Yamaha YZ450F

[j bjork]

I started to get worried when I didn't seem to find any weak cells at first, until I got to 7.
That seem to have very high ir. The voltage went up right away, and the charger stopped at 3mah.
That could explain why it dropped some voltage after charge, and also that it seems to bounce back rather much after discharge.

But I continued with the rest of the cells, and when I got to 10 I got this:

All the others stopped around 3600mah, this just continued charging at 0,6A.
I turned it off and the day after the charger shows this:

 

[j bjork]

It started leaking coolant from the motor recently, from the joint between the center housing and the transmission cover, just above the mounting screw. Took the motor apart and had a look:

Hard to say why but I thought the surface on the central housing was a little darker in that area.
Might be from the transmission cover, where the surface was partly panted.
But I also saw this:

Material from the motor has stuck to the mount, so it must have moved even though it was mounted properly.
That's why I am suspecting that it could have moved in the sealing area too.
Well, I cleaned it up and used a screw with a large head that cover a larger area.
I guess time will tell.

This time I was able to pull off the gear:

Here I found some oddities..
If you look at the upper keyway, you will see that it is not parallel, but the locking key is.
On my finger you can see some metal shavings from the key..
I cut a groove in the locking key so it fit in the groove in the gear, it probably made the gear a lot easier to mount

 

[j bjork]

Ok, went to the track yesterday for a test ride.
After the first lap it was leaking worse than ever
Rode a few more laps and had a look at it today:

So, here is the problem.
There is a maybe 15mm long crack in the center housing

 

[bananu7]

Drill the end of the crack, dremel out to V shape, patch with metal epoxy.

But I guess you know how to deal with stuff like that. Do you think it's because of the movement on the motor mounts?

 

[j bjork]

Drill the end of the crack, dremel out to V shape, patch with metal epoxy.

But I guess you know how to deal with stuff like that. Do you think it's because of the movement on the motor mounts?

Yes, I think it is because it is moving. And especially as it is the motor housing that cracked, I think that is what does most if it.
I don't think an easy fix will cure it, I think the mount on the housing is too weak so it moves and crack.

If you look at the picture with the motor apart in post 227 you can see that it only attach to the thinner material in the water channels. I think it should have been attached to the thicker material between the channels.

I wonder if I can weld not only the crack, but also reinforcements on the housing without burning the stator
Or if I can get it out and in again without braking anything..

 

[j bjork]

The motor is disassembled:

I wonder if I can get the stator out of the housing without damaging it?
There is this screw, I suppose there must be something else holding everything together:

This is where the crack is:

'

The material where the crack is, is just 2,5mm thick. It is actually thinner than the rest of the cooling channels, they are 3mm.

The rotor is 84,5mm dia, 100mm long.
The stator 85,5mm dia, 100mm long.

 

[gustavo88py]

write to the manufacturer.
has to tell you how to remove the stator.
and how to reinforce those supports.

 

[JeroenM]

In order to remove the stator from the housing you have to take out that set-screw, and then heat up the housing to at least 150c with a heatgun or torch.
With this particular motor, you might need to heat it longer or use more heat because the outside of the waterjacket doesn't conduct heat very well to the inside.
If you are lucky, the stator will simply fall out, but some times you need to gently hit the stator with something soft, like the wooden or rubber handpiece of a hammer.

I have done this succesfully on several motors, including QS138.

When you put the motor back together, make sure to position the stator correctly, so that the little dent created by the set-screw is in the right location.

 

[j bjork]

I tried with just heat first, but wasn't successful.
Made a little tool for the press:

Actually, I didn't use the square tubing. I found a better steel plate that covered all of the tube.
Some pressure and a lot of heat, then eventually I had this:

 

[JeroenM]

Excellent job with that tube/ring, it will also prevent the laminations form bending.

Nice to see that the stator was really tight in there, that should allow for good thermal transfer between it and the waterjacket.
How are you going to reinforce the thin wall sections around mounting points?
The designer really should have rotated the mounting points a few degrees, so that they attach to the thick wall sections.

