@whobbs1496 set the channel purpose: To discuss all things climbing
Maybe we can crowd source some analysis of the climber. A good first step may be to break the overall problem into small sequential steps and analyze each piece. I'm thinking about these steps: driving to the area of the rope, seeing the rope precisely enough to approach it for climbing, approaching the rope, manipulating/guiding the rope for connection to the robot, connecting the rope to the climbing mechanism (the hardest part possibly), spooling the first couple revolutions of rope onto the climbing mechanism, the start of the climb when part of the robot is still on the ground, the in-air climb, the end of the climb, pressing the trigger plate, stopping the robot climb while sustaining pressure on the trigger plate & without descending.
What am I missing as far as the process broken into small steps?
*Thread Reply:* Trying out the new Thread feature of Slack. Copying coach's addition to the list of small sequential climber steps: Enrique Chee [10:51 PM] method to remove robot after game is over ?
*Thread Reply:* Coach thought of a post-match step. How about a few pre-match steps: Prepare the rope (including backup ropes), prior to each match place the rope on the field according to the rules. Also another in-match step: pilot drops the rope when there are 30 seconds left.
*Thread Reply:* Another post-match step: Disconnect the rope from the field
One assertion I'd like to make about the climber is that it may be best to mount it to the chassis since it has to hold the entire robot's weight. If placed elsewhere, there will be stresses on connected joints/structures due to torque & those joints/structures may have to be overbuilt (vs. the requirements of their other purposes) as a result.
Here's a link to a collection of pictures of RI3D & FRC team climbers: https://drive.google.com/open?id=0B4PZHoaYufG1WmdMcHRiUVZfRUk
Wow , Chris .!!! Let's get some cad folks to add climber modules to our existing cad file next week. I have ordered paracord , webbing , Velcro , and the rope from the game manual as Chris R suggested .
OK, I moved the climber steps to a spreadsheet & added some framework for analysis. I did not fully fill this out, but I did do a few rows. Here's the Google Sheet: https://docs.google.com/spreadsheets/d/1dnWTJIN4CAGP5R1NSg4AeI0IX9QV2BJsT1ecEZcycqk/edit#gid=0
I encourage anyone interested to go to the sheet and add knowns (including rules), requirements, solution options, etc. I think if we do this pre-work it can help the team hit the ground running when there are resources available to work on the climber.
looked at team update 5 - some good news in my opinion - Velcro binding coating on the back of sew-on Velcro is not a disqualifier & as long as the associated knot is not more than 2" below the davit fingers a loop may be used around the davit fingers as a retaining feature... Rope must...
D. consist entirely of (except for an adhesive applied by the VENDOR as part of the normal manufacturing process for a COTS item and no longer tacky, e.g. a “binder coat”) flexible, nonmetallic fibers sewn, twisted, tied, woven, or braided together except for the last 4 in. (~10 cm) of any cut end (E) which may be whipped (with material that is flexible and non-metallic) or fused only to prevent fraying.
E. be configured such that it engages securely with the FIELD. with a Retaining Feature (RF) that does not extend more than 2 in. (~5 cm) below the DAVIT fingers.
With these and other clarifications, they have really made climbing doable. The most prominent field pieces are the airships - even in district matches they want the spectacle of robots ascending for takeoff! And so they're making it doable. I thought the hard part was going to be connecting the robot to the rope, but Velcro works for that. Now I think the hardest part is stopping at the top so there is sustained pressure on the plate (without burning out the motor & without breaking the dang plate). From looking at Chief Delphi, I believe over 90% of teams will attempt to climb and I believe over 50% of teams will be able to do it reliably - too many points to leave them on the table - better believe teams are investing in this. I personally view this as a must-have. Viva la escalador!
Good to see you are reading rules. Hope the students are . Again, I agree about having a climber but the priority is LIA right now .
agree... just need to reserve some volume somewhere on the robot for the climber... can't afford to achieve LIA in a way that prevents climber IMO - should be on the list of requirements for LIA
@michael_nelson has joined the channel
Worth trying as an option. I like the compactness & that it already has flanges. There are still unsettled factors that could lead us to use something different, of course, but it could end up being the best choice.
