Limitless Speed Runners Motor KV selection

[LibertyMKiii]

I figure now that we have a speedrun section of the site it would be good to post a thread about motor KV selection. Motor selection is a mystery to many and although I don't have it all figured out (I am not the experienced expert) I believe I have a pretty good understanding.

First I want to say that Castle ESCs like the XLX and MMX8s are really the only 2 good options. This is because you will need high amps and data logs to make informed decisions on adjusting your setup.

Running a Typhon as a buggy you would likely be limited to a 6s setup or running a Limitless you could run up to 8s. First things first you must know what power level you plan to run and how large of battery you plan to run. As of today the top 3 batteries for speed running would be the Turnigy Graphene/Panther, CNHL 70c, and SMC True Spec Extreme V2. You will need the best of the best LiPo due to their amperage delivery and reduced voltage sag. On the topic of voltage sag higher mAh values will help. This is why you will see many guys run 2 packs in parallel. A parallel 6,000 mAh setup would effectively give you 12Ah!

Next to begin the discussion on KV of motors you must know that most if not all motors are hand wound. This means no 2 motors are the same. Some are power hogs and some are highly efficient. I recently purchased the SkyRC motor analyzer and if you are serious about speed running this would be a wise purchase. The 2 big things this tool will give you is actual KV and how efficient in amperage draw your motor is.

This first example is a good TP4070 with the D-wind 2080KV
This is a good motor that my friend has done 159 mph with on a 1/8 Mugen. You can see here the amperage draw is “efficient” at 3.6A

The next was another TP4070 with the D-wind 2080kv it would be an ok motor, but it will draw more amps (around 2x the other motor) and run hot! Also notice the KV difference of these 2 motors both sold as 2080KV. While higher KV would pull slightly more amps we wouldn't expect it to be this much higher. Ideally TP Power should have taken this motor back as "bad".

(note I had tried charging my 2s testing LiPo fully to 8.4V and same results came out.)

Back on topic to selecting a motor. The Magic is shooting for your motor to run around 51-55k rpm knowing it wont quite get there. Reviewing logs of people who have truly gone fast, their motor RPMS are in the high 45-49k rpm range to as much as 52krpms.
KV multiplied by the voltage = the motor RPM

The topic of voltage sag here and data logs is key! Without testing your motor, gearing, and battery... the best thing we can do is use typical numbers.
Typical value of a 6s Lipo drops during a speed run to around 19 volts. On 8s it drops somewhere to around 26 volts. These numbers are for 5000 to 6000 mAh packs as described above as "best LiPos" I have seen far worse sag on low end LiPos.

First we can look at a 6s setup. If I want 51k RPM I will take 51,000 divided by 19 volts and come up with 2684 KV. This is why many people like the 2700KV motor for 6s use!
The major key with this setup is that you cannot over gear it. The higher KV motor like this will pull some serious amperage. I would recommend starting conservative and use you data logs as you make adjustments. You would be keeping an eye on RPMs, voltage sag, amperage, and ESC temps. With a Typhon/Limitless differential running GRP tires you could expect as much as 154 mph with 27 pinion and 34t spur gear. (assuming your car gets there safely and has the aerodynamics to do it) Given that you should not go out your first time and do 150mph I highly recommend gearing way down on this setup and work up to that gearing.

There is a mobile app called "RC Car GearCalc" that I use. this one does not have you pick 6s 8s 5s etc. Instead it allows you to input a specific voltage so that you can use it to properly adjust for voltage sag.

Next we will look at 8s!
For this again our around 34 volts will sag to 26 volts. (this could differ depending on the batteries you choose.)
Again shooting for 51,000 rpms divided by 26 volts provides a motor of 1961 KV. This would be why speed runners recommend 2000 KV up to 2200 for 8s use. (This is why I purchased the 2080kv TP 4070 motor)
Also why I am using my BLX 2050KV on 8s! Which that motor is actually closer to 1900kv

Most scenarios we find that staying under 1:1 gearing is ideal for 40mm motors due to the gearing torque loss. Also keep in mind that the gearing creates the amperage load for the ESC and if you over gear you will not reach the desired RPM range. Always start low and work your way up. Read the data logs and adjust accordingly. At a certain point the car will not go faster but the motor heat and ESC heat will go way up. At that point you have reached the limits of the setup. You have 2 options. Try a different motor or in my mind the better option is to improve the aerodynamics.

People are now playing with large 1/5 motors which are around 1000 - 1350 kv. I am not certain yet how all of this will play out as their power band will be different and requires some gearing way beyond 1:1
It is also worth noting that TP Power 1/5 motors are 6 poles and the more poles of the motor will provide more timing stress on the motor as it has less time between each pole to do its "job".

