Banding together for the E-bike cause.........

[mud2005]

The throttle will take your 36 or 48 volts in

I'm runnin on 24V is that a problem?

 

[recumpence]

No. I can set your throttle up for 24 volts. That is easy.

Matt

 

[mud2005]

alright then I'll spend $140 to get rid of my wiring mess. I'll take 1, just let me know when it's ready and I'll send the $
anybody wanna buy a BEC and servo tester

 

[northernmike]

My bikes are overpowered. That being the case, heat is generated, but rarely sustained because the huge wattage they produce is delivered in very short bursts. My KMX trike, for example, will be running twin 3220s. I highly doubt heat will be an issue at all. Heck, it will be tough to make the motors warm on that thing unless I am doing hard stop and go runs.

Matt, this is why you're leading the charge, so to speak.

I spoke to a guy on Bloor St. (the main East-West thoroughfare of Toronto) who was piloting a VERY nicely made "scooter type" eBike.

One of the first questions I asked him was "what's the range like?" .... his reply: 100kms.

He'd paid $1300 for a huge Lithium pack. The bike alone cost him $2200.

So, if anyone is questioning if the economics of large, efficient, cool-running motors are viable?

I think this example, plus the PK's sale, plus the electric KMX business model (nice!) show serious promise!

 

[GGoodrum]

You can cut down the amount of heat generated, for a given power level, by running at a higher voltage. Heat losses go up as a square of the current, so if you could double the voltage, the current used would be half and the heat losses drastically reduced. The sensorless RC controllers available at reasonable prices (under $200...) have 50V absolute voltage limits, which means you can't even use a "standard" 48V LiFePo4 setup, which has a nominal voltage of 52.8V and about 58V right off the charger. I do have a Kontronik PowerJazz, which is good for 63V and 200A, that I'll be trying on my Mariner folding bike, but it costs $450.

What I'm trying to do is to get AstroFlight to add sensors to the 3210s and 3220s, so that these can be used with our "normal" ebike controllers, like Methods' modded 18-FET Clyte/Infineon controllers, but this is going painfully slow. Bob Boucher made the first brushless RC motors 10 years ago with hall sensors because back then there were no sensorless controllers. I'm having a hard time getting him too excited about "going backwards", to put the sensors back in. This isn't a technical issue at all. There's room inside the motors to add a magnet collar and a small PCB. Getting this to come to fruition, however, has not been an easy journey.

If it were to happen, a single 3210, running at, say 90V and peaks limited to 100A, would put out similar power levels as Matt's dual 3210 44V PK setup, A single 3220 setup this way, but with a higher current limit, would blow it away.

Another approach I'm looking at is to develop a better implementation of the add-on board Keywin originally supplied that was supposed to allow sensorless operation. Although these might not have worked all that well, the concept is still valid. This is what all the RC controllers do, which is to try moving the motor back and forth slightly on startup, in order to figure out the current position of the magnets. The ultimate solution, of course, is to get this function included in the 18-FET controller software.

-- Gary

 

[Miles]

You can cut down the amount of heat generated, for a given power level, by running at a higher voltage. Heat losses go up as a square of the current, so if you could double the voltage, the current used would be half and the heat losses drastically reduced.

Not for a given motor speed, though...

 

[GGoodrum]

You can cut down the amount of heat generated, for a given power level, by running at a higher voltage. Heat losses go up as a square of the current, so if you could double the voltage, the current used would be half and the heat losses drastically reduced.

Not for a given motor speed, though...

You are assuming the same wind, though. One of the big advantages of the AFs is that they have a huge selection of winds with different number of turns. You just pick one with half the kV. Same rpm, double the voltage, half the current. :wink:

-- Gary

 

[Miles]

You are assuming the same wind, though. One of the big advantages of the AFs is that they have a huge selection of winds with different number of turns. You just pick one with half the kV. Same rpm, double the voltage, half the current. :wink:

Same rpm, double the voltage, half the current, but what about the winding resistance? :wink:

 

[johnrobholmes]

Bob "forgot" that I wanted sensors on my 3220 that he just finshed for me. He didn't seem too happy that I pressed the issue. I won't have pedals, I can't start up without sensors!

I have a few of the sensorless shenzen controllers here, I need to test them and see if they are really worthy of a hard ride.

 

[MitchJi]

Hi Matt,

So, a tip for you to obtain the smoothest low throttle response and performance is to wrap your BEC in foil to reduce the RF and any hesitation you may have otherwise encountered.

I will be utilizing this trick on my twist grip throttles I am building.

Matt

Any plans for thumb throttles, something similar to the PK throttle?

 

[recumpence]

Not right now.

The twist throttles are easy because I am stating with a Magura throttle. The PK throttle was a time consuming part to make (8 hours in that throttle). But, maybe that will be the next thing. :wink:

Matt

 

[CNCAddict]

You are assuming the same wind, though. One of the big advantages of the AFs is that they have a huge selection of winds with different number of turns. You just pick one with half the kV. Same rpm, double the voltage, half the current. :wink:

Same rpm, double the voltage, half the current, but what about the winding resistance? :wink:

Yep, half the Kv and motor resistance is 4x higher. So resistive losses are identical. There is no free lunch with motors. High voltage does reduce losses in the wiring setup and interconnects, but makes very little difference in the motor.

Bob "forgot" that I wanted sensors on my 3220 that he just finshed for me. He didn't seem too happy that I pressed the issue. I won't have pedals, I can't start up without sensors!

Do you have a freewheel on your bike?? It should only need a tiny push to get going even if you don't.

 

[GGoodrum]

High voltage does reduce losses in the wiring setup and interconnects, but makes very little difference in the motor.

That is just not accurate. heat losses go up as a square of the current, so for a high-performance setup, where you are pushing against the limits of what a motor can deliver, running low voltage and high current wastes a lot of power to heat. that is going to keep the motor from reaching the same power level that you could with higher voltage and less current.

-- Gary

 

[Miles]

It is right Gary,

Voltage goes up, current goes down, winding resistance goes up - it all balances. Same amount of heat generated in the motor.

 

[johnrobholmes]

Right. Two identical motors spinning the same RPM making the same power will have roughly the same heat generated even if they have different KVs and working voltages. Copper fill differences between winds may give a bit of change, but on the whole all winds have similar efficiency.


Sensorless would be quite the bear to start up with a passenger on a steep incline! If nothing else I would like to avoid nasty peak startup currents, so sensored is the way to go for my motorcycle project.

 

[Tiberius]

It is right Gary,

Voltage goes up, current goes down, winding resistance goes up - it all balances. Same amount of heat generated in the motor.

Yes,

For the same motor, speed and power, it just comes down to the volume of copper.

Generally for a given motor, you can make it more efficient by running it faster and gearing it down more. To do that, you need a higher voltage but will end up with lower current. That may be where the idea of higher voltage/lower current comes from. But its the gearing that needs changing, not the Kv (windings) of the motor.

Nick

 

[Tiberius]

Matt,

This is all good news. Are there specs on the Astro Flight motors that you can post or link to?
I have a fiendish project in mind......

Nick

 

[Miles]

Here you go, Nick.

http://www.astroflight.com/pdfs/3210WEB.pdf

http://www.astroflight.com/pdfs/3220.pdf

N.B. there are lots of "typos"......

 

[liveforphysics]

It is right Gary,

Voltage goes up, current goes down, winding resistance goes up - it all balances. Same amount of heat generated in the motor.

This is if you neglect field rise time.

I know on paper it's just a simple matter of multiplying total power in by rated efficiency, and it gives you the same amount no matter what combination of amps and volts you use to get a given power value. Due to the different rate of field rise time, I think it makes more of a difference in efficiency while a motor is in rapid acceleration. For steady state running, the controller would be able to get timing correct, and I don't see it making as much of a difference.

Statically, it's a simple 1 step math problem. Dynamically, it's pretty damn complex to see the differences between high/low voltage setups. It's going to be very close either way for inside the motor. All the losses in the rest of the system definitely favor the high voltage system.

 

[mclovin]

It is right Gary,

Voltage goes up, current goes down, winding resistance goes up - it all balances. Same amount of heat generated in the motor.

This is if you neglect field rise time.

I know on paper it's just a simple matter of multiplying total power in by rated efficiency, and it gives you the same amount no matter what combination of amps and volts you use to get a given power value. Due to the different rate of field rise time, I think it makes more of a difference in efficiency while a motor is in rapid acceleration. For steady state running, the controller would be able to get timing correct, and I don't see it making as much of a difference.

Statically, it's a simple 1 step math problem. Dynamically, it's pretty damn complex to see the differences between high/low voltage setups. It's going to be very close either way for inside the motor. All the losses in the rest of the system definitely favor the high voltage system.

On that note.......Hopefully this isn't too far off topic but.......do you know how to characterize the transient response for an electric motor when the load changes? If that is too complex to describe here then is there some second order (or whatever) approximation for it? That could be very helpful to know when trying to develop a 2 speed shifter design.

 

[Tiberius]

Here you go, Nick.

http://www.astroflight.com/pdfs/3210WEB.pdf

http://www.astroflight.com/pdfs/3220.pdf

N.B. there are lots of "typos"......

Mmm, nice. Thanks Miles.
Its so good to see some data for a motor, rather than just X Watts at Y Volts.

Now all I need is someone to tell me they are waterproof.....

Nick