Review – XM20A ESC – F396 MCU

The DYS XM20A is a relatively new speed controller for multirotors. The processor used by XM20A is the Silabs F396 which can run at higher speed than the previous generation ESC’s of F330 (for example the FVT/ZEUS Littlebee 20A or RG20A).

Many thanks to 65Drones.com and DYS for sending us the XM20A ESC for this review!

Why XM20A, or F39X MCU?

The most interesting difference between the XM20A and the Littlebee’s is the processor (MCU) used.

The F396 (XM series ESC) runs at 50MHz while the F330 (LittleBee) runs at 25MHz. One limitation of the F330 MCU is, when your motor is spinning pass certain speed, it will start to struggle to keep up and desync. To be a bit more specific, the F330’s eRPM limit is at around 200K, while the F39X is 360K (eRPM is the product of input voltage, magnetic phases (poles/2), motor KV and efficiency, source). For a 2000KV motor on 6S for example, the motor eRPM can go up to 240K where you might see motor/ESC de-sync.

Since the F396 is twice as fast as the F330, it also has the ability to process the flight controller commands twice as fast as the F330 which ultimately leads to better and more stable control.

When talking to Justin Nishida, he added:

When using Raceflight and Multishot with the xm20 and say even a naze32 (rev5) you can run 4k/4k (looptime / ESC refresh rate) without having any desync issues. The f330 can’t properly handle refresh rates above 2.6k so you are limited to 4k/2.6k at best and even then you are pushing the ESC to its limit, so most just run 4k/2k to stay on the safe side. Many people that have flashed their littlebee 20A with multishot (or other older generation ESC) aren’t even aware that their ESC’s aren’t processing as fast as the xm20 with multishot.

This is probably why, FVT Littlebee is trying to catch up, and released a new version just recently with the better F396 MCU – the Littlebee 20A Pro.

Specification and Versions

XM20A is Rated at 20A constant and 25A burst. This speed controller supposedly has better performance on high kV motors that runs on high eRPM.

  • Input voltage: 3-4s Lipo
  • BEC: NO
  • Board dimensions: 24x14x4.4mm
  • Out of the box weight: 4g excluding cables
  • Default Booloader and Firmware: BLHeli 14.X

Note that the XM20A has more than 2 versions available on the market currently: the V1 doesn’t come with motor wires soldered and V2 does. However there is no hardware difference between V1 and V2 apart from this.

The V1 we received also comes with a plastic protective case. DYS ensured us the case should not cause any worse overheat issue, because of the holes on the case (as long as you have good airflow).

xm20a-esc-dys-plastic-case-protective-inside

xm20a-esc-dys-v1-plastic-case-open xm20a-esc-dys-plastic case closed

xm20a-esc-dys-v1-f396-front-processor xm20a-esc-dys-v1-f396-back-fet

The one we received from 65Drones are probably pre V1: different label, doesn’t have “stick out” solder pads, and the motor wires are pre-soldered on. However apart from the lable and wires, they all appear to have the same hardware.

xm20a-esc-dys-pre-V1-heatshrink

This is how XM20A ESC looks like without heatshrink (naked).

xm20a-esc-dys-pre-v1-front-naked xm20a-esc-dys-pre-v1-back-naked

Naked XM20A (yellow signal wire) comparing to the littlebee20A (white signal wire) side by side.

xm20a-esc-dys-compare-littlebee-20a-naked xm20a-esc-dys-compare-little-20a-naked-top

Performance Testing By Konrad

This test was carried out by Konrad.

We are mainly comparing thrust results between the FVT Littlebee 20A ESC and DYS XM20A ESC. We tested Cobra 2204 2300KV motor as well as the DYS1306 4000kV on Diatone 3045 (Biblades and Triblades) and some propellers cut down to 3” using 3S lipo.

Both ESC’s were using latest BLHeli 14.4 with identical settings.

A 3-position switch was used to send 0%-50%-100% throttle signal to the ESC, to ensure consistent throttle signal throughout the test.

