@Kenneth Wiersema has joined the channel
@Paul Vibrans has joined the channel
@James Sovick has joined the channel
@Kenneth Wiersema set the channel purpose: Scissor
@Cruz Strom has joined the channel
@Tanner Nalley has joined the channel
@Kaedric Holt has joined the channel
@Enrique Chee has joined the channel
@David Ard has joined the channel
@Riyadth Al-Kazily has joined the channel
@Justice James has joined the channel
@Mark Tarlton has joined the channel
@Will Hobbs has joined the channel
@Randy Groves has joined the channel
@Rose Bandrowski has joined the channel
@Peter Hall has joined the channel
@Harper Nalley has joined the channel
@Binnur Alkazily has joined the channel
@Kirsten Martel has joined the channel
@John Sachs has joined the channel
@Declan Freeman-Gleason has joined the channel
@Darwin Clark has joined the channel
@Jack Chapman has joined the channel
@Bill Bandrowski has joined the channel
@brian_bonifaci has joined the channel
@Dana Batali has joined the channel
@Violet Advani has joined the channel
@Ethan Rininger has joined the channel
@Chris Rininger has joined the channel
I recall someone saying the number of heights we'll be able to make our lift go may be quite limited by the characteristics of pneumatic actuation, which seems contrary to the goal of precise placing with a high placing bot. Is something like this a possible alternative? tsubakimoto.com/power-transmission/linear-actuator/zip-chain-actuator/zca/ Here's a video: https://www.youtube.com/watch?v=ysrjC5-zjS4 And here's a lighter duty one used by an FRC team: https://www.youtube.com/watch?v=ED-X4P6jX70 Product video: https://www.youtube.com/watch?v=ZSxFwPN-u9g
@Samantha Rosen has joined the channel
The solution to positioning is to add a toggle lock on the piston rod with a small air cylinder operating it. It is not quite as elegant as the Tsubaki device, more like the thing that holds screen doors open.
I'd like to see a student figure out what those height positions are. One for the exchange, one or two for the switch, one for the portal(?), multiple for the scale (depending on which tier and position of scale), and one for climbing....
Agree. Also, to be effective at precise high placing (similar to our humans play the game video), what are all the positions we need to be able to place precisely at? I think the scale alone may have up to 6 desired positions (which may or may not be able to be consolidated - need to do the math or get out the tape measure). Scale platform low first cube row, low second cube row, level first cube row, level second cube row, high first cube row, high second cube row
The issue in our application is not the push chain, which I have made, but the mechanism that pushes on it. It is slower than air cylinders.
Specs on the rigichain by Serapid are 300mm/sec which seems pretty fast
I just think the ability to have as many positions as we want + ability to adjust via software rather than hardware will end up being advantageous
The dart actuators would work for the flipper at the top of the scissor lift. Unfortunately they are out of stock.
1 ft / sec means 5-6 sec to lift one cube from ground to scale...Or perhaps less given the multiplicative effect of the scissor? How does this compare with pneumatics?
It seems like all the questions and conversation here go back to the fundamentals of systems delivery... Step 1: What are our detailed functional requirements / desired specs (written down, understood, and agreed upon across sub-teams). This includes the specific lift stops including heights, the speed we wish the lift to be able to go from down to fully extended, how fast we need to be able to adjust the height of one or more lift positions (e.g. if the actual field measurements turn out to vary from expectation & we need to quickly adjust), projected max weight the lift needs to raise, etc. Step 2: Design and build in the context of the detailed requirements / desired specs... The pitfalls of doing step 2 without having done step 1 can be huge.
I am trying to achieve 1 second from lowest scissors position to highest scissors position and 1 second for however far the flipper must go. When Step 1 is complete, I hope that what the mechanics produce meets the Step 1 requirements. The faster Chris' team completes Step 1, the sooner we can feel comfortable (or not) with the design as it is progressing.
@Paul Vibrans which dart actuators do we need ?
6 inch or 12 inch ?
None of them.
Do you know the specs for the dart actuators that you need ?
@Cory_Houser has joined the channel
I have to change the design to use Dart actuators. The design effort is better spent elsewhere.
@Paul Vibrans we could do both. Use an actuator to set the stop/limit for the pneumatic piston. Drive team could set the position while driving (since actuator takes longer), then they hit the up/down button to quickly activate the pneumatic/scissor-lift. Just a thought....
We could, but the device that sets the stop does not need the dynamic load rating or power of an actuator. I have a stop design for the vertical that is way simpler and we can test it on the prototype. The flipper can use a motor driven linear actuator that we can make from parts about as easy as figuring out how to connect to a COTS actuator for much less money.
@Brian McKinnon has joined the channel
@Jules Blythe has joined the channel
Hey Paul. Do we need to 3d print the pulley. Is it possible to order instead. The estimate for the 3d print is about an hour and with the amount we have to print, we would have to wait 5 meetings to get the pull fully printed.
@Terry Shields has joined the channel
We can machine the pulleys from plastic bar
The CNC should work for that btw.
