29mm to 35mm to 50mm -> My evolution to the Cromotor

[methods]

No sir, you are incorrect.

Peak power occurs at 50% efficiency. You are not running peak power at 89% efficiency - you are running some arbitrary power. IF you were to apply more load your power would go up and your efficiency would go down - therefor you are not running at peak power. You are running at peak performance.

Just because you run a 45lb scooter motor in your bike that allows you to float around far from peak power does not mean that we dont operate ours at peak power. A perfect example is when a rider points a bike straight up a very steep driveway, stops, then starts again. If the motor runs up against the current limit then you are going to be running at the peak power which will be close to 50% efficient - and that is why it overheats so fast.

Any time you are accelerating hard you are getting close to peak power. That should not be mixed up with the constant power it takes to keep you cruising down the road.

-methods

 

[methods]

I wouldn't be surprised if while it was really hot like you guys were running that max power would be 20-30% lower than a run starting with a cold motor.

John

You are confusing the absolute peak power output - which... yes would be reached on a cold motor and be an absolute value.... with the fact that a BLDC motor produces its peak power (the peak power that it can produce at that time given temperature and other enviornmentals) at a 50% efficiency level.

These are two different things. One being an absolute value and the other being a rule of operation

I dont really like talking with you John. If you wish to step into my threads and take me to task on technical details I will probably spend a great deal of time picking apart what you say and arguing with you.... so you better be right and not in just in a "thumb in the wind" sort of way. As you may have noticed I left your threads.

-methods

 

[liveforphysics]

Just to clarify, when you increase current until your efficiency drops to 50%, you only reach peak power for that one RPM point. (at this RPM, no amount of additional current pumped into the motor will make a bit more power output)

If you increase the RPM, you can get more power output. For example, if the cromotor at 25mph is outputting 10hp when current is increased until efficiency drops clear down to 50%, if you were to do the same thing at 50mph, and go back to increasing the current until the efficiency drops to 50%, the peak motor power at 50mph would be roughly double, something like 20hp.

Nobody is going to design a system to run the motor power at this fairly absurdly low efficiency point, because the continuous power of a motor is TERRIBLE at this point (like, on fire in 20-30seconds bad). It's helpful for something like an electric dragster to know where to stop feeding in more current because it's not going to make you any faster.

The point of peak continuous power is a whole different number, and it generally happens in a pretty high efficiency part of the power curve (depends on many factors, including ambient temps, airspeed across the motor etc).

 

[methods]

Just to clarify, when you increase current until your efficiency drops to 50%, you only reach peak power for that one RPM point. (at this RPM, no amount of additional current pumped into the motor will make a bit more power output)

To be consistent lets call this the saturation point.
Most people are running on a current limit so in that case the peak current is already fixed and the variable is load - you can keep adding load until you reach the peak power for that current limit - which is how much torque and power you would have climbing straight up a driveway. Useful to know not just for drag racing.... as the #1 thing members ask me is about short burst hill climbing ability.

If you increase the RPM, you can get more power output. For example, if the cromotor at 25mph is outputting 10hp when current is increased until efficiency drops clear down to 50%, if you were to do the same thing at 50mph, and go back to increasing the current until the efficiency drops to 50%, the peak motor power at 50mph would be roughly double, something like 20hp.

Nobody is going to design a system to run the motor power at this fairly absurdly low efficiency point, because the continuous power of a motor is TERRIBLE at this point (like, on fire in 20-30seconds bad). It's helpful for something like an electric dragster to know where to stop feeding in more current because it's not going to make you any faster.

The point of peak continuous power is a whole different number, and it generally happens in a pretty high efficiency part of the power curve (depends on many factors, including ambient temps, airspeed across the motor etc).

To be consistent I would call this the continuous rating of the motor and it is really just the thermal limit of the motor. I think I was reading the other day that this number is commonly anywhere from 65% eff and up depending on how good the cooling system is.

So that is what a member wants to know... what is the peak power they will have climbing a very steep hill for a short period of time (what we were finding on the dyno) and what power level can they maintain on average - since of course it is a super dynamic number that will develop an average over time.

-methods

 

[Alan B]

A Torque vs Current graph is quite useful also and makes the different saturation points visible (there seem to be more than one). Justin did that for the 9C. I have not seen it for the Cromotor.