The spool seems a bit narrow for velcro to attach to paracord/velcro . I think we need a wider spool, greater surface area.
Agree more surface area = easier/faster catch the rope
would have to adapt 3/8 to 1/2 - I assume that's doable
I did a little math at lunch, and depending on diameter of climber cylinder I think something closer to 35:1 may be better than 70:1. If we happened to use the 3.1:1 winch we bought, then a ratio of closer to 10:1 or 12:1 would probably be better. Do we have the right gear kit(s) for the versaplanetary to support different gear ratios for different options?
Looking at Q&A, the premise of a knot that comes untied (like our slipknot) was discussed and not rejected - requirement is location of knot must be the same - no problem, we can mark the rope
Q175 Inspection Policy If Rope Is Untied Then Retied Identically If the actions of a ROBOT result in a knot in a ROPE being untied, and the rope is then retied in the same way it originally was, would that rope have to be reinspected? asked 18 days ago by FRC 639 ROPE Answer If the ROPE were re-tied in an identical manner, it would not need to be re-inspected. If there were any change in knot location, materials, etc, it would require a new inspection.
Note that if we somehow extend the rope (pull on the slipknot) and then release it, we would be in violation. What kind of foul is that? (ie, point cost?)
(and is it possible for the other team to pull on our rope to release the slip knot, resulting in a foul against us?)
From Q152, as long as the rope length from the retaining feature to the end is within the acceptable range (5'3" - 8'), there's no foul. The default rope is 7'2" and goes from the retaining feature to the ground, so an 8' rope would be laying on the ground for 10 inches. If we wanted to have the rope go to the ground with a slipknot in it, we'd have those 10" to incorporate into the knot. 10" would allow for over 1 1/2 revolutions around the 1 1/4 cylinder with little load, which should be enough.
We'll definitely keep looking at rules for things we've missed & anyone who sees risks, please keep them coming.
Won't that be two ropes tied together? Is there a rule that might prevent that?
9.104.D from manual: [Rope must] consist entirely of (except for an adhesive applied by the VENDOR as part of the normal manufacturing process for a COTS item and no longer tacky, e.g. a “binder coat”) flexible, non-metallic fibers sewn, twisted, tied, woven, knitted, crocheted, intertwined, or braided together except for the last 4 in. (~10 cm) of any cut end (E) which may be whipped (with material that is flexible and non-metallic) or fused only to prevent fraying.
2 ropes tied together can compose 1 rope is my read
Yep, it definitely says 'tied'. Cool! And I think that is a very elegant way of creating an extendable rope.
According to this https://frc-qa.firstinspires.org/qa/22, a rope can be composed of two or more different kinds of rope, straps, or whatever other material meets the spec - some team will surely prank this by trying to pass inspection with 17 different rope remnants tied together
There have been so many rules clarifications and Q&A items on climbing and the team-provided ROPE in particular. I've attempted to collect everything in one place so it is easier to get a broad understanding of all the relevant rules, constraints, etc. Here it is...
@chrisrin pinned their PDF All Things Climber from Rules, Q and A, etc. + field pics.pdf to this channel.
In the next meeting or two we will be making design decisions for the final climber. If you want to be in on that, please review the "All Things Climber" compilation of rules and Q&A above before the next meeting. There are lots of rules, nuances, clarifications, etc. on this (covered thoroughly in the above), and I would like NOT spend a lot of time educating folks on that & instead spend the time on design, build, and test.
I thought the gear handler was to be moved closer to the launcher to provide more space for the climber.
Thanks Dana - looks pretty cool. We're thinking of using a variable length rope (i.e. it slips) to allow low-load spooling, but we're not ruling out something like what the 1598 robot has. I'm a little concerned about how much space the pivoting mechanism might require.
Here's a summary of the design decisions already made and still to be made for the climber. Intention is to use this to facilitate a discussion with the students and then close in on decisions so we can build the final climber. If anyone has additional options or pros/cons for the existing options, go ahead and send them my way.
