Yamaha YZ250F

[j bjork]

Finishing up one of the batteries:

I opened a new tube of glue, thinking it was black.
It wasn't..
I went with it anyway, it should not be very visible when it is done

When mounting it I realized there was no room for the big washers,
luckily it was not too hard to get them off and go with small ones instead.

I decided not to put the display in the front cover of the battery, at lest not now.
It should be possible, but it would take a pretty advanced box to make a decent install.
I might be able to 3d print something in the future.

So what is needed to change batteries is take the seat off, the left side cover, 2 connectors, the front cover, then the battery itself.
That is 2 screws for the seat, one for the side cover and 4 for the battery. Well there are two dzus for the front cover too.

For the seat and battery I plan to use the same screws, If in a hurry it should be possible to just use a cordless drill with a 12mm socket. The problem is the 8mm head on the screw for the side cover witch makes it necessary to use 2 tools to change battery.
Maybe I should try to find or make a m6 screw with 12mm head.
I was thinking of some kind of snap in place solution, but a different screw is just so much easier..

On the other hand I will probably never be in such a hurry to change battery that I cant change tools.
It is more the thing that it should be quick and easy. I kind of want the bike to be comparable to a factory build

 

[j bjork]

If anyone wonders:

I did make a m6 screw with 12mm head :wink:
So now it is one tool needed to change battery.

Moment of truth for the batteries:

Will they keep tight under pressure?

No the pressure is gone in a minute.
I dont find any leaks around the seal or rivets, but I find a little from the on/off button for the bms.
Annoying, I really covered it in sealant to avoid that.
Still, it is a small leak. It dosent seem to be enough.

I have tried around the connectors and not found anything, but then I really fill up the charge connector and find pretty much leak there.
Same thing, I really covered the wires to avoid leaks, see this old picture:

There you can see both the charging wires and the button for the bms.
It is a pit were the wires exit, that I have completely filled wit sealant.
The conclusion I make is that it has to be leaking trough the wires themself.

That means I should do the same test on the thickest wires:

Yea, there was the biggest leak..

So now the question is should I call it good enough or drill drain holes?
I do have a ventilation that should flow easier, preventing water to be sucked trough the wires..

 

[thoroughbred]

how about schrader valves at the lowest drain point and one on top. Then you can drain the case and purge the case with nitrogen, argon or helium after any wet riding :thumb:

 

[Elektrosherpa]

So now the question is should I call it good enough or drill drain holes?

I would let it as it is.
I am not going to do scuba diving with my bike :lol:

 

[j bjork]

I think it would be very hard to find room for a schrader valve in the bottom, it would probably be possible in the top
I think I will just leave it as it is, and if I worry that I might have gotten water inside I can always drill a drain hole and see.

I might go diving, I have more or less done it with the ktm :wink:

 

[matt912836]

definitely reminds me of when an oil cooled hub motor would leak through the phase wires themselves no matter how good you sealed it, its leaking through the strands of wire themselves

 

[j bjork]

Yes, I think that is what happens here too.
I hope it wont a problem as I have a ventilation so there shouldn't be any vacuum (or pressure) in the battery.

I ran in to another problem. I was mounting the display, and connecting it.
When I tested it it didn't work, it only said connect bms.

No obvious wiring problem, so I tested with the other battery. Works just fine.
I dont want to cut up the battery to check connection, so I start taking measurements and compare to the other battery.
I dont find much useful measurements, but at least I get 10kohm between 2 wires, but not the same ones on both batteries.

I start looking at old pictures, and find this from battery 2:

And I find this of the wiring for battery 1:

The only wire that is in the same position is the red
I re-positioned them in the connector outside of the battery, and now both batteries work fine

 

[j bjork]

This is how the dash panel turned out:

It started with the box on top, actually the holes for the buttons is what was the rear mounting holes.
But it was just to ugly, the cover with the light in got too far forward.
I had to get it closer to the handle bars, then the box for the display had to go under to be able to get it further back.

I think it looks a lot better with a front light.
The brake line is a little short though, I had to put the brake grip as far to the middle of the handlebars as possible and moved it up as long as possible at the lower mounting point on the forks.

It seems to be the same line for both YZ and WR at a online shop (mx and enduro bike), so it might not be longer.

 

[tylerwatts]

Yes, I think that is what happens here too.
I hope it wont a problem as I have a ventilation so there shouldn't be any vacuum (or pressure) in the battery.

I ran in to another problem. I was mounting the display, and connecting it.
When I tested it it didn't work, it only said connect bms.

No obvious wiring problem, so I tested with the other battery. Works just fine.
I dont want to cut up the battery to check connection, so I start taking measurements and compare to the other battery.
I dont find much useful measurements, but at least I get 10kohm between 2 wires, but not the same ones on both batteries.