It might indeed. We should order some 3" diameter PVC round bar from McMaster-Carr. It is catalog number 8745K63 and a 1 foot length should be enough for the number of pulleys we need.
3-D printing might get better consistency between parts.
I just checked my math and discovered we need more than one foot of PVC bar so a two foot length of it should be ordered. McMaster-Carr only sells it in even foot increments.
I would like to help the programming team get more information about how the scissor lift and associated mechanisms will be controlled. I spoke with @Paul Vibrans today during the meeting, and have a vague idea, so I'm going to repeat what I heard, and hopefully get an affirmation or clarification.
First, the scissor lift will be raised by opening a pneumatic solenoid valve. To reach different levels, the pneumatic solenoid is opened, allowing air to raise the scissor. A feedback mechanism (either discrete magnetic switches on the cylinder, or a potentiometer connected to the mechanism) will indicate height as it rises. The controller will stop air flow (and not allow the air to escape) when the desired height is reached.
Second, the "flipper arm" at the top of the scissor lift is connected by cables to a lead screw, directly driven by a CIM motor. I assume that there will be limit switches at each end of the lead screw to prevent mechanical damage (and reach home position correctly). Will there also be an encoder to allow for controlled positioning?
Does that seem correct? Are my assumptions of sensors on the system for position feedback already part of the design? If so I'd like to get the programmers some similar sensor hardware to develop code to.
Oh, and I don't know what (if any) mechanism control will be in place for a "grabber" at the top of the scissor lift. I assume there is a control required to "clamp on" to a cube at the bottom, and release the cube at the top. Will that be a simple pneumatic system (single or double solenoid)? Or is it motorized?
Thanks for starting this conversation.
The description of the scissor lift is correct as far as it goes. There will be potentiometer feedback on the scissor angle that can be translated into height. There will be potentiometer feedback on flipper angle. The air cylinder for raising the scissor will have an up solenoid and a down solenoid if the control valve has a closed center. There also will be a stop cylinder that can halt the rise of the scissor at any programmed location but it can not halt the descent. It will have an actuation solenoid but no de-activation solenoid as the cylinder is spring return. The scissor elevation cylinder will have as many as three (3) magnetic switches that can be used to set pre-programmed heights for the various game elements. I expect these to be switch height, scale low height and scale high height. The grabber has a single air cylinder that will have two solenoids in the control valve: grab and release.
Thanks Paul. I will work with the pneumatics team tomorrow to try and draw the schematic for this. The concern I have is that the scheme calls for non-standard solenoid valves (ie, center off position). I am also concerned about how we vent the pneumatics, and what that means to the scissor lift state. If there is a closed-off valve holding some air pressure that keeps the lift up, then venting the system with the pressure release valve may not lower the lift. While I don't think that is dangerous, it may not pass inspection, and it may also mean that the robot is difficult to move off the field if the scissor lift is left in the up position at the end of the match. Stay tuned for more questions and potential for control system restrictions on the design...
Please remember we also need a lift height for when the robot is butted up against the tower and the arm on the lift needs to catch the hook onto the ring. May or may not be able to reuse one of the heights mentioned
Maybe go to a higher height, swing arm over the ring, then let lift descend... hopefully would work
The scissor in Stronghold looks good but you can't tell from the photo if there is a cable inside carrying the load or not. Ours is certainly not designed to carry the weight of 3 robots. Please look at a traction machines as an alternative to a winch for providing the primary line pull. A winch plus a traction machine in combination can provide the high tension with two speeds. It is a common solution used with hydrographic winches. I can explain on Sunday.
Thanks - watched some videos but I can't say I fully understand. I think to keep the solution simple we're likely to go the route of two motors into a shifting gearbox, and one or two winch spools.
It depends on what you think is simple.
@Chris Mentzer has joined the channel
Thanks Paul !!
I will print for meeting today
@Andrew Peterson has joined the channel
Up stop is part of the scissor. The flipper wire attach lug is on the arm between the scissor and the grabber.
The linear actuator is part of the flipper but is mounted below the scissor.
Did the flipper arms, the ones that support the grabber, get slotted for the lift hooks?
Not yet, I'm waiting for mill time
Can someone send me a photograph of the front support for the scissor lift?
You might be able to come in after school today @Enrique Chee, I can be there to get you the material
I won't be there afterschool but I will have Kenneth get it for you afterschool.
Is 3:00 PM a good time? When does after school begin?
3:10 is when school gets out, I'll probably need a few minutes to get the parts ready
I have the part for you @Paul Vibrans. You can pick it up now before 2:55 to avoid school traffic.
@Paul Vibrans You are welcome to come now.
@Paul Vibrans@Cruz Strom@James SovickHave we manufactured the stop parts for the pneumatics?
We can remove 0.4 pounds from the robot by drilling lightening holes in the PVC pulleys. See the attached drawing. I can make a drill jig so the process can be fast. We need to undo the flipper wires to install the climber hooks so swapping drilled for undrilled pulleys can be done then.