Reminder: Read the "All Things Climber" rules and Q&A summary if you want to be in on the climber design decision discussion
My review of the posted cad picture tells me the climber must be made with the motor and winch drum parallel and connected by a chain drive. There are two good choices for the drum: 1/2" pipe (0.84" OD) and 1-1/2" pipe (1.90" OD). There is one choice for the rope: 1/4" diameter. The drum axis should be angled so a perpendicular to the axis at mid length passes through the robot center of gravity. The assembly should be mounted so it can take up 5.3 inches of rope with the big drum or 2.7 inches of rope with the small drum before lifting the whole robot weight.
Gear ratio for the big drum is 70:1. Gear ratio for the small drum is 35:1. The drum absolutely needs a brake for lowering. It is very easy to incorporate in the drum design. I can provide details on Sunday.
Thanks Paul. I'm not sure that's the final CAD. As you're indicating, the final CAD will potentially eliminate some options for some of the decisions we need to make. As far as making the design decisions, the plan is to have a collaborative discussion with students, making sure nonviable options are crossed off (& explain why) and the viable options and associated pros/cons are identified. Of course, sometimes it is really a set of decisions & how they work together (e.g. combination of drum size and rope diameter), and we'll get into that as well. I will add 0.84 OD to the list of drum options. As far as gear ratio decision by option, I'd have to look at your math, the motor power curve, etc. to understand, and it would be good for the students to also see that. At least one team on Chief Delphi indicated they are having success with CIM+1.25 diameter+35:1, which would climb faster than 70:1, so again... I'd like to see your rationale. As far as the absolute need for a brake for load, I need to understand that more. If the ratchet takes the load & if teams are allowed to hold the robot, disconnect the rope, and lower the robot with the rope still around the winch I don't understand the "absolute" nature of the need.
and one additional thing to keep in mind: the electronics haven't been considered in CAD drawings that I've seen. There was a plan to "go vertical" which I interpretted to mean to consume 4-5 inches along a side (from top to bottom). NB: This is pure speculation on my part.
I have a motor spreadsheet and a design calculations spreadsheet but I am limited in my knowledge of Slack to distribute them. My computer is not connected to Slack and I want to keep it that way.
@paul_vibrans : or you could email it to someone for them to post it to slack
Here's a Chief Delphi thread that covers very well rope length, various field dimensions, etc. https://www.chiefdelphi.com/forums/showthread.php?t=154450&highlight=davit+measurements
@paul_vibrans: you were talking about the need for a fairlead including rollers on on or both sides. Would it make sense to ask the coach to order something we could adapt faster that custom building? Would something like this be adaptable? https://www.amazon.com/gp/aw/d/B00426L7DM/ref=mp_s_a_1_31?ie=UTF8&qid=1486789565&sr=8-31&pi=AC_SX236_SY340_QL65&keywords=fairlead+roller&dpPl=1&dpID=41YLgwr6EWL&ref=plSrch
@connor_weiss Here are instructions for the versaplanetary we need to assemble https://content.vexrobotics.com/vexpro/pdf/VersaPlanetary-v2-User-Guide-20161123.pdf
The fair lead from Amazon looks way too big and heavy for our application. It is probably for a 2000 lb rated winch. It is certainly intended for wire rope. I have had second thoughts about the fair lead if we have the right shape for at least one drum flange.
@paul_vibrans - thanks for sharing that. Much more complete analysis framework than my back-of-napkin quality spreadsheet 🙂. Based on what I'm reading on Chief Dephi, teams using CIMs for climbing are achieving sub-5 second climbs with no issues. We'll need to prove it through testing, but I believe we'll have more current available than assumed in your calcs & be able to use a more aggressive gear ratio accordingly. I have an idea that will allow us to (mostly) use the very small diameter 550 paracord rope, which should also help. One thing you pointed out that we'll need to mitigate is the risk of the rope dragging across something & increasing load... that'd slow us down.
Torque and current are directly related. RPM and current are inversely related. To use more current you will slow down. This assumes constant voltage. Everybody that is testing at this stage is probably using fully charged batteries with output voltages near 12, which gives either 20% more speed or 20% more torque. Also testing may be going on with the bumpers off.