I start looking at old pictures, and find this from battery 2:

And I find this of the wiring for battery 1:

The only wire that is in the same position is the red
I re-positioned them in the connector outside of the battery, and now both batteries work fine

Phew that was good you could find and fix it. And now the batteries match so can be simply swapped. This is a great build!

cheers
Tyler

 

[j bjork]

Thanks
I was out on a small testdrive in the woods yesterday, but was a bit disappointed with the performance.
I knew it had been better, and I think the only changes have been to the kp/ki settings.

I think I started with a kp of about 3,5 and had changed to 1,5. With that I had been able to get good current control in the logs, but sluggish performance it seems.
I started turning up, and the logs started getting worse. The ki setting dosent seem to make much difference.

However, it made a lot of difference on the road
With a kp of 7 I had wheelspin on asphalt and pretty violent wheelies, it felt very nice and a little scary :wink:
I also had cut outs and overcurrent faults.
It feels good to know that the motor seems to have some potential, I got worried about the torque curves in the controller. Now is the question if I can get the controller to deliver like that without problems.

 

[dougf]

Looks awesome mate. I usually find the best way to stop water getting inside it is to make sure water cant pool around the joint, connector, glue etc.

Attached are my torque settings. Not sure if it's any help being from a different drive.

Doug

 

[larsb]

Thanks
I was out on a small testdrive in the woods yesterday, but was a bit disappointed with the performance.
I knew it had been better, and I think the only changes have been to the kp/ki settings.

I think I started with a kp of about 3,5 and had changed to 1,5. With that I had been able to get good current control in the logs, but sluggish performance it seems.
I started turning up, and the logs started getting worse. The ki setting dosent seem to make much difference.

However, it made a lot of difference on the road
With a kp of 7 I had wheelspin on asphalt and pretty violent wheelies, it felt very nice and a little scary :wink:
I also had cut outs and overcurrent faults.
It feels good to know that the motor seems to have some potential, I got worried about the torque curves in the controller. Now is the question if I can get the controller to deliver like that without problems.

There are some guides how to do basic PID tuning online if you search for it. Like, set every parameter low or zero, start with tuning one at a time until system reaction is good. Problematic since it would be so much easier with a dyno, but still..

 

[j bjork]

Looks awesome mate. I usually find the best way to stop water getting inside it is to make sure water cant pool around the joint, connector, glue etc.

Attached are my torque settings. Not sure if it's any help being from a different drive.

Doug

That looks worse than my tuning, over 20 settings to get right

I have problems with my rather few settings:

This is what the manual says about it.
I wonder what the ones in the blue circle are that isn't even mentioned..

LarsB, I was looking at some of those instructions last time. I dont think I got a lot wiser, but maybe I should take another look.
I think it usually was a third variable in the examples that I dont have.

I want to tune it in a dyno, but I would like to know how the controller works first, what settings are relevant etc.
But it seems almost impossible to find out other than by trial and error.
I actually asked today at a motorcycle workshop, but they were fully booked in the dyno over a month.

I was hoping to be finished in the winter, then there would have been a lot better possibilities to negotiate dyno hours..

 

[dougf]

I guess APT have a set of params for all riding modes. Looking at your manual there aint no way my settings are helpful :wink: .

Doug

 

[j bjork]

My 4mm ventilation from the primary reduction didn't work, I got oil dripping from it.

Now there will be changes, 8mm and a fuel filter as oil stop.
So far it seems to work fine, and it dosent look like any oil has got to the filter.
Maybe it would have been enough to upgrade to the 8mm tube.

I have been tuning a little more, I tested to turn up the torque map but it didnt seem to make a difference.
But then I made a breakthrough, I realized it was usually about the same speed I got the overcurrent.
I tested to not give full throttle, and I noticed it started accelerating pretty much by itself at higher speeds.

I then turned down the fw current a lot, thinking I probably dont need 500Arms :wink:
I lost a lot of top end, but it didnt cut out with 7Kp.

I think most of the important settings are in this picture, I know what about half of them are for..
I someone can tell me about "Iq max headroom, frequency/mod index control, max drive (and brake) mod index, percentage minimum allowed saturation of Ls" I am very interested
I am not sure about "openloop start FW" either, but it seems to usually be left at 0.

Some more control gains that I dont really know what to do with

 

[j bjork]

First real test:

I have some work to do. It feels strong just cruising on very little throttle, but on a fast throttle blip the response just isn't there.
The regen brake reacts very strange too, but i knew that. On just a little brake it locks the wheel for 2 sec or so, then it brakes very little and pulses.

I didnt like the front brake either, hopefully just a matter of bleeding it.
Some suspension tuning will be required. I adjusted the sag in last minute, which meant max preload on the rear spring.
It sure needs more rebound damping now.

 

[dougf]

Looks better covered in mud 8) . What did your total weight end up at? I upgraded to a 270mm disc because the stock yammi brakes are a bit garbage. I think the 2t 250 and 450 came std with a 270 but the 125 and 250f had a smaller one. Could be wrong though.

