Can I use a crawler for testing inertial reference units?

[GeneM]

I have a project at work where I need a very slow, very smooth RC car/crawler. Before I design my own gizmo, I thought I'd check to see if I can buy something that comes close and am hoping you guys can help. Is there a RC crawler with a max speed of something like 1 mph or less and the ability to control the acceleration rate so the crawler can be set to smoothly go from 0 to max speed in 50 to as much as 500 seconds? Or something I can modify? I have pretty good mechanical, electrical and programming experience. I also have a full machine shop with 3D printers if I have to customize something.

The goal is to test several 9-axis inertial measurement units by mounting them on the crawler and driving a preset path with known accelerations and deacceleratons along the path so we can compare the measurements against the known accelerations. The path will be mostly flat, no rocks.

I hope that makes sense. If not, please ask questions.

Thanks

[Xx Losungen xX]

Hardly seems like the appropriate platform. Seems like you'd want to limit introducing variables such as sloppy drivetrain, suspension movement etc.

Surely such a set of tests is worthy of an equally worthy test platform no?

[GeneM]

If you know of a platform that might be a better fit, I'd be happy to hear about it.There are a whole bunch of reasons I can think of why a RC crawler might not work. And probably a lot more that I haven't thought of. That's why I asked. But if a crawler gets me close for a few hundred bucks, that is well worth it. And maybe I could ditch the 4WD drive train later and add my own. For example, the Losi night crawler looks like it might come close and is something I could maybe put my own drive train on. But it already has the radio, steering, chassis, battery, front end and all the other stuff I need and I won't have to design any of that. If I was to build something from scratch, I'd use a Maxon brushed motor with a big gearhead and a timing belt drive to a pulley mounted directly on the axle with a Elmo sevo controller that gives all kinds of control including limiting the acceleration to whatever value I want. Is there a better starting place than the Losi?

[crawlinwithacat]

Originally Posted by Xx Losungen xX

Hardly seems like the appropriate platform. Seems like you'd want to limit introducing variables such as sloppy drivetrain, suspension movement etc.

Surely such a set of tests is worthy of an equally worthy test platform no?

While I somewhat agree, personally The Night crawler is the first rig I thought of for his purposes. Those axles have a gear ratio of 25:1. So compared to a stock scx10ii at 3.75:1 or a stock trx4 at 3.09:1 (Not including portal reduction but that's rather negligible here) the Losi axles are far superior in terms of getting very low speeds like he's after. Coupled with a gear reduction unit on the motor it might be okay.

In terms of it being a good test platform or not, I would think that would depend on how accurate these sensors (I assume thats basically what they are, if not please correct me) need to be compared to a known value, how accurate they need to be to each other, and how sensitive they are. If you could tell us those things that would be really helpful in determining if a crawler is really the right choice of test platform. The other thing that would be good to know is how they measure acceleration and deceleration. (I am likening these units to accelerometers in my mind here) Is it a high resolution thing where every tiny bump and jerk would register as a spike on a graph or is it lower resolution where you'd see a more averaged curve during acceleration on a graph?

I'm sure I probably sound silly compared to what you know but these sorts of things would be good to help us point you in the right direction. Cheers!

[GeneM]

The gear ratio information is helpful, thanks. IMUs have 3 gyroscopes, 3 accelerometers and 3 magnetometers with one of each sensor aligned in the X, Y and Z directions relative to the sensor body. I'm actually only interested in velocity which I can calculate from the accelerometer measurements vs time. I need the other sensor to convert the accelerometer measurements, which are relative to the X, Y and Z axis of the sensor body, to the earth's East, North and down directions.

I'm not too worried about high frequency bumps due to suspension or whatever because I have to low pass filter the signals to get rid of as much high frequency noise as possible. Also we'll be running simple back and forth linear paths maybe 100 yards long and also run circles maybe 50 feet in diameter on a relatively smooth parking lot, hence the need for RC or something similar.

Really what I'm looking for is a low speed (0 to a few mph max), steerable platform that I can start with. I will need to control the acceleration at some point but so far I haven't found any RC radios, receivers and ESCs that allow me to set the maximum acceleration. Is there such a setup?

If a crawler is too fast or even too noisy to see the lowest level measurements I'm interested in that's OK to start with and I'll just ditch the geared 4WD drivetrain and add my own belt driven drivetrain in the future. If the shocks are a problem, I can replace them with solid links.

In any event, thanks for the help.

