FrankenWheel {{Landwheel L3-X Mod}}

Time for an update guys, here we go. Hope some people find it interesting because this took a ton of work! I am enjoying riding it so far and it has given me a lot more kick.

The upgrade: BMS + charge port, additional 3P of batteries, better mid enclosure, bluetooth module.

First thing is first - here’s some more information about the batteries I received. @KaramQ was kind enough to pass off 3 batteries from old landwheels & the first generation of the revel kit batteries that used the same cells as the l3-x. Upon receiving them during the summer, I didnt’ have much time to a whole lot with them except disassemble them into their pouch form.

This procedure was annoying as ever, but at least this time I knew what I was doing since I had done it before. Open the battery case, hack out the cables and lipo pouches, desolder the circuit board, rinse and repeat. Here’s some pictures of the disassembly:

These early version Revel U batteries had the same exact layout, BMS, and pouches as the old L-3X ones; this makes sense because the battery enclosures and interface were exactly the same as well. Given that I had 3 batteries and had already purchased the BMS, I was set on going with 8s. After opening them all up and removing the dead cells, I looked back in my inventory from the first time I had disassembled the original batteries, remembering that I had set aside the last few leftover good ones from before. To my dismay, I was only a couple short of being able to do 8s7p, or 10s6p, so I had to go with 8s6p, which isn’t bad considering I started with 8s3p from the original pack I had already made.

I don’t really recommend what I have done here as it is kind of dangerous, but here are the cuts I did to get the batteries out. These pouches are pretty well secured into their enclosures with silicon? adhesive on most of the sides. To get around this, you can cut the plastic out and bend the case to remove the cells. In addition to the silicon, the entire inside bottom of the enclosure is covered in double sided tape.

Step 1, cut out the latch area, pretty easy since the cells aren’t real close to it.

You can see that I have started to do the second and third cuts, branching out to the sides, along the vertical walls. These are very dangerous cuts because sometimes the batteries go right into the corner near the screw posts. I actually nicked a little bit of once of the cells when I was cutting and freaked out but fortunately it didn’t pierce all the way through.

After the branch cuts, you can slide down the sides, along the wall, to separate a lot of the silicon from the base and the cells. It is a pretty simple cut but you still need to be careful - this might be a good time to mention that I made all of these cuts with a dremel and the plastic cutoff wheel.

Here’s a top view of what it looks like with the front and the sides removed. After this step, you can peel the baseplate away and bend it down, sort of forcing the battery off of the double sided take and ripping away the remaining silicon.

Here’s a view of the second battery, with all of the cuts made at once. This is the one I nicked a bit I believe, since I was trying to make it go a bit faster.

Again I don’t super recommend doing this if you aren’t experienced with your dremel because sometimes the cutoff wheel catches in the plastic and goes deeper than you intend.

After separating the case from the cells, you can then remove the BMS carefully, slowly peeling away the adhesive so as to not cause weird pulls on the cells. Following that, the cells can be separated with the floss technique that I mentioned before.



For this part it is quite easy to cut your fingers on the floss, even though I was using screwdrivers to hold it, I found a few cuts on my fingers after separating all 30 cells.

Here’s all the good cells laid out and the extras, labeled with their voltages and eventually their polarities. You can see that they were mostly even, but with some variation within a reasonable amount. Their dead cells were seriously dead, like 1.5v or lower.

At this point I put the project on hold until I got to school and had some more time. (AKA until I found something to procrastinate on and do this instead.)

3 Likes

Next steps! I’m at school and found a few Saturdays to procrastinate. Many thanks to my roommate Roger who let me take up the kitchen table for almost a month of school insanity while I tried to work on it between classes and in my freetime.

Anyways, the next steps after labeling all the cells was to do the painstaking process of manually balancing all of the cells and bringing them to a universal charge.

First, I worked on each P group, bringing them to even charges if they were not already at them. I used a 1s charger board that plugs into a USB cable (TP50xx or something I don’t remember, I bought them a long time ago). It’s pretty safe because it only charges at 500ma. This process took quite some time, requiring the charging of each individual 24 cells when necessary.

Once all the P groups were brought to the same charge, I paired them all up into their 3P groups and set them aside until I was able to solder them.

I didn’t really take many pictures as I was soldering the cells together because it’s pretty routine stuff. I will explain the way I did it breifly - stack the cells together using their adhesive to hold them, then fiber tape the non-terminal end. Check the length needed to span across all three cells and strip a piece of wire to that length, tin it with a good amount of flux and solder. After that, I cut off the piece while it was still hot.

I then fluxed all of the terminals that I was going to solder together, and laid the tinned wire across them. I had the 3P pack stuch in a vise with a towel so as to not puncture the batteries. This process was really annoying, holding the wire in place and adding solder to it while simultaneously trying to not heat up the pouches too much.

Here’s a picture with most of the parallel packs finished off.

