'm wondering if anyone knows of an alternative disk-shaped part to create a contact to the PCB or other ideas to use a PCB as the connection. Lack of a PTC isn't a deal-breaker.

trampa-weldless.jpeg
 
I found pure copper swaging stops - like a little donut cylinder

easily goes to a parallel smooth sided disc shape

with a big pair of Knipex parallel jaw pliers, 16" version I think it was.

Inserting fine stranded tinned copper boat cable first, say 10 AWG or down to 14 if that's all that's needed

gives a very solid primary power pair connection

suitable for physical pressure contact with cell terminals, or

I was thinking maybe using conductive adhesive, potting the ends with Poron caps outermost


 
drdrs said:
Exactly right. They were using a proprietary part with a PTC placed inside the disks on the PCB. That part is probably unobtainium.

I'm wondering if anyone knows of an alternative disk-shaped part to create a contact to the PCB or other ideas to use a PCB as the connection. Lack of a PTC isn't a deal-breaker.

trampa-weldless.jpeg

Recently i had one person asking me about my design on esk8 forum cause this trampa design failed. I think its too complex to fabricate something reliable with pcb's.
 
I've been drilling holes for 1/8" pop-rivets through this (lowest grade) XXX Phenolic board I bought, and noticed de-lam chunks coming off the bottom side once in a while. Hmmm,.. so I took a hammer and beat up a small piece of it and it basically shatters like glass!

That's what XXX will get you I guess. Live and learn. It will be fine for what I have glued around the spacer core, but for the side
compression covers I'm going to have to switch to something better. I'll keep going with the phenolic for now, which is OK for testing
and it will make a good pattern-maker for the new stuff.

I took a piece of the green Garolite FR4 board, (fiberglass based), and smacked it up good with a hammer and it just bent a little, but never broke. Super tough. So I guess I'll probably go with some of that. Much more $$$, but necessary.

Now I have a big pile of XXX to make things out of. :lol:
 
drdrs said:
They were using a proprietary part with a PTC placed inside the disks on the PCB. That part is probably unobtainium.

These are a few of a number of pages a google search finds for "PTC discs":
https://fuse-china.en.made-in-china.com/product/xBNmCOKJAwYt/China-USB-Drive-Thermostatic-Ceramic-Heating-Element-Filter-PTC-Disc-Thermistor.html
https://www.diverseelectronics.com/products/Ceramic-PTC-Thermistor-20025-4-7Ohms-20-140V-Disc-Type-p9537/
https://www.ntcsensors.com/Disc_Type_Thermistor/

I don't see enough details on many of the pages to know what they're actually selling, but there are multiple physical sizes of PTC discs there, as well as varying resistance/etc specs.


THe problem with most of the ones I see is their high resistance. You'd need one that is very very very low resistance under normal conditions, less than the cell's resistance if possible, so it doesn't interfere with current flow unless there is a problem, but will go to a high enough resistance to prevent a problem when it does happen. It must also be fairly insensitive to heat below a certain point, so that it doesn't begin to increase in resistance just because the pack starts to get warm, or else the PTCs themselves will begin generating heat in an increasing cycle just because of the heat they themselves are generating, and effectively shut the pack down. Additionally, PTCs nearer the core of the pack will be more affected by this than those near the edges, as heat builds up in the pack. There was another thing I thought of while typing but I can't remember what it was now.

Most likely these types of companies will make anything you want, if you order enough. ;)


AFAICS, though, PTCs are only really useful to keep a completely dead-shorted cell from causing a fire by pulling so much current from other paralleled cells that it or they overheat. In other cell-protection cases I can imagine, the current might be too low to trigger them without having such a sensitive PTC that it could interfere with the normal current flow from the cell. But I haven't done any math on it, so I could be wrong.
 
I have the copper temporarily riveted to the non-spring side plate. Or rather, the rivets are just setting in place, (non swaged).
I have to take it apart to cut the strip sides for thru bolt clearance yet. The other side of the rivet is recessed, and will be epoxied
closed. I'll use a few more rivets in the final assembly

Here's a picture of what I was thinking for bolt distribution for starters, but welcome any suggestions. I should probably alternate the placement of the two centermost lines of bolts, 11P makes it uneven. Not sure about the end bolts, (yellow),just figured the end buss should get a little more clamp assurance.

I can drill more holes and the bolts can go through the center of the strips too,.. as they'll be doing on the other side.

Thru bolt placement 1.jpg

I was thinking each bolt should compress about 4 sets of 4 dimples around it's axis. Hopefully that's reasonable.

Here's a batteries eye view of the box in it's vertical position. Side by side battery arrangements are not the best way to go for space, but it does make the thru bolt placement easier.

Side view spring side.jpg
 
Thanks!

I'm thinking I might have too many bolts in that last picture though. I'll have to think about this a bit. Originally wanted to do it
with 12 bolts.
 
Think I figured out how to attach the flex-plate strips, using a washer or a fiber washer inside the relief holes on the strips.
Pop-rivet the centers. I tried forging the back side of the rivet like the tops, (steel and alum.), with a set of punches, and it worked out pretty good. Nice and flat on both sides, and the strips can still flex freely. (waiting on the thru-bolt pattern)

Flex plate rivets.jpg

Also decided to bond some cheap T-nuts to a large washer and the phenolic plate, for the back side of the box. Comes out flush
on the inside and allows for bolt insulation all the way to it. Found out that a 10-32 threaded nut fit's a 5mm bolt better than the 5mm nut does. Probably use a drop of blue Loctite as well.

Still pondering the bolt pattern, and can't really move on until that's done. I think a more staggered pattern is better, and this one
gets down to 18 bolts, which is better, but still a little uneven. I think the strips are throwing me off, and I'll have to try to look past them.

