crf450r 45kw race bike build

rivvs said:

no I changed to the xlite master.

Click to expand...

Ok, my thoughts would be to check the signals tab, you may be able to select the option that determines pack voltage from the sum of the cells rather than the overall pack I2C2. It would only be a temporary workaround.
Potentially the firmware isn't correct? Did you confirm the firmware with Kevin when you were setting it up?

bunya said:

Ok, my thoughts would be to check the signals tab, you may be able to select the option that determines pack voltage from the sum of the cells rather than the overall pack I2C2. It would only be a temporary workaround.
Potentially the firmware isn't correct? Did you confirm the firmware with Kevin when you were setting it up?

Click to expand...

Yes I had asked him and he tested it with my master xlite before sending it and it worked. I set him an email and he said maybe try swapping the polarity on the isospi. I will try. Because it seems like the bms do not see the slave board since every cells are at 6.53v

rivvs said:

Yes I had asked him and he tested it with my master xlite before sending it and it worked. I set him an email and he said maybe try swapping the polarity on the isospi. I will try. Because it seems like the bms do not see the slave board since every cells are at 6.53v

Click to expand...

Yeah right, that makes no sense. The only time I've seen a cell go out of range is when I had a loose balance wire connection. But that can't be the case for every cell...

A few checks that you may have already done but doesnt hurt to ask:
Have you wired the slaves in order from most negative to most positive?
Have you shorted all the unused pins to the most positive cell on that slave?
I had to double check my isospi wiring because the wiring from the master to the slave is different than the wiring between slaves. One was a crossover and the other was straight through. The documentation from Ennoid explains it but I needed to check it against the silkscreen on the pcb. I only had to do this because the supplied connections were too short for my install and I needed to extend the wiring.
Good luck getting it sorted.

bunya said:

Yeah right, that makes no sense. The only time I've seen a cell go out of range is when I had a loose balance wire connection. But that can't be the case for every cell...

A few checks that you may have already done but doesnt hurt to ask:
Have you wired the slaves in order from most negative to most positive?
Have you shorted all the unused pins to the most positive cell on that slave?
I had to double check my isospi wiring because the wiring from the master to the slave is different than the wiring between slaves. One was a crossover and the other was straight through. The documentation from Ennoid explains it but I needed to check it against the silkscreen on the pcb. I only had to do this because the supplied connections were too short for my install and I needed to extend the wiring.
Good luck getting it sorted.

Click to expand...

Yes the wiring is good it worked with the previous bms. I took the slave out and they seems defective. One has the led always on and the other the led flash between high and low intensity

rivvs said:

The other the led flash between high and low intensity

Click to expand...

Yeah mine doesn't flash, not sure what's happened there. Surely your run of bad luck must come to an end after this issue!!

Si Ennoid repaired the slave board and now it works. Only problems Is that it will not charge because the bums need to see voltage at input and the charger need to see voltage at the output so non of them turn on. I charged without the bms and did a test ride. It is fast down low with lots of torque but the top speed seems slow. Motor got to 45*c and controller to 40. I need to find a way to get the charger to work

Nice to finally get a decent testride
What is your gearing, settings and top speed?
I have 7,44-1 and a top speed of maybe 105km/h.
I have a little field weakening, but not enough to make a difference.
I think I had considerably higher top speed with the fardriver, it probably is possible to get a lot more with more fw.
But I dont really need it, and I am worried to blow the controller

j bjork said:

Nice to finally get a decent testride
What is your gearing, settings and top speed?
I have 7,44-1 and a top speed of maybe 105km/h.
I have a little field weakening, but not enough to make a difference.
I think I had considerably higher top speed with the fardriver, it probably is possible to get a lot more with more fw.
But I dont really need it, and I am worried to blow the controller

Click to expand...

I have 6.18-1 gear ratio. I think with tuning I should get more speed. I think something was getting hot because after several acceleration it seemed to get slower then after taking it easy it seemed to regain power. Also the cooling on the battery seems to be working since the side plate got warm!

Today I went to a sand pit to do some testing. I see a power output of 39kw on Vesc tool. But the problem is that the controller overheat in less than 2 minutes.
I lowered the phase amp to 600 and 300 batt and it still overheat. I mounted a pc fan to the controller and still the same thing. Maybe setting are bad and generate more heat

I'm trying to think which incorrect settings would generate more heat in the controller, basically any of them would do in the motor but assuming it's sticking to those current limits only a few variables should change the controller heat very much. Switching frequency and deadtime come to mind, lowering the switching frequency a little should be pretty safe as long as you don't go super low, I think 30khz is the default could try 25 or 20. Deadtime you shouldn't mess with unless you know what you are doing (which I don't). As always with changing settings make sure things are good before pushing things.

Besides that additional cooling is always possible, a water cooled loop, denser finned heatsink with more powerful fan (like a server heatsink which are often flat, high performance, and very reasonably priced on ebay). But I would stick a thermistor on the case, get the controller warm and make sure the difference between the reported vesc temperature and case heatsink isn't too far off, that is so say make sure the thermal contact between the FETs and case is good and heat is moving as it should.

I was at a sandtrack today, I have a log of a lap that might be interesting to you:



I suppose your 39kw was a top note, not a 2min run?
Anyway, my top notes seem rather similar in that case I think.



Here you have my mosfet temps, the motor curve looks rather similar.
But a little warmer as you can see.
This lap is about 17min 30sec and 9km long.



Here you can see some phase currents in the screen and something about efficiency in the numbers.
Dont know what to make of the numbers though, I think I am usually between 150-200wh/km.
But this track took more power than usual, but it still looks like normal efficiency

What do you have the controller overheat temp set to? 48C Mosfet temp is not even really warm and the motor is hotter anyway, considering the power your are moving that's not bad at all. Ideally of course you should consult the motor and controller manufacturer for their recommend operating temps but like 70-80C for the controller is pretty common and somewhat conservative temp limit setting. I think the VESC default is like 85 start 100 max, maybe like 70 start and 85 max or in that range feels a bit more conservative.

But if you can easily add additional cooling with a more efficient heatsink or fan then that never hurts as the cooler everything is the more efficient it is. You could also wire the fan up to a thermal switch so it only runs when it gets hot so you could run a powerful fan. I've always figured if I needed it that's the route I would go, maybe using a little control circuit to control the fan speed but thinking about it what would be simpler is to just get two thermal switches at two temps, like idk 40C and 60C then wire the lower temp one with a resistor so the fan comes on at a lower speed than goes full power at the higher temp. For instance get one or two if you can fit of those flat server heatsinks with the really dense fins and then one of the high power blowers from the same server and they could probably move a lot of heat if needed but still be very flat to fit easily. Just thinking what I would do.

As for the motor it looks like QS temp recommendations have quite a range depending on if you are considering max or recommended but 70C would be quite conservative. Looks like QS recommends like 70-120C and like 150C max but honestly I would set it to start limiting power above like 70-80 and max of like 100 to be safe. And again cooler is always better if you can cool things down.

Sand sucks a lot of power so you have to figure efficiency is going to be lower in it.

The one other thing that comes to mind is there is a VESC setting to adjust the motor parameters slightly based on the motor temp, so a hot motor will behave a little different than a cold motor. Haven't tried it myself but you notice issues with tracking, power loss not related to the temp limits and speed above the sensorless transition but still lower speeds and while the motor is hot could be worth trying but if that's not an issue.

So I did another test with a 230cfm 2amp fan. The overheating happened but like a minute later. The overheating is set at 65c. The fan was on from the beginning. So next I think I will need to play with the gearing. Maybe I am asking too much from the set up. I tried to do a log but it did not save any data.
Here is the GoPro video from the test.

There are cases when the controller circuitry does not correspond to the resistance of the motor windings (phase resistance is too low), then the controller will heat up in an avalanche-like manner, and so quickly that the fan will not have time to cool it. Then you need to look at how well the heat transfer from the MOSFETs to the radiator and, accordingly, the cooling fan is organized in the controller.

I cant see where the controller is mounted, but in older pictures I see a controller that looks to sit right under the seat with the heatsink facing up. Is it still the same? It dosent look like there can be much air-movement there.

Have you tried riding with the seat off?

j bjork said:

I cant see where the controller is mounted, but in older pictures I see a controller that looks to sit right under the seat with the heatsink facing up. Is it still the same? It dosent look like there can be much air-movement there.

Have you tried riding with the seat off?

Click to expand...

I changed that it is mounted sideway in the air box with heatsink facing the number plate now.

I said earlier the gear ratio was 6.18:1 but just realized that I got a 48 sprocket on the rear and that makes it 5.7:1. Maybe my gearing is too high

That sure dosent help, I would try to get close to 8:1 as a start. The bike would be a lot peppier too.
If you think you are loosing power at higher speed you can always gear up later.

Going to try with dual fan on the controller inspired by graphics card cooling! Also get a 52 sprocket for the rear hopefully see some improvement.

rivvs said:

So I did another test with a 230cfm 2amp fan. The overheating happened but like a minute later. The overheating is set at 65c.

Click to expand...

Is there a reason why you set the temperature limit so low?
I have never seen anyone trying to run a high power controller with such a low temperature limit, usually it's 90+ degree.

In my opinion your limit right now is not reasonnable.

Dui ni shuo de dui said:

Is there a reason why you set the temperature limit so low?
I have never seen anyone trying to run a high power controller with such a low temperature limit, usually it's 90+ degree.

In my opinion your limit right now is not reasonnable.

Click to expand...

It is the recommendation from 3shul. I asked hackey and he said I can up the limit to 70 degree. It is at 65 riight now

Maybe he meant 70 as the "cutoff start" and not the "cutoff end" which would be more reasonable and still pretty conservative. The 70 start and 85 end I recommended would give you a fair bit more headroom and still be well below what most use/recommend. If nothing else at least increase the cutoff end to 70C and you could lower the start cutoff. Even 5C warmer means those fans can remove more heat. Ideally the start cutoff should start pulling back power so it only rarely gets to the cutoff end or never depending on your use and preference. Yes you'll lose power when it gets warm but it won't cut out.

But yes also gearing will probably help a lot, honestly I would just set the gearing to something around your fastest speed you normally ride to match your motor unloaded speed based on voltage and KV (so not including any field weakening or MTPA) and then with MTPA on you should still have plenty of torque near that RPM but don't really have to use it to spin the motor beyond it's unloaded speed where there is diminishing efficiency.

It seems like mine is starting limiting at 85 degrees. I have never really bothered with the temp numbers, so this is how Hackey left them.


The same numbers on motortemp, so I would guess it is the default settings.

scianiac said:

Maybe he meant 70 as the "cutoff start" and not the "cutoff end" which would be more reasonable and still pretty conservative. The 70 start and 85 end I recommended would give you a fair bit more headroom and still be well below what most use/recommend. If nothing else at least increase the cutoff end to 70C and you could lower the start cutoff. Even 5C warmer means those fans can remove more heat. Ideally the start cutoff should start pulling back power so it only rarely gets to the cutoff end or never depending on your use and preference. Yes you'll lose power when it gets warm but it won't cut out.

But yes also gearing will probably help a lot, honestly I would just set the gearing to something around your fastest speed you normally ride to match your motor unloaded speed based on voltage and KV (so not including any field weakening or MTPA) and then with MTPA on you should still have plenty of torque near that RPM but don't really have to use it to spin the motor beyond it's unloaded speed where there is diminishing efficiency.

Click to expand...

Maybe I was not that clear on my settings. As a French speaker sometimes it is hard to explain things in English. The cut off start is at 65 the cutoff end is at 85. Hackey told me I could set the cutoff start to 70. I feel when I get to 65 as the bike becomes a lot slower then as it get hotter it gets slower and slower until the will not move faster than walking. And my hearing was based on the unloaded speed which is 4100rpm per Qs motor. This gives 55mph which is quite slow for motocross. It needs at least 60 to 65 mph to be competitive. So with the new gearing I will need field weakening

You only get 4100 unloaded? That sounds strange, I would expect something like 5500