I was thinking speed on the other side of the gearbox (slower motor with more torque + more aggressive gear ratio could mean faster), but I will defer to your experience. Good news is the coach acquired for the team a new versaplanetary v2 gear box with a bunch of different gear kits that will allow many different ratios between 3:1 and 70:1, so we can flexibly try some different options and make a choice.
Chris, can you make sure someone print out the new instructions for Vex gear box before they assemble the new one. The one in our binder is for a different version from 2 years ago. Here is the link:
After it is printed out , we will 3 hole punch it and put in binder for reference. Thanks
Here are Paul's climber concept drawings based on his calculations + the assumption that motor and winch drum will be in parallel
There is some confusion in the programming team about how the code will end control of the climber when it reaches the top. We were originally under the impression that the driver would turn the motor on and off with a button on the joystick, and would observe the climb visually. When the robot reaches the top of the rope, and the light illuminates, then the driver would turn off the motor. The design I observed included a ratcheting wrench to keep the rope from unwinding.
Dana mentioned that Paul suggested using the motor to maintain position on the rope (ie, always keep some voltage on the motor after reaching the top). I'm concerned about this for a couple of reasons. The primary reason is that we don't have encoder position feedback of that motor, and therefore won't be able to control position by varying the current to the motor. Instead, we will just have to pick a voltage (percentage) and leave it applied, resulting in some forward force pushing up against the switch. This percentage will need to be fairly high to account for variable voltage drain through a match, and therefore will dump excess current into the motor that would make it hot. (If we do need this solution due to the lack of a ratchet, then please heat-sink the motor and controller.)
My second reason for concern about using the motor to hold us in position is that motor current will be cut off by the field management system at the end of the match. The system needs to not "drop" the robot to the ground. Even if the gearbox back-winds fairly slowly, the motor will be generating power as it falls, and I am concerned about what that energy might do to the motor controller.
@adrien_chaussabel has joined the channel
@rose_bandrowski has joined the channel
With the climber design we have today it won't drift down when the power is off.
Someone needs to make dimensioned drawings of the climber parts by tomorrow's meeting to give to the people who make them.
@bodbaird @whobbs1496 @alexlf @lucasrininger Re: the parts drawings Paul mentioned, who would you recommend? Paul gave Bo the draft CAD, and I believe they were saved in the centralized repository where everything else is located. @paul_vibrans please note that Bo indicated there is an opportunity to make the drum longer if we want (space under the gear I think)
I can help out with manufacturing if needed, if that's what you were wondering
If the drum under the gear can still grab the hanging rope then more length is useful. Otherwise it is not likely to have rope wind on it.
I can have the drum under the gear. It will mean making a new climber with the same components just in a different orientation.
I will consult you guys to confirm my design but the way the things are now they will not fit and mount well into the frame.
Paul reminded me on Sunday that any increases to the distance between motor and drum need to be done in increments of 3/8 due to adding pitches to the chain. Also, he said the starting distance should be slightly less than the drawing indicates, also because of a need for a bit of slack in the chain.
The gear can't block the dangling rope from touching the drum.
@paul_vibrans: Need a guard of some kind on the chain? Seems like a good idea
We need to keep the rope out of the chain. Keeping fingers out of the chain is good too.
Looking at the load ratings guide for the versaplanetary gearbox. Here's a link: https://content.vexrobotics.com/vexpro/pdf/VersaPlanetary-Load-Ratings-Rev4-20161121.pdf
I'm not positive, but I think it may be wise to add a bearing to support the 1/2 inch hex shaft coming out of the gearbox. It seems like the current configuration has the sprocket pretty far out on the shaft. If all are confident we're fine, then so am I - just wanted to raise it as a possible concern. 35:1 is the max recommended 2-stage gear ratio with simple load when driven by a CIM. I think teams commonly push the gearbox more than that without problems (though I have heard of failures).
In the gearbox manual it has distances from the base of the output shaft that is recommends based on the torque on the shaft. An easy solution (I haven't seen the design) would maybe be to put a bearing mounted on something as close to the base of the output shaft as possible.
if you scroll up to Feb 12, you'll see a couple drawings in PDF files
We have room to add a bearing and we have the correct bearings in inventory.
