View Full Version : Andymark rhino drivetrain discussion
AngryCalculator
10-01-2016, 12:29
My team saw and bought two modules almost immediately after kickoff. I have seen many people mention how they are not totally necessary to over come the defenses. Post your thoughts here!
they are delayed by about 2-3 weeks, in case anyone is wondering...
MrJohnston
10-01-2016, 13:34
We discussed numerous possibilities for wheels/traction/etc. last night - including the Rhino... There are less expensive ways to do the same thing, though we do think this is a great product....
Zebra_Fact_Man
13-01-2016, 08:22
Anyone know how fast these threads are rated for / how fast you've got them to go? Anyone cross 15 ft/s?
Our team has gone against them, but also discussed creating our own with the 2-3 week delay. I don't think it would be that hard for a team to CNC frames and assemble the parts.
techtiger1
13-01-2016, 08:39
Just FYI belts/ pulleys typically have a 2-3 week turnaround time from most if not all suppliers. 1251 my team has done tank treads in both 2010 and 2012 so we've done quite a bit of prior searching.
Robomarfa
13-01-2016, 12:14
Our second-year team purchased these as soon as they were available on Saturday and were just told they will ship by Thursday. We are a young team with limited machine shop capabilities. Rhino drive looks like it can handle any of the defenses easily depending on robot height and mechanisms.
If we were building a high goal shooter, we might want something more stable (no pivot point) and more maneuverable. For weakening defenses, should be great.
Frustrated at the whole "out of stock by the end of the first weekend" trope rearing its head again with FRC suppliers.
Look, I know it's difficult. It's a small market and it's hard to predict demand, etc. But this can't be the best that can be done. It's hard/impossible to teach good engineering practice when you're forced to rush to make your drive choice in the first day lest you not be able to purchase the parts you need until halfway through build season.
I understand that 2 or 3 cim motors are needed for each side of the Rhino drive.
Is each cim motor connected to its own motor controller?
How would one program this in Java?
Thanks.
JPBlacksmiths
23-01-2016, 18:12
8 inch pneumatic work fine
nandeeka
23-01-2016, 19:30
I understand that 2 or 3 cim motors are needed for each side of the Rhino drive.
Is each cim motor connected to its own motor controller?
How would one program this in Java?
Thanks.
Yes, each CIM motor is connected to its own motor controller. The programming of multiple motor controllers is no different than controlling one motor per action in Java. You just have to make sure that every time you set the speed for one motor, you also set the speed for the other motor, in order to ensure that they are always going at the same speed.
IronicDeadBird
23-01-2016, 19:32
These COTS treads ought to make for an interesting shift in what drive bases can do. I can't wait to see how people respond to the "threat" of them appearing.
we cnc ed our own out and made our own belt. it is fast and i will try to get a fps. looking at the AM bumper kit and going to try to replicate something like it.
https://www.youtube.com/watch?v=YlRQYMbeK2Y
Could I use PWM splitters to run the motor controllers of 2 or 3 CIM motors at the same time?
nandeeka
23-01-2016, 22:03
Could I use PWM splitters to run the motor controllers of 2 or 3 CIM motors at the same time?
Yes, I think you can, though my team has never done it before. In that case, I think you only have to control one motor controller.
R53 Each power regulating device may control electrical loads per Table 4-4. Unless otherwise noted, each power regulating device shall control one and only one electrical load.
one controller per motor is how i read it
Greg Needel
24-01-2016, 00:01
R53 Each power regulating device may control electrical loads per Table 4-4. Unless otherwise noted, each power regulating device shall control one and only one electrical load.
one controller per motor is how i read it
Correct. It is 1 motor per motor controller. You can use a PWM Y-Cable (like provided in the KOP) to provide control signal to 2 motor controllers at a time.
Could I make a PWM Y-cable to control three CIM motors or really three motor controlers for the Rhino drive? Is there a better way to wire and program Rhino drive?
Greg Needel
24-01-2016, 01:16
Could I make a PWM Y-cable to control three CIM motors or really three motor controlers for the Rhino drive? Is there a better way to wire and program Rhino drive?
The simple answer to this is yes (although it is recommended you use the same type motor controller for all 3 motors, 1 motor controller per motor).
Alternatively you could use 3 different PWM ports on your roborio, and make them function the same in code.
I thought using one motor controller for three CIM motors was not allowed.
orangemoore
24-01-2016, 01:31
I thought using one motor controller for three CIM motors was not allowed.
