Dyno, Homemade, Using a Novak Sentry Data Logger, Continued, The Experimental Thread.
04-12-2009, 08:52 PM
#16
Run Line would be the actual path a good racer takes on a track. Here is some more things to think about as Dyno equipment heads South. Make the clearance from your outer board on the ends to the 12 foot semicircle about 9 feet instead of 11 or 12. It has to be near the runline. The actual runline will be within inches of the end boards of the oval. The 173 foot runline was measured on a wheel device. He actually walked a path similar to your drawing but the faster line was a little more diamond shaped. More pointy in the corners so to speak. Run time for a fast lap was 4.1 seconds with a 13.5 motor and 4 cell. Oval cars can exceed the G's pulled by touring cars if there is high grip. Occasionally we had high grip and pulled the 4.1 second laps.
Throttle position is very dependent on car setup. On my car it was always dropping throttle slowly off at mid straight. Off throttle just momentarily to plant for the tight corner, medium to 3/4 throttle very steady through the corner, roll to full throttle starting just before corner exit reaching full throttle at mid straight or sooner depending on traction. I do a pan road course corner exactly the same except there is a slight pause on corner exit to straighten wheels, then throttle roll on. Otherwise there is a corner exit spin. We can tune this corner exit spin completely out on the oval with wedge and other tricks.
One car was set up with so much traction, or so planted so to speak that toward the end of the run you could hold full throttle all the way around if you followed your pictured line. This was about .2- .3 seconds off the pace, though.
One very fast car could increase throttle at mid corner and roll it on all during corner exit. He is the guy geared 5 teeth over my ratio (64 pitch gears). He and his buddy burned motors up. Both could run a fast lap with a smoking hot motor. My theory is this deadened the car and reduced corner exit problems and throttle instability that might cause a spin. Pan cars are delicate beast to keep the pointy end going forward. It may also have been the thought that the hotter you run it the more power you have.
Top speeds by radar were about 32 mph by the end of the straight. Average speeds were about 29 mph based on a fast laps and the run line distance.
So what part of the course affects gearing (or what are the motor restricted parts of the race). Well on the oval for me it would be from corner exit to mid straight. Everywhere else you are on part throttle. For a stock truck on dirt it was about the first 10 feet out of a slow corner. You make more gains in the slower parts of the course.
My wide pan runs the oval course in 3.7 seconds with 2sLiP and Novak 3.5 R. you can still reach full throttle by mid straight.
Throttle position is very dependent on car setup. On my car it was always dropping throttle slowly off at mid straight. Off throttle just momentarily to plant for the tight corner, medium to 3/4 throttle very steady through the corner, roll to full throttle starting just before corner exit reaching full throttle at mid straight or sooner depending on traction. I do a pan road course corner exactly the same except there is a slight pause on corner exit to straighten wheels, then throttle roll on. Otherwise there is a corner exit spin. We can tune this corner exit spin completely out on the oval with wedge and other tricks.
One car was set up with so much traction, or so planted so to speak that toward the end of the run you could hold full throttle all the way around if you followed your pictured line. This was about .2- .3 seconds off the pace, though.
One very fast car could increase throttle at mid corner and roll it on all during corner exit. He is the guy geared 5 teeth over my ratio (64 pitch gears). He and his buddy burned motors up. Both could run a fast lap with a smoking hot motor. My theory is this deadened the car and reduced corner exit problems and throttle instability that might cause a spin. Pan cars are delicate beast to keep the pointy end going forward. It may also have been the thought that the hotter you run it the more power you have.
Top speeds by radar were about 32 mph by the end of the straight. Average speeds were about 29 mph based on a fast laps and the run line distance.
So what part of the course affects gearing (or what are the motor restricted parts of the race). Well on the oval for me it would be from corner exit to mid straight. Everywhere else you are on part throttle. For a stock truck on dirt it was about the first 10 feet out of a slow corner. You make more gains in the slower parts of the course.
My wide pan runs the oval course in 3.7 seconds with 2sLiP and Novak 3.5 R. you can still reach full throttle by mid straight.
Last edited by John Stranahan; 04-12-2009 at 10:14 PM.
