12 kw rc motor

[katou]

Idea I had, use torque screwdriver, make custom adapter to fit to shaft. Change torque setting upwards until motor stalls.

Not sure if it will work though.

Katou

 

[olaf-lampe]

Idea I had, use torque screwdriver, make custom adapter to fit to shaft. Change torque setting upwards until motor stalls.

Not sure if it will work though.

Katou

I believe the torque will be some 10 times higher than of any torque screwdriver. More comparable to an impact wrench.
But these woudn't work as counterpart.
The best way is to couple two motors together and use one as generator to recharge the battery. This way you can add a smaller* charger to the battery, which only needs to deliver the energy wasted by losses.
With this method you can do some longer tests with continuous voltage level.

[*] Would still be at least a 2kW charger. Bigger is better.

Here is an example of a huge motor testbench. Don't ask for expenses
The coupling should allow a little angular and radial disalignment.
Example: http://www.maedler.de/en/Catalog/P/1643/1622/1649/486.aspx

It's a big investment, also add expenses for measurment and datalogging stuff. But if you want to become a professional reseller ( outside the e-bike DIY-scene ), you have to provide some reliable data. It's for your own certainty and to evaluate warranty limits. Maybe you can find an independent engineering office, which could do the testing for you?
-Olaf

 

[mdd0127]

Here's another cheap dyno idea. Get an old exercycle with a friction brake. Use a known hp motor to determine the amount of friction braking at each adjustment. You could calculate the load with electrical measurements from the motor, temps, determine rpm. Not super accurate but should work ok. The one I found has a calorie meter. I'm guessing it senses brake friction and wheel speed to calculate calories. I haven't messed with it yet. Just a thought. I hope this makes sense. Tired.

 

[gip_mad]

Here's another cheap dyno idea. Get an old exercycle with a friction brake. Use a known hp motor to determine the amount of friction braking at each adjustment. You could calculate the load with electrical measurements from the motor, temps, determine rpm. Not super accurate but should work ok. The one I found has a calorie meter. I'm guessing it senses brake friction and wheel speed to calculate calories. I haven't messed with it yet. Just a thought. I hope this makes sense. Tired.

An exercycle is designed for the ~150W of human power, it will catch fire instantly with 100 times that power! :lol:

 

[dangerzone]

Here is an example of a huge motor testbench. Don't ask for expenses
The coupling should allow a little angular and radial disalignment.
Example: http://www.maedler.de/en/Catalog/P/1643/1622/1649/486.aspx

It's a big investment, also add expenses for measurment and datalogging stuff. But if you want to become a professional reseller ( outside the e-bike DIY-scene ), you have to provide some reliable data. It's for your own certainty and to evaluate warranty limits. Maybe you can find an independent engineering office, which could do the testing for you?
-Olaf

It might be cheaper to put the Collossus into a motorcycle frame and put it on a dyno test platform. I guess a 17' motorcycle wheel is close in diameter to a 26' bicycle wheel so Hal could get all dyno charts: torque, power and wheel speed under load. It's just an idea, why spend thousands of Euros on developing a testing method instead of using a professional one for a hundred or two..?

 

[mdd0127]

Here's another cheap dyno idea. Get an old exercycle with a friction brake. Use a known hp motor to determine the amount of friction braking at each adjustment. You could calculate the load with electrical measurements from the motor, temps, determine rpm. Not super accurate but should work ok. The one I found has a calorie meter. I'm guessing it senses brake friction and wheel speed to calculate calories. I haven't messed with it yet. Just a thought. I hope this makes sense. Tired.

An exercycle is designed for the ~150W of human power, it will catch fire instantly with 100 times that power! :lol:

Maybe the cheap ones are. You should see the one I found on the side of the road! I'll post up a pic when I dig it out from under the junk pile. I can torque that friction brake down far enough that I can't turn the crank with a 4ft breaker bar, I'm guessing that it's got over 800ft/lb of static holding force. Reducing it to half that pressure and spinning it up for 10 seconds should be enough of a load to test that big motor through it's RPM range. The load feels pretty linear too. When you crank it, it feels like it takes generally the same amount of force to get it going as it does to keep it going. None of this is science. I've taken no real measurements. I just saw the thing on the curb and thought, hey, I can throw an alternator on that thing and have a human powered generator. The dyno ideas have come since.