I hope this is helpful and again I am not a proven expert on motor selection, but I believe this is at least a good ballpark for you!

 

[SrC]

The Sky RC motor analyzer doesn't lie. My Arrma 2050 motors vary, and consistently around 1900 kv like you said. I have 2 same analyzers. They can vary on readings all else being constant but are close enough I also find that no 2 same motors are built the same like you noted. Due to winding variations. I don't belive that most motors are hand wound. . I see potential with my 2250 Hw sensored 4 pole on my lim. Agree that Castle is the only way to go regarding ESC'S and serious speed running.

 

[LibertyMKiii]

The Sky RC motor analyzer doesn't lie. My Arrma 2050 motors vary, and consistently around 1900 kv like you said. I have 2 same analyzers. They can vary on readings all else being constant but are close enough I also find that no 2 same motors are built the same like you noted. Due to winding variations. I don't belive that most motors are hand wound. . I see potential with my 2250 Hw sensored 4 pole on my lim. Agree that Castle is the only way to go regarding ESC'S and serious speed running.

I am not sure about HW motors, but I know that commonly used speed run motors are hand wound. Leopard, Neu, Cordova, and TP Power motors all advertise that they are hand wound.

Another motor test to add to this thread. This is a Leopard 5692 manufacturers claimed KV of 1340 actual is about 1300. Come on Castle with the 12s on-road ESC!

 

[jondilly1974]

I am not sure about HW motors, but I know that commonly used speed run motors are hand wound. Leopard, Neu, Cordova, and TP Power motors all advertise that they are hand wound.

Another motor test to add to this thread. This is a Leopard 5692 manufacturers claimed KV of 1340 actual is about 1300. Come on Castle with the 12s on-road ESC!
View attachment 86935

Just run two 6S in series ????‍

 

[SrC]

Depends on the can size and model, but Leopard motors are not all handwound. Correct me if I am wrong here. I agree that tighter and neatly hand wound motors (stators) are the best. But like machine winds, they can also vary from can to can. There are many variables that will cause KV inconsistencies from same model motors. Some motors even test out to higher than advertised, but I find that more rare. They are usually lower than spec.
BTW, I find that my Analyzer is more consistent for comparative testing when using a regulated power supply, versus a lipo.

 

[LibertyMKiii]

BTW, I find that my Analyzer is more consistent for comparative testing when using a regulated power supply, versus a lipo.

This is good to know, what sort of differences are you seeing? I assume the KV varies, but current requirements of the motor remains the same?

 

[SrC]

If you can keep the analyzer power supply constant, you will find test results more accurate from motor to motor. The Amp draw reads will vary too much with differing volts. That will throw off results. In a perfect world I always feel 5.5-6.0 amps is best and most efficient. But I don't see that in recent motors of late. If the mechanical timing is adjustable at the motor can you can up the amp draw to the range stated above. But most motors are fixed mechanical timing. Just my observations.
Been using this analyzer for several years now. Im on my third one.BTW, my10yod reversed polarity on my last one. It friedn split second. Be careful with reversing polarity to this analyzer.

 

[LibertyMKiii]

If you can keep the analyzer power supply constant, you will find test results more accurate from motor to motor. The Amp draw reads will vary too much with differing volts. That will throw off results. In a perfect world I always feel 5.5-6.0 amps is best and most efficient. But I don't see that in recent motors of late. If the mechanical timing is adjustable at the motor can you can up the amp draw to the range stated above. But most motors are fixed mechanical timing. Just my observations.

Could you explain what you see in the real world when running a motor that on the analyzer ran in the 5.5-6.0 range versus one that shows up like mine in the 2.8 - 3.8 range?

When I tested the 2 identical TP4070 motors and had one that tested at 3.6A versus the other at 6.5A. The 6.5A motor ran very hot and seemed to be doing the same work less efficiently (consuming more power). With the limitations of our ESCs and LiPos my understanding was that a lower amp motor would be better all things being equal?

 

[SrC]

Again if the voltage is kept constant there shouldntbea different amp draw from two identical motors. I believ one of those motors is not well built. (if the timing is Fixed)

 

[LibertyMKiii]

Again if the voltage is kept constant there shouldntbea different amp draw from two identical motors. I believ one of those motors is not well built. (if the timing is Fixed)

I agree, and yes the timing is fixed on those motors.
I had tested that 6.5 amp motor on the SkyRC analyzer from 7.4v up to 8.4 volt and it remained at 6.5Amps. I would have expected it to raise a little bit.

