That is an interesting looking mower. What model is it? It looks like a slope mower of some sort but to find one that is all electric seems special.
I can’t tell the exact model of the FlySky radio transmitter/receiver in that picture but it looks like the 6 channel model. Technically the bare minimum number of channels is 4 but it in the real world most users prefer 10 to 16 channels to allow more things to be added (like a remote kill switch, remote arm/disarm switch, expanded function control, blade control, etc). Also I do know that some of those FlySky models can be upgraded to operate with 10 channels.
1 Like
I have a similar slope mower that I intend to convert soon and document via video. It should be very straightforward. The project keeps getting put on the back burner for any number of reasons, but I need to get it done!
1 Like
I bought this lawn mower through AliExpress By the company Vote Machinary, I’m adding the datasheet below for your knowledge.
About the radio control I confirm that it is the FlySky of 6 channels, good to know that about increasing the capacity of controls by the channels. Initially I was thinking of dealing with a PPM splitter to be able to perform more controls. Maybe increasing the quantities of channels is enough for this demand. Thanks again.
1 Like
I feel your difficulty, due to the other demands.
I will sending the evolution of my project here, and certainly you could help me by having a broad knowledge about Pixhawk, ArduSimple F9P and Mission Planer parameters
Congrats on the crawler mower - its looking like the start of an interesting project.
We’re now able to mow large paddocks with ours. Accuracy is +/- 50mm mostly.
some observations and challeges on our project.
- Crawler mower - 700mm wide, 16HP, 24v 500w motors, converted from RC to Ardupilot
- 2x F9P, Qiotek Zealot flight controller, RFD900x, corrections via NTRIP and mission planner
The chinese mower is excellent including the Loncin 465cc motor.
Challenges have been
a) black box ESC’s no documentation and not strong enough. Replaced these with VESC’s huge win.
b) Seals on roller bearings on tracks - way too tight and binding. Have removed, but problematic.
c) Machine is slow. 0.7m/s - it takes a long time to do a lawn. We’re looking to speed it up. 1m/s+
d) Machine is great at mowing slopes, its goes anywhere.
e) Please add front and back protection - it will eat your toes. Bumpers will keep your feet away…
Hardest part of project - Dealing with Deck height control and short grass, long grass. Mower will stall out with easily if it gets some long grass. Until you get that sorted, you have to watch/listen for stalled motor and “silent pretend mowing…” at least until the battery runs down.
Here is an example https://youtu.be/ZLtWCLwtgT4?t=162
To make it work for us - it needs to be " Eyes Off " - so no stalling, no runaways, no running kids over etc.
Enjoy your project.
1 Like
Excellent project. Have been mowing large areas with autonomous mowers for several years now and have often thought about monitoring engine RPM and reducing forward speed when it drops off by 10% or so and resume forward speed when it clears itself. Would be very useful for us. Perhaps a winter project…. Good luck!
PS - Those bumpers are an excellent idea.
PPS - DATE CORRECTION! MOWSTOCK 2024 - March 15-17, Church point, LA
Dear Paul, thank you very much for the ideas and experiences shared!
Replacing parts is really worrying, as it is a product from China, spare parts are usually a headache.
Regarding the ESC, I had added this model to do some tests https://www.dimensionengineering.com/datasheets/Sabertooth2x32.pdf
I made a test cut and actually due to the height of the cut, it touched the ground and tripped the circuit breaker. This will be one of the challenges.
Can you give me some light regarding the power supply circuit, as the circuit works with a 12Vdc battery and the PixHawk with a 5Vdc power supply, for this it uses a power supply circuit. For this, are you using a circuit with a battery suitable for the Pixhawk or did you use a 12V to 5V voltage step-down source?
Again thanks for all
re Parts from China. The ESC’s tha came with the mower were labelled “mower speed controller” and nothing else. Really just a black box. The engineering on our mower has been to a high standard, well made and thought out, with no items (apart from ESC) being hard to source or work around. The Loncin motors are very common, the oil filter is OEM same as KTM so really easy to purchase locally, cheaply, and thats been consistent with the rest of the kit.
Power has been drama free for us. The petrol motor drives a 24v alternator which charges 2x 12v deep cycle batteries in series. We have pulled 12v from one - which then goes into a power module - for 5v for the flight controller, and a bit of info on volts/current.
I’ve used a few other 12v - 5v regulators for RFD900x (after the link dropped when i beeped the horn 
) The existing wiring from factory is all fuse protected, with good sized relays for starter motor and ignition cut out etc. Im aware i have a ground loop and no optoisolation from the VESC’s which apparently causes lots of problems with blown electronics. Relatively low voltage and currents set in the VESC’s im guessing is helping here. Cheers.
2 Likes
I’ve got it working well enough to be able to go and do some other stuff while the mower is doing its thing, which was very not possible without the stall protection script (and the RTKfix checker)
2 Likes
Yesterday at night, after sent the message I was looking for more information about the voltage regulator and the answer was there. Thanks for the info.
Very nice! How are you monitoring “stalling”?
RPM sensor - a hall sensor above the motor with a magnet glued onto the spinny bit.
No load RPM is around 2500 - if rpm drops below a threshold - 2350 then drop the speed and let the revs recover, avoiding the stall.
local RPM_revs = RPM:get_rpm(0) or 0
if RPM_revs < RPM_min_revs then vehicle:set_desired_speed(0.2)
2 Likes
So is the hall sensor feeding directly into the FC? Any signal conditioning?