04-13-2009, 10:56 AM
#17
My run line 145 feet, Ellipse shaped
This is probably the most reallistic shape of my run line. It is slightly more diamond shaped (with rounded points, laying on the side), but would have about the same perimeter distance. The length of the ellipse is 145 feet. I can see a fudge factor appearing in my gearing calculation. I used a 173 foot run line as that is what was measured by the wheel, but I run a shorter 145 foot run line and make the corners a little harder. This suited my car better. Thats a 16% factor for tire slip. The rear tire turned farther than my actual runline. This fudge factor is adjustable and included on the Fantom dyno software. Other guys with more oval experience run a line closer to system theory's drawing. At the carpet oval nats, (the majority of the videos are no longer viewable,) the run line is more like this one and painted sharply and darkly in the carpet. The mod guys in the video I posted run a little fatter run line.
So why discuss oval first. Well it is a little more subject to measuring and calculation due to the more simple course.
This is probably the most reallistic shape of my run line. It is slightly more diamond shaped (with rounded points, laying on the side), but would have about the same perimeter distance. The length of the ellipse is 145 feet. I can see a fudge factor appearing in my gearing calculation. I used a 173 foot run line as that is what was measured by the wheel, but I run a shorter 145 foot run line and make the corners a little harder. This suited my car better. Thats a 16% factor for tire slip. The rear tire turned farther than my actual runline. This fudge factor is adjustable and included on the Fantom dyno software. Other guys with more oval experience run a line closer to system theory's drawing. At the carpet oval nats, (the majority of the videos are no longer viewable,) the run line is more like this one and painted sharply and darkly in the carpet. The mod guys in the video I posted run a little fatter run line.
So why discuss oval first. Well it is a little more subject to measuring and calculation due to the more simple course.
Last edited by John Stranahan; 04-15-2009 at 10:21 AM.
04-13-2009, 05:26 PM
#18
Thanks Matt. The parts arrived in good condition.
Here is a picture of the complete dyno. It is setup on an old oval chassis. I have installed the standard type of T-plate and upped the center shock oil to 60 weight. Now when this 3.5 starts there is smooth barely noticable movement. Vibration and noise are very low compared to a Fantom or Robitronic Dyno run. This is because of the motor suspession. A simple T-plate clamped to a table would probably be sufficient. I have tested it with the Steel Fantom Flywheel at 55,000 RPM. Everything was smooth. To start a run I rotate the Astroflight servo tester know full clocwise. I have positioned this knob out of the shrapnel path in case a flywheel or motor shaft should fail. The servo tester hooks up the speed control radio sensor harness, and it has an extra output to feed the Sentry, Battery Eliminator Circuit Juice, from the speed control.
All solder joints were checked. Heavy gauge wire is used throughtout (except at the motor which has 14 gauge) as well as high quality connectors; the motor will receive maximum amps and volts from the battery. The novak sentry is the little clear receiver sized looking device. It has an amp sensor ($25) visible on top of the battery attached to a short 12 gauge sensor harness that is home made. Volts are also sensed here on the battery side. This amp sensor only flowd current in one direction so it has to be hooked up with the + toward the battery. Other than this harness which both Matt and I assembled without trouble there is little difficulty in the setup.
The software and owners manual is available at TeamNovak.com.
The RPM range advertized is 500-99999 RPM. That is why we are getting erroneous readings on the first few tenths with the heavy flywheel. I may try the magnetic sensor just to see if there is any improvement. The simple sensor, to install, plugs in to the sensor wires of the motor.
A battery charger will be part of the system. I will measure at what voltage I start the tests. It will be on the Chargers first blinking green light (90% charge). Results will be somewhat specific to this battery in spite of this. I do have another brand of battery to test.
Tests to come.
Here is a picture of the complete dyno. It is setup on an old oval chassis. I have installed the standard type of T-plate and upped the center shock oil to 60 weight. Now when this 3.5 starts there is smooth barely noticable movement. Vibration and noise are very low compared to a Fantom or Robitronic Dyno run. This is because of the motor suspession. A simple T-plate clamped to a table would probably be sufficient. I have tested it with the Steel Fantom Flywheel at 55,000 RPM. Everything was smooth. To start a run I rotate the Astroflight servo tester know full clocwise. I have positioned this knob out of the shrapnel path in case a flywheel or motor shaft should fail. The servo tester hooks up the speed control radio sensor harness, and it has an extra output to feed the Sentry, Battery Eliminator Circuit Juice, from the speed control.
