Go Back  RCU Forums > RC Cars, Buggies, Trucks, Tanks and more > RC Car General Discussions
Reload this Page >

Another brushless 'myth', lower KV = more torque

Notices
RC Car General Discussions This forum is for all general discussions related to radio control cars. Check forums below for more specific categories if applicable.

Another brushless 'myth', lower KV = more torque

Old 02-17-2009, 12:07 AM
  #76  
mattnin
Senior Member
 
mattnin's Avatar
 
Join Date: Jan 2006
Location: grand junction, CO
Posts: 834
Likes: 0
Received 0 Likes on 0 Posts
Default RE: Another brushless 'myth', lower KV = more torque

Well I want to thank Access and this thread for inspiring me to get a motor dyno. There is a saying, "there is a fine line between hobby and insanity" and I believe I crossed the insanity threshold a long time ago :P Actually, I got a really great deal on an old brushed motor dyno that I will have to convert over to work with brushless, and I am very confident I can do so.

So here soon enough and since this will be a modified dyno, I will be able to do testing on 540 size can motors up to 40,000 rpms and 5S Lipo voltage limit. If I ever come across any 3S-5S systems, I will test them out. However, I am really interested in testing peak power and torque values for the stock brushless motors, namely the Novak 13.5 vs. the Hacker 13.5 vs. Orion et. al and using a Tekin RS Pro speed control that has adjustable timing.

This will be complementing my already vast collection of professional RC racing gear, although I suck at racing haha, but getting better! I am still 1st in points series in offroad stock truck this season by pure luck though It seems that weekly I get accused of having a mod motor in stock because there is no one hands down faster than me on the straights, but I am doing a lot of little tricks to my truck such as giving it perfect 50/50 side to side weight balance and perfect crossweight balance and tire balance. And having a dyno so I can gear to peak power is gonna make me that much faster I hope Yea they all used to laugh at me when I was there trying to tweak the perfect cross-weight (I am the only one that does it) and balance my tires. Supposedly it was said they go out of balance every race but they stay surprisingly in balance week to week.

Here is my list of insanity gear so you racing freaks can drool over :P
1 text book "Race Car Vehicle Dynamics" by M&M (this book is used to teach F1 Engineers at Purdue)
1 custom made tweak board with 4 scales

1 TI-89 graphing calculator with a custom program that calculates crossweight percentage, crossweight difference, Left to Right CG position, Left to Right CG difference, Front to Back CG percentage, Total Weight
1 tire balancer
Lots of gear oil and time making everything smooth
Pretty much copying Cavalieri's LIPO T4 setup (I really did modify some settings, I swear!)
1 Novak Sentry Data Logger. This is a particular valuable tool. I am currently using it to make a G-G diagram, and I was just taught how to make a "Total G vs Time" diagram which will be what I start doing based on this weeks races

This graph shows the actual data I logged during the last point series race where I won by half a lap, although it was a close race up until the last 45 seconds and then he wrecked twice right in a row and I think he got caught up a little in some lap traffic, damn close race though and everybody hootin and hollerin made me real nervous on the stands

1 custom stick radio, notice the throttle and steering sticks are swapped


And soon, 1 motor dyno :-) Stay tuned to my posts for the latest!
Old 02-17-2009, 12:47 AM
  #77  
mattnin
Senior Member
 
mattnin's Avatar
 
Join Date: Jan 2006
Location: grand junction, CO
Posts: 834
Likes: 0
Received 0 Likes on 0 Posts
Default RE: Another brushless 'myth', lower KV = more torque


ORIGINAL: Access

I wonder if there is a similar tool one could download for free, for running simulations like that.
(1) "Motor 1: Simulation: 7.5 Volt"
(2) "Motor 2: Simulation: 7.5 Volt"
If he were to increase the voltage on motor (1) to the point where both motors maxed out at the same RPM, then what would the torque curves end up looking like? If the reality is the motor can't provide more than 90 units of torque, it would look like the picture I uploaded enough (except the dotted lines/theoretical should be crossing at some point).
I bet something like this can be done testing motors of different KVs and 2S vs 3S Lipo, and using virtual timing or physical timing to get the KV required for the test. My bet would be that they would have the same torque, but the higher voltage setup will have more power and the torque curve would have less slope and be more efficient too, but that is just a guess.
Old 02-17-2009, 01:47 AM
  #78  
srt10
My Feedback: (20)
 
Join Date: Jan 2008
Location: Frederick, CO
Posts: 5,971
Likes: 0
Received 0 Likes on 0 Posts
Default RE: Another brushless 'myth', lower KV = more torque


ORIGINAL: dacaur

Access is right in only one thing, which is that any brushless power system can have x amount of torque, the thing he seems to be unwilling to admit is that the battery power required to get there is often unacceptable, even if the motor/ESC combo can handle it, when compared to higher volt/lower KV systems, or even simply lower kv systems.