 

[SrC]

6.5 amps seems too high. But every motor tends to be different. (same can) RPM will vary with volts, that alters KV, The amp draw is the resulting from the motors preset timing/ the quality of the winding from pole to pole.

 

[fpvmiller]

@LibertyMKiii Have you ever gotten a hold of any hobbystar motors to see what they do in real life?

 

[LibertyMKiii]

@LibertyMKiii Have you ever gotten a hold of any hobbystar motors to see what they do in real life?

I have not tried them myself.

 

[Arrmabitmad]

So what is the benefit of running batteries in parallel vs series... Is more amps better on a speed run or more volts...

I ask this as im currently building a limitless, a have a max6 combo with a 4985 1650kv motor just wondering to wire the esc for double the amps or volts...

 

[LibertyMKiii]

So what is the benefit of running batteries in parallel vs series... Is more amps better on a speed run or more volts...

I ask this as im currently building a limitless, a have a max6 combo with a 4985 1650kv motor just wondering to wire the esc for double the amps or volts...

ESCs choke on Amps and batteries only can deliver so much. Higher volts and lower amps is always the best recipe.

With that said the reason some use parallel LiPos is to fight voltage drop. As noted in my original post your 8s 34v can drop to 26v or lower. Parallel 8s would reduce that however there is a trade off of a significant weight addition.
Many chargers and battery manufacturers don't offer 8s packs so you would potentially need 4x 4s packs?

Many speed runners instead have found the SMC 6400 mAh pack is a good compromise in the middle.

 

[jondilly1974]

To understand motors, you need to know they output in watts. Watts are amps x volts. If the voltage sags, as it always does, then amperage has to increase to make up the same wattage output. Increased amps creates more heat and risk of damaging fets. The ESC will shut down to prevent this damage. So running parallel packs keeps voltage up as much as possible to keeps amps down as much as possible.

 

[LibertyMKiii]

When going really fast an important factor is torque to push through that air resistance. Motor torque has to do with the reaction between the stator and the windings. This torque is applied from one pole to the next. You can think of this as turning a bolt with a wrench in 1/4 turns (with a 4 pole motor). Imagine a 40mm motor is using a standard wrench and a 56mm or 58mm motor is using a long breaker bar to turn that bolt. With that visual you could imagine how the larger diameter motor has significantly more torque. It is worth noting that the length of the can also plays a part. A 40mm motor could have the same torque of a 56mm motor if the 56mm is shorter and the 40mm is longer they may equal out. The torque is multiplied by the area which is the length.

Slightly switching topics but still on the subject of torque:
I keep thinking about the low KV with gearing into the negative, versus a higher KV motor with a gearing that wouldn't restrict torque as much. Rather than buy a motor and do testing which would take months of work. I decided to stop being so lazy and crunch some numbers.

In this test scenario we have two motors one is a TP Power 4070 in a 4150 KV wind the other is a 2080 KV wind.
The scenario both will be on the same voltage and amperage. The only difference is the gearing Final drive (FD) to reach the same theoretical speed. Lets call that speed 130mph.

I have highlighted torque value for the 5.0 final drive for the 4150KV motor to be at this theoretical speed, also the 2.5 Final drive for the 2080KV to reach this speed. Kt is the brushless motor torque value, which is multiplied by the amperage. I chose 400 amps as a base value to share between the 2 motors.

These torque values are less that 1% from each other telling me there is no difference in these two approaches.

Motor	KV	KT mNm/A	amps	Kt at 400a	2.5 FD	3.7 FD	5.0 FD
4070​	4150​	2.301​	400​	920.4​	2301​	3405.48​	4602​
4070​	2080​	4.591​	400​	1836.4​	4591​	6794.68​	9182​

The key difference would be if you get outside of the testing scenario, the 2080kv can run on higher voltage and could potentially use the lower gearing FD of 3.7 or 5.0 depending on what the speed goal was, and ultimately go significantly faster as noted by the higher torque values. This would also come along with higher wattage/power outputs of the motor.

I am not saying the 2080KV is the motor to go with, but simply wanted two motors that had technical specs available, and where one was double the KV of the other for a good data comparison.

The only thing I cannot determine in this would be if either motor produces this torque value better at a given RPM as motors do have a torque curve. My suspicion is that the large 1/5 scale motors do very well in the lower rpms and the 40mm motors like the 30k to 50k range.