All solder joints were checked. Heavy gauge wire is used throughtout (except at the motor which has 14 gauge) as well as high quality connectors; the motor will receive maximum amps and volts from the battery. The novak sentry is the little clear receiver sized looking device. It has an amp sensor ($25) visible on top of the battery attached to a short 12 gauge sensor harness that is home made. Volts are also sensed here on the battery side. This amp sensor only flowd current in one direction so it has to be hooked up with the + toward the battery. Other than this harness which both Matt and I assembled without trouble there is little difficulty in the setup.
The software and owners manual is available at TeamNovak.com.
The RPM range advertized is 500-99999 RPM. That is why we are getting erroneous readings on the first few tenths with the heavy flywheel. I may try the magnetic sensor just to see if there is any improvement. The simple sensor, to install, plugs in to the sensor wires of the motor.
A battery charger will be part of the system. I will measure at what voltage I start the tests. It will be on the Chargers first blinking green light (90% charge). Results will be somewhat specific to this battery in spite of this. I do have another brand of battery to test.
Tests to come.
Last edited by John Stranahan; 04-15-2009 at 10:21 AM.
04-13-2009, 07:20 PM
#19
Welcome John, that is the least I could do for all the help you offered. Looking forward to the results
Matt
Matt
04-13-2009, 09:43 PM
#20
Well here is the first output of the dyno. I used the RPM sensor that plugs into the motor sensor harness. I used the Fantom Steel flywheel. The flywheel had a inertia of 9.89 10-5 kg m^2/s2. This is plugged into the spreadsheet in the appropriate place. Here is the major output and a graph attached.
Max Power 438 W
Max RPM 55,500 RPM
Max Torque 2545 g-cm
This is very realistic of the numbers you would get with 2s LiPo and the LRP X11 3.5, Nickel plated rotor. Now here is the rub. If you look at the amps graph in green it is a rather peculiar shape. There might be a little robot inside the LRP speed control screwing with me and limiting torque on first startup. I'll update this graph later. This is the typical type and form of the output that you will get with this dyno. It looks really super to me. I know it is accurate.
Now the exiting stuff. How does it feel to have 55,000 RPM in the room. Well it is quite tolerable now. I have no instinct to flee. I held the chassis in my right hand rotated the knob with the left hand, kept the flywheel shrapnel zone away from me. This thing is so compact and pleasant to use. No bulky case. No 50 lb power supply. etc. Now I'll see if I can get some precision out of the beast.
How does it feel to do a second run with this bad boy motor, the steel flywheel and the 2 s LiPo without a long rest. Well the run finished with a puff of smoke exiting the flywheel path. The leads from the speed control to motor were fortunately only 14 gauge. They were limp with heat and hot to the touch. OK never again. About 15 minute wait should be OK. Note also this was a second run and likely would not have that race start mess at the beginning. It may not be possible at all to go full power without melting things. Fortunately Matts Sentry is safe. I will have data files available for the curious.
be sure and click on the pic a second time to get a clear view. Use full screen.
Max Power 438 W
Max RPM 55,500 RPM
Max Torque 2545 g-cm
This is very realistic of the numbers you would get with 2s LiPo and the LRP X11 3.5, Nickel plated rotor. Now here is the rub. If you look at the amps graph in green it is a rather peculiar shape. There might be a little robot inside the LRP speed control screwing with me and limiting torque on first startup. I'll update this graph later. This is the typical type and form of the output that you will get with this dyno. It looks really super to me. I know it is accurate.
Now the exiting stuff. How does it feel to have 55,000 RPM in the room. Well it is quite tolerable now. I have no instinct to flee. I held the chassis in my right hand rotated the knob with the left hand, kept the flywheel shrapnel zone away from me. This thing is so compact and pleasant to use. No bulky case. No 50 lb power supply. etc. Now I'll see if I can get some precision out of the beast.
How does it feel to do a second run with this bad boy motor, the steel flywheel and the 2 s LiPo without a long rest. Well the run finished with a puff of smoke exiting the flywheel path. The leads from the speed control to motor were fortunately only 14 gauge. They were limp with heat and hot to the touch. OK never again. About 15 minute wait should be OK. Note also this was a second run and likely would not have that race start mess at the beginning. It may not be possible at all to go full power without melting things. Fortunately Matts Sentry is safe. I will have data files available for the curious.
be sure and click on the pic a second time to get a clear view. Use full screen.