BLHeli Settings and configuration

XM20A:

BLHeliSuite-Setup_XM20A-esc-reflashed-default-settings

LB20:

BLHeliSuite-Setup_FVT-littlebee-20a-esc-settings-damped-light

Cobra 2204 2300KV Motor

Testing Setup:

  • 4S 1300mah 40C SLS
  • DIY thrust stand – kitchen scale and RAYTRONIC WM-100
Props   Throttle % Thrust g Peak Current A Power W Efficiency g/W
Gemfan 5045 (non bullnose) Littlebee 20A 50 310 3.81 63 4.9
  100 922 18.40 289 3.1
XM20A 50 292 3.60 59.4 4.9
  100 915 17.93 285 3.2

DYS 1306 4000kV Motor

Testing Setup:

  • 3S 1800mah 40C SLS
  • All props are “bullnose”, including the cut-down props
Props % Peak Thrust – g Peak Current – A Max Power – W Efficiency – g/W
Diatone 3045×2 LB20A 50 114 2.78 34.7 3.3
100 301 14.02 167.7 1.8
XM20A 50 120 2.98 37.4 3.2
100 297 13.18 159.4 1.9
DAL 3045 triblades LB20A 50 130 3.09 38.4 3.4
100 324 13.29 159.1 2
XM20A 128 2.88 35.5 3.6
313 12.88 150.3 2.1
3030 (6030 cut down) LB20A 50 97 1.95 24.4 4
100 297 9.07 111.2 2.7
XM20A 50 102 2.06 25.7 3.9
100 284 9.86 107.6 2.6
3040 (5040 cut down) LB20A 50 104 2.16 27 3.9
100 292 9.89 119.5 2.4
XM20A 50 104 2.16 26.9 3.9
100 295 10.27 123.3 2.4

Disclaimer: Please note that the thrust reading should be taken purely for comparison between the ESC’s, than actual power of this motor and ESC combos. This is a kitchen scale after all.

Both ESC’s performed good enough to push this 2300kv motor to its thrust limits. XM20A is marketed as perfect for higher KV motors, where the littlebees was said to have sync issues in some specific configurations (however we didn’t have any problems with our Littlebees running 4000kv motor on 3S lipo).

I noted small difference in performance on mid-throttle on all tests – in favour for the littlebee.

I repeated the tests 3 times on fully charged lipos for each ESC, the results were within a margin of error.

I also recorded a Damped Light test. Both ESCs seem to have very similar braking power:

You will see a slight stutter on motor startup on the littlebee, but it rotates smoothly on the rest of throttle range. I fly those motors with LB20 with motor timing set to med-high and that solves the issue.

Performance Testing by Oscar

I also run my own tests as well just to make this review a bit more reliable. Konrad and I were using 2 completely different thrust stands, different XM20A ESC’s, even different KV motors, to avoid possible duplicate test results.

However we received a similar conclusion from the data: the littlebee 20A gave slightly more thrust, and also run a bit more efficiently, for most of the propellers we tested.

EMAX RS2205 2300KV Motor

Testing Setup:

Props % Peak Thrust – g Peak Current – A Max Power – W Efficiency – g/W
DYS 4045BN LB20A 50 262 4.6 74.5 3.52
  100 630 17.3 280.3 2.25
XM20A 50 252 4.7 76.1 3.31
  100 619 18.1 293.2 2.11
GF 5030 LB20A 50 300 3.8 61.6 4.87
100 790 15.5 251.1 3.15
XM20A 50 299 4.0 64.8 4.61
100 779 15.4 249.5 3.12
GF 5045 LB20A 50 368 5.6 90.7 4.06
  100 875 21.8 353.2 2.48
XM20A 50 339 5.3 85.9 3.95
100 892 22.9 371.0 2.40
GF 5045 Tri LB20A 50 405 6.2 100.4 4.03
  100 1004 26.3 426.1 2.36
  XM20A 50 395 6.6 106.9 3.69
  100 997 27.3 442.3 2.25
HQ5040 Tri LB20A 50 402 5.8 94.0 4.28
100 975 24.8 401.8 2.43
XM20A 50 389 6.0 97.2 4.00
100 980 24.6 398.5 2.46

EMAX RS2205 2600KV Motor

Battery: 4S 1800mAh (16.2V – 16.6V)

