Joined: 21 Jul 2012
Posts: 129
Location: Shopshire, England
Equipment
Envy _________________ Online store - www.ranglebox.com
Pulsar and friends.
Sun Feb 19, 2017 2:39 am
Nick Experienced Roboteer
Joined: 16 Jun 2004
Posts: 11802
Location: Sydney, NSW
I salvaged a nifty constant force coiled spring mechanism from a monitor base. The problem with using those springs is keeping them under control while the bot is being serviced - if you ever lose control of the end of the spring it will unwind everywhere. Elastic isn't perfect but it is cheap, light and easy to manage. Torsion springs around the arm's pivot point have also been discussed but I don't recall any pictures of them in actual use; elastic always seems to be the easier option.
Just thinking about Hooke's law, a torsion spring would potentially provide a better flip as it will need less force to compress it at the beginning of the lifting stroke, where most of the flipping action occurs. The other advantage is that I could probably wind a torsion spring on the lathe. Still going to try elastic first though _________________ Australian 2015 Featherweight champion
UK 2016 Gladiator champion
Sun Feb 19, 2017 8:28 am
Nick Experienced Roboteer
Joined: 16 Jun 2004
Posts: 11802
Location: Sydney, NSW
Sportman:
This is a rough mock-up of how the bungee cord will be installed:
There will be pulleys where that rod is and the bungee goes almost to the very back of the bot. having the bungee right up the front of the flipper arm gives it the best mechanical advantage for retraction and I'm hoping that I can get away with a smaller diameter cord. The pulleys will have a fairly large diameter and ball bearings to make sure the cord rolls over them easily - bungee cord REALLY doesn't like moving around sharp bends! Next step is to make some pulleys and test with some light weight cord that won't rip the MDF frame apart.
Laser power loss:
The laser was having trouble cutting 9mm MDF so I ran a full end to end power check; its rather depressing!
The 11.49% loss after the first two mirrors is not too bad - about what other users get for average quality mirrors. The real shocker is the losses at the output head. A 20.69% loss after mirror 3 either means the mirror is completely fucked or the beam is so off centre that its partly hitting the housing; that can be tested by adjusting mirror 2 up & down while taking power readings.
Next, there is the 20.69% loss in the lens (its suspicious that the two losses are near identical, so I repeated the tests and got the same results). The top side of the lens is quite dirty but most of the power loss is likely due to the beam being off centre.
Finally, there is the 'after the cone' measurement, which is the power that arrives at the table. The cone is where the air assist comes in and shouldn't have any loss at all. The 12% loss means that part of the beam is hitting the metal cone - it could even be that all the beam is hitting the cone as aluminium is an excellent infra-red reflector.
Fixing that is going to take several steps and much power testing along the way; it should be possible to reduce the power loss to under 20% or about 70 actual watts at the table. Going from 30W to 70W will make for some impressive cutting! _________________ Australian 2015 Featherweight champion
UK 2016 Gladiator champion
Mon Feb 20, 2017 1:57 pm
Nick Experienced Roboteer
Joined: 16 Jun 2004
Posts: 11802
Location: Sydney, NSW
Laser update:
I found part of the problem, the 3rd mirror is not even close to being in-line with the opening in the head. This is how it looks from the laser beam's POV:
I was aiming the beam at the centre of the opening so part of the beam was missing the mirror. To aim the beam properly, the entire mirror has to be removed:
Now that the beam is aligned with mirror 3, the output power has climbed to 65 watts, a solid 6.9% improvement. The mirror is still losing 13.8% so it must be defective. Mirrors can be made of several materials, these a probably silicon. Two better materials for reflecting infra-red are copper and molybdenum so I will order a set of copper mirrors. In the mean time the lens needs a clean and more power testing. _________________ Australian 2015 Featherweight champion
UK 2016 Gladiator champion
Mon Feb 20, 2017 11:31 pm
Nick Experienced Roboteer
Joined: 16 Jun 2004
Posts: 11802
Location: Sydney, NSW
Laser u7pdate:
Centering the beam after mirror 3 made a massive improvement; after cleaning the lens the power output at the table is 58.5 watts, up from 30 when I started the alignment. The lens is still losing 7.47 % power so its worth upgrading to a lens with better coating. At the current power the laser is easily cutting 9mm MDF and can probably go faster. Its time to buy some more MDF and cut out a prototype of an ant arena . _________________ Australian 2015 Featherweight champion
UK 2016 Gladiator champion
Tue Feb 21, 2017 11:35 am
teamshock
Joined: 08 May 2012
Posts: 5
Location: Reading, England
Having read through your efforts with the laser cutter I had a go at improving the one at work. I spent a couple of hours cleaning and realigning the mirrors and the results are so much better! We're cutting at over 3x the original speed which makes cutting pieces for several hundred students much faster- I completed a class set of ply pieces in under 10 minutes!