Good luck welding

 

[j bjork]

At first I planned to weld reinforcements on the outside of the housing, but I am afraid it will distort it.
My new plan is to make a thick plate that fits between the transmission and center housing, that fits the mounting points on the bike. Then take off that thickness from the center housing on a lathe or mill.

 

[j bjork]

I think I have a winner:

Now I have to decide how thick to make it, and in what material.
I should be able to go up to 20mm if I want to, but it would be heavy obviously.
I could maybe shave off some more material in that case

 

[j bjork]

I tried to move forward in repairing my battery.
The problem is that I have 1mm copper bus bars that I have to make something stick to.
I tried welding 0,1mm nickel to 1mm copper with my k-weld, but hasn't been very successful.
I have used a 12v lead acid, but it only gives me about 1100A .
When I tried a 4s LFP I got overcurrent.
I then made this 4s prismatic battery that gives about 1800A, at least when it is 3,8V/cell:

I tried a few welds, then the fuse blew

 

[eee291]

You can't just change the battery and use it right away, you have to do the calibration procedure again. Also it doesn't matter if you increase welding current it won't stick with the kweld no matter what.
I think there are some other welders that can weld nickel to copper. If heat isn't a problem then a quick and dirty fix would be to solder the 2.

 

[j bjork]

The problem with the solder approach is that it is welded to the other 18 cells:

It takes a lot of heat to solder on a rather big copper plate
I also have to decide what to do with the leftovers from the plate I took off:

Leave it and weld through, it is nickel plated copper after all?
I tried to grind it off:

The positive side is not a big problem. I can peel it off, and leave the weld spots sticking up.
Or grind them down like in the picture, and maybe damage the corrosion protection
The negative seems harder.. I am afraid to damage the cell if I try to peel it off, and there will be damage from grinding too.

 

[A-DamW]

The problem is that I have 1mm copper bus bars that I have to make something stick to.
I tried welding 0,1mm nickel to 1mm copper with my k-weld, but hasn't been very successful.

You may want to have a read of this thread:

Spot welding Copper / Copper Nickel / Copper Stainless Steel Sandwich

With all the new high power Tabless power cells on the market, and even higher power and higher capacity cells promised in the near future, it seems the electrical bottlenecks will not be the cells themselves, but the parallel and series connections between cells and P groups. I have built 3...

forum.esk8.news

Seems the magic happens with copper flux + stainless steel for higher spot welding resistance.

Copper flux:

Flux for Copper-to-Copper Spot Welding (100g)

This Flux for Copper-to-Copper Welding is specifically engineered for lithium battery builders who want to spot weld copper with confidence. If you're working with copper tabs, terminals, or busbars, this is the secret weapon that helps you achieve strong, low-resistance welds without pushing...

diy500amp.com

 

[eee291]

You could extend the nickel so it reaches the nickel from the other cell.

 

[j bjork]

You may want to have a read of this thread:

Spot welding Copper / Copper Nickel / Copper Stainless Steel Sandwich

With all the new high power Tabless power cells on the market, and even higher power and higher capacity cells promised in the near future, it seems the electrical bottlenecks will not be the cells themselves, but the parallel and series connections between cells and P groups. I have built 3...

forum.esk8.news

Seems the magic happens with copper flux + stainless steel for higher spot welding resistance.

Copper flux:

Flux for Copper-to-Copper Spot Welding (100g)

This Flux for Copper-to-Copper Welding is specifically engineered for lithium battery builders who want to spot weld copper with confidence. If you're working with copper tabs, terminals, or busbars, this is the secret weapon that helps you achieve strong, low-resistance welds without pushing...

diy500amp.com

Thanks, I went through that thread. It was probably pretty much the same things I've seen here too. That flux sure would be interesting to try, but it seems a bit far away from here. I wonder if it is about the same thing as you use when tin caulking (I'm not sure that is the right word ) a car body? You know, the old way to fill imperfections instead of plastic putty. Then you had a special flux that left a tinned surface on steel.
Not that I have any of that anyway..

