Go Back  R/C Tech Forums > General Forums > Radio and Electronics
power versus torque at rpm >

power versus torque at rpm

Like Tree4Likes

power versus torque at rpm

Reply

Old 12-27-2018, 09:19 AM
  #31  
Tech Adept
Thread Starter
 
Join Date: Jul 2018
Location: Florida
Posts: 105
Default

Originally Posted by waitwhat View Post
The 5.6A (not what every manufacturer recommends btw) thing is just something that is repeatable based on the test equipment available to the guys that care to use a motor analyzer. It is a safe starting point based on what that manufacturer (Trinity I think) knows based on their motor architecture and the tolerances to which their motors are constructed. They also provide gearing guidance too. Even still, these are just starting points. They let you know that some small adjustments might be necessary.

I love the way a car drives with dynamic timing. It's incredible how much you can increase the performance of the motor, while wasting less energy as heat. I get the appeal of the simplicity of a blinky setup. Dynamic timing is something that was just too difficult for most people to get right. Set it up wrong and you could burn up your stuff very easily.

Because I use an aluminum gearbox the heat from my motor dissipates into the chassis. After a run my chassis is warm to the touch in the back.

There are possible outliers that will impact the perfect gearing and timing. For example, the largest track where I frequent usually has a flowing layout with a few slow corners. If they did a track layout with one long straight and a bunch of tight corners and short straights, the taller gearing would be slower on track. Similarly, this track had a main water pipe burst in the winter, which meant they couldn't water the track. This made the dirt dry out, which made the grip decrease significantly. The track surface got really rough because of the cracks in the clay. These outliers shift the optimal gearing and timing, and it is something that would be hard to quantify and predict the new optimums. Sometimes there are jumps that have really short run ups, and making that jump consistently is one of the most important aspects of a good lap time. In that case the timing and gearing might not be optimal for any other section of the track, but the small trade off everywhere else is worth being able to make the jump throughout the race.

​​​​​Based on the weight limit you mentioned earlier I would say you run touring cars. Touring cars would probably be the best platform to extrapolate data. The grip level is much more consistent than off-road, which makes calculating your target load much easier. Until you mentioned that weight I had no idea what type of car you are running. I think you could use the telemetry data to make your adjustments more effectively.

I track my mAh used very carefully. One time I switched from the really soft stock A-arms for hard arms. I could drive the car so much harder that I used 100mAh more than usual in a 5 minute run. That is probably enough of a load increase to warrant a timing or gearing change, just from different parts that moved the physical limit of the car.
Try some thermal paste between the motor flange and bulkhead. It helps allot if the flange and/or bulkhead has minimal contact surface. I found a thin aluminum disk Maybe 30mm between the motor and the bulkhead helps.

dirt is certainly more difficult. We cant do it in rc until the electronics become better but something I have done in the past....if a motor has a gear train that functions like a press and you use it to squish marshmallows the current will rise slowly. The current will rise slowly per degree of motor rotation. Its called gradient control or yield control in custom automation. So the opposite is true as well when pressing a rock.

the slope of torque/angle is more accurate than a torque transducer if the system is setup properly and it is especially good at calculating dynamic torque during acceleration. It works for determining a robot in water or on land and potentially could work on traction limited rc cars to provide data for planning or dynamic changes in the esc. Its also sensor less traction control.

i would help an aspiring esc mfg to develop a teach cycle for an esc to record a reference lap and do the calculations automatically for a manual end bell adjustment. With a little more interaction from the operator it could spit out a peak and efficient range of fdrs. Outside of the rc world optimal commutation angle is calculated by the ESC and its dynamic. After this is done its easy to cripple it and manually dial it in for a specific track (on road).
Bry195 is online now  
Reply With Quote
Old 12-27-2018, 01:39 PM
  #32  
Tech Fanatic
iTrader: (23)
 
Join Date: Apr 2005
Location: north-west Indiana
Posts: 914
Trader Rating: 23 (100%+)
Default

Originally Posted by Bry195 View Post
i would help an aspiring esc mfg to develop a teach cycle for an esc to record a reference lap and do the calculations automatically for a manual end bell adjustment. With a little more interaction from the operator it could spit out a peak and efficient range of fdrs. Outside of the rc world optimal commutation angle is calculated by the ESC and its dynamic. After this is done its easy to cripple it and manually dial it in for a specific track (on road).
This would be great!
trilerian is offline  
Reply With Quote
Old 12-27-2018, 08:38 PM
  #33  
Tech Adept
Thread Starter
 
Join Date: Jul 2018
Location: Florida
Posts: 105
Default

Originally Posted by trilerian View Post
This would be great!
It would be great for the MFG too because they can control which motors can generate the automatic report.
Bry195 is online now  
Reply With Quote
Old 12-29-2018, 08:53 PM
  #34  
Tech Adept
Thread Starter
 
Join Date: Jul 2018
Location: Florida
Posts: 105
Default


I got the dyno back and the eddy current load working. It adds about 50Nmm across all the rpm max. I can dial it down if i need to.
Bry195 is online now  
Reply With Quote
Old 12-29-2018, 10:22 PM
  #35  
Tech Adept
Thread Starter
 
