Portable snow tank for towing skiers

Joined
Dec 7, 2021
Messages
54
I’m starting a new build and want to document it. I’m struggling with where to post this one, so I’m going to start here. If it needs moved let me know.

The kids and I moved from the mountains a couple years ago into a very flat part of the Midwest. I used to snowboard but can’t where we are, so to help get through the winter here I’m working on a… snowbot? The idea is something I can control in the snow with a nunchuck, and that has a tow rope behind it (like a boat for waterskis) and to use it for pulling me so I can snowboard on the golf course near my house or wherever I can find.

The design is conceptually straightforward. It’s a track-drive skid-steer powered by two motors. I’ve built a few other things so this seems doable, although I’ve never worked with rubber treads so this will be a new frontier for me.

I’m looking at two QS mid drives (the 3kw V 3’s) and a pair of VESCs. I’m thinking VESC primarily because I’m familiar with them.

I’m a mediocre C programmer and have some experience with precision machine work, and have access to a fiber laser, so I’m planning to do the fabrication and coding myself. For the brain I’m thinking an ESP32 since I can repurpose that for some other software stuff later. For now I’m not looking to do anything crazy with computing. Just basic joystick drive.

I’m not sure how to size motors for snow, especially given it will be pulling a 200 lb person on a snowboard. I’ve also never built my own battery pack but expect I will need to for this one. So those and tank treads are my learning milestones for this project.

My ideal is to keep this under 100 lbs but given the motors alone are around 90 I’m already kind of giving up on that. I’d like it to be portable but it has to work too.

I started work on the controller today. I want to build my own, and all the nunchucks I’ve seen lock the two motors together anyway (you can’t accelerate them separately to steer). I’m curious to see how ESP-NOW works and may try that.

Any suggestions are welcome. I’m in brainstorming and design mode now but will post pictures as cads come together. I’m thinking aluminum for the structure and bending some metal for the outer body. It needs to be able to take a beating.
 
that is badass

with regards to the weight I think it will need be over 100 so it has enough mass to tow riders as well as having some mass for the torque, but then again if it is used in powder conditions with a paddle/pow style track 3"+ nubs that may be less of an issue

I just my mountain/powder sled and haver been thinking about doing an electric conversion on an older sled as they do have plenty of room for batteries

Also I recommend using a bungee type rope for towing skiers as I've tried it with tractor tow straps and doing that once too many could lead to either some very sore arms/shoulders or possibly injured, I'm trying out a snobunje this season, its a great small company out of Wyoming and seems to be great quality on first inspection :thumb:

Looking forward to see what you come up with and more than happy to help out with any suggestions or brainstorming
 
With QS-mid you'll need reduction, possibly two stages (unless you mean the one with internal 1:2,35 reduction) to get the rpm down, unless you want to go really-really fast.

IMHO two Cyclone 3k/4k motors with it's internal reduction of 1:6 would be enough but not sure about pairing those with VESC's (hall sensor signal noise).

I have seen enough available Arduino code for skid-steer bots, plus some code examples for Arduino VESC control.
 
I’m new to working with stuff this big, so I didn’t know about Cyclone. Checking out their website it looks like I’d save a lot of money going with their motors and controllers. The circuit I’m building will just output a standard pwm signal so I don’t need anything special. I’ll research cyclone on the forum - any opinions on the quality?

As for the snobunje, that’s a good idea. I’ll check it out. I wonder if some old dynamic rock climbing rope would work too (I may have some lying around).

Sent from my iPhone using Tapatalk
 
Philosoraptor said:
I’m new to working with stuff this big, so I didn’t know about Cyclone. Checking out their website it looks like I’d save a lot of money going with their motors and controllers.


Why not use 2 hub motors in the trackwheel to drive the track. Two high kV 9 inch "1000W" hub motors would do fine.

Also see this thread and my posts in it with links:
https://endless-sphere.com/forums/viewtopic.php?f=35&t=110224&hilit=snow
 
With PWM you'd need some kind of feedback of the actual RPM or you need to correct your steering constantly to compensate one motor going faster than the other.
 
Hall senators give you that feedback, right? I should have been clearer that I do intend to use sensors on my motors and controllers.

Also, what do you mean by “hall sensor signal noise?”

Sent from my iPhone using Tapatalk
 
Snow? Tank? Uphill? You want 4wd, and four tracks. You must account for slip.

You do not see many sno vehicles with just a pair of tracks, no? Usually they have four....