You can get 300hp out of a 4 cyl engine (low volt/high KV), or you can get 300hp out of an 8 cyl engine(high volt/low KV). but the 8 cyl will run cooler, be more reliable, and last longer.
Access, you said that motor/esc heat was the # 1 problem when trying speed runs. Well, that heat comes from the # of amps running through the motor, not the voltage, so if you up the voltage, you lower the amp draw, to up the voltage, there comes a point you need to lower the KV, or the motor wont have the necessary torque without drawing huge amps.

What it comes down to is this. In any given setup, a lower KV motor will produce more power/torque at the wheels with less battery power than a higher KV motor. The reason it is doesn't have to work as hard trying to get to high RPM's, so runs cooler and pulls less amps, so gets less voltage sag.

What I have been saying, and what everyone else is trying to tell you access, is that lower KV motors are chosen because they can produce more torque with a given amp draw, and give you a realistic RPM to work with. Put a 10,000KV motor in ANY rc vehicle, and you will either 1) need $500 worth of batteries to run it, since it will pull so many amps, (but the motor is still unlikey to ever reach full rpm), 2) melt your batteries, or 3) gear it way down to prevent the ESC going thermal, motor meltdown, and/or battery meltdown etc...

What you (access) are saying is like telling someone the above mentioned 300hp 4 cyl is just as good as the 300hp 8 cyl. Its not. While the motors might produce the same power, there are a lot of trade offs to get there. In the case of a brushless motor, the trade off is amp draw.

Your power system doesn't heat up because of voltage, it heats up because of amperage. You could run 600 volts @ 1 amp through the wires on your ESC and they would be ice cold. but try running 200 amps at ANY voltage, and they will heat up FAST.

Look at ANY 1/8 scale electric conversion, and you will NEVER see one with a high KV motor. Why? because they run on high voltage systems. you will never seen one that runs on 7.4 volts because it would have a stupid high amp draw, and run very very hot. higher voltage/lower kv is the future.

Airplane guys figured this out a looooong time ago. You can get more thrust with less amps using a lower KV motor. Yea you could use a higher kv motor and get the same thrust, at a much higher amp draw, but why would you?

I will say this. If you have a magical battery that can supply unlimited amps, then sure a high KV motor will put out just as much torque as a lower KV motor. But until that day comes, lower KV will give you more torque on any given setup, (unless you you gear it stupid low)

Being that I have a Brushless E-Revo and tried different KV motors, I have to say I agree with this above...

Give me the Big Block aka lower KV.

mattnin - Like your Losi, looks fast!

Old 02-17-2009, 06:18 AM
  #79  
Access
Senior Member
Thread Starter
 
Join Date: Mar 2005
Location: , CA
Posts: 4,900
Likes: 0
Received 0 Likes on 0 Posts
Default RE: Another brushless 'myth', lower KV = more torque

"So here soon enough and since this will be a modified dyno, I will be able to do testing on 540 size can motors up to 40,000 rpms and 5S Lipo voltage limit. If I ever come across any 3S-5S systems, I will test them out. However, I am really interested in testing peak power and torque values for the stock brushless motors, namely the Novak 13.5 vs. the Hacker 13.5 vs. Orion et. al and using a Tekin RS Pro speed control that has adjustable timing."
If you have different batteries sitting around, like 6-cell NIMHs, 7-cell NIMHs, and LiPos, try testing the same motor with this array of different batteries (or batteries of different capabilities).
Old 02-17-2009, 08:04 AM
  #80  
Argess
 
Argess's Avatar
 
Join Date: Jan 2006
Location: Pleasantville, NS, CANADA
Posts: 2,355
Likes: 0
Received 1 Like on 1 Post
Default RE: Another brushless 'myth', lower KV = more torque

I'm not sure if this matters for the discussion here, but I think that graph may be for permanent magnet brushed motors. At least that's what the shape of the torque curve seems to tell me. I can't seem to find one that compares different turn brushless motors. Anyone?

