Yamaha YZ450F

[Pilip]

I don't think it will really work like this, first, the idea is to fling the oil off the rotor so the rotor will only get as warm as the oil that's picked up, heat won't conduct from the windings to the rotor. Second the rotor is already as hot as that anyway since it's inside with no way to conduct heat other than through the air (ok technically some small amount though the shaft I guess) which is doing the same thing as the oil but worse at moving heat to the water cooled case. And 3rd the rotor is also being heated by magnetic forces and you can overheat the magnets just from that so in some cases that oil may be cooling the rotor. And forth of course removing heat more efficiently in total will keep everything cooler since inside of the motor is more or less one thermal system.

Now will the potting compound also work well, yes and maybe better or worse than oil, no idea. So far I think I haven't seen any EV motors that are potted, some are exterior cooled like the QS we are talking about here and some have oil sprayer systems. Now why is that I don't know, maybe the oil spray works better since they can remove and cool the oil directly, many also use square hairpin windings so the copper is really packed in there and has a lot of contact with everything. Probably also since they can spray the end turns which have a lot of surface area they can suck a lot of heat out. With a wire wound motor though I bet the oil also helps to fill those wire gaps inside the stator better when compared to a hairpin motor.

I think though if you could ensure everything in the motor is compatible with silicone oil, which is probably the most compatible oil anyway and can be found in extremely low viscosities and is highly temperature stable (don't want it getting thick at low temps) may be fairly safe to try. If there is like too much viscous losses or something you could remove the oil, like even if you just got most of it out it would solve the viscous drag issue and the left over would only help conduct heat or you could probably wash it out if you absolutely needed.

Potting on the other hand seems much much higher risk, you mess it up and the motor is done. Really heavily varnishing the windings is the middle ground, a tried and true idea, I've never seen it go wrong and you probably get a fair amount of the gains from the potting by basically gluing all the windings into a block so they can conduct heat to each other and to the stator with no air getting in the way. I did this on my Leaf motor with just a million coats, the stuff is so thin is just soaks right into the windings but you have to repeat the application about a 100 times to get any real buildup.

Look, filling with oil works well with an outrunner when the oil removes heat from the stator, transfers it to the rotor and then to the external environment, in this case the temperature of the stator (windings) and magnets is equalized, which is also not positive, since the rotor with magnets is an intermediary for transferring heat from the stator, but it works. In the case of an inrunner, God himself ordered the heat from the stator to be transferred directly to the finned and cooled motor housing, so it is better to fill the stator with compound and this is done in industrial machines. This will require a vacuum unit and equipment for a specific motor. This is not a new solution. Winding varnish is used for other purposes - to glue the wires so that the insulation of the cores does not wear out during vibration.

 

[scianiac]

Look, filling with oil works well with an outrunner when the oil removes heat from the stator, transfers it to the rotor and then to the external environment, in this case the temperature of the stator (windings) and magnets is equalized, which is also not positive, since the rotor with magnets is an intermediary for transferring heat from the stator, but it works. In the case of an inrunner, God himself ordered the heat from the stator to be transferred directly to the finned and cooled motor housing, so it is better to fill the stator with compound and this is done in industrial machines. This will require a vacuum unit and equipment for a specific motor. This is not a new solution. Winding varnish is used for other purposes - to glue the wires so that the insulation of the cores does not wear out during vibration.

I mean yeah you are probably right potting the windings probably performs best of the 3 options (excluding pumping oil through a cooler) but have you ever potted a motor? I have a ton of experience working with casting resins in vacuum and pressure systems and even if I was to try it I would be super careful and still consider it a possibility that I would ruin the motor. In an industrial setting I still don't even think it's that easy (so it's expensive) which is why it's not super common and seems to be mainly used in extreme condition motors, military and aerospace, medical.

 

[j bjork]

I mean yeah you are probably right potting the windings probably performs best of the 3 options (excluding pumping oil through a cooler) but have you ever potted a motor? I have a ton of experience working with casting resins in vacuum and pressure systems and even if I was to try it I would be super careful and still consider it a possibility that I would ruin the motor. In an industrial setting I still don't even think it's that easy (so it's expensive) which is why it's not super common and seems to be mainly used in extreme condition motors, military and aerospace, medical.

What do you think makes it so difficult to pot the windings?
As you say, there probably is a reason it is so rarely done.

I think it should more or less be possible to make an insert instead of the rotor, maybe tape the inside of the stator or something to get some extra protection against getting potting on the surfaces.
Also make covers around the end turns etc. to not fill up the entire space with potting.