 

[olaf-lampe]

It might be cheaper to put the Collossus into a motorcycle frame and put it on a dyno test platform. I guess a 17' motorcycle wheel is close in diameter to a 26' bicycle wheel so Hal could get all dyno charts: torque, power and wheel speed under load. It's just an idea, why spend thousands of Euros on developing a testing method instead of using a professional one for a hundred or two..?

You're right about that. It's too expensive only for one test, but there are institutes ( don't know for Croatia tho ) that have all the equipment and they provide unbiased data.

How long will the test be in your case? Not long enough to get useful temperature readings IMO.
You would need a huge charger to provide enough energy, when you do not recuperate through a generator.
Or use huge batteries from a forklift maybe? Some are 48V, rarely more.

For most people the motor is THE deal, no matter if there are any relevant data shown.
It's the hype, that sells the motor. ( I'm infected too, but out of money )
If Marko and HAL want to sell it after the hype is gone, they better have arguments.

Only my 2 cents as always
-Olaf

 

[liveforphysics]

If you wish to dyno, one of my best buddies owns a dynapack 4000 series that I'm pretty much given free reign over when I want to use it. It's a hub dyno, so it's made to have the wheel removed and the hub bolted to the dyno module. It would be pretty easy to bolt-up a motor adapter.

http://www.dynapackusa.com/specs.htm
http://www.dynapackusa.com/product.htm

The only potential problem I see is it may not have the torque needed to make it turn. We dyno'd my Honda Insight Hybrid just for kicks, and it required major revs and clutch slip to get the dyno turning so it wouldn't just stall the engine. But with electric, no stalls... so maybe it would work perfect.

Anyways, it's something to think about if you want me to give testing a shot when I get a motor. I think it would only take a few hours to fab up the adapter plates and something to hold the motor stator rigid to the dynopack. The problem is, i have a 220 amp BLDC Kelly, but it's not going to be able to push this motor to it's limits.

 

[SilverSurfer]

It might be cheaper to put the Collossus into a motorcycle frame and put it on a dyno test platform. I guess a 17' motorcycle wheel is close in diameter to a 26' bicycle wheel so Hal could get all dyno charts: torque, power and wheel speed under load. It's just an idea, why spend thousands of Euros on developing a testing method instead of using a professional one for a hundred or two..?

You're right about that. It's too expensive only for one test, but there are institutes ( don't know for Croatia tho ) that have all the equipment and they provide unbiased data.

How long will the test be in your case? Not long enough to get useful temperature readings IMO.
You would need a huge charger to provide enough energy, when you do not recuperate through a generator.
Or use huge batteries from a forklift maybe? Some are 48V, rarely more.

For most people the motor is THE deal, no matter if there are any relevant data shown.
It's the hype, that sells the motor. ( I'm infected too, but out of money )
If Marko and HAL want to sell it after the hype is gone, they better have arguments.

Only my 2 cents as always
-Olaf

Yeah, Im not so worried about the test, as whatever issues it might have, gives me the need/oportunity to solve problems and improve upon the original. I just assoon buy it now and get started. Where I look forward to the testing results is in knowing the ideal drive ratio. Saves alot of time playing with sprockets. Nothing a great CVT couldnt solve, oh yeah, dont have one....

 

[dangerzone]

It might be cheaper to put the Collossus into a motorcycle frame and put it on a dyno test platform. I guess a 17' motorcycle wheel is close in diameter to a 26' bicycle wheel so Hal could get all dyno charts: torque, power and wheel speed under load. It's just an idea, why spend thousands of Euros on developing a testing method instead of using a professional one for a hundred or two..?

You're right about that. It's too expensive only for one test, but there are institutes ( don't know for Croatia tho ) that have all the equipment and they provide unbiased data.