I always wanted to build a dyno for my nitro engines. An inertial one, not a brake Hp unit. It could be used for electric motors as well. The formula comparing mechanical power to electrical power is correct. V x I = RPM x T x Constant is what's used with full size brake hp dyno's. I admit, it is a little simplistic for me to apply it to brushless motors as it doesn't take into consideration changes in motor impedance over rpm, efficiency, etc., so it's only true under certain situations unless we further descibe the current/impedance with more mathematical rigour.
Old 02-17-2009, 08:16 AM
  #81  
mattnin
Senior Member
 
mattnin's Avatar
 
Join Date: Jan 2006
Location: grand junction, CO
Posts: 834
Likes: 0
Received 0 Likes on 0 Posts
Default RE: Another brushless 'myth', lower KV = more torque

Well I just realized as I woke up this morning that any flywheel type brushed motor dyno ran using a brushless speed controller in conjunction with a Novak Sentry will gather all the necessary data that any of the best brushed motor dynos collect. I resigned to the fact last night that I wouldn't get efficiency data because the brushed motor dyno wouldn't be able to read voltage and amperage, but the Novak Sentry can do that up to 100 amps, so we are back in business!

Access, thanks for the idea, and please keep them coming. I'm sure that other test you spoke of is possible to actually do as well. I am sure that once word gets around at the track that I will be doing this, I will end up testing many different configurations, and one particular configuration that I think many people will be interested in is dyno'ing the stock motors using the RS Pro at various virtual timing levels. I'd also like to see the difference from one stock motor to the next from the same company.
Old 02-17-2009, 08:21 AM
  #82  
mattnin
Senior Member
 
mattnin's Avatar
 
Join Date: Jan 2006
Location: grand junction, CO
Posts: 834
Likes: 0
Received 0 Likes on 0 Posts
Default RE: Another brushless 'myth', lower KV = more torque

Now that I think about it even more, a Novak Sentry and a flywheel is enough to gather all the data necessary. Although the Sentry only reads 10 times a second, however that might be enough.
Old 02-19-2009, 10:57 PM
  #83  
mattnin
Senior Member
 
mattnin's Avatar
 
Join Date: Jan 2006
Location: grand junction, CO
Posts: 834
Likes: 0
Received 0 Likes on 0 Posts
Default RE: Another brushless 'myth', lower KV = more torque


ORIGINAL: Argess

I'm not sure if this matters for the discussion here, but I think that graph may be for permanent magnet brushed motors. At least that's what the shape of the torque curve seems to tell me. I can't seem to find one that compares different turn brushless motors. Anyone?
I found some dyno tests of some brushless motors courtesy of John Stranahan.

These are graphs of a Novak EBX 13.5 running through a Fantom power supply, so only 5V.





Novak 3.5 Large rotor vs small rotor, green large rotor, blue small rotor



I thought I would do one more test while this 3.5 was hooked up. This time is used a 4 cell pack. This pack can deliver the huge amps this motor needs to perform at its best. It is not like a 6 cell and at this point I don't recommend running the dyno with a 3.5 and a six cell pack. It is starting to sound like a small jet engine already due to the 48,000 RPM reached. Note that without using the Fantoms power supply you can get no efficiency numbers. The dyno cannot measure input power.

In the red corner the small rotor peaked about 140 W. I got very smooth data an did not need a trendline. In the green corner the large rotor 123 Watts. This large difference did not require replicates on this smooth data. you get more power with the little rotor. This is a repeat of the conclusion with the 13.5.

Pic: Note the white scatter shield has a beveled cut so it's supported by the black knob as well as a strip of tape. Aim the exposed part of the flywheel into a safe area. Note that pile of papers; that is two years receipts for RC stuff. Damn. Don't tell the wife.
John




My test setup will be nearly identical to Johns, except that I will be using a Novak Sentry to get efficiency by calculating wattage from the battery vs, wattage realized at the flywheel.
Old 03-03-2009, 04:47 PM
  #84  
mattnin
Senior Member
 
mattnin's Avatar
 
Join Date: Jan 2006
Location: grand junction, CO
Posts: 834
Likes: 0
Received 0 Likes on 0 Posts
Default RE: Another brushless 'myth', lower KV = more torque

Seeing as we don't yet have a brushless dyno, I figured I'd share how I made a dyno out of a Novak Sentry datalogger, a Fantom Aluminum Flywheel, a Tekin RS Pro, spektrum 3001 receiver and Sanwa Exzes radio.