FYI: I probably will not be at the meeting tomorrow until 6:15.
As far as work on the climber, it seems like a couple things can be done in parallel: 1) manufacture parts other than the drum (because drum length may change) and the housing side plates (because distance between motor and drum may change) and 2) finalize the CAD so dimensions of those two excluded parts above are finalized & the parts can be manufactured. Part of finalizing the CAD is also adding a chain guard to prevent rope getting caught and adding a bearing to support the hex shaft coming out of the gearbox... sound right?
We can make a drum if we start with the longest anticipated length. It is easy to shorten. The rope guard is not difficult to make from sheet metal or sheet plastic.
@bodbaird I saw that the design went back to motor and drum inline & 1/16" wall aluminum tubing with OD of 1 1/4 (same stuff as the intake rollers) as the drum. I originally bought a length of 1 1/8" oak wood dowel that will fit inside that tubing if sanded down a bit (for added strength). If the drum length is very short, then we may not need it, but if the drum length is longer then it would be easy to add the dowel inside the tube for peace of mind. The rope we're using is 550 paracord (except for a small length of derby rope at the bottom for catching the velcro), and the small diameter rope combined with relatively thin-walled tubing could potentially cause a failure. I believe the moments of greatest force may be the very ends of climbs when the climber reaches the top - could be several hundred pounds of force with the weight of the robot and the stall force from the motor. I have no intuition about how significant a risk it is. Anyway, let me know if interested in the oak dowel.
*Thread Reply:* I checked WCP's manuals again on the aluminum tubing, and it is clearly designed as an easy-to-use material for rollers rather than heavy weight bearing, though teams are obviously using it for climbers. I strongly recommend putting the sanded dowel inside the tubing - should take very little time to spin the dowel and sand it down a bit with a sander block so it can be inserted.
Will there be any opportunity to do a quick and dirty mount of the final climber onto an old robot to test climbing at full robot weight? We only tested the prototype at 30-40 lbs with the knowledge from history and other teams that a CIM appropriately geared is strong enough to lift the robot, and it would be good to test the overall solution (climber + rope) at full load without putting the competition robot at risk.
Very nice!! Thanks for shooting the video and sharing!
(will aborted before it reached the top, so no current spike at the end, initial spike is motor-on)
Interesting. I wonder if we should put a 30A breaker on that line, for additional safety. Looks like it would start tripping when the robot reaches the top. (I wonder if the 40A breaker was tripping...)
The breakers are thermal only which means they won't trip right away at 40 amps. Typical trip curves indicate one minute delay before tripping at rated current.
A close look at the motor current shows two current levels. The first is when the winch is lifting halvf the robot (and the ground is supporting the other half). The second current level is when the robot is hanging entirely from the rope. If the voltage is being limited for speed control purposes, current greater than 40 amps may not be possible. Shutting off the climber automatically at 35 amps looks like a reasonable control setting.
In all seriousness, thanks Dana for collecting and sharing the data, and I agree with Paul that auto stop based and current seems viable
I think optimum control would be to start at a reduced speed to engage the rope then go to full speed/voltage when the current rises as the load is applied the auto shut off at 35 amps or whatever stall current is with a depleted battery.
We can easily make it so there are two different speeds for the climber in the code
And use the current as the change but it requires testing
Everything you need to know about the climber and rope system and important rules to know
@chrisrin commented on @connor_weiss’s file Climber Inspection/General knowledge and Prematch checklist: FYI. Also reminder to inspect the rope every time before climbing, especially the stress points where the rope bends across the wood of the practice field piece. If there is damage, switch to the backup rope tied around one end of the frame.
I recommend making some more monkey's fist ropes as backups. We should probably assume we'll need to change the rope out between events, and maybe even one rope per day would be a good estimate. And if not monkey's fists, then maybe some other stopper at the end of the rope? (I think the monkey's fist is the cooler way to go, and maybe someone can teach others on the team how to do it.)