How long will the test be in your case? Not long enough to get useful temperature readings IMO.
You would need a huge charger to provide enough energy, when you do not recuperate through a generator.
Or use huge batteries from a forklift maybe? Some are 48V, rarely more.

For most people the motor is THE deal, no matter if there are any relevant data shown.
It's the hype, that sells the motor. ( I'm infected too, but out of money )
If Marko and HAL want to sell it after the hype is gone, they better have arguments.

Only my 2 cents as always
-Olaf

How long..? I guess a day or two to get all the relevant data and charts done...

I myself wouldn't use batteries but a transformer with scalable voltages, currents, loads... Then put it to test the limits, the motor might give in to all the torture eventually so it would be smart to anticipate and calculate the 'red zone' of the motor before harsh abuse. But I don't think it's the hype that sells the motor, it is rather the price, the weight and what it can do. The competition of the Strecher RS990.40, Lehner 3080, Predator 37/6, Cyclon Big Boss and Hacker A200-8 which are all 15kW peak motors are at least TWICE more expensive than the Collossus. The cheapest of those competitors is 599Euro so for that much money I'd rather buy two Collossuses...

The only problem I see is a controller capable of taming such a motor, any ideas..?

 

[olaf-lampe]

The competition of the Strecher RS990.40, Lehner 3080, Predator 37/6, Cyclon Big Boss and Hacker A200-8 which are all 15kW peak motors are at least TWICE more expensive than the Collossus. The cheapest of those competitors is 599Euro so for that much money I'd rather buy two Collossuses...

The only problem I see is a controller capable of taming such a motor, any ideas..?

How about using the same controllers from the above mentioned competition?

I've read a few pages back, someone wants to use the colossus with a 100V100A Infineon controller.
Actually this is not an ideal combination. You might as well use a Turnigy motor instead and save a few 100$. Plus weight/size is also better.
A 300-500A controller would be a better match. Voltage depends on your application ( my estimation: 36-150V )
I believe some of the Infineon tuners are getting sweaty hands, by dealing with a motor that can easily draw all the power their controllers are supposed to deliver. I foresee desoldered powertrace reinforcements or even burnt PCBs.

IMO we'd need to develope a whole new powerstage for an infineon controller with solid powertraces separate from the controller PCB.
Marko and HAL surely are very interested to support such a development, but I haven't read a lot from them the past days?
-Olaf

 

[liveforphysics]

I've read a few pages back, someone wants to use the colossus with a 100V100A Infineon controller.
Actually this is not an ideal combination. You might as well use a Turnigy motor instead and save a few 100$. Plus weight/size is also better.
A 300-500A controller would be a better match. Voltage depends on your application ( my estimation: 36-150V )
I believe some of the Infineon tuners are getting sweaty hands, by dealing with a motor that can easily draw all the power their controllers are supposed to deliver. I foresee desoldered powertrace reinforcements or even burnt PCBs.

IMO we'd need to develope a whole new powerstage for an infineon controller with solid powertraces separate from the controller PCB.

-Olaf

100% agreement with all of that.

I think we should spread the work-load out a bit though. Mark and Hal are working on the motor. Bigmoose and myself are working on a controller design (planned to be 100% open source with DIY build instructions), as well as Jeremy Harris working on another controller design. Both in the multi-hundred amp level, which should hopefully do some justice to this monster.
The more people building high current BLDC controller designs the better!

 

[dangerzone]

The competition of the Strecher RS990.40, Lehner 3080, Predator 37/6, Cyclon Big Boss and Hacker A200-8 which are all 15kW peak motors are at least TWICE more expensive than the Collossus. The cheapest of those competitors is 599Euro so for that much money I'd rather buy two Collossuses...

The only problem I see is a controller capable of taming such a motor, any ideas..?

How about using the same controllers from the above mentioned competition?

I've read a few pages back, someone wants to use the colossus with a 100V100A Infineon controller.
Actually this is not an ideal combination. You might as well use a Turnigy motor instead and save a few 100$. Plus weight/size is also better.
A 300-500A controller would be a better match. Voltage depends on your application ( my estimation: 36-150V )
I believe some of the Infineon tuners are getting sweaty hands, by dealing with a motor that can easily draw all the power their controllers are supposed to deliver. I foresee desoldered powertrace reinforcements or even burnt PCBs.