The following equations are what I am using. If there are any errors, please do not hesitate to respond.

For all math, RPMs must be converted into Rad/s.
ex. 3713 RPM, W = 3713RPM (2pi Rad/1 revolution)(1min/60seconds) = 388.82 rad/s
Torque = I (moment of inertia) x a (angular acceleration in rad/s)
Moment of inertia I = 1/2*MR^2
power (kW) = (torque (N*M) *2pi*rpm)/60000
angular acceleration a = (W-Wo)/t, W must be in rad/s

The moment of inertia for the fantom flywheel is,
47.59E-6 calculated from the weight, 79g and diameter 69.42mm

The test below is only a proof of concept. The tests are not ideal because I didn't top off the battery first, and I didn't let the motor run long enough so I didn't see a max RPM. Plus, I need to get my girlfriend to make an excel spread sheet for me because I did all the math by hand. Also, I do not yet have the current sensor for the Sentry and that should be in by the end of the week.

The following is a Hacker 13.5 dynoed @ 5 degrees virtual timing boost on the RS Pro, and a 5200 mah LIPO battery.



Old 03-03-2009, 05:50 PM
  #85  
Access
Senior Member
Thread Starter
 
Join Date: Mar 2005
Location: , CA
Posts: 4,900
Likes: 0
Received 0 Likes on 0 Posts
Default RE: Another brushless 'myth', lower KV = more torque

Once you're able to measure current, then you can do efficiency calculations also.

How much inherent error do you think there is in the process? Once you get the process down, run the same test twice or even several times and see what the standard deviation is. It looks like, especially in the 21000+ RPM range, there could be some significant error. But the general shape of the curves is a good sign. Need to get down into the lower RPMs, though.
Old 03-03-2009, 06:35 PM
  #86  
mattnin
Senior Member
 
mattnin's Avatar
 
Join Date: Jan 2006
Location: grand junction, CO
Posts: 834
Likes: 0
Received 0 Likes on 0 Posts
Default RE: Another brushless 'myth', lower KV = more torque

I think there is some error, we'll see how my other tests go. I am worried that the Sentry's resolution isn't high enough. For instance, check out the voltage readings. It shows 7.742V and the next step it registers is 7.837. So I imagine it only has 256 bits of resolution or less. I don't know. So I assume it would have that same lower resolution for all its readings, RPM, Current, etc....

Plus, the Hacker 13.5t motor revved up so quick, so it would be very hard to show lower RPMs. Literally after the first .1s, the flywheel was already at over 7000RPM! It appears to almost nearly have reached its max RPM in nearly 2 seconds. So this means only 20 inputs will be of value. The Sentry only takes a reading every .1s, but so does the Fantom Dyno. I might get better information if I lowered the voltage actually. The Fantom Dyno has a 5V output for testing, and so I may run the Tekin RS Pro off of that, and a 6V receiver pack.

Or.... I may just figure out how to collect the data myself using a PLC or some other discrete device hooked up to the sensor harness and using the motors hall sensors to record RPMs, and take voltage and current readings. I would like to take a reading every 0.01s and have a resolution of 3 decimal places for current and voltage and be able to take a more precise RPM reading as well.

I really do feel my math is correct. I have spent several days with my physics book, sample 19t brushed dyno data from the fantom, measuring the flywheel and so on and so forth, and I feel I have a correct methodology. I received a PM on rctech that I should consider the moment of inertia for the rotor, and I agree that is important to add as well. So soon enough, I will include that. That also means that the power ratings I show below will improve. I also plan on calculating the 'friction factor' and adding that to further increase the power. I can determine that by spinning up the motor to max rpm, and having drag brake turned off, and taking readings while the motor decelerates. That deceleration that I measure will become the friction factor, and will be added to the total power of the motor.