I believe Cruz is the expert on that knot. And Chris made a form for it.
Yes the plan is to create more ropes. I've calculated the rope needs to be 76" long from the bottom of the stopper knot to the bottom of the rope. I think we should create maybe 3 to 5 ropes.
Cruz is most familiar with our rope - he could teach others if he has time
I also recommend we have ropes of slightly different lengths be inspected... try the 76", if a bit too long move to 75.5", etc. Of course, the slipknot loop can also be made a bit larger or smaller to fine tune the length, but I believe doing that will require rope reinspection
I just realized meetings are early this week & I cannot attend... We currently don't have a final set of ropes. @CruzStrom @connorweiss @Sam Rosen Since you have experience, will you please make sure there are at least 3 ropes ready for the weekend? Thanks
If not enough monkey's fist knots can be tied, another option is just a small bowline loop at the top that will go through the two davit fingers and then wrap around the outside of each davit finger before securing the pin.
@CruzStrom @connorweiss @Sam Rosen They did a stress test on the rope tonight, letting the motor continue to run for several seconds at the top of the climb putting additional strain on the rope, and the rope broke. So we need the rope at the top to be stronger to mitigate that kind of scenario. Here's what I propose: Let's do a 3-strand braid of the same paracord at the top of the rope (from the bottom of the stopper knot down 30 inches), and also let's use the new 750 paracord (in coach's room). That would practically quadruple the strength of the rope at the top of the climb compared to the single 550 paracord. Cruz, do you think it would work to tie the monkey's fist with the 3-strand braid? Might only need one knot instead of 3 nested ones. Any other ideas?
Do you mean make a monkey's fist out of a braided rope"
Or we could just use a tight loop instead of a stopper knot as I've mentioned before or some other stopper knot
clarification: when I say braid I mean braid the paracord
I will pick up the paracord tomorrow and try to work on the braid at lunch.
Should I cut 2 pieces of rope off and use those 2 and the main section to tie it?
If you have other ideas from your experience fire away - this seemed like a fairly minimal change to the overall rope design. Could also use a different stronger rope at the top, but we would need to acquire it, etc.
I will try the braid. I just don't want to cut the rope too long or too short.
Maybe we should just go with a small bowline - would be strong enough, and no problem with predicting length (just not as cool :))
Cruz can you tie one rope with then new 750 pound test rope and a monkeys fist for tommorow?
We can also try the braid with a bowline if we think it is necessary
I should be able to. The monkey's fist would be best and I will try to make it tomorrow.
Most importantly, did anybody get a video of the fall and crash?
I think that it broke at the top because they stress tested it at the top.
One thing that would be a good idea before marking the resettable slipknot on the new rope: spend a couple minutes tweaking the rope length with the robot. Have the rope swing and the robot approach, and then determine what length best results in the rope coming to rest centered in front of the robot. Do this a few times without actually climbing. And then mark the slipknot on one of the ropes. That length becomes the baseline. In the first couple matches this weekend observe closely the same thing with the rope & robot on the actual field and adjust as needed. I have used measurements to date to determine rope length, but I think this would be better.
It broke at a knot that won't be there in competition, The connection to the back up system.
Someone should contact NET Systems on Day Road to see what kind of Spectra rope they have that might work. They once had 12 strand braid in small sizes that was much stronger than nylon and could be spliced easily.
Paracord is stretchy. Spectra is not stretchy. Paracord is better for shock loads.
Agree spectra is very strong but I have heard knots come untied more easily with spectra than other types of rope, so using it will mean checking knots regularly. Not a deal breaker, but something to keep in mind. If there is a local source it is worth trying.
NET Systems on Bainbridge Island might donate a 1 pound spool of 1/8" inch Spectra if you ask. If John Adams in sales has not retired, he is the person to contact.
I have spliced 12 strand Spectra with great success in a very high load application. It was 3/16" diameter however. I regularly loaded it to 1000 pounds.
I am heading to NET Systems on Day Road right now (noontime). I will let the team know what I find out shortly.
