Using Castle Creations data logs to test motor KV

[LibertyMKiii]

Could you advise what I might be doing wrong? I recently tested a TP Power 4050 motor with a delta wind rated at 3840kv.
On a SkyRC motor analyzer it tested at 3817kv. This falls in line with what I expected as many motors test slightly lower than the rated KV.

When I did a test on the MMX8s connected to 2s LiPo the logs are indicating a much higher KV value. For reference the motor is free spinning in the air. (graph smoothing is disabled) I tried the 1:1 gearing option and No Gearing/Direct drive option and both netted the same results. I also tried with 10 degrees timing and 0 degrees timing, but also came out with the same results.

34,256 rpm / 7.9V = 4,336 KV

Maybe @robert@castle can shed some light on this subject.

 

[SrC]

Yeah something is amiss. At least at first glance. I think I follow you....

The Castle log shows 8.3 volts. What were the volts when tested on the Analyzer.?
I would charge the lipo to identical volts before testing both ways.
But the log appears erroneous in its math.
Do you have an RPM gun? And check it while running up with the ESC against the log reads.
Where did you get 7.9 volts from? Log is 8.3v. Its not the answer, but is an inconsistency nonetheless?
I would not expect the log data to be accurate all the time. But it should be closer than it appears? A software error. Is there some FW update for your ESC. Not familiar with that ESC specifically.
Try repeating tests. How are you throttling the ESC on the bench? with a Servo tester directly or using a Tx/Rx?

My analyzer is usually close also and a drop lower than kv specs say, like you observe. All motors. That is a viable reference. As long as voltage is consistent, Kv reads will be.

 

[LibertyMKiii]

Yeah something is amiss. At least at first glance. I think I follow you....

The Castle log shows 8.3 volts. What were the volts when tested on the Analyzer.?
I would charge the lipo to identical volts before testing both ways.
But the log appears erroneous in its math.
Do you have an RPM gun? And check it while running up with the ESC against the log reads.
Where did you get 7.9 volts from? Log is 8.3v. Its not the answer, but is an inconsistency nonetheless?
I would not expect the log data to be accurate all the time. But it should be closer than it appears? A software error. Is there some FW update for your ESC. Not familiar with that ESC specifically.
Try repeating tests. How are you throttling the ESC on the bench? with a Servo tester directly or using a Tx/Rx?

My analyzer is usually close also and a drop lower than kv specs say, like you observe. All motors. That is a viable reference. As long as voltage is consistent, Kv reads will be.

Even at 8.3V it is 4,127 KV
The 7.9V is taking values during peak rpm. The 8.3V value is in the log before the motor was spinning/pulling the voltage down with the "load" applied.

I had tried the SkyRC analyzer at different voltages from 7.2 up to 8.3V and the KV remained nearly the same. Only the rpms and amps changed as expected with more or less rpms because of the voltage difference.

Throttle is applied via the transmitter and got the same result on 3 different tests. I tried with the timing adjustments also but that made no change to peak rpm logged.

To make this more odd during an actual use with the car the RPM seems correct. My 58 mph speed should have been 17k rpm based on my gearing and it lined up perfectly with the gearing speed calculator. It seems just when used to detect the motor KV when there is no load it is overshooting the rpm reading.

 

[SrC]

Do you have a servo tester for a direct Thr. control at the ESC?
I know what i am saying sounds silly......and maybe off base here. But the radio's Thr. parameters may be skewing the max thr. output somehow.
I know that my HW ESC's gives erroneous reads at times. I know that HW logs are not as robust as Castle's. Just saying. Try clearing and resetting the Castle ESC and its log data. It straightened out my HW bad reads when I did this. Has Castle reached out to you yet?

I must assume you are testing using Sensored mode at the ESC. (Full) Sorry if a stupid question. My HW relies on Full Sensored mode, compared to Hybrid sensored mode at the ESC, settings for accurate reads for the most part. For example, I have 2 possible Sensored type variables. FWIW.

 

[LibertyMKiii]

Do you have a servo tester for a direct Thr. control at the ESC?
I know what i am saying sounds silly......and maybe off base here. But the radio's Thr. parameters may be skewing the max thr. output somehow.
I know that my HW ESC's gives erroneous reads at times. I know that HW logs are not as robust as Castle's. Just saying. Try clearing and resetting the Castle ESC and its log data. It straightened out my HW bad reads when I did this. Has Castle reached out to you yet?