 

[John in CR]

Just to clarify, when you increase current until your efficiency drops to 50%, you only reach peak power for that one RPM point. (at this RPM, no amount of additional current pumped into the motor will make a bit more power output)

Exactly, max power out. Maybe I misunderstood Methods post where 50% efficiency and max current coincided. I figured the motor was just really hot and that accounted for the lower power than some of the other runs, but if that wasn't peak power out for that run then there's more current to give.

If you increase the RPM, you can get more power output. For example, if the cromotor at 25mph is outputting 10hp when current is increased until efficiency drops clear down to 50%, if you were to do the same thing at 50mph, and go back to increasing the current until the efficiency drops to 50%, the peak motor power at 50mph would be roughly double, something like 20hp.

Is that a dyno run kind of setting, ie where rpm is fixed? Or are you talking about the fact that every run other than the max power run crosses through 50% efficiency at 2 points with peak power and peak efficiency somewhere in between.

Nobody is going to design a system to run the motor power at this fairly absurdly low efficiency point, because the continuous power of a motor is TERRIBLE at this point (like, on fire in 20-30seconds bad). It's helpful for something like an electric dragster to know where to stop feeding in more current because it's not going to make you any faster.

Actually, I'm not so sure I agree with this, because our bikes are such light loads for the big power motors. I figure that I must have been getting fairly close to max power with my big diameter hubbie at 425A/550A battery/phase limits. Running the bike at max power settings wouldn't be an issue, because I simply can't present the motor with enough load running my 20s pack voltage. There isn't a steep enough hill. I stopped at the base of a 30% grade segment and accelerated up it significantly faster than my red bike you rode when you were down accelerated on flat ground. I have phase current restricted enough that I don't dump crazy power into it at the low efficiency rpms, so I only ever see peak power in for a split second and only during acceleration. I'm scared most about blowing my SteveO controller, so I never do more than 3 consecutive hard launches, and the ventilated motor never gets close to hot unlike the ventilated controller.

The point of peak continuous power is a whole different number, and it generally happens in a pretty high efficiency part of the power curve (depends on many factors, including ambient temps, airspeed across the motor etc).

That's why I chose the 100A motor test report for my big high efficiency hubbie as the example in my previous post. It's the one they used to rate the motor as a 6kw motor. The 700W+ of heat it has to shed to achieve that 6kw peak power out continuously at 89% efficiency is pretty reasonable given the surface area of the motor and being bolted with a large flange to an AL rim, and I'm sure their big 300lb+ scooters could easily present a continuous 6kw load at 750rpm. On my bike though I can't find a hill that's steep enough and straight enough to recreate that load. The 8% grade hill on the highway doesn't do it, because I'm able to maintain at least 55 up it. That's why even sealed up in stock form I don't have heat problems at double the current of the factory controller.

 

[liveforphysics]

A Torque vs Current graph is quite useful also and makes the different saturation points visible (there seem to be more than one). Justin did that for the 9C. I have not seen it for the Cromotor.

Yes. I would like to make one of these graphs for the Cromotor so we can quit with the guessing and know.

 

[methods]

Ok - I will sell you a CroMotor for $800

Baby needs a new pair of shoes....

-methods

 

[hjns]

That's why I don't even use temp sensors, since I have enough load to tell when power drops off from hot windings. I back off immediately and the windings cool quickly as that heat spreads to the iron of the stator, so as long as I keep the load light and rpms up the motor is able to shed the heat without a failure and get down to reasonable temps pretty quickly.

Quick shedding of heat? Hmm. Not in my cromotor.
I pushed my cromotor up a steep hill for 4 intense minutes (see here), making approximately 50A continuously at 80V during those 4 minutes. After that I drove fast but without stress, like 20-30A continuously. Initially, the temperature sensor on the windings of my cromotor showed an initial increase from 30oC to 80oC, continuing going up to 100oC even during my descent down the steep hill (admittedly with heavy regen), and then the temperature did not change but remained at 100oC during the rest of my ride. This is the stock cromotor configuration as I have not yet filled the hub with oil or vented it with big holes.

 

[liveforphysics]

That's why I don't even use temp sensors, since I have enough load to tell when power drops off from hot windings. I back off immediately and the windings cool quickly as that heat spreads to the iron of the stator, so as long as I keep the load light and rpms up the motor is able to shed the heat without a failure and get down to reasonable temps pretty quickly.