[magic_yeti]

How about something with MOA Axles? This would eliminate a lot of drivetrain slope.
Nice Project btw.

[Xx Losungen xX]

What surface are you testing on?

[GeneM]

I just googled MOA axle and that looks interesting. If I was to add my own drive train, it would be MOA but with a brushed servo motor with a gearhead on the motor and a belt drive to the axel as I think that is much smoother than the geared differential style drive in most (all?) RC vehicles. And it's what I'm used to building.

As far as the surface, it will be a concrete or asphalt parking lot. No rocks. I probably don't need the suspension of a crawler but they do seem to be the lowest speed RC vehicles out of the box.

[HumboldtEF]

Originally Posted by GeneM

I will need to control the acceleration at some point but so far I haven't found any RC radios, receivers and ESCs that allow me to set the maximum acceleration. Is there such a setup?

Most (non RTR) radios can do this, you could adjust the throttle travel setting which can set the max throttle level available.


The Traxxas TRX4 comes to mind because its radio has a cruise control feature where you can get up to a speed of your choice and hit a button to activate cruise control.
There are a few other options for setting the speed of an RC. you can use the throttle trim to dial in low speeds, this would be done as you drive with some button presses. I believe you could also use a "stick radio" with the return spring removed, this should allow you to set the throttle stick where you want it.





If you need to slow an existing RC down you might consider adding a planetary gearbox in addition to the existing tranny.


MOA's dont have a ton of room and may not be the best platform to build on IMO. I would think one of the 1:10 scale, scale trucks would be the best platform to work off of.


If you need to slow the truck down with a planetary gearbox in my experience the simplest truck to do this to (from my collection) is the HPI Venture. if you aren't using a body the extra length added by the planetary gearbox isn't an issue.

[crawlinwithacat]

Originally Posted by GeneM

I'm not too worried about high frequency bumps due to suspension or whatever because I have to low pass filter the signals to get rid of as much high frequency noise as possible. Also we'll be running simple back and forth linear paths maybe 100 yards long and also run circles maybe 50 feet in diameter on a relatively smooth parking lot, hence the need for RC or something similar.



If a crawler is too fast or even too noisy to see the lowest level measurements I'm interested in that's OK to start with and I'll just ditch the geared 4WD drivetrain and add my own belt driven drivetrain in the future. If the shocks are a problem, I can replace them with solid links.

In any event, thanks for the help.

Low Pass signal filter. Perfect! It will certainly help a lot with cutting out the noise in your data like you said. I would say that with some tweaking you should be able to get pretty close to the measurements you want. I was thinking the shocks might be a problem as well and thought you could replace them with straight links but looks like we're thinking on the same page there.

As for planetary gearboxes/ gearheads, I know that RC4WD offers a handful of them as well that would probably be worth a look. I suppose in theory you could hook half a dozen of them in line and get some rediculously low rpm out of it lol. I don't have any experience with them but i am sure someone on here could give you some good advice.

[ferp420]

Originally Posted by GeneM

Really what I'm looking for is a low speed (0 to a few mph max), steerable platform that I can start with. I will need to control the acceleration at some point but so far I haven't found any RC radios, receivers and ESCs that allow me to set the maximum acceleration. Is there such a setup?

.

some of the higher end escs out there will let you set max throtle to a point

castle creations has a pc program and i 5hink they have apps to that adjusts tbe throtle curve and you can set full trotle to any point from 0 to max
they also have a blue tooth adaptor so you can change the settings easyer

they also have them in different sizes so you could get say a mamba micro rather cheap and take advantage of the programing it dosent sound like you need alot of power so a micro should work just fine

[QuesoDelDiablo]

The one big thing I haven't seen addressed is controlled acceleration. If you need something that accelerates with a known/defined jerk then an esc out of an rc won't do that. You might need to look at do something Arduino driven to control throttle if you need to properly define your rates.

ESC's and/or radio programming can define a Max throttle input, but nowhere in their functionality is any allowance made for a controlled acceleration curve or a max acceleration rate.

I suppose you could also do it empirically, just set the max throttle and collect data over n passes to determine the acceleration rates. But even with that, you're only looking at an average acceleration rate because you can only collect data at V=0 and V=Max.


A crawler is definitely the right way to go, but I think you're going to need gear reduction (that's already being discussed) and programmable throttle inputs.

[GeneM]

Thanks to everyone for all for the great information.