@Swebre2023, RPM sensing can be done with simple pulse count on an autopilot GPIO pin, and a hall sensor config probably doesn’t really need much filtering (as compared to the software based FFT config that I currently use, which often bites on harmonics, which makes it only useful as a curiosity).
I think maybe we can try this out at MowCon with a little of @ktrussell’s expertise and some scripting by yours truly.
1 Like
No signal conditioning or troubles encountered. Im running thin non-shielded 3 wire standard RC-PWM cable of approx 300mm between the top of the ICE motor and directly into the flight controller. I was expecting noise - but haven’t had any troubles, so haven’t upgraded to shielded cables or baluns etc.
Challenges / observations -
- Hall sensors are polarity sensitive - getting the magnet right way up boosts performance
- Height from magnet to sensor - im running “close as possible” - which looks like 3 to 4mm
- Im using a cheap hall effect sensor from a cheap aliexpress 37in1 kit or similar- they have LED’s on the small supporting board - makes troubleshooting and confirmation the hardware/sensor is working so much easier - ie LED flashes when it senses the magnet. My sensor looks like this. But sealed up with hot melt glue, and attached to the top of the mower with a holder epoxy cemented to the mower. The module has Negative, 5v Positive and a pin that goes from HIGH to LOW when the magnet is near.
- I built a test rig with an Arduino Nano, then later developed the scripts with a spare flight controller wired to the hall sensor with a magnet on a USB fan. Simulating slow downs on the fan was much easier than running the mower…
- Allow some time to figure out the ardupilot pin mapping for your FC. Setting pins to GPIO took me a while, and a bit of testing - and shout out to Yuri and the team in this forum to unravel the coding conventions.
In summary - the RPM sensor has been highly reliable - for us, essential for the mowing of paddocks where livestock seem to hang around in specific locations - generating huge grass growth from their emanations. The RPM sensor info is stored in the binary log file - helping for troubleshooting etc.
https://ardupilot.org/copter/docs/common-rpm.html
https://www.aliexpress.com/item/1005006497628639.html
1 Like
Sincere Thanks Paul! Excellent write-up.
Was wondering if I could do something simple like getting pulses by wrapping some wire around the plug wire. (Have used several small tachs that just had a ground and a few turns of wire around the plug wire.). Will borrow an oscilloscope and see what voltage pulses that gives and if usable.
Thanks again.
I had a little time today and raided my stash of electronics parts for an A3144 Hall effect sensor.
Here it is encased in some adhesive lined shrink tube, wired to a DuPont/servo connector with a 10k pull-up resistor bridging 5V and signal:
Here’s a simple bracket I fashioned from some aluminum angle:
Installed alongside the ignition coil. You can see the flywheel ignition trigger magnet passing by it in this shot:
I used the following parameters to enable GPIO RPM sensing on the Cube Orange’s AUX2:
SERVO10_FUNCTION,-1
RPM1_TYPE,2
RPM1_PIN,51
And here’s the active telemetry showing results!
4 Likes
I chose to go with heavy duty hall sensor LITTELFUSE 55100-3H-02-A which is connected almost directly to cube orange FC (through the ressistor as per the sensor’s datasheet). 3D printed bracket for it and disks with 10 magnets in each. The only reason disks are this big is only because of the sensors placement, mounted it in the existing hole of the frame. Works well, gives pretty accurate and high resolution RPM readings.
5 Likes
I am considering purchasing a Crawler, small 4 wheel drive mower to use in very tight mowing areas that have a lot of obstacles. That way I would have the zero-turn for big areas and a mini-mower for use as my string trimmer. I am interested to see what others think about the usefulness of building a small but very agile machine that could still cut taller grass. I want one with wheels on it for my application because my property is pretty flat, I think it would be more nimble, and I worry about tracks tearing up grass when you pivot. Parts of my property have numerous fence lines, trees, and other obstacles which makes mowing with the bigger zero-turn hard to do in some areas and I end up mowing further away from items that I worry about hitting. I am getting better at not hitting things, but in very tight areas you always worry, especially if you are not right on top of things.
Here is an example of what I am looking at. Joyance New Robot Lawn Mower With 4 Wheels High Quality Gasoline Remote Control Mini Lawn Mower - Buy Multifunctional Remote Control Crawler Lawn Mower Small Lawn Mower For Wasteland Mower On Road Slope yongkang Upspirit Hand Push Electric Lawn Mower Factory smart Robot Lawn Mower Suitable For Lawn Up To 600m2 With Lcd Display Product on Alibaba.com
Any thoughts on how close one could mow to objects with a small nimble mower like this, that was setup with the Ardurover RTK system on it? It looks small enough, with the right wrap-around bumper, and due to the fact that it is not that heavy, if it bumped or rubbed on a few objects, it wouldn’t matter.
Easily within a foot, so long as it’s well tuned and provides repeatable path following.
I routinely do so with my big mower, though I tend to keep a wider berth on big/expensive obstacles because I make such frequent/experimental updates in the interest of testing. But the latter is more of a “me” problem than a system issue.
A word of caution on large/expensive things from Alibaba/AliExpress/BangGood/etc: keep an eye on the shipping costs, need for a freight handler, and/or need for a customs broker/expediter. Added fees can quickly overcome the “good deal” you think you’re getting. I know @swebre had a fairly negative experience in that regard a year or so ago.
2 Likes