Join Date: Jul 2018
Location: Florida
Posts: 105
Default

I had a great question about something I touched on. Inertia mismatch. Precisely about how to match rotor inertia to drive train inertia. Has someone or can someone calculate the summed inertia of any buggy or touring car? Even if its just generic rods and shafts with bulk diameters we can apply 98-99 percent rotating efficiencys to the bearings and ill Have to dig into belt and spur gear losses.

i think solidowrks can do this.
Bry195 is online now  
Reply With Quote
Old 12-31-2018, 10:52 PM
  #36  
Tech Adept
Thread Starter
 
Join Date: Jul 2018
Location: Florida
Posts: 105
Default

If you break your track telemetry down into torque/rpm segments each point must sit under the brown curve but the entire lap (average of each segment) needs to sit under the green curve.

This is very close to the torque and rpm for a 21.5 monster max, lightly fan cooled on an 80F day and sea level. This assumes 2 bolts on aluminum bulkhead (tc6.2) to motor.

friction and inertia need to be close to a tc6.2 and the fdr is 4.05. Sorry you want orange and red curves on torque rpm curve not green and brown on efficiency curve.


Bry195 is online now  
Reply With Quote
Old 12-31-2018, 11:18 PM
  #37  
Tech Adept
Thread Starter
 
Join Date: Jul 2018
Location: Florida
Posts: 105
Default

For clarification:

below 10000 rpm the end bell was used to get as much space as possible under the curve (you see the intersects for all of the curves).
this was done by using the dyno to make rough adjustments to the end bell (something like 42-50 degrees.
A scope was then used to align the current at the motor to a Hall effect pulse.
this resulted in something like 47 degrees at the end bell. This is the first curve.
the second curve (very wide)...I set the boost start point, rate, and rate increase to widen the curve as far as possible.
this resulted in 1 pull at 130 degrees but that was enough for me. This is my peak curve.
i then went to the track and adjusted boost and fdr until no matter how hard i tried i could over heat the motor and it stabilized at 72 degrees c.
I never moved the end bell and I will convert the numbers the esc was pushing through the motor if someone needs it in terms of degrees so that you can use it to set your end bell.
I left the track with stable timing numbers from a thermal perspective.
i then made the third run which is the continuous duty curve.
i am now breaking down the track into torque/rpm segments per lap and the running averages for the entire session.
This will show that specific segments exceed continuous while others were well well under continuous but the average will land on the orange curve.
Once I calculate the difference between the dyno and the track i can give you a curve that is just motor data. You can then add drivetrain data for your setup on top of that.
Bry195 is online now  
Reply With Quote
Old 01-01-2019, 04:23 PM
  #38  
Tech Apprentice
iTrader: (1)
 
Join Date: Dec 2018
Posts: 81
Trader Rating: 1 (100%+)
Default

How many degrees of boost timing was used to get temperature stability on track?
waitwhat is offline  
Reply With Quote
Old 01-01-2019, 09:54 PM
  #39  
Tech Adept
Thread Starter
 
Join Date: Jul 2018
Location: Florida
Posts: 105
Default

Originally Posted by waitwhat View Post
How many degrees of boost timing was used to get temperature stability on track?
good question. I have Sanwa equipment and i feed it numbers through the transmitter. The numbers dont have units and i havent done the math to calculate what they mean yet.

short...medium...long
boost rate=0 for short, 20, and 60 for medium and long.
i did curves for manual changes on the end bell so i was thinking that I would see if I can use that to scale the sanwa units.
For each of those curves the start rpm was about 10k and accelerated by 20 (to stay with the curve) and no units again.
maybe sanwa will tell me.
Bry195 is online now  
Reply With Quote
Old 01-01-2019, 10:01 PM
  #40  
Tech Adept
Thread Starter
 
Join Date: Jul 2018
Location: Florida
Posts: 105
Default

i did take my local track telemetry and averaged the torques and rpms. It sat right on the continuous curve. i see that the average rpm is 10k with a standard deviation of 2250 rpm. Its close but if i change the gearing to move the rpm to the center of the efficient part of the curve or torquey part of the continuous curve Ill have less average waste heat so i can push the car further into the peak.
Bry195 is online now  
Reply With Quote
Old 01-01-2019, 11:15 PM
  #41  
Tech Addict
iTrader: (7)
 
gubbs3's Avatar
 
Join Date: Aug 2006
Location: MN
Posts: 657
Trader Rating: 7 (100%+)
Default

I might recommend adding a time weighted average to the track telemetry data. including only standard deviation to the average rpm is a good baseline but I would speculate the track configuration may have a significant impact on torque needs at different rpm. In racer terms, a flowing layout vs a tight layout (long straights linked with tight 180s). Both layouts could have the same average rpm and but the standard deviation in rpm will be asymmetric and the ideal timing setting may not necessarily match the peak on the bell curve. This would require on track testing against lap times, as most changes which need testing. The repeatability you desire would be found in identifying the rpm range on track where the longest duration of high load exists and adjusting the timing to put the most efficient torque curve into that rpm range.
gubbs3 is offline  
Reply With Quote
Old 01-02-2019, 12:38 AM
  #42  
Tech Apprentice
iTrader: (1)
 
Join Date: Dec 2018
Posts: 81
Trader Rating: 1 (100%+)
Default

When I would go from a perfectly adjusted blinky setup to one with dynamic timing I lower my motor timing ≈10 and I lower my gearing 10-15%. I add the maximum amount of dynamic timing (55) over a wide rpm range (3k-20k). This gives a very smooth and linear power band, while decreasing temps from a blinky setup that is on the limit. You can adjust the rpm upper value to tune how aggressive the timing comes on.

