Logobot Bumpers

[Robert Longbottom]

Hi,

I've been having a playaround with an idea for some bumpers.  The actual bumpers look ok, but the way they interface with the microswitches seems a bit naff.

Have a look in the Bumper branch on git, in the sandbox is BumperDev.scad (and MicroswitchHolderDev.scad).  It's not quite complete because there is no way to attach it to the base yet, but that could be done by adding some lugs on the microswitch holder to interface with a hole in the base.  I've also temporairly added it into the main LogoBot model so you can see everything together to check ground clearance  - temporary because it's not using attach(), probably needs a seperate BumperAssembly.scad, instructions, etc, etc.

Some screenshots:

[Jamie Osborne]

Ground clearance does look like it'll be an issue from those pics. Can the bumper be shorter and moved up at all?

Jamie

<fjhafegi.png>

<idjebigg.png>

<ejjbjdhd.png>

<eedijiac.png>

--
You received this message because you are subscribed to the Google Groups "swindon-hackspace" group.
To unsubscribe from this group and stop receiving emails from it, send an email to swindon-hacksp...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Gyrobot]

A good start, Maybe wrap the bumpers around the sides so the front and rear bumper meet in the middle for better side collision detection.

[Robert Longbottom]

Yes, maybe.  That's just a parameter that can be changed so easy to tweak.

--

[Robert Longbottom]

I think its actually okay, though it is quite close.  But I've had a better idea for the microswitch connector that will be neater and give it more ground clearance, so I'll model that and see how it looks.

[Damian Axford]

Cool...  random thoughts/ideas:

• Could try changing the sliding joint for a curved "spring" so the bumper is part of the same print as the microswitch holders (hope that makes sense, can draw a picture if not)
• Can probably reduce the thickness of the bumper wall to 1mm (approx 2 perimeters)
• what ground clearance does that give?  I would have thought 5mm would be sufficient, although it could probably be more to save plastic/weight...  most collisions are likely to be with walls, other robots - generally things more than a few mm high.
• I like Steve's idea of wrapping the bumpers round the sides...  if only cos it would look cool

[Robert Longbottom]

Just pushed some updates, that give better ground clearance as the result of some lower profile mircoswitch holders.  Also changed the bumper wraparound value to bring them further round the body.

I chose 2mm as the bumper thickness thinking it wants to be reasonably strong for bashing into things :-)  But thats just a parameter too.

I like the idea of the "spring" connectors if they can be made springy enough.  I'm thinking it doesn't have much grip on most surfaces (or get up and go given how it struggled with the a slight incline on that chair), so you need to hit something hard enough that you can push against whatever is required to trigger the microswitch.  So there is a spring in the microswitch and adding more resistance with a spring in the bumper might just be too much.  We can try though.

[Damian Axford]

have added some springy bits to the Bumper branch... for illustration.  I suspect at 5mm high, they'll be a bit stiff...  prob either reduce that to around 3mm or make them a single-perimeter thick.  I'd be inclined to try single perimeter first.

I'm expecting the spring in the microswitch to have negligible resistance (as they're tiny microswitches), but agreed that the bot hasn't got much oomph, so the bumper needs to be very easy to trigger. :)

[Jamie Osborne]

Would it cost much more (££ or space) to have more powerful steppers (or servos with wheel encoders)? Would be a shame to limit what we can do due to choice of drive.

Jamie

On 2 Dec 2014, at 17:11, Damian Axford <dam...@axford.me.uk> wrote:

have added some springy bits to the Bumper branch... for illustration.  I suspect at 5mm high, they'll be a bit stiff...  prob either reduce that to around 3mm or make them a single-perimeter thick.  I'd be inclined to try single perimeter first.

I'm expecting the spring in the microswitch to have negligible resistance (as they're tiny microswitches), but agreed that the bot hasn't got much oomph, so the bumper needs to be very easy to trigger. :)

[Robert Longbottom]

Ah yes, that's the kind of thing I thought you meant.  That will look nice if we can get it to work.

I printed a springy thing from thingiverse on Mon night to test.  That was 5mm high I think and two perimeters and it seems quite flexible.  I think two perims would be better if it works because it will be easier to print and stronger.

Will have more of a play in the bumper branch.

On 2 December 2014 17:11:19 Damian Axford <dam...@axford.me.uk> wrote:

have added some springy bits to the Bumper branch... for illustration.  I suspect at 5mm high, they'll be a bit stiff...  prob either reduce that to around 3mm or make them a single-perimeter thick.  I'd be inclined to try single perimeter first.

I'm expecting the spring in the microswitch to have negligible resistance (as they're tiny microswitches), but agreed that the bot hasn't got much oomph, so the bumper needs to be very easy to trigger. :)

[Robert Longbottom]

Right, some more updates pushed to the bumper branch to have springy mounts and smaller mounting plates for the microswitches.  It makes a bumper and microswitch holders a single piece to print.  At the minute I'm assuming that we will attach this using some small bolts that go through the microswitch and into some (custom) holes in the main base.