I'm doing pretty much what eee291 said, but with copper/nickel plated steel sandwish.
Now my k-weld has gone a little crazy though.
It set itself to Auto, I have always used it in manual with the foot pedal.
And I cant change it back. When I go to the manual/auto menu it starts showing weld statistics after maybe 1-2 seconds.
If I am really quick I can change before it starts, but it starts anyway, and it seems like I cant get out of it without disconnecting the welder from the battery.
When I connect it again it is back to Auto..

 

[eee291]

I have a pulse welder, it requires argon to make clean welds but I've tried even without gas and it actually makes some solid welds using nickel plated copper, I haven't tried pure copper yet but I'm sure it can do it without any issues.

It is too weak for welding large aluminum tabs though so I want to get rid of it.
I'm just going to go with a Tig welder that has a pulse function.

 

[Zambam]

That flux sure would be interesting to try, but it seems a bit far away from here. I wonder if it is about the same thing as you use when tin caulking (I'm not sure that is the right word ) a car body? You know, the old way to fill imperfections instead of plastic putty. Then you had a special flux that left a tinned surface on steel.

Probably not the same flux for car body work.

This is the right stuff. https://www.aliexpress.us/item/3256...st_main.5.488b1802VE2ipi&gatewayAdapt=glo2usa

Read the description! There's a lot more info since I bought mine (about storage and other info). I have not used mine (still do not have a spot welder) so my tube is probably all dried up.

 

[j bjork]

Zambam, thanks I think I will add one of those with my next order

Here comes a proper update with pictures:

I had to make these special copper pieces to do a serial connection, because of the honeycomb layout.

Then I made a parallel strip too.
I am a little worried about the welds, especially on the battery to the left.
That was when the welder decided to change to auto, and shot off a weld when I wasn't ready.
The contact wasn't good enough, and it shot trough the nickel plated steel. I hope the cell didn't get damaged.

Doing testwelds with 0.1mm stainless and 0.2mm copper. I get good welds with 25J

Here it is with the 25J to the left, 20J to the right. That also welded the copper to the can, but very small dots.

I was able to get off the leftovers and clean the cells pretty good. Also added protective paper to the positive cells.

So this is what I did, instead of solder strips to the old copper bus plate I made a new one of 0.2mm copper and 0.1ss.
A bit scary to weld, as it is hard to tell that you are in the right position. I used the old bus plate to draw squares through the holes, but also a protection ring. With that I could get a pretty good position everywhere where I could see a little bit of the cell.

The welds on the positive side are so ugly
I started with 4, then decided I could just as well do 6. But then I was afraid to spread them too much so I just put them wherever there were room.
I also did them a little hotter, I ran 30J. Except the first one that I did 35, but that was too much.
On the negative I did 25J.

 

[Zambam]

You only used 25 & 30J from Kweld for 0.2 copper and 0.1ss sandwich when max is 500J. It'll be interesting if you can weld 1 mm copper and 0.1ss sandwich when you get a tube of the copper flux.

 

[j bjork]

The max setting is 500J, but in reality you cant get anywhere near that as the pulse time goes over 200ms which is max.
I think I can do about 150J, but it seems to vary.
I am curious to try 0.3mm copper with 0.1mm ss, if that works reliably I think that is what I will do on the next battery

 

[Zambam]

What grade ss did you use?

 

[j bjork]

304
This is it if the link works:

https://www.aliexpress.com/item/1005008958075941.html?spm=a2g0o.order_list.order_list_main.47.746c18028WfTBt

I saw some strips of 301 too, I don't know if it matters:

https://www.aliexpress.com/item/1005006759023644.html?spm=a2g0o.cart.0.0.277d38dax0EMBO&mp=1&pdp_npi=5%40dis%21SEK%21SEK%2024.99%21SEK%2022.25%21%21SEK%2022.25%21%21%21%402103890917589020081922294e9870%2112000038212213265%21ct%21SE%21139590896%21%211%210