I must assume you are testing using Sensored mode at the ESC. Sorry if a stupid question. My HW relies on Full Sensored mode, compared to Hybrid sensored mode ESC, setting for accurate reads for the most part for example. I have 2 possible Sensored type variables.FWIW.

I do not have one of those, but calibrated the radio and ensured it does not go beyond 100% (radio is capable of going to 120%)
I did notice that with my futaba full throttle was 1.94ms and this one is 2.012ms so you may be on to something there.

I suppose I will email Castle creations, I was hoping Robert would chime in to educate us all, but he must be busy. Either way if I find some answers I will share back in this thread.

 

[SrC]

You got me curious with this Thread. I like a good challenge every once in a while. ?
Good luck.

 

[robert@castle]

I do not have one of those, but calibrated the radio and ensured it does not go beyond 100% (radio is capable of going to 120%)
I did notice that with my futaba full throttle was 1.94ms and this one is 2.012ms so you may be on to something there.

I suppose I will email Castle creations, I was hoping Robert would chime in to educate us all, but he must be busy. Either way if I find some answers I will share back in this thread.

Power can't go over 100%. If you look in your log the "Power Out" value is the percentage of power put to the motor. If the radio goes beyond the calibrated point, the ESC just outputs 100% power.

There are lots of different variables in motors that can affect the KV. Even 2 motors that should be exactly the same can have a few hundred KV difference between them. We test our motors to have + or - 5% from the rated KV. That reading shows about 12% above the rated KV, so it would fail our testing, but I don't know what range TP considers acceptable. By the way that TP sells dozens of different winds per can size, I'm willing to bet they just publish the theoretical numbers from their motor design software and don't correct their charts for real world performance.

I'm trying to work out why our logs would be reading so much higher than the SkyRC. The TP motor might have too high of an idle current(We call it IO Current) for the SkyRC tester's internal ESC since that it intended for 1/10th scale motors and can't put enough current in the motor to hit 100% speed. Looking at your log, it looks like the IO of that motor is 15 amps. Our 2028 1100kv motor has the highest IO of any of our motors and we test that it is under 5 amps at 7.4v. The TP motors we have looked at have a lot of current draw at lower RPM because they take a lot of current to overcome the strength of the magnets and they just generally have absurdly high currents in general. We have seen peeks of over 800 amps for a few microseconds while just doing our startup beeps on 8s.

If you have other larger motors, I'd be curious to see how KV results compare on the SkyRC vs Castle Logs. It might also be worth testing smaller, easier to run motors on both to compare the results and see if the values are more comparable.

 

[LibertyMKiii]

Power can't go over 100%. If you look in your log the "Power Out" value is the percentage of power put to the motor. If the radio goes beyond the calibrated point, the ESC just outputs 100% power.

There are lots of different variables in motors that can affect the KV. Even 2 motors that should be exactly the same can have a few hundred KV difference between them. We test our motors to have + or - 5% from the rated KV. That reading shows about 12% above the rated KV, so it would fail our testing, but I don't know what range TP considers acceptable. By the way that TP sells dozens of different winds per can size, I'm willing to bet they just publish the theoretical numbers from their motor design software and don't correct their charts for real world performance.

I'm trying to work out why our logs would be reading so much higher than the SkyRC. The TP motor might have too high of an idle current(We call it IO Current) for the SkyRC tester's internal ESC since that it intended for 1/10th scale motors and can't put enough current in the motor to hit 100% speed. Looking at your log, it looks like the IO of that motor is 15 amps. Our 2028 1100kv motor has the highest IO of any of our motors and we test that it is under 5 amps at 7.4v. The TP motors we have looked at have a lot of current draw at lower RPM because they take a lot of current to overcome the strength of the magnets and they just generally have absurdly high currents in general. We have seen peeks of over 800 amps for a few microseconds while just doing our startup beeps on 8s.

If you have other larger motors, I'd be curious to see how KV results compare on the SkyRC vs Castle Logs. It might also be worth testing smaller, easier to run motors on both to compare the results and see if the values are more comparable.

Thanks for the responce!

The RPMs seem to be spot on in the speed runs, which is really what I am most concerned about.
I sold my larger motors. I used to have a Leopard 5692...

 

[jondilly1974]

Give me instructions and I’d be glad to test some motors I have on hand.

TP 4050 3200kv
TP 4050 2300kv
Castle 1717 1650kv
Spektrum 1250kv
Leopard 4274 2200kv
Hobbywing 4268 2600kv
Arrma 2050kv
Arrma 2400kv

 

[LibertyMKiii]

Give me instructions and I’d be glad to test some motors I have on hand.

