Chief Delphi

Chief Delphi (http://www.chiefdelphi.com/forums/index.php)
-   Technical Discussion (http://www.chiefdelphi.com/forums/forumdisplay.php?f=22)
-   -   Drivetrain Question (http://www.chiefdelphi.com/forums/showthread.php?t=109089)

Ether 12-10-2012 11:09

Re: Drivetrain Question
 
Quote:

Originally Posted by tr6scott (Post 1190218)
Jags short the motor leads when off... Victors do not do this...

By "off", do you mean "unpowered"? If so, can someone please confirm this?



AlexH 12-10-2012 11:16

Re: Drivetrain Question
 
if the jag was in brake mode it would short the leads if it was in coast mode it would not.

Ether 12-10-2012 11:32

Re: Drivetrain Question
 
Quote:

Originally Posted by AlexH (Post 1190223)
if the jag was in brake mode it would short the leads if it was in coast mode it would not.

The question on the table is this: Will an unpowered Jag short the motor leads, regardless of jumper position.



Jon Stratis 12-10-2012 11:32

Re: Drivetrain Question
 
Quote:

Originally Posted by 1502 (Post 1189950)
The second goal of disconnecting the wheels for the motors is a matter of safety. Whatever way we end up doing it, there will be a deadman switch that works like one off of a treadmill. We like to keep our carts safe, and we're pretty well know for that. ::safety::

A thought for this...

Disconnecting the motor from the wheels with the deadman switch will lead to the cart coasting to a stop, potentially over a long distance carrying a lot or momentum (150 lb robot + 50 or more lb cart...) - in other words, it can defeat the purpose and still run over a little kid. Instead of going that route, use some Jaguars to control the drive train. Set the Brake/Coast header to Brake, so when it stops, the motor actually helps the cart slow down faster. Hook all the limit switch inputs up to a limit switch sitting behind a slot. Slide a card into the slot will close the circuit, allowing the cart to move. If the card is pulled out, the circuit will open, and the Jaguars will immediately stop, with no programming needed on your part!

This can separate your two issues nicely, which might let you use an easier solution for pushing the cart around than if the solution had to solve both problems at once.


Oh, another idea I just had... a similar limit switch/slot concept, except the limit switch would activate some pneumatic cylinders, which would push some wooden skids straight down into the floor. Set it up so they can lift the cart wheels + robot completely off the floor. Now, the deadman switch serves two purposes - it stops the cart immediately, and while in your pit it can give you a more stable working surface that won't roll around!

techhelpbb 12-10-2012 11:40

Re: Drivetrain Question
 
A dead man's brake doesn't need to actually be part of the drive assembly.

One can arrest the momentum of a wheeled cart just by creating friction between the surface on which the cart rolls say by dropping a rubber covered shoe down on that surface to create drag (might want to tapper the edges on that shoe so it doesn't stop quite so violently that the robot might come off the cart). It might be much easier to make a brake like this.

JamesCH95 12-10-2012 11:41

Re: Drivetrain Question
 
Quote:

Originally Posted by Jon Stratis (Post 1190227)
A thought for this...

Disconnecting the motor from the wheels with the deadman switch will lead to the cart coasting to a stop, potentially over a long distance carrying a lot or momentum (150 lb robot + 50 or more lb cart...) - in other words, it can defeat the purpose and still run over a little kid. Instead of going that route, use some Jaguars to control the drive train. Set the Brake/Coast header to Brake, so when it stops, the motor actually helps the cart slow down faster. Hook all the limit switch inputs up to a limit switch sitting behind a slot. Slide a card into the slot will close the circuit, allowing the cart to move. If the card is pulled out, the circuit will open, and the Jaguars will immediately stop, with no programming needed on your part!

This can separate your two issues nicely, which might let you use an easier solution for pushing the cart around than if the solution had to solve both problems at once.


Oh, another idea I just had... a similar limit switch/slot concept, except the limit switch would activate some pneumatic cylinders, which would push some wooden skids straight down into the floor. Set it up so they can lift the cart wheels + robot completely off the floor. Now, the deadman switch serves two purposes - it stops the cart immediately, and while in your pit it can give you a more stable working surface that won't roll around!

I very much agree with your first idea about disconnecting drive motors, very good observation.

I think your second idea could introduce a serious safety hazard. Pneumatic components that always return to one state when de-powered can actuate at un-intended times. If they are explicitly switched you might run into other issues. For example, the cart's battery could run low and the pistons could then actuate unexpectedly, causing the whole cart+robot to raise/lower without warning. If the skids default position were 'deployed' down, and there was an issue with the cart, it might come to stop in a hurry in a queue or pit lane, which would be most inconvenient.

The idea of using skids to stabilize the cart for working is good, but I would not rely on them as a safety mechanism, nor would I have them pneumatically actuated.