This is a Snocat we once bought. 4wd, hydrolic steering, articulating in the middle. This would go through 4 feet of deep of mud with four feet of water on top.

Lol this thing was fun. Right up double black diamond slopes. All day. Chrysler 318 and a three speed. Bogies on leaf springs. 6800 lbs. Permanently in 4-low.

I, currently, work on plans for a tracked wheelbarrow, electric. I am reusing a mobility wheelchair. The joystick will be repurposed for the control. 24v, 100lbs, 200lb payload.

Alot of Kelly controllers will work on a joystick.
 

Attachments

  • firetank.jpg
    firetank.jpg
    161.4 KB · Views: 2,552
Hall senators give you that feedback, right?
Motor hall sensors give feedback to motor controller (vesc) for timing. Not (servo-like) position you need for your master controller, which has to keep the speed exactly equal.

Also, what do you mean by “hall sensor signal noise?”
Cyclone motor-specific problem due to it's rotor design, there can be hall signal noise in-between proper hall pulses that in many cases can cause problems with sinewave controllers. Trapezoidal signal controllers (that cyclone sells with it's motors) and few advanced sinevawe/FOC controllers (Nucular) are capable of filtering out the signal noise and work very well.

If the tracks are not wide apart, I believe 2 tracks are good enough.


2 hub motors might work as well - hoverboards usually have small diameter one-sided shaft motors with vulcanized rubber track, which could be modified in lathe to suit your rubber tracks.
Here's one taken apart I got (for my kids go-kart):
205mm diameter, solid rubber "tires".
hub.jpg

I run it from dual flipsky vesc, at 13S8P battery, 2x30A bat Amps.
 
minimum said:
Motor hall sensors give feedback to motor controller (vesc) for timing. Not (servo-like) position you need for your master controller, which has to keep the speed exactly equal.

How do you recommend getting position / maintaining RPM?
 
VESC should be capable of either with custom app or external controller.

Custom app running on VESC CPU('s), keeping both RPM's equal, except when steering. Input through ADC or PPM/UART/CAN.
Or VESC constantly reporting to master controller RPM/Amps/etc, which then controls VESC's. Comms through UART or CAN.
 
Philosoraptor said:
The idea is something I can control in the snow with a nunchuck, and that has a tow rope behind it (like a boat for waterskis) and to use it for pulling me so I can snowboard on the golf course near my house or wherever I can find.

Before going into details of the technical realisation, I would rather research about the physical basics of towing.
If the towing device is much lighter than the object/person to tow, in my understanding it cannot work.
Have you ever tried to tow a truck with a motorbike?
 
Elektrosherpa said:
Before going into details of the technical realisation, I would rather research about the physical basics of towing.
If the towing device is much lighter than the object/person to tow, in my understanding it cannot work.
Have you ever tried to tow a truck with a motorbike?

Traction (friction) plays a big role in this.
You don't even need a motorbike to pull a truck. The rider can do that himself:

[youtube]FGLcPHNxVQU[/youtube]
 
If you want an easier solution you could build a high speed electric winch with a hub motor to pull you across the snow. Such a build can then also be used in summertime for water skiing or wake boarding.


[youtube]kr2-DdBsqFw[/youtube]
 
Sounds like a fun project!

I work as a embedded developer so I would say I know my coding. My strong recomendation is to skip the part where you develop your own motor controller and go stright to the stearing controller part. Buy two qs motors with matching motor controllers, votol / fardrive get them up and running and then use whatever nunchuck or input you prefer and connect to an arduino. Use two pwm outputs and connect to the 0-5v throttle on each motor controller and you're good to go.

Vesc motor controllers are not usually good for these kind of application.

But if you really want to do everything from scratch, then do it but from personal experience, its better to start out with as much off-the-shelf products as possible so you dont give up before you reach your goal.
 
hugok said:
Sounds like a fun project!

My strong recomendation is to skip the part where you develop your own motor controller and go stright to the stearing controller part.

Oh man, there’s no way in making my own controller. Just a simple joystick steering circuit, just like you said.

Why don’t the VESC’s work well? I like them, but have only used 4.12s for 500w motors.

I’m intrigued by these for the price: https://flipsky.net/products/flipsky-75100-foc-75v-100a-single-esc-base-on-vesc-for-electric-skateboard-scooter-ebike-speed-controller

I know a lot of esk8 guys hate flipsky because they’re more prone to fail than a real vesc, and that translates into getting thrown off your skateboard. Or my use case it’s more forgiving, but it still has to work well.

