Castle Phoenix HV ESC 30 for brushless HUB motor (250W)

evaleto

10 mW
Joined
Jan 4, 2011
Messages
29
Location
Geneva
Hi,

I have read a lot on this forum about Castle ESC and ebike motor. It seems clear that a solution with Castle ESC is not easy.

We are building a new ebike prototype based on a fixie bike declination. Our target is this word: minimalist. No wire, minimal battery, minimal motor power and assistance only when it really needs ;)
- 10S2P of panasonic NCR18650A (6.1AH 37V)
- brushless motor 250W 36V 1,6Kg!!! (http://www.ecitypower.com/pdf/Q-F-85SX-Specification.pdf)
- Throttle by arduino and accelerometer
- controller, 36V 8A, 15A peak, Impossible to find smallone with those specs ? Expect this one (http://www.aliexpress.com/store/402602/210093832-318873823/Brushless-motor-controller-for-electric-bike-match-for-LED-display-.html) 85x50x35 mm it's the only one.

But Castle Phoenix HV 30 has the perfect size, it would be perfect to build our proto with this ESC.

Anybody can help us on this subject?

Best regards & Happy New Year ;)
Olivier
 
If you're going for the Cute motor why not just buy the matching controller from bmsbattery. It's a small 6fet that looks very similar to the one you linked.

The Cute looks very attractive except that like the Tongxin it's a two piece main shaft. I feel happier with the small Bafangs as they look to be better engineered. Although 1kg heavier or so.
 
If you're going for the Cute motor why not just buy the matching controller from bmsbattery.
Yes it's a good starting. I was wondering about the tiny size (12x25x51mm) and weight 29g of the Phoenix HV it's really really good! For our prototype the controller size is really an issue.

The Cute looks very attractive except that like the Tongxin it's a two piece main shaft.
What it mean?
 
On a hill, the Castle controller will not control the throttle to keep the amp draw in check. Either the motor, battery, or controller will likely get a hard beating and fail.
 
evaleto said:
The Cute looks very attractive except that like the Tongxin it's a two piece main shaft.
What it mean?

Here, there are some photos of the inside.
http://endless-sphere.com/forums/viewtopic.php?f=3&t=13998&hilit=cute&start=60

There's no shaft running right through the motor. Instead there's a shaft welded to a plate that is bolted to one side and the same on the other side. It's probably fine, but just looks like a potential weakness to me.
 
johnrobholmes said:
On a hill, the Castle controller will not control the throttle to keep the amp draw in check. Either the motor, battery, or controller will likely get a hard beating and fail.
You mean Amp (current) is natively managed by the ebike controller. So I must add a current feedback with a resistor sens (0.02Ohm) to the throttle (via my arduino cpu) and it should be ok.
 
If the castle controller was designed for it, yes. But the 30a version has very little thermal mass and would be failure prone when it gets into low duty cycle needed to keep amp draws low.

Maybe you can give some more details of the vehicle and setup that you will be using so we can know what you need.
 
johnrobholmes said:
Maybe you can give some more details of the vehicle and setup that you will be using so we can know what you need.
We are building a minimalistic bicycle (<=14kg) with an accelerometer as throttle. The idea is to keep bicycle without assistance until it's really needed. The battery is small (<25Km autonomy) The motor hub is the smallest for 250W and only 1,6kg.
This is why we are looking for a smaller controller, we would like to hide it in the box with the battery and the electronics.
 
You will really be better off using an ebike specific controller IMO, unless you are willing to oversize the controller quite a bit. While the small Castle controllers are very tempting, they cause other issues in the system that are not easily addressed.

With the 250w hub motor, I would assume that you will want to run a 10a limit at 36v to prevent motor overload. As such I would say that the HV 85 or ICE HV 80 would be safe at the low duty cycles that will be seen going up hills. The downside is that there is no way to control the phase currents. You may be able to get away with a smaller controller, but it will have a shorter life and higher likely hood of damage by high phase currents.


If you do not already have a good connection for the controllers, I am a distributor for Castle products and can help out.
 
Lyen 6fet Infineon would be a better option they are about the size of a pack of cigarettes so still reasonably small form factor...very cheap to.

KiM
 
evaleto said:
johnrobholmes said:
On a hill, the Castle controller will not control the throttle to keep the amp draw in check. Either the motor, battery, or controller will likely get a hard beating and fail.
You mean Amp (current) is natively managed by the ebike controller. So I must add a current feedback with a resistor sens (0.02Ohm) to the throttle (via my arduino cpu) and it should be ok.


No. A controller doesn't feel battery current at all.
If you want to protect the controller, you need to read phase current.
 
Also,those batteries a about as bad of a choice as you could make.
 
Nonsense, at the power level he wants the cells will perform just fine. They do excellent at 2c continuous and he doesn't want more than 3c burst. It will be perfect for minimalism.


If you want to use an RC style controller in combination with your cpu to limit current, you could know the duty cycle of the controller roughly enough by the input signal you are giving it. While it is a dirty and inaccurate method to estimate motor amps, it could be a useful interface for keeping the controller mostly happy and allowing a smaller controller. I would still recommend a 6 to 8 times power factor buffer for the ESC if you want to have a prototype that isn't likely to smoke on a long hill.
 
Lyen 6fet Infineon would be a better option they are about the size of a pack of cigarettes so still reasonably small form factor...very cheap to.

johnrobholmes said:
If you want to use an RC style controller in combination with your cpu to limit current, you could know the duty cycle of the controller roughly enough by the input signal you are giving it. While it is a dirty and inaccurate method to estimate motor amps, it could be a useful interface for keeping the controller mostly happy and allowing a smaller controller. I would still recommend a 6 to 8 times power factor buffer for the ESC if you want to have a prototype that isn't likely to smoke on a long hill.

If you do not already have a good connection for the controllers, I am a distributor for Castle products and can help out. No I haven't yet, thank you !

I will digest all information and go back with good feedback after testing
Thank you all for your support !!!
 
johnrobholmes said:
Nonsense, at the power level he wants the cells will perform just fine. They do excellent at 2c continuous and he doesn't want more than 3c burst. It will be perfect for minimalism.


If you want to use an RC style controller in combination with your cpu to limit current, you could know the duty cycle of the controller roughly enough by the input signal you are giving it. While it is a dirty and inaccurate method to estimate motor amps, it could be a useful interface for keeping the controller mostly happy and allowing a smaller controller. I would still recommend a 6 to 8 times power factor buffer for the ESC if you want to have a prototype that isn't likely to smoke on a long hill.


John- I've got a stack of them on my desk right now, they don't come close to the datasheet specs unless you get them around 50C to start the test.

http://industrial.panasonic.com/www-data/pdf2/ACA4000/ACA4000CE254.pdf

It's a great cell for a laptop. Terrible cell for a small pack-sized EV.

RC LiPo makes vastly more sense for this application from every perspective, cost, packaging, construction, voltage drop, usable energy density/volume at the pack level, etc.
 
On the point of phase current vs battery current, we were doing a motor dyno yesterday where the ratio was over 1:10 battery to phase current. Pack voltage 59v, pack current 11amps. And out of the controller... 5.2v out at 117amps phase current (a hill climbing load simulation). It was one of the highest controller current bucking ratios I had seen in a real-world situation. Normally you seldom go beyond 1:5.
 
Ugh, I was assuming that the datasheets were at room temp. Lipo would indeed be a better performing way to go if they are such a dog of a cell at 2c in the real world.
 
Back
Top