TP 4050 3200kv
TP 4050 2300kv
Castle 1717 1650kv
Spektrum 1250kv
Leopard 4274 2200kv
Hobbywing 4268 2600kv
Arrma 2050kv
Arrma 2400kv

Most similar to my test would be your TP 4050 3200kv and then compare that to the castle 1717.
Hook it up run the motor with no pinion full throttle for about 5-10 seconds and pull the log to see what the RPMS show.
Make sure logs are set to no smoothing, 1:1 gearing, 4 poles.

 

[LibertyMKiii]

I got a response from Thomas lead support rep at Castle. He indicated that the ESC has dynamic timing unless you run sensored mode. I suppose 10 to 15 degrees timing is capable of increasing in my use case 500+ rpms/kv. Many thing I had seen talked about low end vs high rpm torque, but maybe peak RPMs is effected also?

Quote from Thomas's response:
"When you have the timing set to Lowest (0) the ESC can still add timing. Our timing is dynamic and it changes based on the feedback from the motor. You’re setting a range the ESC is allowed to work within. So at Lowest (0) the ESC runs from ~0° to 10° of timing; at Normal (10) it’s around 5° - 15° of timing. So it’s possible for the ESC to run the same actual timing when it’s set at 0 and 10.
The only want to have a true 0° of timing is to be running in sensored only mode; this has no dynamic timing and is fixed. But it would need to be a sensored capable motor to do so."

 

[robert@castle]

I got a response from Thomas lead support rep at Castle. He indicated that the ESC has dynamic timing unless you run sensored mode. I suppose 10 to 15 degrees timing is capable of increasing in my use case 500+ rpms/kv. Many thing I had seen talked about low end vs high rpm torque, but maybe peak RPMs is effected also?

Quote from Thomas's response:
"When you have the timing set to Lowest (0) the ESC can still add timing. Our timing is dynamic and it changes based on the feedback from the motor. You’re setting a range the ESC is allowed to work within. So at Lowest (0) the ESC runs from ~0° to 10° of timing; at Normal (10) it’s around 5° - 15° of timing. So it’s possible for the ESC to run the same actual timing when it’s set at 0 and 10.
The only want to have a true 0° of timing is to be running in sensored only mode; this has no dynamic timing and is fixed. But it would need to be a sensored capable motor to do so."

I hadn't really thought about motor timing. The SkyRC tester may have zero timing if the motor is running sensorless and we may push it further since we are used to running sensorless and the SkyRC tester is designed to test sensored motors..

To add on to what Thomas said, sensored only mode should only ever be used with 2 pole 540 sized motors. Only use sensorless or smartsense if using a 4 pole motor or a larger motor than a normal 1/10th scale motor(so all of our motors). 4 pole motors do not get a clean reading from the sensors above 5000 RPM, so smartsense changes from a sensored start to a more efficient sensorless drive when motor is above a few thousand RPM.
It isn't really relevant to this thread but I want to post this information incase it helps someone avoid issues someday.

 

[TL01magic]

Could you advise what I might be doing wrong? I recently tested a TP Power 4050 motor with a delta wind rated at 3840kv.
On a SkyRC motor analyzer it tested at 3817kv. This falls in line with what I expected as many motors test slightly lower than the rated KV.

When I did a test on the MMX8s connected to 2s LiPo the logs are indicating a much higher KV value. For reference the motor is free spinning in the air. (graph smoothing is disabled) I tried the 1:1 gearing option and No Gearing/Direct drive option and both netted the same results. I also tried with 10 degrees timing and 0 degrees timing, but also came out with the same results.

View attachment 109986

34,256 rpm / 7.9V = 4,336 KV

Maybe @robert@castle can shed some light on this subject.

Am in the middle of looking for "answers" and see you've also had / noticed too high a KV value for motor being tested.

a castle 1717 1650kv measured 2325kv ?

issue am having is change in how torque control is working; before using b-link torque control worked as expected...after it works like power control.

also b-link shows torque control disabled; when I enabled and send new settings I reconnect and it shows disabled again?!

since removed b-link, and redid motor test using usb castle link and to enable torque control (and adjustment via Aux.) and that's when I noticed the KV difference. Torque control still feels like power control.

"glad" to see it's not just mine


Remarkable my over KV reading is also 500 rpm range.