 

[Pilip]

There are enough examples of filling the stator with windings with heat-conducting compound, even the motor wheels are filled:

 

[scianiac]

I don't even think it's that hard but a project for sure and you have to make sure the plugs and molds are done well, resin where you don't want it won't be fun. And need a thin enough resin and ideally some assistance like a vacuum chamber to get it deep into the windings. I usually vacuum things and then put them in a pressure pot to remove all bubbles but probably not needed for this. If it sounds like a project you want to try I think you should go for, my point was mainly it's a project not and easy test and I'm sure there are some things that need to be carefully considered.

Like I think how I would do it is make a silicone mold that forms the core and one side around the end turns. This silicone mold is hollow and a slightly tapered plug goes inside it to push it against the stator and wall but also make it easy to remove. The mold for the silicone and the tapered plug could both be easily 3D printed. Potentially other just 3D printed methods as well but feel riskier.

 

[j bjork]

I have made final (?) top battery holders in tpu, to replace the gray pla ones:

I also got my new sprockets:

I went from 13/48 to 14/48, off course the chain got to short so I had to replace it too
It seems like a good gearing to run with 900pA as a base setting, but not very good in chain length.
It is almost adjusted all the way already, so I probably will have to change rear sprocket to end up in a good place in the adjustment.
I suppose I will try 14/50 and 15/50.

I was at a rather technical track this weekend too, but faster than the one last week.
Power use:

And motor temp:

This is from the last lap (4) though, but it sure is on the warm side I think.
It actually is very noticeable that it looses "bite" as it gets warmer, both in acceleration and braking.

But I must say, it is so damn nice to ride
The suspension is way too hard, especially on a track like these I have ridden so far, but the bike is so nimble and nice.
You can just step on the foot peg to throw it in to a turn or whatever.

The plan so far is to observe the motor temp on a few different tracks, then maybe try something and se what happens
Same thing with the suspension, I want to test it more before I start to make changes.

 

[scianiac]

You really need to dry your TPU lol.

I think you are on to something, if you keep track of temperature on tracks and with gearing changes that could be a good way to get an idea of efficiency. Testing or even calculating motor efficiency is tricky in ideal conditions let alone changing ones but motor temp kind of averages it all out.

 

[harrisonpatm]

You really need to dry your TPU lol.

And/Or lower nozzle temp by 5-10 degrees

 

[j bjork]

The gearing is mostly to find a good balance, if it is too low I just get wheelspin. If I lower the phase A to tame it, I loose power over the whole range. So I want tracktion and power without going too crazy on the phase A.

I have the waterpump set to start at 70 degrees now, I thought it didn't make a difference as the outside of the motor didn't get warm. But after a while it might do, when the heat evens out. So maybe I should set it back to 50 degrees.

I suppose I should get a filament dryer.. I print at 228 degrees now, I wonder if I can go much lower?

 

[scianiac]

So are you saying that if on the lower side of the gear ratio's you've tried you don't have an issue of top end torque or maybe you do but it's not too bad. Ideally you want it geared as low as possible while still having enough top end to let the motor spin fast to make power. If you are gearing it higher only to avoid wheel spin the ideal solution would be something different, what that would be I'm not sure.

Like in theory you have the same amount of traction regardless of speed so some gearing and phase amp combination will give you the maximum torque you can apply to the ground but in reality that isn't true so teasing out the difference there and how to adjust things is the trick.

Maybe low gearing, high phase amps but a throttle curve to give you more control in the phase current range where wheelspin is an issue.

Traction control would be nice but the only guy working on VESC traction control stopped, I've been wanting to look into it more, if his code is available I don't know how hard it would be update and get it adjusted to work with current firmware.

 

[amberwolf]

I suppose I should get a filament dryer.. I print at 228 degrees now, I wonder if I can go much lower?

I've been saving all the "dry packets" out of anything that has them, and tossing them into the Gamma Vault bucket I use for my filament storage.

They also make the same basic thing (silica "gel") as cat litter, and as plant moisture-retention crystals.

What shall we do with a 3D printer...early in the morning!

A good summary of (nearly) all the different filaments available, their pros and cons. I love these videos because I always find a couple that I've never heard of that I want to try. Anybody else try PCTG yet?

endless-sphere.com

What shall we do with a 3D printer...early in the morning!

A good summary of (nearly) all the different filaments available, their pros and cons. I love these videos because I always find a couple that I've never heard of that I want to try. Anybody else try PCTG yet?

endless-sphere.com

The GV bucket is the same kind I use for dog food storage, like the one next to Yogi's head

typical dry-packets found in all sorts of product packaging


Random google hit for silica cat litter https://www.amazon.com/LMJP-Moisture-Absorption-Unscented-Non-Clumping/dp/B0CYSNC5BJ

 

[harrisonpatm]

I've been saving all the "dry packets" out of anything that has them, and tossing them into the Gamma Vault bucket I use for my filament storage.

They also make the same basic thing (silica "gel") as cat litter, and as plant moisture-retention crystals.

What shall we do with a 3D printer...early in the morning!

A good summary of (nearly) all the different filaments available, their pros and cons. I love these videos because I always find a couple that I've never heard of that I want to try. Anybody else try PCTG yet?

endless-sphere.com

What shall we do with a 3D printer...early in the morning!

A good summary of (nearly) all the different filaments available, their pros and cons. I love these videos because I always find a couple that I've never heard of that I want to try. Anybody else try PCTG yet?

endless-sphere.com

The GV bucket is the same kind I use for dog food storage, like the one next to Yogi's head
View attachment 375695
typical dry-packets found in all sorts of product packaging
View attachment 375698

Random google hit for silica cat litter Amazon.com

View attachment 375700

Silica is good for keeping filament dry. Once it has absorbed moisture, forced drying is needed to expel it, using heat and moving air. Do you have a food dehydrator? Those work well for filament.

 

[amberwolf]

Silica is good for keeping filament dry. Once it has absorbed moisture, forced drying is needed to expel it, using heat and moving air. Do you have a food dehydrator? Those work well for filament.

I have the summer desert sun and breeze.... :lol: Just do it on one of those not uncommon <10% humidity days, when it's still cool enough to not melt the filament. (ok, it probably wouldn't ever *melt*, but it certainly gets hot enough in direct sunlight on plenty of days to soften it)

 

[j bjork]

So are you saying that if on the lower side of the gear ratio's you've tried you don't have an issue of top end torque or maybe you do but it's not too bad. Ideally you want it geared as low as possible while still having enough top end to let the motor spin fast to make power. If you are gearing it higher only to avoid wheel spin the ideal solution would be something different, what that would be I'm not sure.

Like in theory you have the same amount of traction regardless of speed so some gearing and phase amp combination will give you the maximum torque you can apply to the ground but in reality that isn't true so teasing out the difference there and how to adjust things is the trick.

Maybe low gearing, high phase amps but a throttle curve to give you more control in the phase current range where wheelspin is an issue.

Traction control would be nice but the only guy working on VESC traction control stopped, I've been wanting to look into it more, if his code is available I don't know how hard it would be update and get it adjusted to work with current firmware.

I think it was a problem that it "ran out of steam" at too low speed before, and probably still is. But on the tracks I have ridden so far that hasn't really been a problem, they are not that fast. I can also play with field weakening later if necessary.
I have a progressive throttle curve already that I am rather happy with, but it can always be fine tuned for sure.

I wonder how traction control should work without a speed reference
I mean if you don't compare motor rpm to front wheel speed or something.
One option could be to limit the rpm gain/sec to limit wheel spin.

I finally got my new display:

Got communication over can, now I should just install packages in the display and controller and it should work
Hmm, what is this:

Try to install the controller package first, doesn't work.
Ok, maybe I get the earlier package when I have 6.05, tries the install anyway..
I mean there isn't even a 6.06 right?
Now I have a black screen

Contacted support, they say I should wait a few days for the 6.06 to be released..
Then I should get the version for my 3shul controller from Hackey
That basically mean I'm screwed I suppose

 

[scianiac]

Yeah, not merging and including the controller firmware into VESC mainline is one of the main reasons I would think twice about buying a 3shul, getting updates you have to wait on Hackey and that sometimes takes awhile.

And yes the traction control would work like that by limiting the RPM rate of change, now how that works in practice I think that may require some tuning. Like the traction threshold may need to be adjusted based on conditions and possibly at different speeds. At the same time I don't see why we couldn't add a front wheel speed sensor and then use both combined to try and avoid issues like traction conditions, front wheel off ground, etc etc.

 

[j bjork]

The problem is that I don't find anything else that can handle the currents and voltage I want. Waiting for Hackey is very uncertain, sometimes he answers right away. He doesn't say "give me 2 weeks". If he doesn't have a solution ready, he probably won't answer at all.

 

[j bjork]

The new vesc labs looks very promising though, Benjamin Vedder himself involved, made in Sweden..
If they dont help I could even go home and wait outside Benjamins house
What they have coming is a little low in specs still though, if they could combine the 150V and 1000A of their biggest controllers it would sure be interesting..

 

[scianiac]

Now we're getting into too big territory but I did see Hackey say that he sells a VESC brain board you can just stick on a Chevy Volt inverter module. I wonder at what point in the controller size does that strategy start to make more sense, start using EV inverter boards. Those will mostly be in the 400v range but besides a little bit of resistance won't they work fine at 150v?

But yes I also am excited about the VESC VESCs, maybe ask him about testing a beta unit, maybe your locality may interest them in using you for some testing, what better testing than pushing it a little beyond it's current specs lol.

 

[j bjork]

The high voltage inverters from cars are usually big, and not that high current.
I think this is for a rear motor on an Audi etron:

Finally the battery is balanced, but it seems like there is an issue with self discharge on group 4

I'm balancing again to see if it stabilizes..

 

[j bjork]

I have been setting up a track today for competition tomorrow.
A part of it is a field where you take a corner, then hard on the throttle to the next corner where you brake hard.
Then full or a lot of throttle to the next corner..
We did several laps on that field to make tracks for tomorrows race.
This little motor wasn't really up for it

It got a little warm..
I felt like I lost a lot of power, but it doesn't look that bad in the logs:

The phaseA sure goes down some though:

When it was colder I sure hit higher power:

The battery voltages looks pretty good after today, I will check how they look tomorrow:

 

[gustavo88py]

how much temperature??
Motor controller and batery

 

[j bjork]

how much temperature??
Motor controller and batery

The motor and controller are right there in the logs. In the chart is the motor temp, the peak of 107 degrees is marked. You can see that it stays around 100 degrees for the rest of the log.
You can also see at this mark the controller was 39,4 degrees (temp MOSFET)
Here is a chart for controller temp during the same run:

There the peak of 47degrees at the end of the run is marked.
You can also see that the motor was 104,7 degrees at the time.
Here is another log, with some riding and waiting:

Here I seem to have hit almost 43kw even though the motor is almost 90 degrees
Here is the chart of the motor temp from that run:

I forgot to check battery temp unfortunately, that would have been interesting
My guess would be about 40 degrees. It is not in the logs, as I don't have vesc compatible bms.

 

[j bjork]

I have been trying to evaluate my battery the last few days:

Top left is soon after the ride, top right the day after I think. Most cell groups seem to have recovered about .005V
Bottom left, a day later. Seems like some groups recovered another .001V
Bottom right, group 4 and 20 have lost .001V

I know, .001V is nothing. But it is only 3 days, I suppose I should leave it for 30 days and see what happens.
Problem is I want batteries now! I think I will call this good and glue it together..

I am also working on the other battery, and it sure makes me less tempted to rip the good one apart:

I tried cutting with windshield cutting wire, but ended up more or less braking the halves apart with brute force.
Same thing when getting it out of the battery box, when you are just about to give up it starts to loosen a little..

 

[j bjork]

It's getting a little messy now, but now I'm finishing the good battery:

What the cell package looks like. The black is 1mm foam rubber.
The yellow is protective paper over double-sided adhesive 2mm foam rubber.

The second cell pack got stuck so hard in the cover that I could barely get it off.
That's why I made this cut out on this one. Not sure if it makes much of a difference though, it was the edges that got stuck the most. Maybe I should have took off the paper in the middle and left it on the edges

I used some crazy sticky butyl cord I had for the lid, the plan was that it shouldn't be too hard to get it off.
I kind of regretted it when I had all that sticky mess sticking out..

I was actually able to cut it off with one of these black-bladed very sharp knives,
then I glued the edge to get rid of the sticky edge.

When I was finished I checked the voltages again, and it seems like they have dropped rather equally
(The one to the left is a few days ago, to the right is today.)

 

[j bjork]

More battery posts.. Charged it full and ran until empty.
This is at the track a little after the bike stopped.
There sure are differences, but at these voltage levels there isn't much energy left.
Still, I probably could have done a better job matching the cells for each group.
I basically thought they were so close that it didn't matter after testing and just sorted out the weakest ones.

Now back to the bad battery:

This is odd, group 7 that I thought was a big problem doesn't seem to be a problem at all
9 seem to be a weaker group.

Cell 11-15, here we go. 14 sure is weak, seems like one cell missing.

Been able to get the bus bar off the cells, no cell drops to zero as I had hoped
Testing one cell at the time, I sure hope one will be dead..