Try to replicate these settings from your ideal blinky settings and I bet you will be surprised how well it works to keep the motor in the sweet spot.

Lastly, I always test settings without a fan first. You want to make sure that your system is safe in case of a fan failure. Unless you can use a telemetry channel to trigger an alarm or put the car into "limp home mode" based on a temperature reading, I would not rely on a fan to keep stuff from going into thermal runaway.
waitwhat is offline  
Reply With Quote
Old 01-03-2019, 09:33 PM
  #43  
Tech Adept
Thread Starter
 
Join Date: Jul 2018
Location: Florida
Posts: 105
Default

Originally Posted by gubbs3 View Post
I might recommend adding a time weighted average to the track telemetry data. including only standard deviation to the average rpm is a good baseline but I would speculate the track configuration may have a significant impact on torque needs at different rpm. In racer terms, a flowing layout vs a tight layout (long straights linked with tight 180s). Both layouts could have the same average rpm and but the standard deviation in rpm will be asymmetric and the ideal timing setting may not necessarily match the peak on the bell curve. This would require on track testing against lap times, as most changes which need testing. The repeatability you desire would be found in identifying the rpm range on track where the longest duration of high load exists and adjusting the timing to put the most efficient torque curve into that rpm range.
the telemetry is a time stamp with velocity and amps per measurement. Sizing the mechanical and electrical system to the application (different tracks) is exactly as you describe. Its an average. Its almost done for me when export the track to excel. i was just shocked by the fact that the deviation was so small.

on just about any track acceleration plays a big part in calculating the continuous torque requirement just because of the quantity of current needed to accelerate. Short tracks make it worse. i suspect the shorter tracks also require a closer look at the rotor inertia as long as traction isnt limited. But you understand this.
Bry195 is online now  
Reply With Quote
Old 01-03-2019, 09:43 PM
  #44  
Tech Adept
Thread Starter
 
Join Date: Jul 2018
Location: Florida
Posts: 105
Default

Originally Posted by waitwhat View Post
When I would go from a perfectly adjusted blinky setup to one with dynamic timing I lower my motor timing ≈10 and I lower my gearing 10-15%. I add the maximum amount of dynamic timing (55) over a wide rpm range (3k-20k). This gives a very smooth and linear power band, while decreasing temps from a blinky setup that is on the limit. You can adjust the rpm upper value to tune how aggressive the timing comes on.

Try to replicate these settings from your ideal blinky settings and I bet you will be surprised how well it works to keep the motor in the sweet spot.

Lastly, I always test settings without a fan first. You want to make sure that your system is safe in case of a fan failure. Unless you can use a telemetry channel to trigger an alarm or put the car into "limp home mode" based on a temperature reading, I would not rely on a fan to keep stuff from going into thermal runaway.
I will.i havent done that yet. i suspect that using the end bell will miss the difference you see in this set of curves. Notice the efficiency and power difference in the middle of the curve? If I was to run dynamic i would set the end bell to the timing mark that gets me the power or the efficiency at this rpm and then increase the rate of timing only wide enough for the average requirement for the track. Great point on the fan. I setup alarms for everything in the Sanwa over telemetry. 80cand it tells me to pull over. i even have a 3rd channel setup to run a PWM fan up and down based on temperature.its overkill...I know.



Bry195 is online now  
Reply With Quote
Old 01-03-2019, 09:47 PM
  #45  
Tech Adept
Thread Starter
 
Join Date: Jul 2018
Location: Florida
Posts: 105
Default

The blue peak in efficiency or green dip in torque might not be as controllable with just end bell timing but Ill give it a shot. Pulling that bump out of the middle may be possible with end bell timing (I hope) but i have to accelerate the rate of boost on a curve from about 7500 rpm up. I couldnt pull the curve up any higher with the end bell timing i had. Tomorrow Ill see if i can pick the torque up at the lower part of the curve with the end bell.

But all of those intersects offer allot of choices and the differences are significant. Calculating an average requirement for a track is easy and you drop the point on the curve and pick 15% torque or 15% efficiency and everything in between to pick your timing.

by the way that set of curves for the 17.5 are not like the previous set for the 21.5. The middle curve is not tested for thermal stability continuous duty.

Last edited by Bry195; 01-03-2019 at 10:05 PM.
Bry195 is online now  
Reply With Quote

Thread Tools
Search this Thread

Contact Us Archive Advertising Cookie Policy Privacy Statement Terms of Service