If you need to slow the truck down with a planetary gearbox in my experience the simplest truck to do this to (from my collection) is the HPI Venture. if you aren't using a body the extra length added by the planetary gearbox isn't an issue.

As you suggested, I've already been trying to find a crawler that would have room for me to add one of my geared motors. I hadn't seen the Venture so good tip there. A few more questions if you don't mind.

Did you add the motor or did it come with the Venture? What kind of motor is it? I'm guessing you had to build some kind of adapter if it wasn't designed for the Venture?

At this point, I'm leaning toward either the Losi Night Crawler or the HPI Venture. I'm not the optimistic type buy it seems like either of them will give me a decent platform to start with at the higher accelerations I'm interested in even if they won't move smoothly at the lowest accelerations I'm looking for.

Thanks again to all of you.

[HumboldtEF]

I sent you a PM GeneM but I thought I'd copy it here as well in case anyone has any corrections or anything to add.

I'm curious to hear your response to what QuesoDelDiablo asked about how smooth the acceleration needs to be for your calculations because at very low speeds there can be some jerky-ness. I would hope the speed of the RC isn't the determining factor for calculations.

So any motor would work but I've used a Holmes Hobbies Puller pro 400 20T brushed motor https://holmeshobbies.com/motors/bru...crawl-20t.html
but this motor is likely overkill, I'd recommend one of their budget motors the Crawlmaster sport 540 20T https://holmeshobbies.com/motors/bru...t-540-20t.html
These motors dont come with the Venture but the RTR versions of the Venture do include a motor. I used the builders kit which comes with only the base components.

Holmes Hobbies makes very high end motors that have a very smooth start up, I dont know if this would matter to you or not. If not you could save a lot of cash by going for a basic motor. If all you want is a slow motor look for one with a high number of turns.
If you're after a smooth start up too Holmes makes the best motors for that.

If you compare these crawlmaster motors to other motors you will find the turn ratings vary, since the Crawlmaster motors are 5 slot and most other motors are 3 slot you have to understand that a 12T-16T (5 slot) is roughly equivalent to a 27T-35T (3 slot). So while a 16T or 20T motor sounds like it would spin very fast they in fact do not.

To see a chart that has various 3 and 5 slot motors compared scroll down this page a bit to see a chart that compares them: https://holmeshobbies.com/motor-recommendations

The planetary gearbox is a 3:1 reduction unit and I'm not sure of the brand, I just found it on Amazon. I think this was it: https://www.amazon.com/Planetary-Red...6453280&sr=8-5

You can also find 5:1 versions as well. The motor mounts directly to the planetary gearbox and it mounts directly to the transmission, no extra parts required. Search for "Planetary gearbox RC" to find these.

[GeneM]

You guys have been so helpful, I thought I'd give a fairly detailed answer. If it doesn't make sense or you have more questions, let me know.

I definitely need a fairly accurate, known acceleration rate so I can compare what I know the accelerometers should be measuring against what they actually measure. There are 2 ways I'm thinking about controlling acceleration. The one that works with a crawler out of the box is to run at a slow but constant speed in a circle. In this case I'd be measuring centripetal acceleration which is equal to the velocity of the crawler squared divided by the radius of the circle (AccelCentripetal = velocity^2 / radius). The good news with this approach is that the crawler only has to maintain a smooth constant velocity with the steering wheels set at a fixed angle. And I can reduce the acceleration by increasing the radius of the circle I'm driving. If I can drive a smooth, constant 0.1 mph in a 50 foot circle, that would be 13 micro Gs which is right in the range I'm looking for.

The more complex option is to use my own motor and controller which allows me to do the same centripetal acceleration as above and also a straight line linear acceleration test. The motors and controllers I've been using for decades give very smooth low speed control and the controller can be programmed for a maximum speed and a maximum acceleration. For example if I set the top speed at 0.1 mph and I set the acceleration rate at 0.1 mph/ 1000 seconds and I just drive in a straight line, it will take 1000 seconds to go from 0 to 1 mph. which is an acceleration of 4.6 micro Gs.

At this point, I think the best answer is to find a really well built crawler with a good radio and the ability to easily swap out the stock motor with my own motor/gearhead which will be much longer than the stock motor. The HPI Venture that HumboldtEF recommended looks good but the Axial SX10 II might work if I pull out the plastic fake engine parts. I'm going to ask that question on the SX10 sub-forum but if anyone reading this knows the answer, I'd be happy to hear it.