Quick shedding of heat? Hmm. Not in my cromotor.
I pushed my cromotor up a steep hill for 4 intense minutes (see here), making approximately 50A continuously at 80V during those 4 minutes. After that I drove fast but without stress, like 20-30A continuously. Initially, the temperature sensor on the windings of my cromotor showed an initial increase from 30oC to 80oC, continuing going up to 100oC even during my descent down the steep hill (admittedly with heavy regen), and then the temperature did not change but remained at 100oC during the rest of my ride. This is the stock cromotor configuration as I have not yet filled the hub with oil or vented it with big holes.

I look forward to hearing how temps do after oil filling. I think oil filling hub motors is the best thing since sliced bread.

 

[Warren]

Luke,

"I think oil filling hub motors is the best thing since sliced bread."

You know about the oil filled ZEV hub motors?

http://www.zelectricvehicle.com/6.html

Warren

 

[Alan B]

...

Quick shedding of heat? Hmm. Not in my cromotor. ...

Thanks for the real data. Running fast does improve dissipation, but just because the side covers are cool doesn't mean there isn't a lot of heat still stored in the stator.

We need built-in temperature sensors in these motors!

 

[hjns]

Thanks for the real data. Running fast does improve dissipation, but just because the side covers are cool doesn't mean there isn't a lot of heat still stored in the stator.

We need built-in temperature sensors in these motors!

Let me see if I can get a second thermistor up and running on the side cover, outputting directly into the CA analogger. That will give some nice hard data around the supposed heat distribution from windings to side covers. Will take a week or two, because I have to attend a congress in Vienna coming weekend (ERS).

For all my controllers, I plan to have a thermistor on the FETs as well. As the CA analogger has two extra inputs, I can even log it. Nothing beats hard real life data. 8) Which is the reason why I am so happy with Methods' posts. REAL data (when he is not talking about baby socks. Thank God I passed that stage 2 years ago).

 

[Alan B]

A sensor on the side cover would be interesting for certain. How are you going to handle the issue that it is moving?

 

[methods]

A sensor on the side cover would be interesting for certain. How are you going to handle the issue that it is moving?

He should use the wireless temp probe that I include with my arduino shield just for that purpose. Finished they layout and it is ready to go to fab...

Two hall current sensors
four rtd temp sensors
One IR temp sensor
Micro sd memory card for datalogging
4x20 lcd for real time data
Prototype area with I2C, spi, uart, adc, pwm, etc access
Sits on an arduino mega
Free starter sketches.
Gui user interface for settings

We already have three unique applications running on the same hardware set




-methods

 

[zappy]

For the 24 fet I am trying to find a way to fan-cool the case without actually blowing air through the inside. Blowing dirty ass air over the circuit board is a guaranteed fail after enough time unless there is a really thick conformal coating or a filter on the inlet.

Hi meth, I cut 18mm out of my 24 fet case in width and then rewelded it back together and cut 3mm off the blank end cap of the case and 25mm off the wire oulet end and made new ally end plates. For the fan i silicon glued on some fiberglass end plates to the heat sinks so all air was forced down between the heatsinks and the only way out was past all the hot fet legs. The heatsink/fet gap has a similar cross section to the 22cfm Jaycar fan outlet. It was a fair bit of dickin around but it ran cool ,clean and now fits nice and low on top of my battery pack between my cranks, and dosn't rub the front tyre under full bump. I only run 5000w -8000w.
Zappy

 

[methods]

Hmmm... you put a lot of thought into that. Very clever design. I am unclear - where is the final outlet, is it at one of the end caps?

I like that you used a filter and really forced the air by where it needs to go.
Next time if you run your three phase wires off the bottom of the board you wont have to seal them at your heatsink end cap. Might not be able to bring them back up tho with your shortened housing layout... but then you could just exit them right through the end cap from under the board - straight shot.

Nice work shortening up the housing too - that has been a subject we have been spending a lot of time talking about. I would like to secure a smaller housing with larger fins - as the current housing is clearly just "something it kind of fit into". I am surprised that after 2 years we are still using the same over-sized housing.

Thanks for stopping by and contributing.

-methods

 

[johnrobholmes]

I love seeing the controllers getting hacked up more and more. It is a sign that we will soon be producing more parts in countries like the US instead of just assembling chinese parts with beefed up traces and nice fets. Even just a simple anodizing is a great first step to the new era!

