mach 5 said:
36v 1000w brush-less motor (but it only has 2 wires?)
That would probably be a brushed motor.
If you apply power to the wires (like from a car battery) then the motor will spin at about 1/3 of the speed it would have with the original, if it is a brushed. Reversing the wires will reverse the direction.
It will not spin at all if it is brushless. If it is brushless, with only two wires, it means it must have a controller built into it, and the two wires are only power for everything in there. With no other wires, it means the controller must be wireless, so you would need to use the original control off the scooter (probably on an armrest or handlebar) to control it. Or else, open up the motor and replace the controller inside with an external controller.
and lots of Ni-Cd battery packs from emergency lighting, D-SC1500HX5 which is 5 x 1.2v 1.5Ah (6v@1.5Ah) in a row and D-SC1500HX2X2 which is 4 cells (4.8v @ 1.5Ah) in a 2x2 configuration - both wired in series. After reading as many posts here as I could I have decided I need fusible links
If you simply mean a fuse for the pack itself (right at the exit of the pack, at either the main + or main -), then that's usually a good idea, to protect the wiring from a short farther downstream.
You'd need to size this fuse so it won't blow at any current the systme is supposed to experience, but would blow if there was an actual short or other failure that you'd want power to be cut by.
If you mean for each cell, well, you can do that but it's going to be a lot of connections and parts. Not usually necessary, but it could, make a safer pack. There are a number of threads discussing cell-level fusing with various methods, if you look around (mostly in the battery technology subforum), probably mostly in threads with "fuse*" or "fusing" in the title.
and preferably a BMS to make some battery packs.
I've never seen a BMS for NiCd or other nickel chemistries. There might be programmable ones (expensive) that could go down that far, but I don't think you need one.
The problem is I'm not sure how many batteries to include so it doesn't discharge too quickly, do I just parallel then series to 36v 1000w or do I add extra so the NiCd's don't discharge too quickly and go boom?
You do not want to parallel NiCd or other Nickel type cells. The way their chemistry works, it is dangerous to charge them in parallel, and the more parallel cells there are teh more likely to get a fire or explosion out of it.
https://endless-sphere.com/forums/viewtopic.php?p=2670#p2670
When NiXX cells reach full capacity, their voltage *drops*, and the cell heats up to get rid of this excess energy. That means that the current that's still flowing to charge other paralleled not-yet-full cells will now flow thru the already-full cell that's already hot will cause the cell to get hotter and hotter. The other not-yet-full cells also won't get full, becuase *their* current will also flow into the already-full cell, since they're higher voltage than it is.
If the cells are not directly paralleled, but only at the main + and - of series strings of them, the problem is still there, just transferred to the voltage range of the whole string, and actually made worse: a single cell in a string that reaches full charge still drops the voltage of the whole string; if that is lower than that of the other strings, those strings (and the charger) will start dumping current into the one with the full cell, and that cell will heat up. Not only that, but now all the other cells in that string will get full quickly, and the string voltage will drop even lower, and more current will flow, causing even more heat in all the cells in that string.
Even if no cell actually catches fire, it will still heat up way more than it should, and will probably vent. That damages the cell (as does the excess heat), so it has less capacity than before, so it will fill up faster than before and cause this problem sooner each time, until it simply dies. (exactly what happens then depends on what happens inside the cell during this process).
If you do not actually fully charge the cells, and don't use their self-balancing feature, then this is not a problem and you could then parallel the cells. Won't have quite as much capacity (how much less you'd have to test).
But you will need to monitor the voltage of each parallel group to ensure they never go above the voltage that they would then drop to after reaching their full state. (otherwise if one of them does reach the full state, it'll suck the others down and cause the scenario described above).
Chargers with a single cell or series string (no parallel) normally detect the voltage drop of a certain amount and terminate charge when that happens. Some cheaper ones just use a thermistor somewhere in the pack to detect the heat, and when it gets hot enough they cut off charge, but these chargers can damage the packs, especially if they are frequently put back on the charger after only a short use. (many heating cycles).
You might be able to *discharge* them in parallel, but you would have to disconnect all the paralleled sets from each other before charging, and charge each set separately. This makes for a complicated pack setup that requires detachable parallel connections, and some way to ensure that it is impossible to charge it at all unless they are disconnected.
The power and capacity needs are determined by your controller and how long you want the kart to run. Since you will probably have to buy a brushed motor controller (unless you can figure out how to use the one that came with the scooter to do what you want), then whatever that controller's current limit is is how much current you need the batteries to supply.
Some examples with potentially useful numbers:
If you have say, a 25A controller, and run the kart at full power all the time, and need to run the kart for an hour, you need a minimum of a 25Ah battery.
If teh cells are actually still all capable of their full 1.5Ah, that's about 17 parallel strings of cells for that capacity.
At about 1.2v per cell, then to get 36v that's about 30 cells in series.
So at minimum, you'd need 30s x 17p cells, which is 510 cells. That is a LOT of cells to connect up (and monitor).
Now...let's say each cell is 1C capable (we don't know; you'd have to test that unless you can find a data sheet for them). That means each cell could supply 1.5A. Since the controller assumed above needs a minimum of 25A at max, then you'd need 25 / 1.5A = 17 cells. So you'd still need that 510 cell pack even if you didn't need 25Ah of capacity for a long runtime.
If the cells are 5C capable, that's 5 x 1.5Ah = 7.5A each. That reduces the parallel cell needs to 1/5, to 4p, or 6Ah.
If you used a 6Ah pack, at 25A continous, it'd last only 1/5 the time, or around 12 minutes, at maximum power continously.