On Wednesday, 3 December 2014 09:45:12 UTC, Robert Longbottom wrote:

Ah yes, that's the kind of thing I thought you meant.  That will look nice if we can get it to work.

I printed a springy thing from thingiverse on Mon night to test.  That was 5mm high I think and two perimeters and it seems quite flexible.  I think two perims would be better if it works because it will be easier to print and stronger.

Will have more of a play in the bumper branch.

On 2 December 2014 17:11:19 Damian Axford  wrote:

have added some springy bits to the Bumper branch... for illustration.  I suspect at 5mm high, they'll be a bit stiff...  prob either reduce that to around 3mm or make them a single-perimeter thick.  I'd be inclined to try single perimeter first.

I'm expecting the spring in the microswitch to have negligible resistance (as they're tiny microswitches), but agreed that the bot hasn't got much oomph, so the bumper needs to be very easy to trigger. :)

--
You received this message because you are subscribed to the Google Groups "swindon-hackspace" group.

To unsubscribe from this group and stop receiving emails from it, send an email to swindon-hackspace+unsubscribe@googlegroups.com.

[Damian Axford]

looks cool :)  would be great if it could use the pin system - should just need the additional of some "ears" with holes through that line up to the grid.  Otherwise it doesn't tick the no tools assembly box.   

Of course, that might entail a little tweaking to keep the microswitches in place without any through bolts...?

[Robert Longbottom]

Yeah. To be honest I was having that much trouble lining up the micro switches I didn't fancy trying to also line something up with a hole in the base.  But it really shouldnt be that hard.  I'll preserve some more later on :-)

No tools...  I can see I'm going to have to work out how to get my printer to do holes more accurately!

On 3 December 2014 21:48:37 Damian Axford <dam...@axford.me.uk> wrote:

looks cool :)  would be great if it could use the pin system - should just need the additional of some "ears" with holes through that line up to the grid.  Otherwise it doesn't tick the no tools assembly box.   

Of course, that might entail a little tweaking to keep the microswitches in place without any through bolts...?

--

You received this message because you are subscribed to the Google Groups "swindon-hackspace" group.

To unsubscribe from this group and stop receiving emails from it, send an email to swindon-hacksp...@googlegroups.com.

[Robert Longbottom]

Bumpers with push pin connectors;  by a stroke of luck, I mean quality design, the pins miss everything inside the shell :-)

Will print some out to see what they look like.

[Robert Longbottom]

So, prints done.  The bumper looks quite good and like it should work.  Maybe a bit difficult to push from head on.  The problem is that when I pushed the pins into the base the pins snapped even though I was being gentle.  Don't know if there is some trick to printing these things, but they always seem to snap on me!  Anyway, here's some pictures:

[Gyrobot]

That looks really good.

The trick with the pins is to call-up the module pinpeg which can be printed sideways therefor much stronger. Then in your bumper part call up the module pinhole and print the hole (either blind or through).

Ultimately we want to try and keep the same pinpeg size across all of the design so they can be mixed. Or design them sufficiently different so they cannot be accidentally put in the wrong hole.

[Robert Longbottom]

Ah yes, of course!  I will give that a go.  I think it will have to be a blind hole because a pin that sticks out beneath the microswitch may foul the ground below, but that's okay.  What sort of pin length should we standardise on?  I'm thinking you want 4mm (2 x dw) so you can clip together two walls.

--

[jmeosbn .]

Nice! :)

Jamie Osborne

--

[Damian Axford]

looks excellent :)  and agreed on thickness - 2 x dw.   

[Gyrobot]

A 4mm long pin will be tiny and a lot of printers will not print it. I think we almost have to design Fisher price components that are over size and can be bashed in with a large rubber clown mallet.

I am thinking about 10mm long for the shortest, (see Brabeast as an example of pins). This will inevitably mean a little localised boss on a typical component to increase it's thickness to match the length of the pin. Either on your microswitch holder or inside the base plate or a bit of both. Although it would be nice to keep the base plate thickness standard for interchangeability of modules.

On Thursday, 4 December 2014 21:21:23 UTC, Robert Longbottom wrote:

Ah yes, of course!  I will give that a go.  I think it will have to be a blind hole because a pin that sticks out beneath the microswitch may foul the ground below, but that's okay.  What sort of pin length should we standardise on?  I'm thinking you want 4mm (2 x dw) so you can clip together two walls.

[Damian Axford]

that actually made me laugh out loud :)    

would be good to test a few at 4mm to verify, but if they have to go to 10mm, then I prefer the option of keeping the base plate at 2mm thick and adding a boss to mating parts.