Last edited by John Stranahan; 04-13-2009 at 11:48 PM.
04-13-2009, 09:55 PM
#21
New Post up above, First Output in Houston
Getting Sentry Data into Excel
This stumped me for a while. You will install a Novak Program that comes with your sentry or Beta v 1.1 from the Novak site.
You push a button on the sentry until you get red. This clears the memory. You push it again after about a minute to get blinking green. Data aquisition is now ready. You run the dyno with the knob on the Astroflight servo tester. Flip it full as fast as possible. Turn it off. You now have a saved dyno file on the Sentry. You could put maybe 30 -100 dyno runs on the Sentry, but you should have adequate cooling periods.
You download data onto the computer with the Sentry Program
here is the trick
You touch the data tab
You select Print
You get a preview screen
Now you can select to print a file to disk that is Excel compatible.
The rest of the instructions to analyze the data, I have printed before on the previous thread, but they will also be on page 3 of the spreadsheet. We will make a new spreadsheet available once we clean up all the glitches and scaling problems.
It is looking really good with this flywheel.
John
Getting Sentry Data into Excel
This stumped me for a while. You will install a Novak Program that comes with your sentry or Beta v 1.1 from the Novak site.
You push a button on the sentry until you get red. This clears the memory. You push it again after about a minute to get blinking green. Data aquisition is now ready. You run the dyno with the knob on the Astroflight servo tester. Flip it full as fast as possible. Turn it off. You now have a saved dyno file on the Sentry. You could put maybe 30 -100 dyno runs on the Sentry, but you should have adequate cooling periods.
You download data onto the computer with the Sentry Program
here is the trick
You touch the data tab
You select Print
You get a preview screen
Now you can select to print a file to disk that is Excel compatible.
The rest of the instructions to analyze the data, I have printed before on the previous thread, but they will also be on page 3 of the spreadsheet. We will make a new spreadsheet available once we clean up all the glitches and scaling problems.
It is looking really good with this flywheel.
John
Last edited by John Stranahan; 04-13-2009 at 10:33 PM.
04-14-2009, 09:23 AM
#22
Tech Adept
John,
The Dyno data looks great. As you note, the current sensor appears to be in error. With that level of torque and power on the flywheel, and the wires so hot, it is likely drawing in excess of 100 amps during the ramp up to 438{W}.
I attach analysis of your run line sketch in Google Sketchup. The Arc Tool puts the tight turn distance s = 24 feet with radius R = 11 feet. If throttle is steady through this section, you could clock the time and compute cornering velocity in the tight sections "s".
This converts to meters per second with tire radius r in meters. Then the gear ratio G, tire radius r, and cornering velocity v determine motor shaft speed w (in radian per second) upon exit of the turn. This maps to a motor operating point on the Dyno curves at less than full throttle input voltage, defining the point where acceleration occurs after picking up the throttle. The choice of G impacts the slope of the force-velocity curve with a trade-off between more punch out of the tight corners and more top speed down the long segments "S".
The Dyno data looks great. As you note, the current sensor appears to be in error. With that level of torque and power on the flywheel, and the wires so hot, it is likely drawing in excess of 100 amps during the ramp up to 438{W}.
I attach analysis of your run line sketch in Google Sketchup. The Arc Tool puts the tight turn distance s = 24 feet with radius R = 11 feet. If throttle is steady through this section, you could clock the time and compute cornering velocity in the tight sections "s".
This converts to meters per second with tire radius r in meters. Then the gear ratio G, tire radius r, and cornering velocity v determine motor shaft speed w (in radian per second) upon exit of the turn. This maps to a motor operating point on the Dyno curves at less than full throttle input voltage, defining the point where acceleration occurs after picking up the throttle. The choice of G impacts the slope of the force-velocity curve with a trade-off between more punch out of the tight corners and more top speed down the long segments "S".
04-14-2009, 11:46 AM
#23
The sketch has realistic numbers now. Nice
The dyno run with hot wires was a second run that I did not process.
On the first run I think the amp reading is accurate till about 29000 RPM at which point the amps pegged at 108 amps.
the lower amp readings before this point may be the result of speed control "rocket start" control. I am not sure. I know on the non TC model LRP competition speed control that this was a problem. The "Rocket start" is bypased by goosing the throttle slightly before the run letting the flywheel come to rest then starting the real run. If you read the speed control manual it says that it's supposed to boost the start. I am not so convinced and think it may be an attempt at traction control.