You are correct. You have to connect each CIM motor to its own (ideally same) motor controller. However the control signal (ie PWM) can be split using a Y-cable to send the same control signal to each of the motor controllers.
I thought using one motor controller for three CIM motors was not allowed.
It isn't.
What I believe he's referring to is that you want to have all three motor controllers be the same type. Different types have slightly different responses, and having the motors react differently in the same gearbox can produce "interesting" results.
Thanks for the clarification.
Got our Rhino drive train in middle of last week. We have both sides assembled and should be done mounting them on the cross beams today. (We ran out of screws last night and somehow came up 4 short.
To be clear we didn't order the modules from Andy Mark, and instead ordered all of the parts from andy mark with the exception of the metal side plates. We got a local machine shop to cut those out on their waterjet. ( I so want one of those...) :-) We did this hoping to save some time on the order figuring that would be one of the things that would hold up production.
In any case, things actually went together really smoothly.
We are using sheet metal to mount the bumpers. I don't have any pictures, but Team 3238 was kind enough to help us out with the use of their CNC plasma cutter. I'll post pictures of those tomorrow.
http://forums.trossenrobotics.com/gallery/files/1/4/9/9/20160123_165530.jpg
Sorry it is so blurry. I was in a hurry when I took the pictures yesterday and seeing that they aren't in focus on my phone is obviously beyond me.
DB
evanperryg
24-01-2016, 09:56
Anyone know how fast these threads are rated for / how fast you've got them to go? Anyone cross 15 ft/s?
I'm sure it's possible... you wouldn't really even need this kind of speed with how tight this field is. We discussed it the sunday immediately after kickoff, and decided that, with the cost of the rhino drive, if we bought it we absolutely had to use it. This wasn't a risk we were willing to take, on a drivetrain we've never tried before, that has yet to see competitive use.
It turns out my team wants to use Talon SRX with CAN signaling. Here is what I think needs to be done to run Rhino Drive:
1. Each Talon SRX would need a different id.
2. Each Talon SRX would need to be daisy chained.
3. Since my team want 3 motors on each side of the Rhino Drive, we will need 6 Talon SRX controllers.
How do I program 3 motors for left and 3 motors for right? Do I setup RobotDrive1, RobotDrive2 and RobotDrive3 each for 2 motors?
Thanks.
How do I program 3 motors for left and 3 motors for right? Do I setup RobotDrive1, RobotDrive2 and RobotDrive3 each for 2 motors?
If you're using CAN, you can 'slave' two CAN motor controllers to output exactly what a third 'master' controller outputs/is commanded to output.
AdamHeard
28-01-2016, 14:17
Anyone willing to post video of a rhino at weight being t-boned by another full weight robot on carpet?
I found some good sample code.
https://github.com/CrossTheRoadElec/FRC-Examples/commit/f44e015a24f35e05db18c6263125997ec38d3768
Tom Line
29-01-2016, 02:07
Anyone willing to post video of a rhino at weight being t-boned by another full weight robot on carpet?
A whole lot of teams are going to be a whole lot of disappointed if those treads slide off when pushed sideways.....
A whole lot of teams are going to be a whole lot of disappointed if those treads slide off when pushed sideways.....
And a whole lot of teams are going to learn how to play some pretty effective defense.
MrJohnston
29-01-2016, 08:13
The Rhino treads are very well secured and will be extremely difficult to knock off so long as the robot's pit crew keeps up with standard maintenance checks and makes sure that everything is kept tight.... Assuming proper upkeep, it would take a tremendous hit to cause any problems.... No, we don't intend to subject our robot to too many violent hits before competition... :)
At the same time, we recognize that there is a potential weakness here, so are developing game play strategies that will minimize such collisions.... Robot damage happens, no matter how robust your design might be... We even had our chassis bent at a district event last year - due to a violent autonomous collision with a member of our own alliance who's routine went haywire in a very bad way.....
I am pretty confident the tracks will stay on when pushed from the side. The pulleys have a V groove which the the belt has an inverted v that fits in the groove. Is it possible that they could slide off when pushed from the side and the bot is trying to move forward or backward? Sure it's possible, but if the drive just lays off the controls for a couple seconds they won't "slide" off. Keep in mind someone can only push for 5 seconds and then has to back off.
JamesCH95
29-01-2016, 09:06
I too am interested to see how these treads work out with full-weight robots, especially with open-field hits.
What makes me nervous is the turnbuckle used to tension the treads that relies on the eyelets to stay bent closed. The first thing I would do with one of those tread modules is to weld those eyelets closed. My intuition is that it will only take one or two hits to open those up.