[Robert Longbottom]

Well, they wouldn't be 4mm long, just the flat in the middle would be 4mm where it passes through the holes in the 2 components being connected, then you'd have 2 or 3mm at each end for the lug(?).

But yes, small things are terribly fiddly to print and put together.

On 5 December 2014 12:08:05 Damian Axford <dam...@axford.me.uk> wrote:

that actually made me laugh out loud :)    

would be good to test a few at 4mm to verify, but if they have to go to 10mm, then I prefer the option of keeping the base plate at 2mm thick and adding a boss to mating parts.

--

[Jamie Osborne]

can be bashed in with a large rubber clown mallet.

Ha! I also Lolled! But; NO TOOLS!!! ;)

Jamie

[Robert Longbottom]

Bumpers v2, using distinct pins to connect.  Print and attach okay, just need to check microswitch fit, shell fit and how well they will work.  They are a bit stiff from head on, but 45 degree impact should be okay I think.

[Damian Axford]

Looks excellent :). How long is the pin (the bit between nubbins)?

[Robert Longbottom]

Its a pintack(side=true, h=7.8+2+2.5, bh=2);

So the "bit between the nubbins" is ms_height + base_thickenss = 7.8 + 2

I think the 2.5 should really be 3, but I wanted it to be a tight fit.

[Gyrobot]

A great example of a modular no tools approach. Not a Fisher price mallet in sight.

[Damian Axford]

@Rob - been pondering springiness...  I can think of 3 simple ways to make it softer (if front contacts prove too stiff):

1. Reduce the height of the springs
2. Reduce the springs to a single perimeter width
3. Increase the length of the springs (more curl), even so far as using S-bends if necessary

Although options 1 and 2 use less plastic, I suspect option 3 would be more robust - so that would be my preferred option.

@Jamie - getting anywhere with the LED light-pipe idea?  

[Robert Longbottom]

Yes, its tricky. One perimeter this will be difficult to print (depending on your printer) and they need to be strong enough to actually hold up the bumper because there is nothing else support it.

The other option is maybe to move the sprints to the other side of the mount.  This will probably make head on collisions softer and side collisions harder, which might be a better trade off.  However you need to be careful not to foul the wheels doing that.

I'll bring the bumper in on wed and we can have a look - I also want to check that a microswitch actually fits because the holder looks very small.  It would be useful to check that it works with the shell on as well.

Rob.

--

[Damian Axford]

good point on changing which side the springs are on...  gut feel is they're in the right place (given wheels)

I'm in Istanbul this week, so can't bring the shell in for testing - don't think it's worth printing another one yet, as we've not yet made the alterations for the "lid" or a cutout for the LEE light-pipe.   

[Robert Longbottom]

Yeah, I thought you'd said you were out this week.  Never mind, I can check microswitch fit if Steve still has one.

The other thing we could try is increasing the radius of the springs and making it so they attach round the back of the micro switches rather than at the side.  Though that might cause the bumpers to droop.

I did think about printing a shell but arrived at the same conclusion about waiting for the next version :-)

I'm going to try and do a few tidyups and a bit of magic-number removal in the bumper branch then hopefully it can be merged into master.

On 8 December 2014 15:24:11 Damian Axford <dam...@axford.me.uk> wrote:

good point on changing which side the springs are on...  gut feel is they're in the right place (given wheels)

I'm in Istanbul this week, so can't bring the shell in for testing - don't think it's worth printing another one yet, as we've not yet made the alterations for the "lid" or a cutout for the LEE light-pipe.   

--

[jmeosbn .]

@Jamie - getting anywhere with the LED light-pipe idea? 

Hoping to prototype something soon with the white PLA we have incoming.  Probably something like I originally had shown, but with 45 degree inverted V shape at top, in place of an arch.  Want to see how well the LED spreads over something that size (should be fine). 

Jamie Osborne

--

[Damian Axford]

cool - the alternative to a fully printed pipe, is to print a very thin outer shell and then fill with hot glue (embedding the LED into the part at the same time as filling with glue).  This way the glue acts as the pipe and also makes the part quick to print/assemble.   It also might avoid the need to use clear PLA, as a very thin white PLA shell with clear glue filling will probably work just as well.

[jmeosbn .]

Hmm... some experiments are in order me thinks!

Jamie Osborne

On Mon, Dec 8, 2014 at 4:21 PM, Damian Axford <dam...@axford.me.uk> wrote:

cool - the alternative to a fully printed pipe, is to print a very thin outer shell and then fill with hot glue (embedding the LED into the part at the same time as filling with glue).  This way the glue acts as the pipe and also makes the part quick to print/assemble.   It also might avoid the need to use clear PLA, as a very thin white PLA shell with clear glue filling will probably work just as well.