The dyno run with hot wires was a second run that I did not process.
On the first run I think the amp reading is accurate till about 29000 RPM at which point the amps pegged at 108 amps.
the lower amp readings before this point may be the result of speed control "rocket start" control. I am not sure. I know on the non TC model LRP competition speed control that this was a problem. The "Rocket start" is bypased by goosing the throttle slightly before the run letting the flywheel come to rest then starting the real run. If you read the speed control manual it says that it's supposed to boost the start. I am not so convinced and think it may be an attempt at traction control.
04-14-2009, 04:07 PM
#24
Novak SS 17.5 stock motor
I did some test to standardize the Sentry dyno procedure and to determine its precision.
I made 3 runs with the Fantom steel flywheel. Precision was just not there as there were only a couple of points near the top of the Power hump. In other words the data was not collected in sufficiently small intervals.
I changed to Matts Custom Flywheel. diameter is 3.426 inch. Weight is 365.5 g. I put in the new flywheel inertia into the spreadsheet. I got the following 3 values for power in the order I took them.
167 W
163 W
158 W
this allows expressing the value with a standard deviation. 163 W + 2 W
Just outstanding. Now we take a look at these values with a keen eye and look for time trends. It looks like power is going down with time. I did not allow a cooling period that was long enough. It was only 2 minutes or so while the battery recharged to the green blinking light. So part of this deviation is probably the motor storing heat and running a little less powerfully each run. It is probably not all random error. I have mentioned this error previously.
This is what the dyno procedure looks like now with an LRP speed control
1. charge with FMA direct charger to 8.25 Volts while charging (about 90% full) When you remove the charger the battery will stabilize at about 8.19-8.21 V. I repeatedly saw 8.21 volts at the beginning of my dyno runs.
2. Goose the throttle just a little (this tended to get up to 2000 RPM) this gets rid of the Rocket start in the LRP.
3. When the flywheel is completely still (no rocking) go full throttle as quickly as possible. Stop when speed stabilizes.
Collecting the data. Back at the computer after downloading the reslults, I started collecting data for our dyno spreadsheet when I noticed the first disruption in amps Volts or RPM. I collected the previous point as point 0
I collected data points until the voltage quit rising. It drops at first and then slowly rises for the rest of the test. I collected only one point at this stabilized volt. Usually there are 10 or so points with this stabilized high voltage at the end of the run (8.12 V or so).
I am very confident in the power results especially now with 3 replicates so close for stock motors.
Based on previous work with dynos and stock motors there is no reason to suspect error in the amp reading on this dyno run. It starts in the proper range and degrades in the proper manner. Efficiency numbers are not crazy anywhere.
Ways to improve my procedure. It only took a couple of minutes for the charge to refill the battery to the level before the test. I should disconnect it and wait a full five or ten minutes before a retest.
There are some reasons to suspect problems with the amp curve on the previous LRP 3.5 motor. I suspect Rocket Start. I will do a test to eliminate this possibility. There is also a possibility the system just is not capable of delivering the proper amps for the length of time required, and then there is the thought (not mine) that the sensor is not accurate, but it is accurate with the stock motor, so I look to other causes.
GEARING
We are getting to the gearing stuff. First we had to prove the dyno.
Note the stock motor has a power peak about 8000 RPM. The more time your motor spends near this RPM on the Track the faster you will go. If you use only the temp gun to find your gear you will be left of this peak and will be putting out less power and more heat. You will go slower.
I am super pleased with the precision with the large flywheel and stock motors.
OK so what is that hump buisiness at the end. Well on this graph the only true hump at the end is in efficiency. I suspect the motor is just more efficient when running fast an possibly the speed control is able to optimize itself at a more constant speed. Another problem that is occuring is that we are measuring all these things using the acceleration of the flywheel. When it quits accelerating we will be getting spurious or less accurate results.
I did some test to standardize the Sentry dyno procedure and to determine its precision.
I made 3 runs with the Fantom steel flywheel. Precision was just not there as there were only a couple of points near the top of the Power hump. In other words the data was not collected in sufficiently small intervals.
I changed to Matts Custom Flywheel. diameter is 3.426 inch. Weight is 365.5 g. I put in the new flywheel inertia into the spreadsheet. I got the following 3 values for power in the order I took them.