For some reason wiring up balance leads has always been confusing to me, so I chose to do it as I went this time. I slowly soldered together the S groups by joining the P groups along the line with small bits of insulated silicon wire. I added the balance leads right onto there, and later realized after that I didn’r quite have them right, so I ended up switching out the leads within the connector because Ihad already taped it up.

I consider myself to have pretty good soldering skills and I am confident that all of my joints will hold up to the test of road vibrations. As I went along, I taped each balance lead section to the batteries, trying to make sure all the leads didn’t overlap.

In order to wire the BMS to the parallel packs, I made sure that all of the opposing batteries were at the same charge, and then paired balance leads from the ends of the pack, counting down into the center. At this point I made sure to check the charge one the big balance connector, and I found some discrepancies - this is why you always use a multimeter to check before you plug it in and fry your BMS.

After fixing the discrepancy in charge, I did a final check on the balance of all the cells and plugged it into the BMS. Nothing fried so I was happy. Next I had to wire in the rest of the important cables, charge port, pos and negative, the BMS, etc.

I will insert my schematic here later, it’s on my desktop at school.

I made sure everything was well taped up and started adding some foam bits before soldering to the BMS and Pos, Neg.

I skipped a lot of steps here between pictures, but here’s a picture of the BMS with all the leads soldered. Between the last pic and now, I designed a new enclosure with an extension in the center to hold the BMS and wires without crushing them, and support the sides of the cells. I put in the charge port and soldered it into the mix as well.

I quite like this connector for a charge port, but I need to design a small plug to stick in it while riding, because there is quite a lot of open space when the connector isn’t plugged in.

At this point I had to get more screws and then wait for the last piece of the enclosure to print.

Here’s a shot of the finished product - so many hours were put into this. Unfortunately, the troubles weren’t quite over. Somehow I had managed to break my GT2b when I was soldering it, so I picked up a Vx1, and I quite like it for this board.

3 Likes

Nice work Ryan. Help me out wiht the math please. Each cell is approximately 4 volts. You group them into bndles of three in parallell, Still 4 volts, then you put them eight in series. So that would be 32 volts fully charged and I think that’s a 25 volt battery. peak voltage of 30 volts, discharged to 25 volts.

Am I right?

So you can get a VESC to run on a 25 volt battery. But for a landwheel VESC, you would want a 36 volt battery that has a peak charge of about 42 volts but undercharged to 41 volts.

So that would require 10 in series instead of the 8 in series that you assembled.

Am I correct so far?

And the reason that you did 8 in series instead of 10 is that you only had enough cells to do two banks of 3s8p. Thats a total of 48 cells. And you would have done 2x3x10 if you had 60 cells? Correct?

Or you coudl have settled for a single bank of 4p10s but then 10 of your cells would have gone unused.

Am I correct on all the combinations?

So you only need 12 more cells to achieve two banks of 3s10p Correct?

Everything you said sounds correct. I did 8S because I bought the BMS from SuPower before I got the 3 extra batteries, and because from the original 3 I didn’t have enough good cells to make a 10s3p.

So yes, I would need two more sets of 3p per bank to make 10s6p (so 12 cells). However, that would require a new BMS, rewiring of the balance wires, re-soldering of both the positive and negative at both ends, and somehow making more space on the board, which has basically run out.

In otherwords to expand to 4wd I need to use another 2 vescs or a maker x dual or something like that, since the landwheel esc is 10s.

And I have one more set of questions about the BMS wiring. What is the function of the balancing wires? do they charge each cell individually? Do they charge each bank of 3p individually? Probably not.

Do they draw excess voltage from the 3p banks that are charging faster so that the other 3P banks can reach thier peak voltage befor other cells overcharge?

And lastly what is the basic layout of the balancing wires? Is there a positive and negative balancing wire for each bank of cells? or is there a common negative across the series and one positive balancing wire for each of the 3s banks.

Another possibility would be to use two V2 Landwheel ESC’s. Those ran on 8s. You would just need to mechanically link two remotes.

Like I mentioned in the article, I don’t have access to the screenshot that I took of the wiring diagram that I drew until I get back to school tomorrow.

Essentially I treated it as if each 1s6p cell is one cell on the BMS. The SuPower BMS’s have the exact number of wires and use the P-? (I think, I don’t remember if if it is p- or b-)

How the whole charging cycles will go I am not sure yet but on the first charge it seemed to all balance out just fine. I purposely put the BMS right in the center of one of my modular case scales so I can just unbolt that one to check balance leads.

I am hoping that this BMS will have enough UMph to take the leak current from out of a 6p group so they don’t overcharge. I have found the SuPower BMS’s to be extrememly reliable for 4.20v balancing. Every single time I have charged my main board (also lipos) it has balanced perfectly. I am hoping this is the case here as well.