 

[rotauq]

Interesting topic. I could be wrong, but don't think these are valid examples.
2D 4070 (4150kv) is speced for 625A @ 12.1V
4D 4070 (2080kv) is speced for 300A @ 24V
I am pretty sure the kt is only constant when run within spec. And these cannot be compared on fixed voltage and amps. You could fix the required rpm for the FD (and thus fix the voltage per motor due to Kv)

Unlike TP, Lehner motors do offer a ridiculous amount of data @ http://www.lehner-motoren.com/calc/diagramm_en.php
I haven't had the time to deep dive into the math (yet) and analyze it all, but i started a little excel to compare different winding of the Lehner 2260 (closest to the TP 4070). I think expanding this with more examples will give the insight we are looking for:

ps. As for torque curves, there is a difference between Y (low range torque) and D (high range torque) winds. Larger motors due to higher power and torque can't physically handle the same high rpm's so the torque band is mostly limited because of that.

 

[LibertyMKiii]

Interesting topic. I could be wrong, but don't think these are valid examples.
2D 4070 (4150kv) is speced for 625A @ 12.1V
4D 4070 (2080kv) is speced for 300A @ 24V
I am pretty sure the kt is only constant when run within spec. And these cannot be compared on fixed voltage and amps. You could fix the required rpm for the FD (and thus fix the voltage per motor due to Kv)

Unlike TP, Lehner motors do offer a ridiculous amount of data @ http://www.lehner-motoren.com/calc/diagramm_en.php
I haven't had the time to deep dive into the math (yet) and analyze it all, but i started a little excel to compare different winding of the Lehner 2260 (closest to the TP 4070). I think expanding this with more examples will give the insight we are looking for:

View attachment 119591

ps. As for torque curves, there is a difference between Y (low range torque) and D (high range torque) winds. Larger motors due to higher power and torque can't physically handle the same high rpm's so the torque band is mostly limited because of that.

I don't claim to be an expert on brushless motors, and certainly could have made some mistakes. This was my first shot at playing with the numbers.
In this case TP Power did give specs with the same voltage 7.5v between the two motors. With those values I was able to come up with the Kt value.

I can tell you from first hand experience that you can drive a motor well beyond its spec rating for amperage with the right gearing. Also each company appears to rate their motors differently. When I was doing my 2s runs on the Limitless voltage was sagging down around 6v and amps went way above 400A with a motor rated for only 380A. Similarly I have taken the Castle 1515 motor over 400 amps. According to NeuMotors that 1515 1Y motor is rated for 73 amps

 

[rotauq]

I am no expert either, but data analyst for a living who studied mechanical engineering a long long time ago...

I am just glad, more people are searching for the same answers i am looking for. To quote a part of my favorite song:
"You start with assumption,
That a million people are smart,
Smarter than 1"

From my own experience in overvolting, constants would shift some whiles really overvolted. Even with way lower gearing i've never managed to measure the same kv as reached with spec voltage. It was always about more speed for me, so i don't think i actually matched the same FD, which could make this a wrong assumption/conclusion.

Undervolting (and sagging) does increase ampdraw to reach the same power output. Gearing and kv increase the FD and can easily exceed max P and current. Running over/under spec is possible is not a discussion. I just doubted comparing a 12v or 24v rated motor, which is a 100% gap.

I've been playing around with the lehner data last few evenings. Unfortunately at some data the amps range stops quite early. But as that data goes up to ~110k rpm it could be considered beyond what other brands consider as there max voltage in the specs. (With NeuMoters as exception apparently)
General rule is they spec max power and rpm on physical aspects. The winding dictates the Kv and therefor the max voltage on max rpm/Kv. And when max voltage has been determined, Max power / voltage makes the max amps.

Some statements from lehner about their motors and data:
-With the Carbon sleeve option for the Rotor you can exceed the RPM in the user manual up to 30%. This option includes balancing of the rotor.
-The rpm per Volt values we provide are, in opposite to those of our competitors, measured under load. You must calculate the value without load about 10% higher.

ps. I don't own any Lehner motors and I am not fan in particular. They are ridiculously expensive. But they are German, which means whatever they say about their motors and data is based on "German gründlichkeit" and not Chinese marketing.

My initial findings on the data does indeed show the Kt hardly shifts on voltage. Even 3250Kv @33V, which i would consider overvolted:

Efficiency does change on low amps but tend to get back on the same-ish track on higher amps. So on low amps more input power is required to keep the Kt consistent and goes up into wasted heat (less time to max T). In these examples low kv would be slightly more efficient than high kv on same amp&voltage. Assuming gearing efficiency is no factor, a low kv motor wins from a high kv motor by some margin...