IMO we'd need to develope a whole new powerstage for an infineon controller with solid powertraces separate from the controller PCB.
Marko and HAL surely are very interested to support such a development, but I haven't read a lot from them the past days?
-Olaf

Maybe they're busy working on the motor and their regular jobs during daytime, who knows..?

Anyway, your estimation seems right 'cause the controller should be in the 60v-150V range capable of 300A-500A for operation. I guess the controllers of the competition with the usual 60V/300A might not get out the best of the Collossus or might just fry with continuous use.

Solid separate powertraces of the FETs would ease the resistance/heat of the controller, although a 24FET Infineon with separate powertraces and cooling for the FETs (4115 or similar up to 150V) might do just fine. The FETs are usually mounted on an aluminum rectangular profile so a fan could blow air through it when the temperature rises, that reduces weight and size of the coolers.

 

[liveforphysics]

You can always keep stacking TO220 package FETs until the cows come home, 24fet, 30fet, 36fet etc, but at some point you've gotta realize it's a silly package to use for high currents. Heck, the legs on the things were originally designed for 40amps continous max, and IR does the fancy thicker leg package on the 4110's to give 75amp legs.

If you look at the TO-264 type packages, they were designed for handling multi-hundred amps right from the start. They have Pd's around 4-6x higher than the TO220 package FETs found in >99.9% of every e-bike controller.

Why are TO220's packages used in >99.9% of all e-bike controllers? Simple. You can get FETs in that package for as low as a dime a piece, and with most controllers doing 20-30amps, they can work out fine for people, and keep the cost of the controller or e-bike kit to a minimum.



Big FET packages = winner for high current situations. This application = high current situation in a big way. This is why we are chugging away on some designs built from the ground up for multi-hundred amp FET stages.

 

[dangerzone]

You can always keep stacking TO220 package FETs until the cows come home, 24fet, 30fet, 36fet etc, but at some point you've gotta realize it's a silly package to use for high currents. Heck, the legs on the things were originally designed for 40amps continous max, and IR does the fancy thicker leg package on the 4110's to give 75amp legs.

If you look at the TO-264 type packages, they were designed for handling multi-hundred amps right from the start. They have Pd's around 4-6x higher than the TO220 package FETs found in >99.9% of every e-bike controller.

Why are TO220's packages used in >99.9% of all e-bike controllers? Simple. You can get FETs in that package for as low as a dime a piece, and with most controllers doing 20-30amps, they can work out fine for people, and keep the cost of the controller or e-bike kit to a minimum.



Big FET packages = winner for high current situations. This application = high current situation in a big way. This is why we are chugging away on some designs built from the ground up for multi-hundred amp FET stages.

What's their impendance/resistance..?

 

[liveforphysics]

You can always keep stacking TO220 package FETs until the cows come home, 24fet, 30fet, 36fet etc, but at some point you've gotta realize it's a silly package to use for high currents. Heck, the legs on the things were originally designed for 40amps continous max, and IR does the fancy thicker leg package on the 4110's to give 75amp legs.

If you look at the TO-264 type packages, they were designed for handling multi-hundred amps right from the start. They have Pd's around 4-6x higher than the TO220 package FETs found in >99.9% of every e-bike controller.

Why are TO220's packages used in >99.9% of all e-bike controllers? Simple. You can get FETs in that package for as low as a dime a piece, and with most controllers doing 20-30amps, they can work out fine for people, and keep the cost of the controller or e-bike kit to a minimum.



Big FET packages = winner for high current situations. This application = high current situation in a big way. This is why we are chugging away on some designs built from the ground up for multi-hundred amp FET stages.

What's their impendance/resistance?