167 W
163 W
158 W
this allows expressing the value with a standard deviation. 163 W + 2 W
Just outstanding. Now we take a look at these values with a keen eye and look for time trends. It looks like power is going down with time. I did not allow a cooling period that was long enough. It was only 2 minutes or so while the battery recharged to the green blinking light. So part of this deviation is probably the motor storing heat and running a little less powerfully each run. It is probably not all random error. I have mentioned this error previously.
This is what the dyno procedure looks like now with an LRP speed control
1. charge with FMA direct charger to 8.25 Volts while charging (about 90% full) When you remove the charger the battery will stabilize at about 8.19-8.21 V. I repeatedly saw 8.21 volts at the beginning of my dyno runs.
2. Goose the throttle just a little (this tended to get up to 2000 RPM) this gets rid of the Rocket start in the LRP.
3. When the flywheel is completely still (no rocking) go full throttle as quickly as possible. Stop when speed stabilizes.
Collecting the data. Back at the computer after downloading the reslults, I started collecting data for our dyno spreadsheet when I noticed the first disruption in amps Volts or RPM. I collected the previous point as point 0
I collected data points until the voltage quit rising. It drops at first and then slowly rises for the rest of the test. I collected only one point at this stabilized volt. Usually there are 10 or so points with this stabilized high voltage at the end of the run (8.12 V or so).
I am very confident in the power results especially now with 3 replicates so close for stock motors.
Based on previous work with dynos and stock motors there is no reason to suspect error in the amp reading on this dyno run. It starts in the proper range and degrades in the proper manner. Efficiency numbers are not crazy anywhere.
Ways to improve my procedure. It only took a couple of minutes for the charge to refill the battery to the level before the test. I should disconnect it and wait a full five or ten minutes before a retest.
There are some reasons to suspect problems with the amp curve on the previous LRP 3.5 motor. I suspect Rocket Start. I will do a test to eliminate this possibility. There is also a possibility the system just is not capable of delivering the proper amps for the length of time required, and then there is the thought (not mine) that the sensor is not accurate, but it is accurate with the stock motor, so I look to other causes.
GEARING
We are getting to the gearing stuff. First we had to prove the dyno.
Note the stock motor has a power peak about 8000 RPM. The more time your motor spends near this RPM on the Track the faster you will go. If you use only the temp gun to find your gear you will be left of this peak and will be putting out less power and more heat. You will go slower.
I am super pleased with the precision with the large flywheel and stock motors.
OK so what is that hump buisiness at the end. Well on this graph the only true hump at the end is in efficiency. I suspect the motor is just more efficient when running fast an possibly the speed control is able to optimize itself at a more constant speed. Another problem that is occuring is that we are measuring all these things using the acceleration of the flywheel. When it quits accelerating we will be getting spurious or less accurate results.
Last edited by John Stranahan; 04-14-2009 at 11:00 PM.
04-14-2009, 07:34 PM
#25
Gearing with runline and average speed
I mentioned that for racing purposes you get the most out of your electric motor when you gear it so that it is putting out maximum or near maximum power more of the time. A dyno can help with this as can an experienced racer.
Gearing the 17.5 for 2s LiPo using runline and average lap
One thing you can take to the bank is that the maximum power on this motor is in the neighborhood of 8200 RPM. I just read this off the graph in the previous post. (The graph could be printed, mark the top of the power hump. Go straight down read the RPM estimating one more digit off the axis at the bottom.) I did some calculations that opened my eyes a bit. I used the maximum and minimum runlines we determined above. I'll call these the oval guys run line (system theories sketch 173 ft) and the road race guys runline (more like my sketch 145 ft. Nothing fancy except converting units. Both runlines are inches from the endboards.