I have 4 pieces of 1/8" Spectra, each piece is 12 foot long. Donated by NET Systems. I will bring them to school this afternoon -- I'll be there by 3:30. Paul, John Adams has retired from NET Systems. The gentleman that helped me today is Xzalivar White (yes, I spelled his first name correctly). His only request is to send him a picture of the Spartonics robot. It would be great to also show the robot climbing on the cord. 1/8 inch Spectra comes in 600 ft. spools. Price is approx. 30 cents per ft. Luckily they had a few pieces laying around so we did not have to get involved with their production inventory.
*Thread Reply:* Thanks, Terry! We'll be sure to get some photos/videos for him.
Splicing Guide - Single Braid Eye Splice Lock Stitch https://www.jamestowndistributors.com/userportal/document.do?docId=911&title=Splicing+Guide+-+Brummell+Braid+Eye+Splice+%28Lock+Stitch%29
A fid is needed for splicing. NETS may loan one or give the dimensions for Cruz to make one from rod.
I missed the first match. Did the robot climb?
Yes we've climbed each time (I'm pretty sure)
I we are 3 for 3 on climbs. There's some concern about misalignment & risk of bearings failing I heard, but I think they'll figure it out.
The bearing next to the break on the Velcro drum had popped out earlier, we glued it back in with loctite red
It was only partially out so the climber was still functional
That is a brake. Your English grade just went down. Otherwise, great job. When can I expect to see a pressure of 30?
What happened in that last match that the robot didn't climb?
Just didn't have time, no malfunction of the climber that I know of.
I heard a ball got into the drivetrain causing lack of control
Is that what happened? Didn't know about that.
@Cruz_Strom: I think we should switch to sewn thread to mark the reset point on the slipknot rather than sharpie because an inspector could interpret rule I-04 D to mean sharpie marks are illegal. One of the Q&A's asked if it ok to color the rope to help visibility, and the answer was no because of I-04-D. I think it is a safe play to go with thread. https://frc-qa.firstinspires.org/manual/rule/I/04 https://frc-qa.firstinspires.org/search?query=rope%20color.
Next match, different inspector... I still think the thread slipknot mark instead of sharpie is a safe play...
Hey! Climber (& rope) contributors: thank you for doing the work to make it happen!
On the topic of replace climber velcro. The stuff that is on there now is nice and "grabby" because of how it has been gnarled up in my opinion. Plus it is REALLY stuck on there from what I can tell. Maybe instead of a full replace we could measure in an inch from each end of the drum, cut with a utility knife & remove those outer bands, and then just add a fresh 1-inch band of velcro on each end of the drum.
potential timesaver: if you have 1-inch masking tape just put a couple wraps of it on each end, and then use that as the guide for cutting
I just thought of another needed fix on the climber. If we have time, in between the left side of the fold-down part of the climber and the bracket mounted to the frame, we need to put a spacer. We just need something there to hold the climber fold-out part to the right. Currently, we have a ziptie which works because there is not enough pressure to break it but it is a temporary fix and we will need something more permanent. If the ziptie breaks, the climber will slide a spacers distance to the left and after a few climbs, the bolt head on the right will wear down and the right side of the climber would come off the worn down bolt head and would put all of the weight on the motor hex. If that happens, the weight of the robot will bend the motor hex and could drop the robot.
Good catch Cruz. Another thing I thought of on the rope. When marking the resettable slipknot with thread instead of sharpie, let's sew it loosely into the sheath of the paracord only in order to avoid creating a weak point in the rope.
another thought: just use another overhand knot above the loop to mark the reset point
I checked the existing climber design and the tapered shaft connections are in the proper orientation. The taper on the sprocket shaft is just too shallow and tightening the brake pulls the taper through the drive flange a little more each time. We need a new shaft and a new sprocket and the tapered bore of the drive flange needs to be changed to match the new shaft. I believe the sprocket is in inventory as it is the same type as on the climber motor. The bearing next to the brake will need to be changed from cylindrical bore to hex bore. I have drawings.
So when we come into the shop what should we prepare? Hex bearing, new sprocket, metal stock for new drive shaft...
We need to make a new shaft and remanufacture the drive flange. The old shaft can't be saved. I will bring materials on Wednesday.
