[hallkbrdz]

Build progress...

3D printed parts are in hand, look great. Only issue is the dimensions are slight smaller due to ABS shrinkage. So I'm having to bore and smooth the cell connections out, and slightly enlarge the copper plate paths. Certainly not show stoppers by any means. Next battery will have revised designed to allow more space for this issue. FDM 3D printing is somewhat of an art vs CNC.

If you look close, you'll realize a mistake and why one of these is being re-printed. Dooh!

Kelly bluetooth and various connectors have now arrived. The bluetooth will allow connection to a laptop or Android tablet for programming.

Motor coolant pump has arrived. It is a Toyota Prius inverter pump and will run off the tool battery to cool the motor.

Orion2 BMS and battery charger will be ordered as soon as the battery construction is completed (so I can use their wiring test tool without extra shipping).

 

[hallkbrdz]

Top tip: To smooth out 18650 holes on a 3D printed part, I found a 47/64" drill bit works pretty good.

Here's a rendering showing that battery configuration in it's packaging and how it will be situated on the racing kart. Four variable speed BMS controlled 120mm fans will be used for cooling.

 

[wjbitner]

Top tip: To smooth out 18650 holes on a 3D printed part, I found a 47/64" drill bit works pretty good.

Here's a rendering showing that battery configuration in it's packaging and how it will be situated on the racing kart. Four variable speed BMS controlled 120mm fans will be used for cooling.

I kept the controller behind the driver, as they can and will fail. Having all the power wires behind might be both shorter wires, and mildly safer. (one controller had a really neat semiconductor
plasma, despite having a reasonably sized fuse). Excuse the 'no chain guard' and such. I'm testing a battery configuration. How did you get the weight centered in the kart? That will
effect it's ability to corner..
Bill

View attachment 122039

 

[hallkbrdz]

Cool. Wow, that's a lot of batteries. Yes, keeping all the HV behind the driver as much as possible is a good move IMHO. Your custom frame really helps with that.

Here is a view at the same angle as yours showing the controller where you have your motor. There's just a real lack of space to put things, especially with a radiator and pump needed to cool the motor. Also AC inverters are rather larger than DC boxes which does not help. A design requirement was to keep plastic side pods, although I will need to either trim the ones I have down just a bit, or get newer style thinner ones.

Weight wise the biggest component is of course me, which is something I need to work on (LOL). There is no getting around it though, it will not be a perfect left/right balance as the batteries and protective case (aluminum 1/4" base and 3/16" sides) just weighs too much with this arrangement designed for a 2-stroke engine, even with minimal run-time high capacity cells. After this is built I am planning to approach Margay for a custom frame design oriented more like what you have with the motor behind the seat, with the seat centered, and the batteries split on either side of seat and inside the frame (wider and slightly longer frame).

Optimally I'd love to get some Tesla 4680 cells for that as the tab cooling would be much more efficient allowing for very quick re-charging. But for now I'll work with what I have.

 

[hallkbrdz]

Here is the end plate collector showing how a copper strip ties together two levels (for the Z fold). Experimenting to find just the right height and size of the bumps to make the easiest contact with the positive side of the cells.

Here's a look into the battery compartment with the lid off. Each printed part has a quarter inch of heavy foam rubber for shock and vibration cushion. Six layers of 7 for 42x10. Not shown are the Velcro straps used to hold everything together until it is secured in the compartment. The over-rated contactor and first 200A battery fuse are internal (there is a second fuse on the controller). The contactor can only be turned on by the BMS when it is satisfied with the condition of the cells, and no other safety switch is off. The green mil-spec connector is for all the BMS taps, contactor wiring, and temperature probes.

 

[wjbitner]

Here is the end plate collector showing how a copper strip ties together two levels (for the Z fold). Experimenting to find just the right height and size of the bumps to make the easiest contact with the positive side of the cells.


If I understand this correctly, you are relying on a pressure contact from the copper strip to the cells or are you attempting to spot weld them to the cells?
Bill
View attachment 122044

Here's a look into the battery compartment with the lid off. Each printed part has a quarter inch of heavy foam rubber for shock and vibration cushion. Six layers of 7 for 42x10. Not shown are the Velcro straps used to hold everything together until it is secured in the compartment. The over-rated contactor and first 200A battery fuse are internal (there is a second fuse on the controller). The contactor can only be turned on by the BMS when it is satisfied with the condition of the cells, and no other safety switch is off. The green mil-spec connector is for all the BMS taps, contactor wiring, and temperature probes.

View attachment 122045

 

[hallkbrdz]

>If I understand this correctly, you are relying on a pressure contact from the copper strip to the cells or are you attempting to spot weld them to the cells?

Yes, this uses pressure for the cell contacts without any welds. I tried the typical cell construction method first but found the nickel and nickel/copper strips will not carry enough current. To use pure copper I would need a $18k spot welder to weld to the cells. So I then tried a bake solder copper/nickel sandwich approach that proved expensive and difficult to manufacture.

So I re-defined the problem and came up with this instead. The trick here is to get the thickness of the copper just right so once worked (deformed) it provides sufficient resistance to deforming back to flat, while still being slightly softer than the cell steel posts.