First get the average velocity of the car on your runline.
oval guys
173 foot/4.5 second x 12 inch/foot x 60 second/minute = 27680 in/min
road race guys trying to be oval guys
145 foot/4.5 second x 12 inch/foot x 60 second/minute = 23200 in/min
Now get the tire rollout which is Pi x diameter = 2.2 x 3.14 = 6.908 inch/revolution
divide this into your speeds to get the axle RPM desired
(27680 in/min)/(6.908 inch/revolution) = 4006 rev/min (oval guy)
(23200 in/min)/(6.908 inch/revolution = 3358 rev/min (road race guy)
divide these into you RPM at max power from the dyno graph to see what gear ratio you need
8200RPM/4006RPM = 2.0 Oval guys
8200RPM/3358RPM = 2.44 Road Race Guys
With a 100 tooth spur
100/2.0 = 50 tooth pinion (Oval guys, 173 foot runline)
100/2.44 = 41 tooth pinion (Road Race Guys 145 foot runline)
(feel free to multiply this by a factor 1.05 to 1.10 to allow for a 5% to 10% tire slip)
so that is how I would do it. That is how Fantom software does it. This is a real eye opener for me as both runlines can exist on the same course. I am calculating 9 more pinion teeth by running wide over running tight. The observed difference was on the order of 5-6 teeth. The 173 foot runline is probably not achievable. It is shorter.
Maybe Southwest tour can gives us some numbers from the velodrome and I can apply this method to it.
I'll run through this method with a dyno output from a 4 cell 13.5 run. I have done this previously on my CRC thread.
Here is a link to the Power vs RPM graph of a Novak 13.5 run on a 4 cell pack. Again that RPM where max power occurs is a very reliable piece of data.
http://www.rctech.net/forum/attachme...bx-13-5002.jpg
OK I ran the same two run lines using the 13.5/4 cell graph I get 26 and 31 tooth pinions for the short and long run lines. This was the observed difference in number of teeth between me and some guys with more oval experience running the wide line. Mistery solved.
I mentioned that for racing purposes you get the most out of your electric motor when you gear it so that it is putting out maximum or near maximum power more of the time. A dyno can help with this as can an experienced racer.
Gearing the 17.5 for 2s LiPo using runline and average lap
One thing you can take to the bank is that the maximum power on this motor is in the neighborhood of 8200 RPM. I just read this off the graph in the previous post. (The graph could be printed, mark the top of the power hump. Go straight down read the RPM estimating one more digit off the axis at the bottom.) I did some calculations that opened my eyes a bit. I used the maximum and minimum runlines we determined above. I'll call these the oval guys run line (system theories sketch 173 ft) and the road race guys runline (more like my sketch 145 ft. Nothing fancy except converting units. Both runlines are inches from the endboards.
First get the average velocity of the car on your runline.
oval guys
173 foot/4.5 second x 12 inch/foot x 60 second/minute = 27680 in/min
road race guys trying to be oval guys
145 foot/4.5 second x 12 inch/foot x 60 second/minute = 23200 in/min
Now get the tire rollout which is Pi x diameter = 2.2 x 3.14 = 6.908 inch/revolution
divide this into your speeds to get the axle RPM desired
(27680 in/min)/(6.908 inch/revolution) = 4006 rev/min (oval guy)
(23200 in/min)/(6.908 inch/revolution = 3358 rev/min (road race guy)
divide these into you RPM at max power from the dyno graph to see what gear ratio you need
8200RPM/4006RPM = 2.0 Oval guys
8200RPM/3358RPM = 2.44 Road Race Guys
With a 100 tooth spur
100/2.0 = 50 tooth pinion (Oval guys, 173 foot runline)
100/2.44 = 41 tooth pinion (Road Race Guys 145 foot runline)
(feel free to multiply this by a factor 1.05 to 1.10 to allow for a 5% to 10% tire slip)
so that is how I would do it. That is how Fantom software does it. This is a real eye opener for me as both runlines can exist on the same course. I am calculating 9 more pinion teeth by running wide over running tight. The observed difference was on the order of 5-6 teeth. The 173 foot runline is probably not achievable. It is shorter.
Maybe Southwest tour can gives us some numbers from the velodrome and I can apply this method to it.
I'll run through this method with a dyno output from a 4 cell 13.5 run. I have done this previously on my CRC thread.
Here is a link to the Power vs RPM graph of a Novak 13.5 run on a 4 cell pack. Again that RPM where max power occurs is a very reliable piece of data.
http://www.rctech.net/forum/attachme...bx-13-5002.jpg
OK I ran the same two run lines using the 13.5/4 cell graph I get 26 and 31 tooth pinions for the short and long run lines. This was the observed difference in number of teeth between me and some guys with more oval experience running the wide line. Mistery solved.
Last edited by John Stranahan; 04-15-2009 at 10:31 AM.
04-14-2009, 10:42 PM
#26
@John: impressive work, very nice.
