Yamaha YZ250F

[dougf]

Ok. The APT controller also carried on a bit after throttle release. Very noticable with wheelies. The fardriver i got is a lot better. My model and firmware are ND1961800B_463_14A.2.3..... But obviously im using an encoder so its probably different. I also found some motor auto learn params;

LD: 1151uH
LQ: 373uH
FAIF: 0.0226

I also did autolearn on the stand with rear wheel spinning.

While you had the Oscope connected did you look at how fast the throttle signal dropped when you let it snap shut?

I nurse a deep dislike for arduinos in general. What you want to do is possible but hugok is right, a comparator is probably what id use. I would not use the PWM output to drive the controller, i would use an electric potentiometer because its more what the drive expects...that or a DAC.

Doug

 

[j bjork]

It seems the name on mine is different, I suppose it is ND961800_63_ARA05 for me?
I tried to compare to what the computer says, but didnt find anything like that there.

Here my LD, LQ and FAIF says something different..
But when looking at the computer it makes more sense:

Still, FAIF seems to be completely different?
What is it anyway?

I didnt try how fast the throttle closes on the scope.
I mean I do it manually when driving, so how fast the spring can close it dosent seem very relevant?
But I can test it just in case, to see that there isnt some weird issue.

 

[j bjork]

This is what the throttle closing looks like:

It bounces a little when it closes, and it is up at 1,1v:

Except for that it closes at 53ms:

As the full range is not used, (it is set to 1,1-3,9v) it is actually a little faster.

The little bounce could potentially activate the controller as it is just on the edge of my set 1,1v threshold.
But in reality I never let go of the throttle and leave it to return to 0 by itself.
I can test to rise the threshold a little, can be interesting to see if it changes anything anyway.

 

[j bjork]

This is probably the throttle modification thread I have seen before:

https://electricbike.com/forum/forum/kits/golden-motor-magic-pie/70584-guide-to-hall-sensor-throttle-operation-testing-and-modification

If you dont want to read it I can say that is mostly about modifying the start or end voltages with potentiometers, and not really what I am after.
It does show that a pot throttle is not really linear, and should have a little curve that could be beneficial to the jerkiness at low throttle. So trying my half-broken domino could be interesting.

 

[dougf]

50ms till zero. Some of that will be the time it takes for the grip to rotate. Dont think thats the problem. Sort of has to be drive ramp.

Dont know what FAIF is. Hopefully fardriver can get you some different firmware to try.

Doug

 

[speedy1984]

This any help?It's the 96v 850a model but this chap claims he sorted his throttle with these adjustments.Looking at your ratio limit settings that are very different from hisHe pulls very nice slow controlled wheelies that indicates he's doing something right.

 

[tylerwatts]

@j bjork I wonder if this can help you sir

https://www.facebook.com/groups/electricmotorcyclebuilds/permalink/2549767471827275/

cheers
Tyler

 

[j bjork]

It seems doing a new self learn did it, even though I couldnt see much of a change

I did a little testdrive yesterday, and I thought it felt a lot better.
So I did a bigger testdrive today, and I can confirm it is a different bike

The delay at throttle release is gone, or at least very close to gone.
I tried the regen at throttle release again too, and it almost worked :wink:
The delay before it takes throttle again was not gone, but it was a lot less than before. Almost usable, I might try it on a track.

Thanks for the suggestions, I see something very interesting in the rated rpm.
He has 4 pole pairs set up, but a 5 pol pairs motor. Still He gets a rated rpm of 5500.
His motor should be about 60kv, and mine too.
But I only get 2800 rated rpm :?

Yes, there is probably a lot of fine tuning in those maybe pA at different rpm settings?
I basically turned them up to see what happened. I dont like to limit power at different rpm:s though, I want to limit the output at small throttle twists without limiting the output at big throttle twists.

 

[speedy1984]

Great news :thumb:

 

[dougf]

Awesome mate, hope you're grinning from ear to ear . Was that an autolearn without the chain on? Maybe I should try it and see if its any different. Did that FAIF value change?

Doug

 

[j bjork]

Yes, it was a very good feeling when it worked
Yes, autolearn without both chain and sprocket (I dont think the sprocket mattered, it is just easiest to take it all off on my bike)
No, no change in LD, LQ or FAIF on the phone app. I didnt know that the computer showed different values at the time, so I didnt check on that.
The "rated speed" went up a little, to 2831rpm from 2815rpm. (it actually went to something like 2854rpm on the first try, but I then tried one more time. Maybe I should have kept that one)

But I think it is a bit strange, if it is based on kv I think it should be about 6000rpm.
What do you have?

Anyway, calibration changed a little. I think phase A zero coeff went from 2077 to 2078 and line zero coeff from 1171 to 1161.
But these values seem to change a little from time to time, so it might very well not be related.

Dui mentions that after he got another firmware to get rid of his throttle release delay his slide regen works.
I dont know if that means that the delay before the throttle kicks in sometimes after regen is gone or not for him.
Mine is not gone, but a lot better.

 

[speedy1984]

On my v3 138 rated rpm was around 4500 ish but changed it to 7500rpm in the app.I also did the auto without the chain and sprocket on when i set the controller up.My throttle seems very good after changing the voltage values slightly but not a lot tbh.I use a domino so maybe that helps some also.

 

[j bjork]

Interesting, I tried changing the rated rpm but it just went back to what it was.
Maybe I should try again, sometimes the changes dont stick on the first try.

Did you notice any difference when you changed the rated rpm?

Anyway, today I took the bike to a track and rode it.
It was a track I have never been before, fast and rather sandy.
Very little difficult parts, a rock climb that you could just take with speed and some parts with skull sized rocks with some distance from each other that didnt take much precision riding.

Anyway, I started in mid power, sport and 200 in throttle rate.
It was so damn aggressive it was barley rideable
I changed to linear, it got a bit mushy with a bit rubber band effect, I then changed to hi on the 3-speed switch.
It was better after that. I dont think it is right, it should just change the amount of torque and power, not the throttle feel :?
Maybe there are some settings I have missed?

After that I changed the throttle rate to 224, kept linear throttle and hi.
I kept it like that, and I think it felt pretty good. I will try sport and lower throttle rate another time.

 

[speedy1984]

Hi,no i changed the rpm in the 3 speed switch settings.Was the first ride on the build this afternoon and could tell there was more top speed with the added rpm.Also no throttle run on or delay,but i did do my setup without the chain on in the first place.Ive made minor throttle voltage changes also.Rides fantastic if honest.

 

[j bjork]

I was talking about the rated rpm, I suppose that has to do with how the field weakening kicks in.
I understand that raising the max rpm will lead to higher top speed :wink:

It seems that most are satisfied, or at least dosent write about it, but quite a few seems to have throttle or/and regen issues.

My observations from my latest ride yesterday:

The throttle release regen still dosent work, going from throttle "off" to "on" is terrible with delay and unpredictable power on.

The throttle control is pretty good in "linear", throttle rate 224 and I think I need to be in high too. There can be a slight delay in throttle release at high power, I felt a little when I accelerated hard and leaning forward and still got a power wheelie.
When I let go I could feel a very slight delay.

It is powerful, but a little lame in character in linear throttle. In sport and 224 in throttle rate it is too aggressive, so I tried 128 in throttle rate. Aggressive with a delay, not usable :?

I am in contact with siaecosys, I will see if it leads somewhere.
The first time the bike got properly dirty (and properly power washed) I think:

 

[speedy1984]

I was talking about the rated rpm, I suppose that has to do with how the field weakening kicks in.
I understand that raising the max rpm will lead to higher top speed :wink:

I was just pointing out that it actually does raise top speed not just a number in the app that does not actually work

 

[Eastwood]

The bike looks great! Nice and muddy too
Hopefully siaecosys can get these controller issues resolved.
Would love to see a video of the bike in action, Shredding the earth 8)

 

[j bjork]

I done some more self learns and checking values.
In the end I decided to take off the primary reduction again and do it with absolutely no load, just in case.

I made a rubber seal this time.
The silicon works very good, but it is a mess every time to get rid of.

Things got rather wet inside of the covers after the wash, I decided I had to seal the delay circuit and 12v converter.
It dosent look very nice, but at least it wont be seen very often :?

After the wash the bike braked and wouldent move, just send a beep code.
Couldent connect to the controller either.
I blew out the water of the leftover contacts, then it worked.
I decided to cut off the anti theft contact and seal the wires.

I also decided it was time for a dirt shield, I dont think I will have any heat problems anyway.

 

[tylerwatts]

Looking great J, nearly there. I hope you get the controller issues sorted out

cheers
Tyler

 

[j bjork]

I got a new firmware from siaecosys

After a lot of struggling I was able to load it, but I still dont know exactly how I did it :?

When trying on the computer I get this:

I suppose I should press the left, highlighted button.
I get some beeps from the controller, but nothing seems to happen with the firmware.
I tried both buttons, none helped.

I also tried the phone (android) but wasnt able to get the app to load the file.
Tried to open the file, and select the app to do so. But the phone just said that I didnt have any apps compatible with the file..

Finally when I tried to share the file with an iphone I was instead able to open it with the app

The file was named ND961800_63_ARA.2.4
But in the controller it shows as ND961800_63_ARA.*.4

 

[Eastwood]

So now you loaded The file from the iPhone are you able to change the settings or?

Have you tried to rename the file and see if that helps. Change it to the name it shows up in the app,
ND961800_63_ARA.*.4
Maybe that will help with loading the file from your computer.

 

[j bjork]

I was able to do the firmware update from the phone after I got it to load the file.

I have been testriding today, it fixed the throttle release regen issue.
So now that works, it reacts to throttle after a release every time now.

But it dosent seem to do anything to the throttle behavior, there is still a "rubberband effect".
As long as I am at 224 in throttle rate it is barley noticeable, witch is fine in linear. But way too aggressive in sport.

 

[j bjork]

I got talked in to competing again at "ränneslätt"
This time not in the electric class, but the big classes.
That means the full length track, probably 20-25km. And the full time 150min + up to 60min (depending on what time you finish the last two laps)

Here is a little from last year:
https://www.youtube.com/watch?v=tIaxtPhjy4Q
https://www.youtube.com/watch?v=DphyJxZbfyM

I thought it was tight with 90 or so starting at the same time, this will probably be over 600

I should have split that module to 3 peaces with 4 cells each, and built 3 batteries with 28s.
Instead of splitting it in two with 6 cells and build two 30s batteries :?
Well, I am not sure even 3 batteries would be enough.

I asked if it was possible to get electricity for charging, but didnt get any answer. I suppose that means no..
So now I have set up my own charging:

You know those generators that dont make a lot of noise?
This is not one of them :|

Tested with 33A charging, the batteries dont seem to mind:

The temp went up about 4 degrees
I can adjust up the voltage a little, but I have to be careful.
I dont want a controller that refuses and sets overvoltage when I changed batteries
(the controller reads 1v higher than the bms, so this would be about 120,6v)

 

[tylerwatts]

Wow that race looks intense J! The swappable batteries sound good. Will you build 3 new ones and use these on another build maybe?

cheers
Tyler

 

[j bjork]

Heh, I dont think I will build 3 new batteries..
If I still had modules I would maybe build one more just like these I have.
But if I would have started over with this build I would have built three 28s batteries instead.

I will probably buy a ND1081800 when (if) it becomes available, but it is really annoying that it would only be for a few extra volt