I may have discovered something from all this however. The fantom dyno has been rumored to show incorrect power ratings for its motors, and showing poor efficiency data. When I tried calculating everything, my data was close, but well above the fantoms. So I figured maybe fantom uses a different moment of inertia, so I tried solving for that using the RPM and torque they gave and I got a number. Then I tried using that number to calculate torque for different RPMs, and it never added up. So I have no clue how they calculated torque. There could be an argument made that maybe fantom has a deceleration figure for friction and air resistance, but that would only increase the value of power I am getting. Below is an example of fantoms data, and the data I calculated:

fantom #'s
time .1s rpm 3713 torque 1107 gm*cm watts 42.24 amps 73.1 effcy 11.6%
time .2s rpm 6113 torque 998 gm*cm watts 62.67 amps 64.2 effcy 19.5%

my #'s
time .1s rpm 3713 torque 1886.162gm*cm watts 71.95 amps 73.1 effcy 19.59%
time .2s rpm 6113 torque 1219.649gm*cm watts 76.57 amps 64.2 effcy 23.85%

I have a theory that maybe fantom didn't correctly convert torque from N*M to gm*cm the proper way, or something, because I just cannot come up with their torque numbers for some reason.
Old 03-03-2009, 06:50 PM
  #87  
Access
Senior Member
Thread Starter
 
Join Date: Mar 2005
Location: , CA
Posts: 4,900
Likes: 0
Received 0 Likes on 0 Posts
Default RE: Another brushless 'myth', lower KV = more torque

ORIGINAL: mattnin
Plus, the Hacker 13.5t motor revved up so quick, so it would be very hard to show lower RPMs. Literally after the first .1s, the flywheel was already at over 7000RPM! It appears to almost nearly have reached its max RPM in nearly 2 seconds. So this means only 20 inputs will be of value. The Sentry only takes a
Heavier or larger flywheel perhaps? One with more rotational inertia, that is.
Old 03-03-2009, 07:06 PM
  #88  
mattnin
Senior Member
 
mattnin's Avatar
 
Join Date: Jan 2006
Location: grand junction, CO
Posts: 834
Likes: 0
Received 0 Likes on 0 Posts
Default RE: Another brushless 'myth', lower KV = more torque

Yes, I need to find a larger or heavier flywheel. I just don't know where I can get one. I suppose I can get the metal shop here to make one on their lathe. Maybe I'll try that.
Old 03-04-2009, 10:10 AM
  #89  
mattnin
Senior Member
 
mattnin's Avatar
 
Join Date: Jan 2006
Location: grand junction, CO
Posts: 834
Likes: 0
Received 0 Likes on 0 Posts
Default RE: Another brushless 'myth', lower KV = more torque

I just ran another test with the Sentry dyno with a fully charged lipo battery and I believe ideal test conditions except no current reading yet, and I finally figured out excel too! :P actually it is open office, but here it is.

This test is with 0 boost on the controller.



Old 03-04-2009, 03:45 PM
  #90  
mattnin
Senior Member
 
mattnin's Avatar
 
Join Date: Jan 2006
Location: grand junction, CO
Posts: 834
Likes: 0
Received 0 Likes on 0 Posts
Default RE: Another brushless 'myth', lower KV = more torque

I am getting a heavier flywheel made at the local machine shop so I should start getting better data real soon.
Old 03-04-2009, 04:06 PM
  #91  
mattnin
Senior Member
 
mattnin's Avatar
 
Join Date: Jan 2006
Location: grand junction, CO
Posts: 834
Likes: 0
Received 0 Likes on 0 Posts
Default RE: Another brushless 'myth', lower KV = more torque

If anyone wants to do this, I will show exactly the equations and how to do it (while my battery is topping off again).

First, convert RPMs to rad/s as I shown in the first post as an example.
2nd, calculate the moment of inertia (I) of your flywheel using
.5MR^2
3rd, find the angular displacement.
0 (theta) = .5(Wo +W)T, Wo is the initial angular velocity in rad/s, and W is the current angular velocity in rad/s, and T is time
4th, find the angular acceleration.
a (alpha) rad/(s^2) = (W^2 - Wo^2)/(2*(0 theta))
5th, find torque
t (N*m) =I*a
6th, find watts
power (watts) = (torque (N*m) x 2pi x RPM)/60
7th, convert torque from N*m to gm*cm if you want to
Multiply t * 10197.162

I won't go into all the math, but you can make a full equation by entering in the equation for theta into the angular acceleration formula, and then and then entering in the angular acceleration formula into the torque formula, and then simplifying.

