Speed vs Efficency - something strange going on?

[mingonn]

Hi all,

Trying to work out what is happening to my setup, up until recently I have been running with the speed limited to 22 km/h via a jumper on my controller but am now running unlimited at 30 km/h.

Strange thing is that I do not seem to be going through significantly more battery capacity, the table below shows what has been going back into my lipo pack, I took it pretty easy when I first switched to unlimited thinking I would be chewing through my battery but the last tow trips I have ridden normally...at 30 km/h that is :lol:

Date Dist mah avg ma/km
15/07/2008 12 3800 317
16/07/2008 12 4500 375
16/07/2008 24 6900 288
21/07/2008 12 3224 269 (Removed speed limiter)
21/07/2008 12 4400 367
22/07/2008 12 4100 342

Is my motor efficiency better at higher RPMs and offsetting the additional drag or acceleration current draw or is it that my controller less efficient lower throttle settings? Either way I am loving the no cost speed gain!!

Cheers
Derek

 

[Ypedal]

between 20 and 30 km/h there is not alot of difference in wind resistance.. but after 30 km/h it really starts to add up.

What motor is this ?

 

[Drunkskunk]

Controllers are less efficent at part throttle than at full, and the jumper is limiting it to part throttle. there is also little wind resistance below 30kph, so it isn't much harder to go 30kph. I imagine the controller's inefficancy is slose to the wind resistance loss in your case.

 

[TylerDurden]

I concur.

 

[johnb]

My understanding is that controller loss is small compared to motor power and is mainly down to the on resistance of the FET's. That wouldn't be affected by the throttle setting.

 

[ZapPat]

My understanding is that controller loss is small compared to motor power and is mainly down to the on resistance of the FET's. That wouldn't be affected by the throttle setting.

AT 100% throttle, a well made hall-based controller will have it's mosfets ON all the time during each motor phase commutation, resulting in dramaticly less time spent in the analog domain by the output fets (100's of Hz VS 20k or so hertz). This effect is the cause of much of the controller's losses, depending on how the controller is designed. You still have the resistive losses of the FETs of course, since this is a static characteristic (DC).

FET selection, switching frequency, and ON/OFF transition time of the output are what determine how efficient your controller is to a great extent.

 

[Link]

Yes to all of the above. Switching the FETs eats up a little bit of power, and 30kph isn't a particularly significant speed. It's possible that what you lose to wind resistance you gain in having the FETs on all the time. The slower the FET, the bigger difference it makes.

 

[mingonn]

Ran full throttle for the last 2 km this morning, 27-29km/h. Controller was barely warm and batteries were slightly warm, the controller and batteries were a lot warmer after the same run with the speed limiter on.

Should the controller get warmer at lower throttle settings if you keep everything else constant eg wind and rolling resistance?

Will put another post up with details and pictures of my ride later today.

Cheers
Derek

 

[ZapPat]

Ran full throttle for the last 2 km this morning, 27-29km/h. Controller was barely warm and batteries were slightly warm, the controller and batteries were a lot warmer after the same run with the speed limiter on.

Should the controller get warmer at lower throttle settings if you keep everything else constant eg wind and rolling resistance?

Will put another post up with details and pictures of my ride later today.

Cheers
Derek

Warmer anything means loss of efficiency, so this is normal, and also confirms that at full throttle a controller is more efficient. Note that this is not necessarily only a little bit of the controller's losses we are talking about, but with a well designed PWM circuit these switching losses would account for close to half the controller's losses, the other half being the resistive FET losses. As link points out, a bigger FET that has low resistive losses will also have higher switching losses because of longer switch times. The trick is to choose the FET with the right balance for the job...

It doesn't explain about your batteries being cooler at full throttle though? Maybe they don't like being pulsed either?

 

[John in CR]

I don't know if it bears any relationship to the topic, but my only controller failure so far occurred with my daughter riding my version 1 bike in the park at maybe 5mph. She kept pulsing the throttle at very low levels, and it overheated and popped after only 10min of that torture.

 

[TylerDurden]

This is my primary reason for wanting to build a dyno.

The7's rough test using just a brake on the wheel suggested significant losses during partial throttle... not just in the controller, but in the motor also; as rise and fall time of switching incurs more low voltage I2R losses.

 

[mingonn]

My take home message from all this is get to full speed ASAP and ride full throttle, pedaling to stay at full speed up the hills. Doubt this model would work above 30 km/h where wind resistance would start to undo any gains.

Derek

 

[johnb]

My take home message from all this is get to full speed ASAP and ride full throttle, pedaling to stay at full speed up the hills. Doubt this model would work above 30 km/h where wind resistance would start to undo any gains.

Derek

I don't think we have got to the bottom of this yet. If I want to go further I use less throttle, works every time, and others report the same.

I think that if the loses were down to the controller then it would have caught fire by now. That's just a guess, perhaps someone would do the maths.

You mentioned hills and this could be more to do with it. If you are slogging up hills at low throttle then the motor will be running less efficiently. As a rule of thumb, you want to keep your speed at least (say) 75% of the unloaded speed at any particular throttle setting for reasonable efficiency.

 

[mingonn]

Hi John,

I generally compensate for hills by pedaling to keep at maximum speed (speed on the flat). I figure the motor should be at peak efficiency this way and pulling the same current as on the flat, with my legs making up the difference. I am not sure if this would make a difference though, I don't feel I am putting in any more energy now.

I still not sure how I can ride 9 km/h faster than I have been and still only use up the same battery capacity. This mornings charge was 3950 mah no still no big change.

Loving the speed though.

Derek

 

[TylerDurden]

I don't think we have got to the bottom of this yet. If I want to go further I use less throttle, works every time, and others report the same.

I think that if the loses were down to the controller then it would have caught fire by now. That's just a guess, perhaps someone would do the maths.

You mentioned hills and this could be more to do with it. If you are slogging up hills at low throttle then the motor will be running less efficiently. As a rule of thumb, you want to keep your speed at least (say) 75% of the unloaded speed at any particular throttle setting for reasonable efficiency.

From Doc's thread:
http://www.endless-sphere.com/forums/viewtopic.php?f=4&t=4983&p=74439&#p74439

I would concur about the high heating at low speed when climbing.

From 0 to 27 km/h, the battery current is limited by the current limit at 60A.

The average was 23 km/h as said by DoctorBass.
At 23 km/h, the battery current was limited to 60A. There would be no difference if the throttle was 100% or slightly lower (say 95%) when the max possible torque was required for a certain slope.

Using 95% throttle, the curves are exactly the same as that of 100% for speed below 23 km/h.
Using 80% throttle, the curves are exactly the same as that of 100% for speed below 13 km/h.

 

[ZapPat]

My take home message from all this is get to full speed ASAP and ride full throttle, pedaling to stay at full speed up the hills. Doubt this model would work above 30 km/h where wind resistance would start to undo any gains.

Derek

I don't think we have got to the bottom of this yet. If I want to go further I use less throttle, works every time, and others report the same.

I think that if the loses were down to the controller then it would have caught fire by now. That's just a guess, perhaps someone would do the maths.

You mentioned hills and this could be more to do with it. If you are slogging up hills at low throttle then the motor will be running less efficiently. As a rule of thumb, you want to keep your speed at least (say) 75% of the unloaded speed at any particular throttle setting for reasonable efficiency.

Probably a certain amount of both effects going on here: higher controller efficiency due to much lowered switching losses at 100% duty output, and also that the motor is probably now operating closer to it's sweet point. Both effects together are probably compensating for your slightly higher wind drag.

 

[johnb]

Hi John,

I generally compensate for hills by pedaling to keep at maximum speed (speed on the flat). I figure the motor should be at peak efficiency this way and pulling the same current as on the flat, with my legs making up the difference. I am not sure if this would make a difference though, I don't feel I am putting in any more energy now.

I still not sure how I can ride 9 km/h faster than I have been and still only use up the same battery capacity. This mornings charge was 3950 mah no still no big change.

Loving the speed though.

Derek

I think you need to tell us what motor, controller and battery you are using. Gross weight would be useful too as well as average speed.

 

[mingonn]

Hi John,

This is a pic of my ride. All up weight including rider is roughly 90 kg or 200 lbs. 70kg (me) + 18kg (bike and motor) + 1kg (44.4v 3800mah RC lipo)

Docs say the motor is 200w 36v for a 26" wheel but mine is in a 20" wheel. Motor housing is shiny aluminum.

Got it off ebay here http://cgi.ebay.com.au/200watt-moto...ryZ98083QQssPageNameZWDVWQQrdZ1QQcmdZViewItem

Going to get a bafang and convert my Dahon 20" folding bike and let my wife use this one to commute. That should knock 3.5 kg off my bike weight and give me lots of juicy torque should come in at 15.5 kg bike, motor and batteries!!!

Cheers
Derek

 

[johnb]

I don't know the motor. I'm wondering if that motor is running inefficiently when restricted.

Could you tell me the average speeds, restricted and unrestricted (I assume these are 22km/h and 30 km/h) and the unloaded speeds (wheel off the ground). We should be able to get an idea of efficiency from that.

 

[ZapPat]

I don't know the motor. I'm wondering if that motor is running inefficiently when restricted.

Could you tell me the average speeds, restricted and unrestricted (I assume these are 22km/h and 30 km/h) and the unloaded speeds (wheel off the ground). We should be able to get an idea of efficiency from that.

How would you go about calculating efficiency of the motor from just this? Without proper current and voltage measurements at the same time, I can't see how...

 

[johnb]

How would you go about calculating efficiency of the motor from just this? Without proper current and voltage measurements at the same time, I can't see how...

For much of its range, you can work out a motor efficiency from back emf divided by voltage applied.

 

[ZapPat]

How would you go about calculating efficiency of the motor from just this? Without proper current and voltage measurements at the same time, I can't see how...

For much of its range, you can work out a motor efficiency from back emf divided by voltage applied.

Please give more details with equations used if possible too! I do not know of a method to mesure efficiency with voltage and back EMF, it would be good to know.

 

[mingonn]

Hi John,

The unloaded(wheel off the ground) speed is 34 km/h. Am starting to take the extra speed and power for granted and my daily power consumption is creeping up slightly as I am relying more on the motor to accelerate and climb.

Derek

 

[johnb]

Hi John,

The unloaded(wheel off the ground) speed is 34 km/h. Am starting to take the extra speed and power for granted and my daily power consumption is creeping up slightly as I am relying more on the motor to accelerate and climb.

Derek

In that case you will be getting very good efficiency at 30 km/h, probably around 80%.