Wed Feb 22, 2017 7:51 pm
Nick Experienced Roboteer
Joined: 16 Jun 2004
Posts: 11802
Location: Sydney, NSW
That's good to hear, was it a Chinese import or something better? It is surprising how badly a laser can be and still do any work at all. _________________ Australian 2015 Featherweight champion
UK 2016 Gladiator champion
Wed Feb 22, 2017 8:02 pm
Nick Experienced Roboteer
Joined: 16 Jun 2004
Posts: 11802
Location: Sydney, NSW
Progress:
The pulleys to guide the bungee cord are done:
Grinding cutter is definitely not my forte but the finish came out OK after some smoothing with a Scotchbrite pad. the pulleys are 33mm diameter by 11mm wide and juuuust fit between the frame and the buffer tank. The bungee rolls over the pulleys very smoothly and while I can't test with real flips, I can't see why the bungee will behave differently at high speed.
The metal for the frame is arribing tomorrow so I am getting th CAD ready for a quote and recalculating the weight of the frame. _________________ Australian 2015 Featherweight champion
UK 2016 Gladiator champion
Sun Feb 26, 2017 8:34 pm
Nick Experienced Roboteer
Joined: 16 Jun 2004
Posts: 11802
Location: Sydney, NSW
Bungee cord retainer:
After messing around with bulky aluminium clamping ideas for the bungee, I found a really sneaky solution. The cylinder needs a 14mm plastic spacer under it, so I incorporated a clamp using a wedge right into the spacer:
The ends of the cord enter the slot from each side and two wedges hold the cord, tightening as the cord tries to pull out. The part is printed in Polymax PLA and should be more than strong enough with four perimeter layers and 50% infill.
Flipper frame:
The metal arrived for the frame and for the first time ever, I wasn't happy to see new titanium sheet. The sheet was described as 1/8" or 3.125mm thick and what arrived is 3.28mm thick. That means every slot & tab in the CAD drawing has to be modified and the weight goes up by 150 grams, even after adding extra holes _________________ Australian 2015 Featherweight champion
UK 2016 Gladiator champion
Mon Feb 27, 2017 8:54 pm
Nick Experienced Roboteer
Joined: 16 Jun 2004
Posts: 11802
Location: Sydney, NSW
Progress:
The bungee retainer from yesterday has some promise but the plastic needs more friction and the channel needs to be narrower to get enough grip on the cord - back to the Wanhou for another three hours
Meanwhile, the drivetrain is progressing well:
The motors and gearboxes are all ready; I went with smaller 250W out-runners and 38:1 gearboxes (mostly because they were lying around). while the motors have less than half the power of the ones in my featherweights, the extra reduction will provide plenty of torque and the low speed control should be excellent.
Once the ESCs are mounted & wired up, its time to see how fast the bot moves. _________________ Australian 2015 Featherweight champion
UK 2016 Gladiator champion
Tue Feb 28, 2017 5:57 pm
Nick Experienced Roboteer
Joined: 16 Jun 2004
Posts: 11802
Location: Sydney, NSW
Mutant Pacman:
The new cylinder spacer / bungee retainer is done and reminds me of a certain retro game The cord is very securely held and I can probably hammer in a longer wedge for even more holding power:
This is how it looks in position:
Its light too - 36 grams for the spacer and 56 grams for the cord. _________________ Australian 2015 Featherweight champion
UK 2016 Gladiator champion
Wed Mar 01, 2017 11:16 am
DumHed Experienced Roboteer
Joined: 29 Jun 2004
Posts: 1219
Location: Sydney
I've always found tying a knot in the cord also works well _________________ The Engine Whisperer
- fixer of things
Wed Mar 01, 2017 11:50 am
Nick Experienced Roboteer
Joined: 16 Jun 2004
Posts: 11802
Location: Sydney, NSW
True, that was plan B if the 'Pac-tainer' didn't work. There isn't much room around the cylinder and there is a chance of the knot moving and jambing the arm up. _________________ Australian 2015 Featherweight champion
UK 2016 Gladiator champion
Wed Mar 01, 2017 11:57 am
Nick Experienced Roboteer
Joined: 16 Jun 2004
Posts: 11802
Location: Sydney, NSW
First time ever
In over a decade of bot building, this is the first one to be severely underweight; I just spent a couple of hours weighing and checking all the parts again and so far Mr T is only 10.9Kg. With only a few plastic parts and the wiring loom to be made, I can't see it going over 11.5Kg unless the frame is beefed up. To start that process, I made yet another prototype:
The new frame is 20mm wider at the back to give more clearance between the armour and the CO2 tank. The panel in front of the electronics is now angled forward to give more room for the wiring, the flipper arm has lightening holes - not that I need to save weight, I just like the look of them . Every single freakin' tab and slot has been resized for the right material thickness.
Some other options might be to use thicker HDPE armour or to reinforce the corners with titanium strips to deflect saw blades - with around 2Kg of spare weight there is plenty of room to experiment. _________________ Australian 2015 Featherweight champion
UK 2016 Gladiator champion
Fri Mar 03, 2017 9:29 pm
pilleya
Joined: 31 Mar 2016
Posts: 91
Location: Sydney
Looks very neat, especially the little shock mounts for the electronics. Did you account for the weight of the CO2 in your calculation?
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum