Yamaha YZ450F

[j bjork]

Another note about the script, it seem to have an eco mode if you activate both sin and cos.
But I cant see what it does? That probably mean that the script wasn't written for me, we never talked about an eco mode.
At the same time it was probably changed for me, and mistakes can have been made.

Something else, the logs on the flipsky display kind of works now:

I'm not so sure about the 99km/h max speed, but the rest seems reasonable.
This was a shorter, winter-adapted section on the less sensitive parts of the course.
So 4.5 ish km is one lap, and 9.x km is two laps.
Then we get to the log data:

Lets say that I am a bit skeptic to these numbers..
The temps seem right when I read them on the screen when stationary.
I haven't looked at the screen when riding.

 

[scianiac]

Idk 104 megawatts looks about right to me.

For some reason that script looks like a kinda strange way to go about it, not that I really have any idea what I'm talking about but what is the point of even calling sin and cos voltages when it's just ADC4 and ADC5, just because that's what the pins are called? It looks like it's reading those ADC pins here:

(setvar 'sinv (get-adc 4))
(setvar 'cosv (get-adc 5))

and setting them to the variables named sinv and cosv (which seem to be just made up variable names and could be anything). I assume if 5v is applied to those pins it should read 1 but honestly I don't really know, maybe try changing those "1.0"s to "0.8" here?

; mode1
(if (>= sinv 1.0)
(setvar 'new_pmod 0)
)

; mode3
(if (>= cosv 1.0)
(setvar 'new_pmod 2)
)

; eco

(if (and (<= sinv 1.0) (<= cosv 1.0))
(setvar 'new_pmod 1))

to make sure the 5v is enough to trigger those if maybe it only gets to 0.95 instead of 1 on that ADC output.

There could easily be some other issue how this code is written I'm totally unaware of.

 

[A-DamW]

The idea with this post and the one before is mostly to get help from those with vesc knowledge, so feel free to comment

Are you running this script directly on the controller?

 

[j bjork]

Are you running this script directly on the controller?

Yes, I don't have any other device connected that I am running it trough if that is what you mean?
It looks like this:

Strange thing is now I changed the motor current in mode2 (that is marked on the screen), and clicked on "upload". Then I went to motor settings and clicked "read", but nothing changed. I know I have been able to change the motor settings by making changes in mode 2 earlier.

 

[j bjork]

For some reason that script looks like a kinda strange way to go about it, not that I really have any idea what I'm talking about but what is the point of even calling sin and cos voltages when it's just ADC4 and ADC5, just because that's what the pins are called? It looks like it's reading those ADC pins here:

Yes, I think it is because it is what the pins are called in the wiring diagram.
I tried changing to 0.8, but it didn't help

But it seems like I cant get it to work at all now (see the previous post), so I don't really know.

 

[amberwolf]

Then we get to the log data:

Lets say that I am a bit skeptic to these numbers..
The temps seem right when I read them on the screen when stationary.

What color are the controller and motor glowing?

I think that wattage is probably excessive even to reach those temperatures. Still not enough for a time machine, though.

Unfortunately I have no helpful response to correct the problem causing these excessively high numbers, other than either a sensor calibration issue, or a multiplication factor / scaling issue.

 

[j bjork]

I took a little ride just on the driveway and lawn and logged.
First a little harder:

The temps look alright, but not the power..

More careful, but still not there.

This one looks like it can be right, but I cant say for sure. I guess I should log with vesc tool and compare.
Not sure I will bother though, got some new stuff coming..

The charger is finished, but I dont really like the display for the bms.
I have to start it manually, then it turns off by itself after a while.
I would like it to be on all the time it is powered.

The new wheel is on the way. I'm waiting for some parts I need before I can put in the bearings, then I can start straitening..

 

[j bjork]

Finished the wheel:

It went smoother than expected, I had basically just screwed in all the nipples about the same amount, and then it was pretty good.
So I just tightened them all with 5Nm and then I had about 2mm of horizontal and vertical trow, I thought that would be good enough on a dirtbike.

I also got this:

It is a voyage megan.
It didn't have any seal between the housing and baseplate, so I decided to take the rest apart to check:

There is a cover for the screen that looks to be properly sealed, and the rubber buttons should seal pretty well I think

I 3d printed a seal between the baseplate and housing, seems I didn't get a picture of it though.
I had just mounted a phone holder a week or something earlier, but it had to go now:

This is just the best spot for this display:

But now I had to have wires between the bike and the tank cover, so I made this cable that it hangs from when I change the battery:

I was out riding in the weekend to see how it worked, even if I hadn't set it up properly yet.
Got a flat on the first lap and didn't have an extra wheel

I suppose it is a pinch flat, even if it is just a few small holes

 

[bananu7]

I suppose it is a pinch flat, even if it is just a few small holes

This looks like tyre wires sticking out and cutting the tube. Inspect the inside of the tyre closely to make sure it doesn't happen again right after.

 

[j bjork]

I didn't see any damage in the tire when I changed tube. I think it is just that you can see the outline of the cord on the inside of the tire, and when it has bottomed out, it has left an imprint in the tube.
I hope the michelin "ultra heavy duty" tube will be able to take some more beating, but eventually you will end up with a pinch flat or ripped valve stem anyway..
The only thing that works seems to be mousse except that they wear out instead

It is getting about time to do some finishing work on the bike:

Fitting black radiator wings that will match the battery cover better.

And fit the gap between the battery cover and radiator wing a little better too..
I have sanded down the imperfections after the mold split and some of the scratches on the battery cover here.

I drilled holes for the extra mounting points in the radiator wings before I decided to make another design on the inside instead.
A little annoying, but it wont really matter when it is finished, you will see

 

[j bjork]

I wasn't entirely happy with the locking system I made for the seat:

I have to pull out a pin on the inside of the fender, and it is often dirty there..

I decided to try to put the locking on the upside instead:

I have done a real life test, and I am not that convinced that it is better.
When siting down normally it is not a problem, but you can hurt your ass when if you sit down hard and a a little to the side.
As can happen if you slip on the foot pegs, for example.
And dirt is still a problem anyway:

But now I'm getting ahead of myself, the changing of radiator wings didn't stop there:

So there I was, I got the bike ready in good time for some riding this weekend.
Turned it on, and nothing happened.
To be continued..

 

[j bjork]

Checked for voltage to the controller, ok. Also tried the horn, it worked (separate 12v converter that turns on when the controller turns on) Also tried connecting the "vin" input directly to battery voltage, to make sure there was no problem with the relay etc. Turns out I get 12v output from the controller, but no 5v.
I tried putting 5v on the output to see if it would start, but no.

I ended up stripping the old bike from its controller so I could ride this one

 

[j bjork]

I don't have much news about the controller, more than that Hackey said someone from their company will contact me.
He didn't say when though..

The rear spring was too soft, so I changed it to the first one I changed to on the other bike:

Took some measurements first, the 250:s oem spring: 51Kg/10mm, 450 oem: 52.5Kg/10mm and stiffer spring: 61Kg/10mm.
Not very exact measured on a bathroom scale and with a folding rule, but shouldn't be too far off.
The 250:s oem spring was measured before to 53 and the stiff spring to 62 I think. Should be 63 according to the marking on the box.
I still seem to need pretty much preload to get the sag right though

One thing I have noticed with this bike, it collects dirt much more efficient than the other one.
It fills up skid plate, radiators and all sorts of nooks and crannies that the other one simply don't have.

Oil change is also more of a challenge on this one

When I put on the stickers on the bike I skipped the fork guards as they are the oem ones with "yamaha" and also some scratches, but now I decided to have a go at it.

It actually lifts the look quite a bit when it becomes more uniform I think

 

[Freezer]

Hi J Bjork,
really like following this thread, where in sweden are you? I'm near Falun and just embarking on a first conversion. would be great if i could ask what you would do differently if you were to start again?

 

[j bjork]

I live outside of Linköping.
I am rather happy with this bike as it is, but I would probably go with another motor if I did it again.
It works fine now after my modifications, but I will see in the summer if heat will be a problem.
It was supposed to weight 12,6kg if I remember correct, but it is more like 16kg in reality.

The qs 165 60h is supposed to weight 15,5kg and could be interesting I think. It sure would be nice with a narrower motor, but instead it will take up more battery space. So I am not sure if it is a better choice to go with that size of motor in the end, but I probably would if I did it again
Sotion also have a pair of motors in that size.

I think the connector on the battery is a bit too hard to connect/disconnect, maybe another solution would be better.

Controller.. I am not very happy with the reliability of the 3shul controllers, but I dont really see a better option atm.
I would probably wait as long as possible before deciding on that, and hope vesc labs comes with a proper sized controller or something.

I would defiantly go for other battery cells, now reliance rs50 or tenpower 50xg or maybe something else if I could wait.
Make sure to get grade A cells..

 

[j bjork]

Got the controller back, at first I was told I would pay shipping from them to me.
I understood it as I would get a new controller when they confirmed it was broken.
Instead they repaired mine with completely new internals from what I am told, and I got to pay for the repairs.
It was a very reasonable price though, so it felt alright anyway.

When I installed it on the bike I had some problems though, I got no 12v from the built in converter and no throttle signals.
I found out that the throttle was a wiring problem, they had changed pin 12 I think, from ground to "mode" input.
A little re-wiring and that was sorted.

Still no 12v though, I asked Hackey if it could possibly be some setting or something, which I highly doubt..
He replied the he should check, but after that I get the silent treatment. So I suppose I am pretty much frocked in that regard, and has now done some re-wiring for that too:

It wasn't that much connected to that 12v anyway, I tried not to put too much load on the fragile controller..
I had an external 5v converter for the display running from it (to not put load on the internal 5v), I connected that to my external 12v converter instead.
Then it was the waterpump control that I run from the aux output over a relay.
I changed that relay to a 5v, that I now run from the internal 5v.

So now I changed to the blue relay instead. I don't really like to run that from this 5v, but I cant use my external 12v if I don't connect that to common ground with the high voltage, and I don't want that either

 

[j bjork]

I realized it is about time to start building the third battery for the summers races, we are soon there
I have waited and hoped that some 6Ah cells would come out, but I don't think I can wait that long.
And I don't want to risk getting something that turns out to be problematic, like the early EVE 50PL.

The cells that seem to be the best available for my use are Reliance RS50 or tenpower 50XG.
RS50 has better cycle life and can take faster charging.
50XG has lower ir and lower voltage sag.

At NKON they are supposed to get 50XG about 28/3, and the RS50 about 5/5..
I was about to go for 50XG, but BlueSwordM is claiming RS50 would be a better choice

While I try to decide I started making the box for it:

I also started testing some spotwelding with 0.3 copper an 0.1 ss:

This is probably about 150J, and that seems to be what the k-weld can do.
It stuck well, but didn't leave much copper on the cell when I pulled it off.

I also tried with flux:

Then it did stuck really well, but it was smoking quite a bit. And still at the edge of what the welder can do.
These cells take more to weld than 21700:s, but I still think it would be uncomfortably close to max on the welder.
I decided to go for 0.25mm copper instead.

 

[j bjork]

Race season is starting, first race is on Monday.
I have been running tubes rather successfully this spring, only had one flat so far I think.
But it would be so annoying to get a flat in a race, so I decided to get a set of mousses:

This rear tire and mousse was a real pain to get on, it was fighting me all the way:

Even just to get here was a struggle
I didn't think I'd be able to get it on when I stood there swearing with sweat running down my face, dripping everywhere.
But eventually I got it on.
I was expecting a struggle on the front tire as well, but that went smooth..
The rear is pretty stiff now too, that mousse was probably one size too big for that tire

I have also had a vibration/ rough running between 50-60km/h.
I was in contact with Hackey who tried to make some adjustments in the vesc tool, but it didn't solve it.
He told me to ride a few km and send him a log, and he would get back to me the next day.
That would have been yesterday, he didn't answer then or today. That probably means he didn't see anything unusual, and don't know what to do.
The front sprocket was rather worn, and I noticed that one pin had moved in the chain, as the dirt hadn't stuck on it.
I decided to change it all, and that solved it!

 

[j bjork]

It has been a rather tough weekend..
So I finished the bike with mousses and new chain and sprocket.
I ran 14/48 sprockets before, but I only had 13 and 15 now.
I tried 15/51 to get close to the same gearing, but the chain got too short.
So I had to settle for 15/48, which turned out good for chain length.
The chain was always a bit long with 14/48.

There turned up a training yesterday, I thought I might as well go and test a little suspension with my stiff new mousses and mapping with the new gearing. Roll a few easy laps just to test..
Of course it ended with 5 laps of 8-9km each. So one less than a full session

When I was finishing the preparations for today's race and turned on the bike I heard a "pof" and the display, horn etc. didn't work any more. I thought it was the dc-dc converter that gave up, but then I couldn't turn off the bike..

Opened up the electronics box and found this:

The fuse for the dc-dc was blown. And a protection diode parallel to the relay, there to protect the on/off switch.
I thought something had happened to the relay, so it had welded shut, but it was the switch that was welded

I realized the diode was probably just for 75v, so that could have been a problem. But I doubt I even need the diode?
And if the diode broke, and as a result the switch welded shut. Why would the fuse for the dc-dc blow?

Another theory is the + for the dc-dc got in contact with something, it was very wet and it wasn't very good shielded.
But then, why the switch and diode?

Anyway, repaired and got to bed in decent time, but way too stressed out to sleep very much. But good enough to oversleep
I got to the race in time, and rode pretty good.
Some problems though, there was a long and a short track. My class was supposed to take the short, but just at the split on the first lap I was passing another rider and accentually took the long lap.

I had planned to do 3 laps, then change battery. I didn't make it when I took one long and two short
Got to push the bike to the pit, change battery and do 2 more laps.
After that I changed again, and did one extra long lap..

It was rather dirty even if it doesn't really look like much in the picture:

I don't have any logs from the controller, only these from the display:

These seems hard to see much in, sure you can see some max values.
Seems like I used rather much juice, 150wh/km. I hoped to stay in the 120-130wh/km range.
But part of the track is a soft, wet sand mx track, so I suppose it isn't very surprising

 

[bananu7]

Seems like I used rather much juice, 150wh/km. I hoped to stay in the 120-130wh/km range.

What was the average speed for that?

 

[j bjork]

Probably 29km/h, maybe 30km/h. I think I saw both in the logs, but I don't remember what the display showed after the race.