The simplified equation for torque if the time between readings is .1s is,

t = ((RPM - RPMo)pi)/3 * I, where RPM is the current RPM reading and RPMo is the previous reading. The equation turns the data into rad/s so no worry about converting. t comes out in N*m as well.

So as an example, lets calculate some data from the previous chart @ 3.8 seconds. The moment of inertia of my flywheel is 47.59E-6 kg*m^2 or 0.00004759 kg*m^2.

t = ((11406-6282)pi)/3 * 0.00004759 = 0.2553603 N*m
convert that to gm*cm
0.2553603*10197.162 = 2603.95 gm*cm
find watts
power = 0.2558969*2pi*11406/60 = 305.01W
Old 03-04-2009, 04:54 PM
  #92  
Access
Senior Member
Thread Starter
 
Join Date: Mar 2005
Location: , CA
Posts: 4,900
Likes: 0
Received 0 Likes on 0 Posts
Default RE: Another brushless 'myth', lower KV = more torque

http://hyperphysics.phy-astr.gsu.edu/Hbase/rke.html#rke

This is the link I'm working off of.
"It follows that the rotational kinetic energy given to the flywheel is equal to the work done by the torque."

You can look at it form a mechanics perspective, or a work/energy perspective, either should give the same results.
Some energy is going to be lost in the bearings (over time), hopefully not enough to throw things off.

Once you can get current data you can do a lot more, efficiency and a complete work/energy analysis.

BTW do you have a hacker motor other than the 13.5 you can test? Same battery is fine, just want to see how the torque compares.
Old 03-04-2009, 07:01 PM
  #93  
mattnin
Senior Member
 
mattnin's Avatar
 
Join Date: Jan 2006
Location: grand junction, CO
Posts: 834
Likes: 0
Received 0 Likes on 0 Posts
Default RE: Another brushless 'myth', lower KV = more torque

Yea, I have a Novak 13.5 but I loaned it to James. I will be getting it back on Saturday at the races. I will be doing tests on everything soon with the current tester as well. However, there may be an error with the mathematics brought to my attention by Stranahan. I am working hard on the math, but I am open to suggestions to make sure I get this done right. I am looking for an accurate way to calculate the angular acceleration when the acceleration is not-constant. John has shown me how to get it by fitting a cubic polynomial to the angular velocities taken from the table. But we need to find another way that is more automatic so we don't have to 'best-fit' a cubic poly every time. I guess this means now I will be delving into my calculus mathematics text book yet again! I haven't opened that thing in years!
Old 03-04-2009, 07:49 PM
  #94  
Access
Senior Member
Thread Starter
 
Join Date: Mar 2005
Location: , CA
Posts: 4,900
Likes: 0
Received 0 Likes on 0 Posts
Default RE: Another brushless 'myth', lower KV = more torque

Yeah, if I remember these problems ultimately turn into "area-under-the-curve" questions, so you either need to live with the error within your resolution of 0.1sec, or you need to do the integrate to find the area-under-the-curve. You'd end up doing the same thing for work/energy analysis where if I remember work is the area under the power curve (?)

I think if you can get a relatively finer resolution, fine enough to the point where it doesn't change much from one sample to the next, you'll be okay with what you are doing now. Approximating something to a curve seems like a workaround for not having sufficient resolution or two much space between the samples.
Old 03-05-2009, 07:46 PM
  #95  
mattnin
Senior Member
 
mattnin's Avatar
 
Join Date: Jan 2006
Location: grand junction, CO
Posts: 834
Likes: 0
Received 0 Likes on 0 Posts
Default RE: Another brushless 'myth', lower KV = more torque

A spreadsheet is being worked out right now that will approximate, however very accurately, a correct angular acceleration at each RPM reading. I will paste a message from John Stranahan where he explains better.

System theory-What we have done to avoid some of the sampling problem is to do a linear regression fit of a model to the angular velocity data. A cubic polynomial fits very well. The derivative of this model gives a more accurate angular acceleration at each data point than can be calculated manually. This is used to calculate power. The data is smoothed in the regression. I think accuracy is improved 2-3 % at each individual point. Since the initial points are so few this may be important to distinguish two motors. It's all done automatically in a spreadsheet which is mostly complete. We are adding an input for flywheel and armature inertia. The robitronic has similar imputs for inertia. The spreadsheet will be provided free soon. Here is a copy of the output at present. The data is from a Fantom Dyno run on a 19 turn.