*more clarification:

The whole pack has one common positive and negative, it just looks weird because the positive comes out of the center of the pack. I stacked the batteries into it in a unique way so I didn’t have to resolder the positive lead, and then I tied the negative all the way around back into the negative on my antispark switch.

That is interesting, I didn’t know that they ever ran on 8s.

Since I am using the Vx1 that might be hard to tap into the landwheel ESC

Whats a vx1? Oh, a flipsky vx1 remote.

I was suggesting gluing two V2 landwheel remotes to a stick and joining the two joy sticks together.

However, the V1 ESC had a seperate reciever that coudl be replaced with any reciever. I saw them advertised on ebay with instructions for synching to a different remote.

The original Landwheels had 25 volt batteries that reached a peak charge of 30 volts. The batteries were too small and the voltage dropped like an anchor. So a kid could ride half way around the block or an adult could ride about 10 or 20 feet before the ESC grunted and powered down.

But when you substitute a 36 volt battery in to them, they are actually one hell of a peppy Skateboard ESC. I actually built a 4WD V2 Landwheel that ran on two V4 batteries. It was crazy fast but the brakes were flaky. 3 seconds of 60% brakes followed by 100% brakes.

I also built a travel V2 that ran on 18 volt Drill batteries wired in parallell for 36 volts with a peak voltage of around 38 volts. It was wicked fast and surprisingly had really good brakes.

1 Like

Haha that’s an interesting experience, I didn’t receive my first landwheel till the L-3x, so I never experienced a v2

https://youtu.be/RS5_tFmskoY

V2 Travel board experiment

1 Like

Lol that is dope, surprising that it just straight up works

More V2 Videos:
https://youtu.be/L8m2sFmebEE
https://youtu.be/LheMRbaKarM
https://youtu.be/KCJ5Sc8xyh4
https://youtu.be/B4wh1yvftY4

Yes, except for imperfect brakes, the V2 could have been a smash hit if it were not for the undersized battery.

1 Like

The problem with the drill batteries is that the wires came loose from time to time and I think that’s not too good for the ESC. I think I might have damaged the ESC during one of the disconnections.

1 Like

Finally got around to doing a real range test today. It was super fun to see how much this board has improved by adding all of the new batteries. I have much more consistent power throughout the whole range of the battery life and about the expected amount of range. This board already travels way further than my normal board, so that is pretty dope.

Here’s a screenshot of my ESC monitor data from after the ride. Looks like I managed about 23wh/mile.

I think it is saying that I drew 396wh but I don’t really know how to properly read that. Someone could correct me on that if they get it.

I ended the ride with about 32-35% left in the tank. That is a real-world charge too since I have the 0% set at 3.4v in ESC monitor. Given the nature of the ride, I was absolutely hammering hills into the battery, I think that I could make it much further on a more conservative ride. I believe at this rate I would have been able to get 3-4 miles more. But again if I was on more flat pavement, I would expect at the very least 20 miles. Here’s a link to the Strava track for anyone who wants to look:

I am pretty satisfied with this result and will continue to get more data as I ride the board more. Couple issues I ran into though. The first one being the state of the enclosure. Most of it is fine still, but the very front piece broke along a layer line. I kind of expected this to happen, so I will just reprint it in a better orientation. I had brought some extra bushings and washers with me to try out, which I did not end up doing, but I stuck a stick and a washer against the screw to limp along the last 3 miles back to my starting location. Surprisingly that worked out fine and I made it back alright. It did have a nut on it, I just took it off before I took the picture.

One of the other enclosure pieces cracked from a printing crack, so I will probably re-print that piece too. Every other part of the enclosure did fine though, and I expect it will last just fine. I hope to paint it all a uniform color one day to make it better.

Interestingly the Landwheel hub motors did just fine and provided way more performance than I expected, especially on hills. The acceleration of the board isn’t too crazy, but that is to be expected. One of the hub motors got MUCH hotter than the other though, and I am not exactly sure why. It also happens to be the one with the slightly loose sleeve, so I have to wonder if the lack of contact between the sleeve and the motor had something to do with trapping more heat.

Using my hand to test the heat of the motors, I would say that the one with the normal tight sleeve was around 60 degrees celcius and the hot one was probably around 80 celcius. I know approximate temps based on feeling 3D printer beds lol. @pkasanda maybe you would have some idea why one motor might be getting so much hotter?

Anyways, it was a good ride. I still will take my TB 110’s for comfort any day, but this is cool to have as a second board, especially with the great range.

I realized later in the day that I actually posted this in the wrong thread, so here we go.

I opened up the board yesterday and found the issue within like 10 seconds. Apparently one of the skinny phase wires had broken out of its bullet connector. Soldered it back on and we’re good to go. It’s a wonder how I made it 4 mi before it broke, I wonder how long it’s been underperforming?

1 Like

Did you finally found out the cause?
I have the same issue.

Honestly I am having a hard time even remembering what I was talking about then.

This board needs a revival sometime

Yeah, its long time ago.