Gas station without pumps

2017 November 25

Bugs found in first assembly of robot

Filed under: Robotics — gasstationwithoutpumps @ 11:18
Tags: ,

I got the voltmeter panel built and photographed Thursday morning, but I spent Thursday afternoon having Thanksgiving dinner with my brother-in-law’s family, and I spent Friday doing SolidWorks, so I did not get time to blog. Here is the voltmeter, in a test setup:

The lower-left segments on the ammeter seem to be non-functional, and neither the voltmeter nor the ammeter is very accurate, but it will do as a quick-and-dirty battery check. I’m using a barrel connector to provide the power to the robot, because barrel connectors are easier to work with than Deans connectors, and more robust than Molex connectors (the two connectors used in the Mechatronics course). 

I have a Futaba-to-barrel umbilical to connect the battery to the robot. I’ll also make a Deans-to-barrel umbilical, so that I can use the batteries in the lab, but I’ve not done so yet.  I was planning to use a 12V wall wart for testing, but the wall wart seems to have a 5mm/2.5mm barrel plug, instead of 5mm/2.1mm plug, so it does not make reliable contact. It may be better to replace the connector on the wall wart, so that I can standardize on 5mm/2.1mm barrels in all my projects.

Yesterday was spent doing SolidWorks modeling: fixing a number of known problems with the mechanical design, adding another tape sensor location, modeling the microswitches used for the bumpers, and designing the bumper. I got the pieces laser cut and assembled yesterday also:

The back view shows the voltmeter and the battery in context. I’ve not shown the cat-toy ball race on top, but it just fits within the 4 threaded rods, with room above it for the (undesigned) third layer.

The bottom views shows the M10 ball-end set screws, the wheels, the velcro strap for holding the battery, and the holes for the tape sensors.

The bumper design was not really successful, because I’m not good at visualizing things.

The bumper design as it was intended, below the second layer.

On the left front, I goofed—the between-layer spacer panel blocks the bumper.

I was doing the bumper design in SolidWorks above the second layer, rather than below it, and everything looked clear there, as the spacer panel does not extend past the MDF. But mounting the bumper and the switches above the second layer doesn’t help, because I did not allow for nuts around the threaded rods—the bumper hits the nuts instead of the spacer panel. In a final design with a third layer, there would be no nuts on the second layer (the rods are just clamping the layers together from below the bottom to above the top), but I’d like to be able to debug the bumper without needing a third layer.

I will have to recut the layers, because of the following errors, found after assembling the robot:

  • The motor-mount holes don’t seem to be quite right.  I wonder whether they did not get properly updated after I fixed the motor mounts.  I can force the motor mounts on, but the alignment is not what it should be.
  • The battery-to-panel umbilical wire sticks out past the edge of the robot, probably because I made the heat-shrink strain relief too long.  Fixing this doesn’t require recutting the MDF, but may require making a new umbilical with shorter strain relief.
  • The velcro strap holding down the battery should be moved closer to the wire end, so that the charging connector and the extra (JST?) battery connector can be held back out of the way.  This would also make the velcro a little more accessible.
  • I need to move the front spacer so that it doesn’t interfere with the bumper.  This seems like a cleaner fix than redesigning the bumper to avoid the nuts on an above-layer mounting.  Another alternative I considered was using the threaded rods for the bumper support, rather than separate M3 screws as I’m currently doing.  This would mean redesigning the bumper, and may make assembly a little trickier, but not by much.
  • The bumper may extend slightly too far forward—it just fits within the 11″ maximum, even though a little of the front of the arc was cut off (I was squeezing the bumper cut into a piece of scrap MDF which was just a little too small).
  • The mounting holes for the switches match the holes in the switches nicely, but they are 2.5mm holes, and I have M2 screws. I’d get a little better mounting if I switched to 2mm holes in the MDF, rather than 2.5mm holes.  I’ll probably fix this by adding 2mm alignment circles to the switch models, and using “Convert entities” to transfer them, rather than transferring the 2.5mm holes as I currently do.
  • I’m wondering whether the springs in the switches are strong enough to restore the bumper to its rest position.  They seemed to be when I did initial checks above the first layer (without the nuts in place), but I seem to be getting more friction now.  I might need to add a small spring increase the restoring force.  (Where? what sort of spring?)  I have lots of little tension and compression springs, but not torsion springs, which would be easier to set up.
  • I forgot the screw holes for the tape sensors!!!  I put in the rectangular holes for the tape sensors to stick through the board, but forgot to transfer the hole markings to the base plate.  I can fix this by marking the holes and drilling them manually, but I should update the SolidWorks model to have the holes, because I need to recut anyway.
  • The right-hand tape sensor hole somehow ended up 8mm×8mm instead of  8mm×12mm.
  • I did not provide holes for a cable tie to attach the inductor for the track-wire sensor below the 2nd layer.  I need to decide whether I’ll have just one track-wire sensor (deciding to fire when the signal is maximized) or two at right angles to each other—looking for the zero in the one that is end-toward-wire while the one that is edge-toward-wire is high enough to indicate that the wire is there.
  • I don’t have a mount yet for the beacon detector—I should add a third layer for this, even if I don’t have anything else to mount on this layer yet.  The third layer will need support spacer boards, which can also be used to fix the cat-toy ball race in place.
  • No screw holes for the power board, mux board, or track-wire/Teensy board.  This was deliberate, as I had not yet decided on the sizes for these boards.  Laying out and soldering these boards was this weekend’s task.  I can use M2 wood screws to attach the perfboards, but I think I’d rather use machine screws and drilled holes.
  • I’m not pleased with the ball-end set screws as ball casters—the balls do not turn very freely, even with lubrication, so they act more like rounded-end skids than like ball casters.  The adjustability of the height is nice, but I’m wondering whether I would have been better off with a larger ball caster that was designed for the purpose.

1 Comment »

  1. […] spent most of my day today with SolidWorks, fixing the problems noted in Bugs found in first assembly of robot and adding new layers to the robot. I’m getting a bit better at using SolidWorks, but I still […]

    Pingback by More SolidWorks | Gas station without pumps — 2017 November 28 @ 22:24 | Reply

RSS feed for comments on this post. TrackBack URI

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Blog at

%d bloggers like this: