CroMotor - Dyno Proven Efficiency Data

[methods]

I just made up that title because I knew it would arouse Luke

So I finally finished my long-awaited 24 fet controller run and I was looking for a way to burn them in before releasing Beta units. It just so happens that I know a guy who has a fetish for blowing controllers and motors... so I packed up my Jeep and headed over there. Here was the Setup

CroMotor laced into 20" on crap rubber
24S 20Ah Pack with a top-off charger running
My super-pimped 24fet controller programmed to 150A DC and 300A Phase

All of this was bolted into my KMX trike.





I really thought I had a setup that would be great for exercising the controller. Turns out I was totally wrong... we could not even get the controller warm and the room was reeking of smoking hot laminations :?

EDIT: LOADED UP MORE VIDEOS

CroSmoker Dyno Videos

10hp to 2hp sweep
[youtube]1Ori-OBUMmg[/youtube]

10hp to 3hp run
[youtube]SNeZvyovgTs[/youtube]

10hp to 5hp rip
[youtube]4nhgZeBcgRo[/youtube]

7hp continuous run
[youtube]giw0sW_xCUg[/youtube]

10hp continuous run (original post video)
[youtube]dAj_CM8aoWI[/youtube]

methods clowning on Lukes motor blowing fetish
[youtube]3RZySSdM6qQ[/youtube]

So... yea
We really had a hard time getting the motor to load the controller enough at 100V. I wanted to get in a 15hp run but to do that I think we are going to need to change a few variables. I have a ton of Dyno Data for you guys to sift through We ripped it a bunch of times and just kept cranking up the load - cranking up the load.

So we calibrated the shunt in my CA (8W shunt)
The Videos capture the mechanical output as well as the CA data - so with that you can calculate efficiency and no doubt you can also ferret out the point at which we were saturated.

Here is what the guts of my prototype look like. Notice that I moved all of the heavy wiring to the bottom of the board - this totally cleans up the top of the board leaving room for either very easy troubleshooting or 2X the Electrolitic capacitors. All those "surf boards" you see are my favorite 250V 1.2uF caps in 4P - so 4.8uF each. Notice that they are strung from the Source of one bank across to the Drain of the other bank. This was Big Mooses idea - though I am not sure if I implemented it how he intended for me to. Anyhow - I have more ultra low ESR caps on this controller than anyone has ever had. Forget that Poly crap... these caps are tiny.

Yes I know the leads are not optimized yet... the original plan was to solder the caps right down on the board but they started popping on us like popcorn - too much flexing on the PCB - to these boards decouple the sensitive caps from all that motion and heat.



To give you a sense of scale that is 6AWG stranded on the traces. The phase wires are 8AWG noodle and the power and ground are dual 10AWG noodle.

The hot rail (the high side) has a 70C thermal cut-out. I was eager to try it out but we were hardly even able to warm the controller up even after a few hard runs.

View attachment 1

I know I have already posted these pictures a couple of times but mmmmm... would you look at the color of those housings? It just makes my hubgina lube up.



So... to the point. Attached is some of the dyno data. I am not sure which runs these were (have not even looked at it) but I think there is some good data in here. It wont be quite as useful until I post the Video with the matching CA data so you know power in.

Raw Dyno Data for download
(note - there was significant traction issues so take that into account while evaluating the data. You can probably hear when it grabs in the video)

I am at this moment waiting for the video to email from my phone to my computer. Dude.. dont ask. The USB is broken and I dont have a uSD reader right now so my only option was to email the 20M movie to myself :x

When we finally loaded it down so hard that it was just groaning we managed to eek out 10hp mechanical. I dont recall completely off hand but I think that was with about 15kw going in. Video or it didn't happen right? Uploading...

So I am actually really surprised and disappointed with how how quickly we had to quit due to CroSmoke filling the room :x This motor absolutely needs to be oil filled ASAP if we have any hope of doing some meaningful testing for my controller.

Until then - I dub my controller the "CroSmoker" until proven otherwise.
(though thankfully I was able to shut Luke down before he managed to actually smoke my personal Cro... for which I do not have a replacement)
(Ok - I have 3 replacements but I would rather sell those to you fiends and buy tires for my Jeep)

Lets see who the first is to have the time to pick through the video and plot the input power vs output showing eff.

I will upload the other videos if more data is needed.

-methods

 

[Harold in CR]

Poor motor was really grunting, EH ??? :lol: :lol:

Sure glad to see you finally got the controller working, Methy. Congrats. 8) 8) 8)

 

[methods]

I just realized that I am a retard.

So grabbing a random data point here is the calculation

13.4kw ~= 18hp in
resulting in about 9hp out
So clearly we are right at the 50% peak power point

Current was pegged at 150A - the DC limit - so.... yea. Duh

94V * 150A = 14kw
14kw ~= 18.8hp
Peak power will be at 50% efficiency so peak power will be limited to about 9.4hp

OH... WAIT... 9.4hp was exactly what we were seeing on the screen
omfg the numbers actually agreed with what we measured. Something must be wrong.

Ok - so clearly if I want to do a 15hp run I need to turn up the DC current limit.

That means we need to see 30hp coming out of the controller which would be about 22kW
22kw / 94V = 240A

ugh... let me run that back to make sure

94V * 240A = 22.5kw
which is about 30hp
which would be about 15hp at the wheel at 50% efficiency

Ok.
Well....
240A / 20C = 12Ah
I guess that is a reasonable amount for someone to source. I will do it.

Fetish man.... prepare the dyno!

-methods

 

[methods]

This time I want to do steps

50A
100A
150A
200A
250A

We can then definitely see how the motor responds and when we saturate. Clearly we are not saturated yet if we are still making peak power at 50% and all the numbers add up as cleanly as they do now.

bugger 9.4hp
Even little kids run 15hp or 20hp in their little plastic barbie cars.

-methods

 

[hillzofvalp]

Love it.

Do you really believe oil cooling is going to be a good option for this power level? I am still stuck between water cooling, air cooling, and oil cooling, but th easiest to try first is air and eliminates chance of using oil.

If this motor was kept under 90C with a nice wAtercooling setup, what power could be put through it? Obviously lots of variables, but maybe you can draw up your theory..

 

[hillzofvalp]

50% is still 100% better than a tuned 75cc motor by the way

 

[methods]

As soon as we find the saturation power you will know how much you can put into it (usefully). At that point it is a simple problem to solve - metal is hot and heat needs to go away. That is something that is well understood. The only limitation of any motor (before the saturation point) is your ability to keep it cool.

I am going to go with oil - if not just for the fact that it sounds like such a bad idea. Makes perfect sense tho - we need to get that heat transfered to the case so it can be cooled off.

-methods

Love it.

Do you really believe oil cooling is going to be a good option for this power level? I am still stuck between water cooling, air cooling, and oil cooling, but th easiest to try first is air and eliminates chance of using oil.

If this motor was kept under 90C with a nice wAtercooling setup, what power could be put through it? Obviously lots of variables, but maybe you can draw up your theory..

 

[methods]

50% is still 100% better than a tuned 75cc motor by the way

Well... we can run it at any efficiency we want but with a brushless motor peak power *always* occurs at the 50% efficiency point. Just the nature of the beast.

-methods

 

[hillzofvalp]

I'm tempted to tear apart my motor today for measurements for air/water cooling.... Just got a gear puller. DONE. PROJECT FOR TODAY. Idk if you heard what I was trying to do with my cromotor... But below you will get an idea of my plans to put it in small, originally 200lb motorcycle..

 

[hillzofvalp]

Btw I'm somewhat interested in your controller. I will be running 28S A123. What are the advantages to your controller over say a 150-200A continuous rated Kelly khb?

 

[methods]

Btw I'm somewhat interested in your controller. I will be running 28S A123. What are the advantages to your controller over say a 150-200A continuous rated Kelly khb?

Kelly will give you thrust control but mine will give you the ability to put out double that current. Remember that a Kelly is rated by its phase current, not its DC current.

Here is a crappy graph - see how everything crosses right at 50% efficiency?

-methods

 

[methods]

Ah - this one I like better
Perfectly illustrates what you should expect the motor to do. Remember that the chart will deform when you add in fixed current limit - the thrust and current lines will go linear.

-methods

 

[methods]

Here is a simulation from the ebikes.ca calculator. It is not accurate (probably due to the fact that Justins Power supply was not rated that high) but it is representative. Look at the relationship between efficiency and peak power. It is all really beautiful.

The only ugly part is where it comes off of the 150A current limit and then becomes smooth. If you wanted to make this graph truly beautiful you would have to increase the DC current limit until the thrust like was perfectly smooth.



-methods

 

[hillzofvalp]

Got it. How much are you going to sell the 4110 version 24FET for? Can't I get thrust controller with current throttle? No big deal to me...