I didn't read through all the posts but saw you are unclear about the "initial" starting phase of the LRP speed-controls. Drop me an email if I can be of assistance for your project.
I didn't read through all the posts but saw you are unclear about the "initial" starting phase of the LRP speed-controls. Drop me an email if I can be of assistance for your project.
04-15-2009, 12:42 AM
#28
Reto has answered one occurence when dynoing the 3.5. Active timing is used by the LRP edition speed control. This means the motor timing is changed by the speed control as the motor spins up. The increase in current as the motor runs and possibly the low initial current on the 3.5 is related to this. This effect is less evident on high wind motors like the 17.5
John
John
Last edited by John Stranahan; 04-20-2009 at 10:49 PM.
04-15-2009, 06:44 AM
#29
Flywheel mass
In an ideal world the flywheel mass would not affect peak power. The heavy flywheel would just take longer to spin up. With the hot winds, however, the speed control is using active timing to constantly optimize the motor. This results in possibly the speed control treating each flywheel slightly differently.
I ran the LRP 3.5 yesterday on two flywheels. The Fantom Aluminum and the thinner Fantom steel flywheel with twice the inertia. I am going to suggest strongly to run powerful motors with the aluminum flywheel. When using the steel, the 14 gauge motor leads are probably close to the desolder point. They are very hot through the insulation. In this test 100 amps or close to it are held for a very long 3 seconds. It is just too much. It simulates overgearing.
Happily the power peak occurs later in time with a more powerful motor, so we get plenty of data near the peak. In other words the sampling rate is sufficient with aluminum and poweful motors. So with a 4.5 or 3.5 go with aluminum.
Reto, from LRP electronics, has advised me on settings of the LRP edition speed control to minimize the use of active timing and limit the use of current control. The settings to be used are:
"I'd recommend using power-profile 3 as this is lowest possible active timing and no internal current limiting. So I'd suggest you use these 4 mode settings 2-3-0-0".
I will use these setting on the 3.5 with an Aluminum flywheel and report the results. Next up a 10.5.
In an ideal world the flywheel mass would not affect peak power. The heavy flywheel would just take longer to spin up. With the hot winds, however, the speed control is using active timing to constantly optimize the motor. This results in possibly the speed control treating each flywheel slightly differently.
I ran the LRP 3.5 yesterday on two flywheels. The Fantom Aluminum and the thinner Fantom steel flywheel with twice the inertia. I am going to suggest strongly to run powerful motors with the aluminum flywheel. When using the steel, the 14 gauge motor leads are probably close to the desolder point. They are very hot through the insulation. In this test 100 amps or close to it are held for a very long 3 seconds. It is just too much. It simulates overgearing.
Happily the power peak occurs later in time with a more powerful motor, so we get plenty of data near the peak. In other words the sampling rate is sufficient with aluminum and poweful motors. So with a 4.5 or 3.5 go with aluminum.
Reto, from LRP electronics, has advised me on settings of the LRP edition speed control to minimize the use of active timing and limit the use of current control. The settings to be used are:
"I'd recommend using power-profile 3 as this is lowest possible active timing and no internal current limiting. So I'd suggest you use these 4 mode settings 2-3-0-0".
I will use these setting on the 3.5 with an Aluminum flywheel and report the results. Next up a 10.5.
Last edited by John Stranahan; 04-20-2009 at 10:50 PM.
04-15-2009, 09:40 AM
#30
Hi John,
Great thread. Wouldn't using an ESC that does not have active timing (or boost/hybrid mode as Tekin calls it) be best for the dyno. Seems to me that unless you are using the same speedo with the same settings on the dyno as you are on the track you would get different results. Using something like a GTB would allow you to isolate how the motor by itself performs (which would be useful when comparing same turn motors from different companies). Don't get me wrong, I can see the usefulness of testing different motor/ESC combos, but that just introduces more variables that us less intelligent people will have trouble wrapping our heads around
Great thread. Wouldn't using an ESC that does not have active timing (or boost/hybrid mode as Tekin calls it) be best for the dyno. Seems to me that unless you are using the same speedo with the same settings on the dyno as you are on the track you would get different results. Using something like a GTB would allow you to isolate how the motor by itself performs (which would be useful when comparing same turn motors from different companies). Don't get me wrong, I can see the usefulness of testing different motor/ESC combos, but that just introduces more variables that us less intelligent people will have trouble wrapping our heads around