 

[wjbitner]

Yes, this uses pressure for the cell contacts without any welds. I tried the typical cell construction method first but found the nickel and nickel/copper strips will not carry enough current. To use pure copper I would need a $18k spot welder to weld to the cells. So I then tried a bake solder copper/nickel sandwich approach that proved expensive and difficult to manufacture.

Cool! I pondered the same problems. I ended up soldering my cells, although I really wouldn't recommend it for safety reasons. I look forward to hearing how this solution works. Are you putting temp sensors on the copper plates to detect a poor connection?

View attachment 122052

So I re-defined the problem and came up with this instead. The trick here is to get the thickness of the copper just right so once worked (deformed) it provides sufficient resistance to deforming back to flat, while still being slightly softer than the cell steel posts.

 

[hallkbrdz]

> Are you putting temp sensors on the copper plates to detect a poor connection?

Yes. I've decided to add the Orion Thermistor Expansion Module that allows me to add a thermister for every bus bar, plus air intake and exhaust side to see how effective the cooling is and how fast to run the fans.

 

[Frank]

How thick is the copper and how much current per cell are you carrying?

Sent from my SM-T380 using Tapatalk

 

[hallkbrdz]

How thick is the copper and how much current per cell are you carrying?

Except for the terminal buses which will be 9mm at the terminal, they are complsed of 0.56mm strips. Total distance positive to negative leads between all cells at that thickness is about 4.6m. The area of the bus is 117sq mm, above the recommended 2/0 at no more than 1% loss that is only 67.4 sq mm. Although each cell can provide up to 35A (for 350A total), that will be done rarely for at most seconds at a time. Over the run of a race distance I will average about 8.5A per cell (or 85A total).

Here's a link to the spreadsheet I use for the battery/controller/motor calculations - best opened in Excel due to the macros on the first tab: EKartGo spreadsheet

 

[hallkbrdz]

Here is the current (and continuously more complicated) high-level wiring diagram. Still need to add the water pump and cooling fans. Fun stuff...

 

[Frank]

2/0 is safe but probably way overkill (extra weight). I suspect you could easily run #2. Please keep updating this post, you have an interesting approach to solving the copper problem.

 

[hallkbrdz]

2/0 is safe but probably way overkill (extra weight). I suspect you could easily run #2.

Agreed. 2 AWG cable is what I'm using for all the external links.

 

[hallkbrdz]

Yes... my replacement return 3d-part (with the cell order correct) has been printed and has been shipped. This time in white ABS. So, Monday I can start the assembly? Time to check the cells and get them all about the same.

Hopefully the 32mm frame mounts will show up soon so I can do a final measurement and start cutting on the motor mount and controller mounts.

 

[hallkbrdz]

> How did you get the weight centered in the kart? That will effect it's ability to corner..

Darn you, I keep thinking about that and it's true. So I spent a fair bit of yesterday thinking about this and how I could re-arrange the things that are moveable.

To start with, the seat, which will require me cutting off the right seat support and creating a new one. The upside is that instead of the seat being cocked clockwise about 15 degrees to the frame, it will be only a few degrees off as the steering support is off-center. That will both move the largest weight to the right, and make for a better driving position.

Second, that pushes the motor somewhat to the right, so I'll need to rotate it so the sprocket is in-board. That required a slight re-design on the motor mounts.

Third, I moved the controller from behind the seat to in front of the radiator. I'll need to work on the new support and cover for that. That does help in another way in that I can again use the mechanical water pump (as long as it pumps enough).

Fourth, the radiator needs to move upward so it gets clear air above the controller now located in front of it.

With all that it will still be somewhat heavier on the left side, but at least closer to balanced L-R for a frame not meant for any of this. Remix rendering (so far).

 

[wjbitner]

Darn you, I keep thinking about that and it's true. So I spent a fair bit of yesterday thinking about this and how I could re-arrange the things that are moveable.

To start with, the seat, which will require me cutting off the right seat support and creating a new one. The upside is that instead of the seat being cocked clockwise about 15 degrees to the frame, it will be only a few degrees off as the steering support is off-center. That will both move the largest weight to the right, and make for a better driving position.

Second, that pushes the motor somewhat to the right, so I'll need to rotate it so the sprocket is in-board. That required a slight re-design on the motor mounts.

Third, I moved the controller from behind the seat to in front of the radiator. I'll need to work on the new support and cover for that. That does help in another way in that I can again use the mechanical water pump (as long as it pumps enough).

Fourth, the radiator needs to move upward so it gets clear air above the controller now located in front of it.

With all that it will still be somewhat heavier on the left side, but at least closer to balanced L-R for a frame not meant for any of this. Remix rendering (so far).

View attachment 122099

I look forward to hearing about it in action! My battery packs (which are going into a Miata) weigh in at 37 lbs apiece and have 264 18650s per side.

 

[hallkbrdz]

Since it's rather dead at work this week (spring break), I decided to sign up for an on-line class on EV safety. Should prove helpful for anyone working with EVs:

Phase I

continue.weber.edu

 

[4Foxtrot]

Here is the current (and continuously more complicated) high-level wiring diagram. Still need to add the water pump and cooling fans. Fun stuff...

WoW seriously impressive - do we have a canBUS guru in da house ? 🙏

 

[hallkbrdz]

It's an EV, so it's going to need some screens...

Oh yes, screens...