Thru bolt placement copy 2.jpg

I think I'll need to place all the washers on the plate and see how they physically fit first, and then go from there. I don't have all the parts yet and need to get things ordered.
 
OK, that helped. Trying to see a pattern through the strips was just making it worse. Placing the washers on the spacers makes it easier to see. There's no way to get an even pattern with 16S/11P, and it's always going to be off balance a little somewhere.
Might tweak it slightly, but I think overall this looks about as even as it's going to get. (20 screws.)

Spacing 20H version.jpg

I'm going to go for it anyways, so now I can move forward and get some things done! :roll:
 
That look good enough to me, especially if that surface material being compressed by the washers is rigid and stiff.

A poron layer between its other side and the copper button sheet may help maintain compression pressure and keep everything locked in place laterally under shock/vibration while in use.
 
Great! Yea, the G10FR4 board is plenty stiff, and the pressure won't be all that much I'm thinking. I don't know how much actual
pressure the Poron will deliver to each dimple,.. TBD. But if I have to, I can use 6mm silicone sheet, which will deliver much more.
Just have to try things out.

Washers will be slightly smaller and I plan to make my own countersunk washers for the flat head screws so they'll match the back
and have minimal height. It all gets painted black so it won't be so visually 'loud'.

Time to buy the BMS for it,.. so I was looking for a Daly unit to fit the bill, and noticed that the blue tooth versions will display the
parallel cells to monitor any real time differences in outputs. Just a thought, but that might be a good match for pressure packs.

Don't think I'm going that route yet though, and I think a standard 60v/80-100A unit might be a reasonable choice for the mid-drive. With the new box, and used batteries of unknown quality it's best to check things out as I ramp up outputs. Need to upgrade some
other parts yet too.
 
Thanks for that ZeroEm, the spring washers are a great idea. If they could be applied to each contact I think that it could very well be the ticket. :thumb: (would go well with a rivet type contact idea)

Speaking of springs, I was thinking of a thru-bolt idea that uses a magnet on the end of each bolt in the middle of the pack. The two screws would never actually touch in the middle, but the attraction of the magnets would be the spring pulling them together
OK. kind of half baked idea, but the plus side is that the two sides would be electrically insulated. Just a thought.

Anyway, yea moving slow here too,.. the shop is cold, the days are short, and the hardware I'm buying is obsolete by the time it gets here! Plans keep changing.
 
A little more progress, think I'll call it a Big Block. :wink:

This is what the 8-32 screws look like with 5/16" polyethylene tube on them, and some cheap countersunk washers. Looks like enough insulation,.. hope that's not famous last words.

Thru bolts.jpg

I bonded all the washers on, sprayed some flat black primer & a marker pen to tone it down for the photo. Starting to get an 'atomic' look, or maybe just an old road sign, not sure which. :)

Spring side.jpg

Hard side.jpg

Anyway, still have some tweaks to do yet. These are the test side plates, and I can probably leave the first set of large washers off on the next Garolite versions, for a lower stack height.

Now I can rivet all the strips in place and start wiring things up. The BMS is on the way, and I'm still working on the P connectors layout. Need to see how much actual room there's going to be for them first.
 
BMS finally arrived in the mail, went for the 100A unit to cover future needs, but wasn't quite ready for the wire size. :shock:

BMS fit up.jpg

The mounting placement will be in back, next to the battery box wires, and the old 40A controller is getting shelved. I need to make the mounting plate for the BMS yet.

Depends on how much room there is, but it would be nice to use two 8 pin Deutsch connectors for the balance wires, the BMS has 17 pins. So I probably need an 8 pin and a 12 pin. (Deutsch doesn't make a 9,10,11)

One thought was to run the 17th negative pin along with the Neg. battery wire, so I can use the two smaller 8 pin connectors. :?:
Anybody know if this is taboo?

Modified circuit;
Daly 100A circuit change 1.png
 
I think you could get away with it. Voltage monitoring of cell group 1 might a bit off under load, but I bet not much. Not sure if noise from the FETs switching would affect it much.
 
What nags me is why they even have two negatives in the first place, if they both go to the same place. I can see where there
might be a problem, (in this situation), where the battery is disconnected first and the balance wires are still connected without
a negative.

Well, probably best not to mess with the unknown with so much at stake, just curious. I'll buy some connectors and see what kind
of room I have to work with first. There are other smaller connectors too,.. I don't plan to remove the battery that often.

There's still the possibility of going without the connectors at all, as a sort of dongle, which I don't like much but should explore yet.
The current plan is to make the BMS base plate attach magnetically to the frame, so it can be removed easily for the connections or whatever.

I need to get some of this stuff worked out before I can start soldering up the wires,.. so steady as she goes.
 
That looks like it should work OK. One thing you want to avoid is allowing a condition where the main B+ and B- are applied to the BMS without the balance wires connected. Since you have the B+ connection only on the balance connector, you should be safe. Unplug the + side balance connector first and plug it in last. This will prevent applying half pack voltage to the lower half of the BMS.

All this may not matter depending on how the BMS is designed, but I've had several failures caused by not connecting the BMS in the right sequence.

What can happen is when you apply voltage to a large section of BMS without the balance wires attached, the voltages may not divide perfectly evenly. This can cause one cell to go high enough to trigger the balance shunt. Once this turns on, the remaining channels will see overvoltage and start turning their balance shunts on. The last channel to turn on may have way more voltage across it than the rating of the parts and fail. Usually the FET that switches on the balance resistor fails and shorts, causing that cell to drain constantly, which can damage the pack.
 
Back
Top