Note these two graphs were used for spreadsheet construction. The first shows the good fit of the cubic trendline to the angular velocity data. The second graphs just has connect the dot lines but is generated from the derivative of the first cubic model shown.
More relevant output graphs will be in the final form.

I know that sampling problems may exist in the RPM data. We will just have to trust that they (Novak, Fantom) did an equally fine job of getting that as accurate as possible.
Here is a picture of where we are so far. There are still many changes being made, but it is starting to look really good in my humble opinion.


Attached Thumbnails Click image for larger version

Name:	Ge96312.jpg
Views:	163
Size:	155.2 KB
ID:	1151034  
Old 03-09-2009, 09:19 AM
  #96  
Argess
 
Argess's Avatar
 
Join Date: Jan 2006
Location: Pleasantville, NS, CANADA
Posts: 2,355
Likes: 0
Received 1 Like on 1 Post
Default RE: Another brushless 'myth', lower KV = more torque

Nice job Mattnin. I always wanted to build an inertial dyno, but was thinking of having to do everyting from scratch, so it appeared too much work....and me being lazy.......

I had another look at all this and it appears that a brushless motor is the same as a pm brushed motor for theory, other than the commutating action is done electronically. That being the case, this may be of some interest:

The equivalent circuit for a DC motor can be thought of a series resistor and a “perfect” armature. Hence the voltage equation:

V = RI + KØw

Where R = winding resistance, K is the motor constant, Ø is the flux, V is the battery voltage, I is the current, and w is the rpm.

The torque equation is:

T = kØI

Substituting, we get:

T = (kØ)ˆ2/R(V/k Ø – w)

Now, since K, Ø, and R are all proportional to the number of turns (N), we can simplify this to:

T = aN - wbNˆ2 where a and b are constants

At w = zero, T = aN

Therefore at low rpm’s, a low KV motor with more turns yields more torque that a high KV motor with less turns.

At T = zero, w = a/(bN)

Therefore at high rpm’s, a high KV motor with less turns yields more torque than a low KV motor with more turns.

This is the developed torque. Actual shaft torque is a bit less due to losses, but I chose to ignore them to make things easier.

Now, let’s look at power:

P = wT = waN - b(wN)ˆ2

Maximum power occurs when dP/dw = 0, or when:

w = a/(2bN)

Since max rpm is w = a/(bN), the rpm (w) at which maximum power is made is ½ of the maximum rpm developed by the motor.

So, it’s not a myth….lower KV motors do make more power than high KV motors at low rpms.
Old 03-09-2009, 12:54 PM
  #97  
Access
Senior Member
Thread Starter
 
Join Date: Mar 2005
Location: , CA
Posts: 4,900
Likes: 0
Received 0 Likes on 0 Posts
Default RE: Another brushless 'myth', lower KV = more torque

ORIGINAL: Argess
I had another look at all this and it appears that a brushless motor is the same as a pm brushed motor for theory, other than the commutating action is done electronically. That being the case, this may be of some interest:

The equivalent circuit for a DC motor can be thought of a series resistor and a “perfect” armature. Hence the voltage equation:

V = RI + KØw
Even a DC motor is a dynamic system, not static. Current over time is not constant.
The proper way to model this would be a switch going to three different resistors and inductors.

Look at the circuit I put in the pics below, this is an approximation but it should at least illustrate the basic concept.
It is an approximation b'cos it doesn't take into account the actual winding (Delta or Wye) and it doesn't take into account the back-emf created by the spinning rotor. As the motor spins (the switch moves), the current draw over time is not constant. I do have a scope at home, if you like I can hook up the scope across a current sensing resistor in series with a brushed motor, post a video of it to show what is actually happening here.

Typically, the impedance in these types of circuits has a much larger effect than the resistance. Take something like an '8-ohm' speaker and measure the resistance with a typical multimeter, it's measures much less than 8 ohms. The DC static resistance is very low (if you look at some motor specs), no higher than if you unwound the copper wire around a coil and measured the straight resistance of that. Take a brushed motor and hold the rotor in place, in a short time this becomes a static system and will draw a very high amount of current. Many motors have rated 'inrush currents' which are drawn at startup and even for a small, say a 5V, 1A motor, this can be in the tens of amps or more.