Jon Stratis 12-10-2012 11:57

Re: Drivetrain Question
 
1 Attachment(s)
Quote:

Originally Posted by JamesCH95 (Post 1190231)
I very much agree with your first idea about disconnecting drive motors, very good observation.

I think your second idea could introduce a serious safety hazard. Pneumatic components that always return to one state when de-powered can actuate at un-intended times. If they are explicitly switched you might run into other issues. For example, the cart's battery could run low and the pistons could then actuate unexpectedly, causing the whole cart+robot to raise/lower without warning. If the skids default position were 'deployed' down, and there was an issue with the cart, it might come to stop in a hurry in a queue or pit lane, which would be most inconvenient.

The idea of using skids to stabilize the cart for working is good, but I would not rely on them as a safety mechanism, nor would I have them pneumatically actuated.

Clearly I didn't explain that last idea well enough :) The pneumatics wouldn't return to a specific state when de-powered, they would work the same as pneumatics on FRC robots - they remain in whatever state they were at when power was removed (air leakage might result in the pistons slowly losing their pressure). The switch simply indicates which direction they would be at. If you wanted to do this without needing a controller on board, see the attached rough diagram.

JamesCH95 12-10-2012 12:11

Re: Drivetrain Question
 
Quote:

Originally Posted by Jon Stratis (Post 1190236)
Clearly I didn't explain that last idea well enough :) The pneumatics wouldn't return to a specific state when de-powered, they would work the same as pneumatics on FRC robots - they remain in whatever state they were at when power was removed (air leakage might result in the pistons slowly losing their pressure). The switch simply indicates which direction they would be at. If you wanted to do this without needing a controller on board, see the attached rough diagram.

I see, that would be better that what I was assuming. However, it still requires the electrical system to function for this safety system to function, and is susceptible to an inadvertent flip of the switch, which isn't ideal in my mind.

Of course none of this detracts from a pneumatic jack's usefulness and definite cool-factor :cool:

FYI: plenty of FRC robots have one-way solenoid valves, i.e. return to one state when not powered. It would be dangerous to assume that all robots use 2-way solenoid valves.

Jon Stratis 12-10-2012 12:42

Re: Drivetrain Question
 
True, there are 1 way solenoid valves... My team hasn't done much with pneumatics, and the teams I've inspected have all used 2-way valves, So I don't really have much experience with them. That must be the basis of my assumption :)

AlexH 12-10-2012 14:11

Re: Drivetrain Question
 
simplest and safest way to prevent unwanted actuation of the ram is to physically lock the ram in the extended or retracted position with a pin or something

JamesCH95 12-10-2012 14:57

Re: Drivetrain Question
 
Quote:

Originally Posted by AlexH (Post 1190259)
simplest and safest way to prevent unwanted actuation of the ram is to physically lock the ram in the extended or retracted position with a pin or something

Please explain how that would work procedurally.

1502 12-10-2012 16:19

Re: Drivetrain Question
 
I didn't mention that we would have dual disk brakes that would engage for a few seconds after an error to stop the drive.

My thought was to create a gearbox that only had one speed and a spring return cylinder that disengages the motor. Two of the wheels would have our disk braking system from above.

AlexH 12-10-2012 20:52

Re: Drivetrain Question
 
Quote:

Originally Posted by JamesCH95 (Post 1190263)
Please explain how that would work procedurally.

what i meant was something that works in the same manner as how the pin that holds the receiver and insert ends of a trailer hitch together except used to lock something in place and keep it from rotating.

another idea is you could put a piece of "c" channel that is smaller than the clevis over the ram's rod to physically jam it and keep it from retracting.

JamesCH95 16-10-2012 13:28

Re: Drivetrain Question
 
Quote:

Originally Posted by AlexH (Post 1190293)
what i meant was something that works in the same manner as how the pin that holds the receiver and insert ends of a trailer hitch together except used to lock something in place and keep it from rotating.

another idea is you could put a piece of "c" channel that is smaller than the clevis over the ram's rod to physically jam it and keep it from retracting.

The issue with this is that it relies on a human-in-the-loop to function. My opinion is that safety equipment on a robot cart and similar devices should not rely on human involvement. If someone forgets to install the pin, or student A thinks "student B did it" and student B thinks "student A did it" then you have a situation that is exceedingly dangerous as those around it are assuming the piston is locked out when in fact it is not.

techhelpbb 16-10-2012 13:50

Re: Drivetrain Question
 
One could make a latch system that works like a automotive jack stand.

As you extend the piston to lift the cart it would ratchet it's way down and then prevent reverse movement without manual intervention. In fact they make small cheap automotive jack stands one could probably tweak for this purpose.


All times are GMT -5. The time now is 06:38.

Powered by vBulletin® Version 3.6.4
Copyright ©2000 - 2017, Jelsoft Enterprises Ltd.
Copyright © Chief Delphi