Props % Peak Thrust – g Peak Current – A Max Power – W Efficiency – g/W
DYS 4045BN LB20A 50 299 6.6 106.9 2.80
  100 720 24.1 390.4 1.84
XM20A 50 292 6.9 111.8 2.61
  100 712 25.1 406.6 1.75
GF 5030 LB20A 50 395 5.8 93.2 4.24
100 1008 22.5 364.5 2.77
XM20A 50 387 5.9 94.8 4.08
100 1065 24.1 390.4 2.73

Conclusion

As discussed at the beginning of this review, theoretically, the advantage of using XM20A is the capability of running higher FC looptime and ESC refresh rates more effectively (i.e. raceflight and multishot, resulting in better flight characteristics), as well as handling higher eRPM. We still have to test this in real flights and report back.

Not everyone needs to upgrade to the XM20A though. For example someone who is happy with 2K looptime with oneshot, on some 2300KV motors running on 4S, might be better off to just stay with the good old Littlebee :)

Credit

Many Thanks to Konrad for the helps in testing the XM20A ESC.

13 thoughts on “Review – XM20A ESC – F396 MCU

  1. Henry

    Can little bee 20a pro’s with a max burst current of 30a handle the higher current draw using 5 inch propellers with the rs2205 2600kv?

    Reply
    1. Oscar Post author

      which 5 inch prop? personally i think it’s a bit pushing it. might work with 5030/5040, but i think it will start to struggle with anything bigger… it will be safer to go with 30A ESC’s

      Reply
  2. Stefano Pedersoli

    Hi Oscar, i have a combination with naze32 rev5 + this escs and cobra 2204…when i calibrate them, and testing with Master slider on cleanflight, they are perfect.. But when i use Taranis, any motors spin up slowly or faster then other.. You have a soluction? My naze have barometer and magnetometer

    Reply
  3. Ronnie

    Thankt for a great review!

    Question:
    The XM20A has PWM=18 khz and a Cobra motor 2204 has PWM=8 khz. They do work together, but how can they do that when there is a PWM speed mismatch?

    Reply
  4. Patrick Dalton

    Hello,

    I have these same DYS XM20A esc’s on my most recent build. I’m having a problem with 4s batteries, where when I punch the throttle, the quad will just stop, and fall from the sky. Does this mean I’m drawing too much current, and the ESC’s can’t handle it? This has happened 2 times. I’m afraid to fly with 4s now. 3s batteries havn’t done this to me yet.

    My build:
    Robocat 270
    Naze32 Rev6
    DYS XM20A esc’s
    DYS SE2205 2300kv Race edition motors.
    DYS 5x4x3 props.

    Any advice would be grateful.
    Thanks!
    Pat Dalton

    Reply
    1. Oscar Post author

      could be anything… failsafe? receiver shutdown because power issue? loose cable? FC shutdown? ESC shutdown? if you have blackbox it might be easier to troubleshoot, or somehow mount your quad on the ground so you can witness what’s going on with the electronis

      Reply
  5. Mazaalite

    Hi, Oscar

    I have a broblem with my quad.
    Yaw “shakes” when turning either way.

    My setup is:
    Qav-r 220 frame
    Dys xm20a esc
    Dys se2205 2300kv motors
    Naze32 rev6 fc betaflight

    I have done all calibrations. And changed yaw jump prevention limit to 400, helped a little. But nothing else seem to help. My yaw pid’s are 1,5 / 0,035 / 10. I have played with the pid’s but this is closest that i have got it. Do you have any default settings that i could try to start from? This problem is driving me crazy! Or does the xm20a esc’s require any special settings? I have learned a lot from your articles and thanks for that.

    Reply
  6. Jae

    HI Oscar,
    I am building micro frame (130mm) with 6500Kv motors. I am wondering if I can use any ESC (with Atmel, F330) with this High KV motors, or only use fast ESC like XM20A with F39X processor.
    Is there any difference between ESCs about maximum RPM supported?

    Reply
    1. Oscar Post author

      what’s the number of cells battery you are using with the motor? you can calculate the eRPM see if it’s over the limit.

      Reply

Leave a Reply

Your email address will not be published. Required fields are marked *

Are you Robot? *

I only check blog comments once or twice a week, if you want a quick reply you can post your question on this forum IntoFPV.com... You might get a faster response from me there (multirotor related only).