Hey! Can you please do some destructive testing of the axle? Do a regen mod and put it on the dyno and do full thrust full regen for 15 minutes! That would make me feel so much safer! Edit: you ought to be able to load it up to regen at least 40A... With shunt mod maybe twice that.

 

[hillzofvalp]

And I'll buy the motor remnants from you if the axle fails! Lol

 

[ohzee]

Some very nice data.

Regarding emailing a 20m file it would be a rare ISP to accept a filesize that large , but I could be wrong.
I just know where I work for a cable isp we reject over 10m. Still filedump or some app like that should
work fine.

Anyway look forward to the next test. Glad Luke and yourself are good enough buds to take it to the next level.

 

[methods]

Got it. How much are you going to sell the 4110 version 24FET for?

We are going to sell them for $420 - for no particular reason other than the number popped into my head. I dont really care if people buy them or not and the truth is that I would prefer the cheap-asses to buy from Lyen He can deal with the endless headaches. At lest I know that I have filtered out all of the customers who dont have the patience to save up $420 :wink:

Once I roll out the whole package and show the build thread I think it will become obvious - we build these things by hand ourselves and they have a laundry list of value added features that truly spans a page. These were the best that we could build... and people who have actually done up controllers from scratch will totally recognize the value.


-methods

 

[methods]

And I'll buy the motor remnants from you if the axle fails! Lol

The last thing I am worried about as axle failure. For every one axle failure there will be 30 overheated motors.

-methods

 

[zombiess]

Once I roll out the whole package and show the build thread I think it will become obvious - we build these things by hand ourselves and they have a laundry list of value added features that truly spans a page. These were the best that we could build... and people who have actually done up controllers from scratch will totally recognize the value.


-methods

Thanks for putting it on the dyno . I've been trying for months to get my bike on a friends dyno but it's never happened because he hasn't had it setup. What kind of load did you have the dyno set for to hold that amperage? I've used a lot of dynos but only on a and alcohol powered cars. Always fun when you can adjust the load. I hate non load cell dynos after using adjustable load for tuning. Good thing you made Luke stop. That guy just loves breaking stuff for fun even after he has good data lol.

What you say is true about the controller building. The 20 24 fet controllers I have that I started building aren't getting built very fast at all. Its a massive time sink to build them. When I purchased them I had 4 days a week off work to build them but shortly after I ordered them I got a 40+ hr per week job and lost almost all motivation because I'd rather veg out on the weekend then spend many hours soldering. Your improvements to them make them worth the price. I'm probably going to finish a handful of 4115 and 4110 controllers and sell the rest as kits unless I find a lot of motivation or hire your friend Matthew.

Btw, change out that shit tire for a pirelli ml75 moped tire and you'll hook up much better.

Are you interested in dyno testing you motor at 125+ volts? I can loan you a controller. I would be interested in seeing how both the motor and controller perform. I know from experience that moving from 100v to 125v is a pretty big difference in feel even when keeping the battery amps the same.

 

[Arlo1]

Great thread methy this will help alot of people understand why putting more in doesnt always get more out! I bet for 15hp out you need ~3x what you had for 9 hp!

 

[hillzofvalp]

And I'll buy the motor remnants from you if the axle fails! Lol

The last thing I am worried about as axle failure. For every one axle failure there will be 30 overheated motors.

-methods

Please prove it when u have time... For every 100000 overheated motors, someone ends up in the hospital. One might argue more personal damage from a broken axle. All it takes is Lots of torque, a hill, and a traffic situation requiring more acceleration. Dropping off the bike in front of traffic... Bad situation.

 

[SweKen]

Gotta love methods wild style...subscribed.

 

[gensem]

hahaha, I wanna see plasma!

 

[Alan B]

Nice to learn where the limits are.

The limits being what it can do without damage.

 

[steveo]

Got it. How much are you going to sell the 4110 version 24FET for?

We are going to sell them for $420 - for no particular reason other than the number popped into my head. I dont really care if people buy them or not and the truth is that I would prefer the cheap-asses to buy from Lyen He can deal with the endless headaches. At lest I know that I have filtered out all of the customers who dont have the patience to save up $420 :wink:

Once I roll out the whole package and show the build thread I think it will become obvious - we build these things by hand ourselves and they have a laundry list of value added features that truly spans a page. These were the best that we could build... and people who have actually done up controllers from scratch will totally recognize the value.


-methods

Right on lol