Are we encountering weight problems with the robot?
Well, we’re going to be at 120 if that’s what you’re wondering, not a lot of room for other things. The robot is 110, but with the rest of the climbing stuff, poly carb, camera and bling we’re going to be cutting it close
Last year we used 1/8 inch thick polycarb. Can we go to 1/16 inch this year for half the weight?
I think we can
The motor on the linear actuator can be changed to a Mini CIM to save 0.6 pounds. I think we will have enough torque to move the flipper. At least we should test it to find out.
The new flipper pulleys need to look like the attached drawing. They start with 3/4" thick discs cut from 3" PVC bar with a chop saw. The manufacturing steps are:
1) Cut discs
2) Clean up one face in the lathe and drill out 7/8" diameter.
3) Clean up other face, machine to correct thickness and bore to 1.1245" diameter.
4) Mount on arbor and machine outside diameter
5) Machine counterbore
6) Mount on arbor and machine grooves
7) Mount in drill jig and drill lightening holes
Have you considered mounting a motor on the articulator?
I think a motor would be too heavy on the articulator... But I'm wondering if we considered cable tensioners (springs to keep the cable from loosening when driven against a fixed object), and/or cable guides to keep the cable constrained at the pulleys...
The ultimate solution is changing to #25 roller chain instead of cable.
I think we'd need more sprockets for that right? Since I can't imagine chain going on the pulleys well.
weight of the chain is something to consider here.
That is why we have cable now and it is an incentive to get cable to work. No free lunch.
i would venture the guess that weight of multiple feet of chain would approach or exceed that of a motor
What about using cam belts instead of chain? I would expect weight to be less.
i guess the problem would be in finding belts of exactly the right length to match the existing lift arms.
Don't speculate, calculate. The chain weighs 0.75 pounds. The sprockets weigh 0.82 pounds in addition to what is there. The weight is distributed along the height of the scissor so only part of it is penalized by the mechanical disadvantage of the scissor. Putting the motor to the top of the scissor means adding a 2.16 pound mini-CIM or 2.8 pound CIM (if we need the torque) and a 100:1 gearbox or a 0.8 pound 775PRO and a 300:1 gearbox where the effect of weight on the lifting cylinders is the greatest. It will be like lifting a second cube every time.
Using belts means finding pulleys that we can adapt to our existing shafts or making new shafts that can handle the extra width of the belts relative to "A" plate sprockets. There may be bearing issues, too.
There is also a control issue with the motor at the top. A motor at the top of the scissor works relative to its mounting so the flipper will rotate away from level as the scissor rises. It is not insurmountable but a nuisance.
If you have access to data, calculations are indeed valuable. Your calcs raise good points. I'm curious if you've calculated how much power we need from a motor? Is it possible that a window motor could do the job? (and how much do they weigh?). I'm unclear on the point you make about motor mounting, but I am clear that running a few wires up the scissor is way simpler than a pulley/sprocket scheme, as cool as that is. At the end of the day lighter is going to win the day, so I'm hopeful that simple spooling changes are all that's required.
Our other concern is time
the time it takes to "respool" if we encounter failures of the sort we experienced last meeting. Will estimated that it takes 10s of minutes to rewire the current system if it unravels. This would be unfortunate if it happens during a competition.
the time to get the current system fixed to the point where we're confident #1 won't be an issue
* the time taken away from programming while engineering/fixing this issue
some data from https://firstfrc.blob.core.windows.net/frc2017/2017-motor-information.pdf
bosch seat motor .8lbs, 3115 oz-in of stall torque 22 N-m, 24rpm
denso window motor .85 lbs 1309 oz-in (9.25 N-m, 90rpm
snow blower motor: 1.11, 1600 oz-in 11.3 N-m, 100rpm
are these no-load speeds sufficiently low to support direct-drive of the shaft?
A Bosch seat motor should work. Do we have one?
Rules say “Select automotive motors (Window, Door, Windshield wiper, Seat, Throttle)” Perhaps a different lightweight slow-spinning motor in one of those categories could work? Another idea: could send some inquiries out to other teams to see if anyone has a Bosch or two on the shelf.
If we have either of the others, we could use 3 to 1 chain drive of a few inches. I may have some #25 sprockets to share... Again, only presented as a backup plan pending analysis of the pulley + wire
First and simplest solution to counterbalance the flipper is suitably configured surgical tubing around a pulley opposite the one with the cables. It is a similar approach to what was done with the scissor.
Do we need to assist lifting the cube from the robot starting position as well as from the floor? That is do we just pull in one direction or do we assist in two directions around some neutral point.
One thought on the lift flipper cables coming of the pulleys. In addition to the springs that are already on there at the top, could you increase the base tension a bit using a small turnbuckle inline at the bottom? And then when tension decreases (maybe due to quick direction change or something) there is still enough tension for the cables to not come off?
@Jim Carr has joined the channel
@Justice James has left the channel
@Paul Vibrans has left the channel
@Cruz Strom archived the channel