Watching other climbers, it seems clear others are using more aggressive gear ratios to climb faster... Faster climb could mean finishing a climb in time or allowing for more time before climb for other activities. Would it be work trying at Cheney, given our decent positioning in the PNW? And then if it didn't work out could switch back to 35:1?
The majority of climbers use 775 pros, which are far faster than a CIM to begin with. I've seen climbers from other teams that use 775 pros at about a 50:1 reduction, which I assume would result in a higher rpm after reduction than a CIM at 35:1. Without switching motors, I don't know how much lower we can go on our reduction
Have any calculations been made to figure out the torque we need to climb?
Do we currently use a dual-speed setting for the climbing motor? A slower speed to latch on to the rope, and faster to climb? If not, we may need to implement that if we go to a faster climb rate, as it may be more difficult for the velcro to attach while the take-up spool is spinning too quickly.
I also think we could save a little time by making the slipknot loop smaller (maybe cut size in half) - I think we have a bit more slack via the loop than we probably need. Our original climbs with robot used a rope with a smaller loop. A couple wraps should be enough for the drum to take the load, and I think we have enough rope in the slipknot loop for more like 4 wraps.
@riyadth It would be good to know if we're climbing at full power - I remember testing at 80 or 90%. Folks with all these potential tweaks I'm just look for ways to cut a few seconds that could be otherwise utilized. If there is consensus it's too risk, no problem - just some ideas.
I agree that we should try to get a faster climb
I think we should try to stick with the cim but change the gear ratio
Climber software sets motor voltage to 90% by default, but it looks like it can be overridden on the smart dashboard
I see there is also a "slow" mode from a separate button that sets the motor voltage to half of the smart dashboard value (45% default for slow mode)
These values could easily be changed in the software, but I would recommend designing the hardware to use 100% for the main climb.
One thing to keep in mind is how easily we can stop when we get to the top of the rope. Faster climb means we're more likely to crash into the touch pad and break the rope (or something else).
At the last inspection we weighed somewhere between 95 and 100 pounds
And how much does the battery weigh? Important, as we probably will be climbing with a battery :-)
I think that when we replace the part on the climber, we will be able to climb at the speed that we were able to before. Which means we could shoot after the ropes drop and then climb.
With that fact, I think that we will be fine without changing the motor or the gear ratio.
If we change the motor or gear ratio, we need to make sure to be able to mount it to the same bracket or manufacture a new bracket.
The gear doesn't change size, it just gets some new internals.
Copying Paul's message from Driving Team channel: Paul Vibrans [9:02 PM] I reviewed my climber calculations and it looks like our time to climb can be reduced by about 2.5 seconds if we change the climber gear ratio from 35:1 to 25:1. The motor current goes up to 25A from 20A at 11 volts supply. Margin on stalling remains, particularly after the new parts are installed in the brake assembly. Is changing the gear ratio worth doing?
I'm not sure we have the gear kits for 25:1. I think we have 3:1, 4:1, 5:1, 7:1, and 10:1, but I don't think we have two 5:1. So we might need to choose from 20:1, 21:1, and 28:1.
Using a 21:1 gear will raise the motor current to about 30A and cause more voltage drop. Using a 28:1 gear makes no great improvement over what we have. If we can't do 25:1 I suggest going to 21:1 and if that tends to stall out or miss the rope, going to 28:1 rather than all of the way back to 35:1.
I just checked the climber at 10 volts and 40 amps with a 21:1 gear. It climbs in 5.2 seconds with some torque margin, according to the calculations. I recommend changing the climber gear to 21:1 in Cheney.
@chrisrin then why are we not climbing? Can't tell the issue from the videos
In our last match, we got stuck on the gear peg. Generally, when we don't climb we didn't have time because we were going for an objective that grants us a bonus r.p.
Did anyone do a really close inspection of the rope? Any plan to create a fresh rope to use?
The last 2 ropes were starting to go out of shape and get all chunked up so we made 3 new ones for worlds. One of the last 2 is still functional and will be used for a few matches before it will be switched out.