 

[methods]

We are spending more and more locally for our builds. Pretty soon I think we are going to be able to hire a local assembler full time... as Matthew moves on to better things. The more I look at these Infineon boards the more I realize that I could just lay my own out - just dont think I can do it for as cheap as they are doing it yet. It is only like $20 for those blanks - loaded with SMD parts - if you buy a box of 50.

-methods

 

[johnrobholmes]

It is tough getting smd work as cheap as china. Heck, even the assemblers I use here in the states get boards made at Gold Pheonix. If you were to remake the infineon board you would make improvements at the same time, so it really wouldn't be the same $20 populated board any longer and would have more value. Just popping it out with better power layout to reduce inductance and increase the vias to 6 ounce copper with some additional buss bars would be a great start. Just think of all the time that could be saved VS beefing up traces. In all honesty I am amazed at how much power people have pushed through these controllers already.

I'm hoping that starting up my new local bike shop will help me focus on products we need for ebikes. I'm so stuck working on robotics and off road RC truck parts to pay the bills that I just haven't had many ideas for ebikes lately. I know there is plenty of progress to be made, there just needs to be more people working on it! It probably won't be long until we can make complete motors in the US for a small fraction more than a china made motor, assuming similar quality.

 

[zappy]

from a PM

i love that blower mod. where'd you get it? got a link? how many v/a does it need?

Jaycar electronics in oz part no. yx-2530 9.5cfm or yx-2532 26cfm they just resell them, you could get probably source them locally with a bit of a search. I ran my fan off a 240/110v AC in, small wall plug switchmode power supply 15v out, off my DC main battery pack and it works(i can't explain how). Or i have 4 internal fans inside my hub and i can series all my fans to my 20s 66volt pack. and have a small dash switch. You can see the outlet vent on the end plate it has oiled filter foam over this usally. hope this helps

Next time if you run your three phase wires off the bottom of the board you wont have to seal them at your heatsink end cap. Might not be able to bring them back up tho with your shortened housing layout... but then you could just exit them right through the end cap from under the board - straight shot.

Yes this would be neater big better

I am surprised that after 2 years we are still using the same over-sized housing

yes the 24 fet is nearly a shoe box size case with a dual suspension bike space gets very tight quick. Kelly's controller extrusion is a better finned for cooling design.

Meth is right controllers used to be a problem, his new modded controller is mega and bestbang for buck and puts out huge phase amps. The problem is most current hub motors (as well as weight ,overgearing, mao quality etc) is their ability to shed the heat at the rate they produce it.
Zappy

 

[Alan B]

With a low impedance 24 FET controller, do we need all the cooling on the controller? Seems a little external fan, or just exposing it to the outside air would be adequate unless you are melting bigger motors than the Cromotor?

 

[liveforphysics]

With a low impedance 24 FET controller, do we need all the cooling on the controller? Seems a little external fan, or just exposing it to the outside air would be adequate unless you are melting bigger motors than the Cromotor?

In our dyno testing where my goal (as always) was to release the poor captured magic smoke from the drivetrain of Methy's trike, the CroMotor was making some pretty foul burning smells while his controller was still comfortable to hold your hand on.

An oil cooled cromotor or highly vented/ducted motor might be a different story though, this was just a bone stock cromotor on our test rig.

 

[zombiess]

With a low impedance 24 FET controller, do we need all the cooling on the controller? Seems a little external fan, or just exposing it to the outside air would be adequate unless you are melting bigger motors than the Cromotor?

In our dyno testing where my goal (as always) was to release the poor captured magic smoke from the drivetrain of Methy's trike, the CroMotor was making some pretty foul burning smells while his controller was still comfortable to hold your hand on.

An oil cooled cromotor or highly vented/ducted motor might be a different story though, this was just a bone stock cromotor on our test rig.

So what is your personal opinion of the cromotor? Good or bad? I loved mine before I ever got involved with greyborg on the business side.

 

[Stochastic]

Here are the pictures of the "Face Smasher" I built. It is essentially a hard tail mountain bike with Zero forks in front with 24" wheels in front and back.
It is an incredible beast to ride.




View attachment 4


I built the torque arms out of an old battle axe my brother made in high school... Hand filed to fit the 16mm axle.
Thick solid steel torque arms were the only way to go here. :mrgreen







I'm still in the process of building the 24S4P LiPo pack.

-Matthew