Of course, I might be completely wrong...
I am pretty confident the tracks will stay on when pushed from the side. The pulleys have a V groove which the the belt has an inverted v that fits in the groove. Is it possible that they could slide off when pushed from the side and the bot is trying to move forward or backward? Sure it's possible, but if the drive just lays off the controls for a couple seconds they won't "slide" off. Keep in mind someone can only push for 5 seconds and then has to back off.
You might want to check G22... ROBOTS can only pin for 5s. I think that you would have a hard time arguing that being pushed in an open field counts as a PIN.
Granted, but where is this open field you refer to?? It isn't like the Airial assist field! A quick back up or move forward and your touching the outer defense and bam foul.
Now I'm not saying there won't be defense played hard this year, but the potential for hits like in 2014 are going to be rare. Plus I would think most would be more worried on breaking defenses and or scoring in the tower. But we shall see
JamesCH95
29-01-2016, 09:30
I am pretty confident the tracks will stay on when pushed from the side. The pulleys have a V groove which the the belt has an inverted v that fits in the groove. Is it possible that they could slide off when pushed from the side and the bot is trying to move forward or backward? Sure it's possible, but if the drive just lays off the controls for a couple seconds they won't "slide" off. Keep in mind someone can only push for 5 seconds and then has to back off.
Granted, but where is this open field you refer to?? It isn't like the Airial assist field! A quick back up or move forward and your touching the outer defense and bam foul.
Now I'm not saying there won't be defense played hard this year, but the potential for hits like in 2014 are going to be rare. Plus I would think most would be more worried on breaking defenses and or scoring in the tower. But we shall see
Therein lies the rub.
An open-field pin, or more precisely: an open field t-bone, doesn't require a high-speed impact. All it requires is a defender applying steady pressure to the side of another robot to completely ruin their ability to maneuver. I would encourage you to try this out with two or more robots, it is comically effective and easy to do.
The turnbuckles are in compression to keep the belt tight, not in tension. It would take quite a bit to get to the point where the eye bolts are opening. If things get that out of hand, the eye bolts opening will likely be the least of our worries....
As a suggestion, it would be a good idea to have jam nuts on one end of the turnbuckles to make sure they stay where you want them.
Therein lies the rub.
An open-field pin, or more precisely: an open field t-bone, doesn't require a high-speed impact. All it requires is a defender applying steady pressure to the side of another robot to completely ruin their ability to maneuver. I would encourage you to try this out with two or more robots, it is comically effective and easy to do.
I understand what you are saying, but when I hear (read) T-bone I am thinking full speed violent hit. But I see your point. I know we will be testing that hopefully this weekend. Will try and grab some video to post here.
JamesCH95
29-01-2016, 10:07
The turnbuckles are in compression to keep the belt tight, not in tension. It would take quite a bit to get to the point where the eye bolts are opening. If things get that out of hand, the eye bolts opening will likely be the least of our worries....
As a suggestion, it would be a good idea to have jam nuts on one end of the turnbuckles to make sure they stay where you want them.
I know that the turnbuckles are in compression, which is what concerns me. It may be fine as-is, but the 2 grams of mass it would take to weld it closed seem more than worth it to me.
I know that the turnbuckles are in compression, which is what concerns me. It may be fine as-is, but the 2 grams of mass it would take to weld it closed seem more than worth it to me.
I hadn't thought about them opening up, which is plausible if a hard impact is taken on the front. And now that I think about it I will be welding ours.
I know that the turnbuckles are in compression, which is what concerns me. It may be fine as-is, but the 2 grams of mass it would take to weld it closed seem more than worth it to me.
I would be more worried about weakening the aluminum at the weld. We at AndyMark did extensive hit tests on these turnbuckles at full speed with a fully weighted chassis into a piece of diamond plate that was backed by a full pallet of roller chain, so it had no where to go. We are working on putting up some videos of our testing so that the community can see what we have done to test the modules before we were comfortable selling them.
Michael Corsetto
29-01-2016, 13:43
I would be more worried about weakening the aluminum at the weld. We at AndyMark did extensive hit tests on these turnbuckles at full speed with a fully weighted chassis into a piece of diamond plate that was backed by a full pallet of roller chain, so it had no where to go. We are working on putting up some videos of our testing so that the community can see what we have done to test the modules before we were comfortable selling them.
Jon,
As part of your video tests, can you run a test of the Rhino track system being pushed sideways by another robot (both at 140lbs)? Specifically:
Robot A - Rhino Tracks (~140lbs)
Robot B - Kitbot (~140lbs)
Robot B pushes square onto the side of Robot A.