[Xx Losungen xX]

Why not an onroad solution? I would think it would behave better given the surface and with gearing plus the programming available with current ESCs, speed vs acceleration shouldn't be an issue.

[crawlinwithacat]

Originally Posted by Xx Losungen xX

Why not an onroad solution? I would think it would behave better given the surface and with gearing plus the programming available with current ESCs, speed vs acceleration shouldn't be an issue.

I thought of this as well. Only issue with that would be the gearing to bring the overall speed down, although if that point was ignored, I think it would be better than a crawler IMHO. But if it were me, I wouldn't consider it an option. I mean he's talking a tenth of a mile per hour. That's 528 ft/hr. The only kind of motor that I can think of that could pull something like that off without cogging at such low speeds would be a stepper motor. But you can't run those on normal RC ESC's. (Please correct me if I'm wrong) But even that being said a stepper motor could be a good thing to look into just for the heck of it to see what it's all about.

Originally Posted by GeneM

You guys have been so helpful, I thought I'd give a fairly detailed answer. If it doesn't make sense or you have more questions, let me know.

I definitely need a fairly accurate, known acceleration rate so I can compare what I know the accelerometers should be measuring against what they actually measure. There are 2 ways I'm thinking about controlling acceleration. The one that works with a crawler out of the box is to run at a slow but constant speed in a circle. In this case I'd be measuring centripetal acceleration which is equal to the velocity of the crawler squared divided by the radius of the circle (AccelCentripetal = velocity^2 / radius). The good news with this approach is that the crawler only has to maintain a smooth constant velocity with the steering wheels set at a fixed angle. And I can reduce the acceleration by increasing the radius of the circle I'm driving. If I can drive a smooth, constant 0.1 mph in a 50 foot circle, that would be 13 micro Gs which is right in the range I'm looking for.

The more complex option is to use my own motor and controller which allows me to do the same centripetal acceleration as above and also a straight line linear acceleration test. The motors and controllers I've been using for decades give very smooth low speed control and the controller can be programmed for a maximum speed and a maximum acceleration. For example if I set the top speed at 0.1 mph and I set the acceleration rate at 0.1 mph/ 1000 seconds and I just drive in a straight line, it will take 1000 seconds to go from 0 to 1 mph. which is an acceleration of 4.6 micro Gs.

At this point, I think the best answer is to find a really well built crawler with a good radio and the ability to easily swap out the stock motor with my own motor/gearhead which will be much longer than the stock motor. The HPI Venture that HumboldtEF recommended looks good but the Axial SX10 II might work if I pull out the plastic fake engine parts. I'm going to ask that question on the SX10 sub-forum but if anyone reading this knows the answer, I'd be happy to hear it.

I'm glad that we could be of such help! That is actually an ingenious way to control acceleration. It never occurred to me that you could do it that way but thinking about it, it makes sense! The only problem I see with numbers and times that you are talking about is that the drivetrains in these vehicles aren't really tight enough let alone designed for that, I'd say. Basically slop. I think the only way to get around that would be like with something so slow that it can gently put pressure/force on the gears and not jerk because that would make for non-smooth acceleration. The other thing would be tires and the fact that crawler tires tend to have knobbier lugs although if you got something like an RC4WD Dirt grabber which has a very tight tread pattern that would mean minimal lurching as it rolls over the lugs. Any thoughts?

Side note, it's cool the way that it feels like we are bringing two worlds together. I can really feel the collaboration and willingness to help and share ideas and expertise from everyone who has posted in this thread.

[OSRC]

Sounds like a cool project!

Maybe off the mark, but ever think of a tracked platform like a tank? Very low speed, minimal drivetrain slop, lots of gear reduction, nice flat top to work with... I'd think with dual motors that control not only speed but steering angle it might be easier to work with than a motor and steering servo?

If not, I'd second the worm axles for sure. Even a 5:1 reduction on the motor will really make it creep along, and the worms axles I've had in the past had very little slop.

[GeneM]

I thought about the tank but haven't looked at them very hard. My guess is that a tracked vehicle will have quite a bit more vibration than a rubber tired vehicle. While I can filter out much of that vibration induced noise, the less noise I start with the better off I'll be in the end.

I've also seen some RC tractors and excavators and vehicles like that that probably go really slow but they seem to be in the $1000+ range so maybe not a good choice for my whacky idea for a test platform that has a 50/50 chance of failure.