From what I’m reading they benefit from some upgrades to thermal dissipation (heat sinks, thermal grease). I don’t have to stay with VESC but if it works I would. Any opinions?

I’ve done some reading this morning. I saw some criticism of the cyclones for overstating their wattage. I have a bad habit of overthinking and over analyzing and need to force myself to pull the trigger fairly soon.


Sent from my iPhone using Tapatalk by
 
I think I’m down to a cyclone or 2-6kw geared motor https://www.cyclone-tw.com/product/1/data/15 or a QS 138 http://www.qs-motor.com/product/new-qs138-70h-3000w-power-train-pmsm-ipm-motor-with-internal-reduction-gears/

I’m not a motor pro, but working with other things has taught me to mistrust vendor specs. Given one weighs 3x as much as the other I dont know what to think. I also want to try to finish it this winter, so ordering from china adds time.

Any first hand experiences?


Sent from my iPhone using Tapatalk
 
Cyclone 2-6kW is peak power. With some modification and depending on cooling, you might get up to 8kW peak from these.

QS 3K mid... Folks here have pushed beyond 20kW and still haven't overheated it. So far haven't heard any cooked motor.
 
Another option for a motor: But this guy is really light. For towing you want weight. Like plowing.

https://lightning-rods10.mybigcommerce.com/xl-motor/
 
Philosoraptor said:
hugok said:
Sounds like a fun project!

My strong recomendation is to skip the part where you develop your own motor controller and go stright to the stearing controller part.

Oh man, there’s no way in making my own controller. Just a simple joystick steering circuit, just like you said.

Why don’t the VESC’s work well? I like them, but have only used 4.12s for 500w motors.

I’m intrigued by these for the price: https://flipsky.net/products/flipsky-75100-foc-75v-100a-single-esc-base-on-vesc-for-electric-skateboard-scooter-ebike-speed-controller

I know a lot of esk8 guys hate flipsky because they’re more prone to fail than a real vesc, and that translates into getting thrown off your skateboard. Or my use case it’s more forgiving, but it still has to work well.

From what I’m reading they benefit from some upgrades to thermal dissipation (heat sinks, thermal grease). I don’t have to stay with VESC but if it works I would. Any opinions?

I’ve done some reading this morning. I saw some criticism of the cyclones for overstating their wattage. I have a bad habit of overthinking and over analyzing and need to force myself to pull the trigger fairly soon.


Sent from my iPhone using Tapatalk by

From what I read VESC can't really handle high current for more than a short while and I guess you don't want to be on the limit.

Regarding motor I would go for the QS120 motor. You get more power than you need with two of those.

I have the QS138V3 and as someone said, you can push 30kw for a short while, I run mine at 10kw and it doesn't even get warm.
 
Latest update - I built the joystick today. Since it’s a skid steer and I want to drive with one hand I had to use a thumb stick and write code that translates that into velocity commands for two separate motors. I’m using ESP32s for communication over ESP-now. So far it’s just experimenting with the microcontrollers, I haven’t hooked motors to it. I have the commands sending nicely, so now to cad up a joystick housing.

I’m also going to repurpose some code I made for a ROS project to translate the joystick commands into geometry messages. I e forgotten how that worked and need to review my stuff and remember. But I want you guys to know I haven’t given up, just been busy with life.


Sent from my iPhone using Tapatalk
 
I ordered the QS motors, EM-150’s, and two Daly 150 amp controllers. I’m researching battery pack options. Any suggestions for 72v packs (buy or build) are welcome. I’m looking at pouch batteries in batteryhookup. From what I’ve read they are slow to start in the cold, but are great once they warm up. And fuse recommendations too - I’m still reading up on building those into battery packs.

The cad work is coming along. I’m playing Tetris with the parts and, since this is operating in the snow but will likely get different wheels and be used in summer, I’m trying to figure out both airflow / heat dissipation and water protection. I’m reading up on motorcycle airbox designs for ideas.

I have access to a great machine shop and will be milling / cutting the parts myself as needed for the frame and enclosure. I’m thinking 6061 aluminum.

The one-handed controller has sat on my desk and stared at me. I need to pull some motors and controllers off an old project and do some testing.


Sent from my iPhone using Tapatalk
 
calab said:
Have you seen something like this?
[youtube]EaDE0RiEgdc[/youtube]
Generally that’s the idea, although smaller, apparently heavier, and faster.


Sent from my iPhone using Tapatalk
 
Back
Top