Robot A does not move.
As Robot B pushes Robot A, Robot A attempts to drive forward/backward.
If you could test the above scenario, that would be awesome!
Thanks,
-Mike
If it becomes an issue with the tread being pushed off the wheels with side load, which I doubt will happen with a properly tensioned belt, can the Rhino Teams add a plate to the sides of the wheels, slightly larger than the wheel, so it overlaps the white part of the belt?
This would be similar to the other belt pulleys that are used.
In addition to the plates, perhaps a mechanism to deflate pneumatic tires could also be added? ;)
EricDrost
29-01-2016, 16:34
In addition to the plates, perhaps a mechanism to deflate pneumatic tires could also be added? ;)
Not comparable. Treads coming off a tread drive (ANY tread drive) from base to base interaction is a chassis that cannot withstand the rigors of FRC. Poking holes in pneumatic wheels is damaging and deliberate contact inside the frame perimeter.
Causing a team with under inflated tires to pinchflat under heavy defense is the parallel.
I'd be very interested in seeing the results of the test Mike has outlined.
JamesCH95
29-01-2016, 16:59
I would be more worried about weakening the aluminum at the weld. We at AndyMark did extensive hit tests on these turnbuckles at full speed with a fully weighted chassis into a piece of diamond plate that was backed by a full pallet of roller chain, so it had no where to go. We are working on putting up some videos of our testing so that the community can see what we have done to test the modules before we were comfortable selling them.
The eyelets in the turnbuckles are aluminum? What alloy?
They appear as zinc-plated steel eyelets with an aluminum body, but you would of course know better!
Posting test videos would be a great way to address concerns of their robustness given the long the somewhat sordid history of treads in FRC drives. I look forward to seeing them!
It turns out my team wants to use Talon SRX with CAN signaling. Here is what I think needs to be done to run Rhino Drive:
1. Each Talon SRX would need a different id.
2. Each Talon SRX would need to be daisy chained.
3. Since my team want 3 motors on each side of the Rhino Drive, we will need 6 Talon SRX controllers.
How do I program 3 motors for left and 3 motors for right? Do I setup RobotDrive1, RobotDrive2 and RobotDrive3 each for 2 motors?
Thanks.
Quick example here.
https://github.com/CrossTheRoadElec/FRC-Examples/blob/master/JAVA_Six_CANTalon_ArcadeDrive/src/org/usfirst/frc/team469/robot/Robot.java
Basically uses robot drive which takes four CANTalons. Then create two more CANTalons and slave them.
Use tankDrive instead of arcadeDrive.
Alternatively you could just create six CANTalon and directly call Set() to manipulate the motor outputs directly, or to use the advanced control modes (speed/position/vcomp/currentDraw/motionProfile).
JamesCH95
08-02-2016, 12:07
Any updates on videos or turnbuckle materials? I am still quite curious.
team222badbrad
08-02-2016, 12:19
Any updates on videos or turnbuckle materials? I am still quite curious.
https://www.youtube.com/watch?v=v8oMG5NQS0o
Any updates on videos or turnbuckle materials? I am still quite curious.
James,
Sorry I missed your previous question. I am not sure what I was thinking when I was talking about weakening the aluminum at the weld. The outer body is aluminum, but the threaded rod and by extension the eyelets are steel. We did do some more testing with the chassis regarding t-bone tests, and have that up on the AndyMark YouTube channel, as posted above. I know we had some problems getting the videos to merge properly to post the testing to the channel, but I will see what we can do about getting the tests duplicated or uploaded individually. I know given the discussions happening here (http://www.chiefdelphi.com/forums/showthread.php?t=143290) that we may have more testing videos once the pulleys have been updated.
JamesCH95
09-02-2016, 15:20
James,
Sorry I missed your previous question. I am not sure what I was thinking when I was talking about weakening the aluminum at the weld. The outer body is aluminum, but the threaded rod and by extension the eyelets are steel. We did do some more testing with the chassis regarding t-bone tests, and have that up on the AndyMark YouTube channel, as posted above. I know we had some problems getting the videos to merge properly to post the testing to the channel, but I will see what we can do about getting the tests duplicated or uploaded individually. I know given the discussions happening here (http://www.chiefdelphi.com/forums/showthread.php?t=143290) that we may have more testing videos once the pulleys have been updated.
Cool, thanks!
vBulletin® v3.6.4, Copyright ©2000-2017, Jelsoft Enterprises Ltd.