Doug

 

[j bjork]

I think it weighs 135 or at least 130kg. I will check to make sure.
The disc seems to be about 270mm, but I think I was able to get a small air bobble out.
Ill see what I think next time.
The regen seems to be a matter of maybe turning up max brake power kw (or what it was i found somewhere) and turning down the braking % a lot.
I probably started at 100% and thought it would mean locked wheel on full brake, not as soon as I start moving the brake.
On 50% it was still the same, but 25% seems about right.

 

[dougf]

I think it weighs 135 or at least 130kg. I will check to make sure.
The disc seems to be about 270mm, but I think I was able to get a small air bobble out.
Ill see what I think next time.
The regen seems to be a matter of maybe turning up max brake power kw (or what it was i found somewhere) and turning down the braking % a lot.
I probably started at 100% and thought it would mean locked wheel on full brake, not as soon as I start moving the brake.
On 50% it was still the same, but 25% seems about right.

Ok, maybe the newer YZ's all have 270mm discs. After bleeding mine i hung it vertically from the shed roof overnight to encourage any air to move up.

I dont have any answers for your regen sorry .

Doug

 

[j bjork]

I have been battling the sevcon a lot lately.
Luckily I have had help from a friend that knows pmac controller tuning, but not sevcon..
Still, it has been a great help with someone that has an idea what a setting is supposed to do.
I have sent him a lot of logs that he has been trying to make sense of :wink:

But calculating kp and ki settings hasn't been successful, not known ways of trial and error either (I dont remember the name of the system now). I think I have found a decent setting now anyway, and moved on to field weakening. It seems very hard to get that right too
Even the rpm limit seems hard to get right. I can get a very nice, stable rpm limit. But then it starts to back off power way too early in road tests.
Anyway, I decided that it had to do. It feels very responsive, and there doesn't seem to be a shortage of power 8)
I have destroyed the rear tire from tests on asphalt, and the front tire was in a bad shape before. So time for new rubber:

I also took care of some dents in the rim:

Turned out really nice, and the bike is ready for a new track test:

 

[SlowCo]

Great work!

 

[j bjork]

Thanks, it was a disappointment today too though
Still sluggish throttle response when riding
When trying on the road giving full throttle from slow rolling I cant keep the front wheel down and has to ease off the throttle.
But on the track riding at some speed at low throttle, and fast hitting maybe 3/4 throttle the reaction is not what I want.
I want instant wheelspin and wheelies, but I get a gentle(ish) increase in torque

I shortened the twist by lowering the voltage to get full throttle, and it helped to some extent.
But it gets twitchy and too sensitive at low throttle instead.
( fell down a damn mountain when I wasnt careful enough on the throttle climbing up)

I wonder how the sport mode on the votol works? That is what I want, on a fast twitch it gives all it got.
If twisting slowly it ramps up slowly.

I also feel it doesn't ramp down fast enough, but that might just be that I have no engine brake.
I dont think I am able to get it unless I have a throttle with a switch like a domino.
It seems dominos are out of stock though, but I have one with a broken housing. I think I will try it and see if I really can get motor brake working with it.

 

[dougf]

Looks good mate! Are you saying that it hooks up and pulls hard on sealed surface but on dirt it doesnt? Seems a bit odd. Maybe the controller does a mix of torque control and speed control depending on motor load? Do you have debug logs? I was under the impression that Sevcon offered payed support but I dont know if they do it for individuals.

Doug

 

[j bjork]

The tests I have done on asphalt is not really the same as on the track.
It has mostly been full throttle from a rolling start on a little turned down max phase amp.
That is to observe the current control, I have a lot of those logs.
I am logging by taking the computer in a backpack with a wire between it and the bike.
I have done a few on full phase amp, but those are hard to get good logs from because it wheelies and I back off on the throttle.
Those are just out, do 2 tests fast. Get the computer out and stop the logging.

It might very well not ramp up fast enough then too, but I just haven't reacted to it.
I think the logs would be very big if I took the computer out on a lap on the track, and I wonder if the computer could take it?

When I shortened the twist for full throttle I also turned up the "maximum acceleration rate" from 20000 to 30000 rpm/s, so that might have to do with that it seemed better afterwards.

I suppose it is still limited here?:

The last two settings "speed limit ramp up rate" they are maxed out at 20000rpm/s
I have read that the %/s settings are not used, but I have turned up the ones that seem relevant anyway.

Something I have seen that looks interesting:

"proportional speed limit in drive" that is enabled. I have tested enabled and disabled on the stand, no obvious difference.
If I change "speed limit mode" to "proportional to throttle" instead of "fixed at maximum" it starts limiting rpm.

I guess I could start messing with "throttle input characteristics". There are curved and crawl to choose too, I just assumed that would make it worse as it means less reaction at first.

 

[Ianhill]

You could try shoving aload of little sandbags in the front wheel keep it down under throttle so you can get your launch but it will be tuned in a way that without them its going to lift.

Anti wheelie devices are simple enough to fit to anything with a electronic twist throttle if thats something you could be into for maximum launch speeds.