Another difference is that the brushed motor has a fixed forward advance. The brushless motor can vary the forward advance dynamically depending upon the state of the motor. It typically starts with none and then increases it as the RPMs increase.

I would wait for Mattnin to produce more data before saying too much. One we see what the data shows, then we can try to relate theory to it to try to explain why. I can think up paradoxes no matter which theory I follow so at this point I am real curious to see what the data in a real, operating system actually shows.

Mattnin that data looks really good, especially useful since you can find the efficiency (vs. RPM) 'sweet spot' and then gear the vehicle to stay within that.
Attached Thumbnails Click image for larger version

Name:	Da83492.jpg
Views:	134
Size:	12.3 KB
ID:	1153827  
Old 03-09-2009, 01:25 PM
  #98  
Access
Senior Member
Thread Starter
 
Join Date: Mar 2005
Location: , CA
Posts: 4,900
Likes: 0
Received 0 Likes on 0 Posts
Default RE: Another brushless 'myth', lower KV = more torque

Taking a second look at this

Mattnin, what do you think of the efficiency values you get down in the lower RPMs? Is the dyno essentially just 'holding full throttle' on the brushless ESC as the flywheel spins up? I think you mentioned a Tekin RS Pro? If you, or 'it' (the test setup), held 1/10th throttle, would the system run more efficiently in the 3000-6000RPM range? What if you slowly 'walked' the throttle up to full, and so on. Am I making any sense here?
Old 03-09-2009, 03:57 PM
  #99  
Argess
 
Argess's Avatar
 
Join Date: Jan 2006
Location: Pleasantville, NS, CANADA
Posts: 2,355
Likes: 0
Received 1 Like on 1 Post
Default RE: Another brushless 'myth', lower KV = more torque

The torque equation I posted works out within 2% error for the two motors in the black-background chart you posted, but I admit, this is all for brushed motors. Motor 1 was a 12 turn, 36031 rpm motor capable of 277 n-m and Motor 2 was an 11 turn, 42000 rpm unit capable of 246 N-m.

I didn't need to consider 3 (or more) poles as I was concerned with the ratio of Torque from one motor to another, all things beng equal except the number of turns. So the number of poles cancel out.

For one of these graphs, at a moment in time, you can use current as constant for torque or power being constant......as if you were using a brake dyno vs an intertial one. Or, for example, the constant current at stalled-rotor torque. Or the constant current at max rpm's. However did look for the inductance to be significant. Only way I could see that was if the ESC was generating pulses much higher than the shaft frequency. If it is, it may change the torque equation by a 1 /N (resistor is 1/N and inductance is 1/Nˆ2). But I don't know enough about brushed motors and their ESCs.

The fact is, I really don't know what the ESC is doing. All I can find is that it is turning on the coil for a pulse similiar to what the communtator of a brushed system does. Mabe you can elaborate a bit more....

Anyway, I was waiting for MAttnin to get some graphs from two motors, identical, except for the number of turns, but thought I might be of help in the meantime. I've been wrong before.....I'm sure I"ll be wrong again.....LOL.
Old 03-09-2009, 04:02 PM
  #100  
Argess
 
Argess's Avatar
 
Join Date: Jan 2006
Location: Pleasantville, NS, CANADA
Posts: 2,355
Likes: 0
Received 1 Like on 1 Post
Default RE: Another brushless 'myth', lower KV = more torque

Oh with reagrds to your last post....the dyno graphs should be for a full-throttle run. If you use PWM or a chopper drive to control the speed, the motor will draw what it needs to increase torque up to a maximum on the graph. So if you hold the throttle partly on and you get to a hill of ever-increasing steepness, the vehicle will not slow down until the angle of the hill becomes steep enough the motor hits maximum torque that it is capable of for that speed. Does that make sense?

Thread Tools
Search this Thread

Contact Us - Archive - Advertising - Cookie Policy - Privacy Statement - Terms of Service - Do Not Sell My Personal Information -

Copyright © 2022 MH Sub I, LLC dba Internet Brands. All rights reserved. Use of this site indicates your consent to the Terms of Use.