The motors? I don't think anyone knows yet. Or do you mean the larger package FETs? In which case it depends hugely on the intended voltage range and price you're willing to pay. Jeremy is working with some dPak7 packages that are something like 1.1mOhm if I remember correctly. I'm working with an IXYS TO-264 package rated for 150v that has 2.8mOhm nominal, with a Pd of 1670w. In other words, a 6-fet of them is more powerful than a 24-fet IRF4115 based controller.

 

[dangerzone]

Big FET packages = winner for high current situations. This application = high current situation in a big way. This is why we are chugging away on some designs built from the ground up for multi-hundred amp FET stages.

What's their impendance/resistance?

The motors? I don't think anyone knows yet. Or do you mean the larger package FETs? In which case it depends hugely on the intended voltage range and price you're willing to pay. Jeremy is working with some dPak7 packages that are something like 1.1mOhm if I remember correctly. I'm working with an IXYS TO-264 package rated for 150v that has 2.8mOhm nominal, with a Pd of 1670w. In other words, a 6-fet of them is more powerful than a 24-fet IRF4115 based controller.

Yeah, these larger package FETs.

The TO-264 are interesting cause they could withstand even higher voltages, how much do they cost..? I know a guy who wanted to make a 500V/500A controller based on those FETs but he works so much he has no time to do it. Too bad, that would have been a great multi purpose controller, for bicycles, motorbikes, airplanes, a whole lot of fun stuff... Is there anyone working on such a project so we might buy it..?

 

[olaf-lampe]

I'd prefer the SOT-227 design. These are easy to assemble on a heatsink and have a cooling surface to speak of. Pd is only a little higher usually, but you don't have to solder drain and source on a pcb.
-Olaf

 

[Arlo1]

Hows testing going?
I have the electronics for a inertia dyno.... I build a rc car 4wd dyno I have 460 lb steal drum to build a motorcycle dyno. My rc dyno maybe good for ebikes and scooters too! Just wondering if the lack of test equipment is holding you up?

 

[Tiberius]

On the subject of controllers

I think we should spread the work-load out a bit though. Mark and Hal are working on the motor. Bigmoose and myself are working on a controller design (planned to be 100% open source with DIY build instructions), as well as Jeremy Harris working on another controller design. Both in the multi-hundred amp level, which should hopefully do some justice to this monster.
The more people building high current BLDC controller designs the better!

The answer surely is to modularise it, and then we just have to agree on interface specs between the modules.
A controller system would split into: high power output stage, sensors, control hardware, control software.

Nick

 

[ryan_lirui]

This is really exciting stuff. I really want to get a motor as well when they're ready. I'd be happy to help if you need anything communicated that isn't getting through, I'm located over in China and speak mandarin as well.

 

[HAL9000v2.0]

I have tons of work on my day job so everything was slow. We manage to communicate with factory changes. (after 50 emails) . Next week I will take few days off and I plan to do next:
I will connect two motors together and run one with cca 10kw continues and load another with 3 heaters cca 2,8kw each. The motor motor cool with water and generator motor will not cool. Then measure currents, voltage and temperature on both motors so we can calculate everything from that. Also I did resistance measurement once again and it came out 10,93 Mili ohm. I will do it again and will measure temperature curve also so we can see quality of copper. It is basically all that I will be able to do myself.
I am looking for suitable bike donor to put one in but it is bad time of year to purchase bike...

 

[markobetti]

what about my aprilia af1 ? If you need it , its there....

 

[jandeman]

i would gladly donate my ducati rolling chassis for testing purposes, but it's somewhat far away for me...(1300km?)

 

[etard]

You don't have any Greyborg frames that you could scratch together a mounting bracket for this thing? The SLA's might help keep the nose down. Since it is sensored you could just get a cheap 80cc dirtbike donor and find out what kind of torque this thing puts out. Although, you're a bigger guy so 125cc might be closer to your size. Hell, there's your radiator right there. :wink:

Dual Collosus motors with one sensored 18 Fet and one Castle controller for space savings,with a bucket full of Lipo in a CR 125 frame, blacked out with gold Excel rims and some nice ASV levers. Now it's time to go eat a Zero and most ICE. I bet dual outrunners would just scream though!!