PDA

View Full Version : 3D printable models for the HR-OS1



r3n33
09-03-2015, 03:40 PM
Since I like to share any 3D printed extras I make here is the start of the HR-OS1 add ons.

The first thing I put together was a bench to sit on. I'm not claiming it's super fancy or anything but I have been getting a lot of use from this print.

61466147

You can download the STL and print the bench or the SKP and modify the source in Sketchup. The files are hosted here: http://www.thingiverse.com/thing:997753

... and .. soon to be ready for sharing... A Chappie head :D

6148

Do you have any 3D printable accessories you'd be willing to share? I know I'd like to see them and I'd probably try printing them out too ;)

KurtEck
09-04-2015, 10:28 AM
Cool stuff! I may need to figure out how to order one :D

Kurt

Zenta
09-04-2015, 12:10 PM
Looks great, especially the head.:D
Btw, you sure got some cool multicopters in the background.

darkback2
09-04-2015, 11:30 PM
Love the head! I have to say I really liked that movie. Does the camera fit inside of the head?

r3n33
09-05-2015, 12:01 AM
Ty ty and ty :)


Cool stuff! I may need to figure out how to order one :D

Kurt

Which? The bench, head, or a 3D printer? ;)


Looks great, especially the head.:D
Btw, you sure got some cool multicopters in the background.

It is a bit of multi rotor heaven over here. The Inspire 1 is the first drone I've owned where I didn't design the flight controller and it's absolutely tons of fun. I highly recommend it if your into that sort of thing.


Love the head! I have to say I really liked that movie. Does the camera fit inside of the head?

I loved the movie too! but that might be obvious. Currently the camera isn't something I've taken into consideration but at a glance it looks like I could make room and add a hole for the fov. I've been having a tough time getting the file to print with all the supports required and I'm not willing to babysit a 8+ hour print job even if it's sitting in the living room so I've separated some of the details and printed the bulk of the head upside down. It helped minimize the required supports and print time and also gave me the idea to have some of the parts attach and articulate which I may get around to at some point. Oh and what I was originally thinking, I still have a little adjusting of the width to do and removal of some extra material in the back and then I hope it will fit and allow the head to move fully. Right now it mostly looks good but I have plans to finish it off :)

6152

KurtEck
09-05-2015, 08:28 AM
Which? The bench, head, or a 3D printer? ;)
Maybe all of the above ;)

darkback2
09-15-2015, 09:43 AM
I am currently ordering some ninja flex filament for printing heads. I have access to a makerbot replicator 2x and we use the dissolvable filament for our support material. I've had relative success printing as of late...knock on wood.

DB

r3n33
09-15-2015, 02:18 PM
I love and hate ninja flex :) It's super cool stuff but I have to print super slowly and still have feed trouble because it's so soft. I hope you have better luck with the machine you'll be using. I've never used dissolvable support material.. do you like it?

r3n33
09-15-2015, 02:20 PM
Oh yes.. And the first (released) iteration of the Chappie head is available on thingiverse:

http://www.thingiverse.com/thing:1016597

I haven't gotten around to modeling the USB camera and making room but can see myself getting there eventually ;)

6173

Zenta
09-16-2015, 10:12 AM
That's awesome! Just print one of the antennas red..;)
Btw. What 3d printer are you using and would you recommend it?

r3n33
09-16-2015, 11:35 AM
Thanks Zenta :D I'd definitely like to print an orange right-side antenna when I get a hold of the filament. I forgot to mention it here but the two big antenna are moveable and attach via M2.5 screws.. Up to 8mm in length or so. I think I used 2.5x5mm temporarily. Anyway, it will be nice and easy to swap for a colored part.

At home I've been using a Solidoodle 4 for almost two full years. It was fairly inexpensive at the time and right off the bat received a couple hundred dollars of upgrades ( LCD Interface, SD Printing, E3D hotend ). Since then it pretty much just works without any major problems. I've replaced a couple cheap cooling fans along the way and learned cheap filament produces crap prints. All in all it has been a very trusty lower-end machine. So 1-2 years ago, yes, I'd definitely recommend it. Now a days I'm guessing the selection, features, and prices have all changed.

r3n33
10-05-2015, 05:15 PM
I don't know about anyone else but my battery is always trying to slide out and I've finally had enough. So I made this quickly printable and easily installable "battery catch". It is designed to attach to the two un-used M3 holes on the left side of the robot and prevent the battery from falling out. Typically my battery doesn't fall out the right side because the wiring and plug are routed through the center which keep it in place unless the bot leans left.

6186 6187

Download the STL and the Sketchup source here: 6188

r3n33
10-06-2015, 11:25 AM
I've been re-working the Chappie head model and now that it's quite a bit better you can make one too. Possibly the best feature is the ability to mount the supplied HR-OS1 camera inside and you have two options here: one with a 24 degree downward offset to allow the bot to see it's feet and play soccer and one with the camera pointed straight ahead like the original (beta) mounting calls for. Mounting holes were added to the top allowing the part to be securely attached to the F3 bracket on the head servo. The next thing I did was attempt to reduce the weight and I brought it down to 60 grams or about 70 with the antennas attached. This is about 22% less than the original design but still not light enough to work with my existing tune... awww darn. Regardless, if you are a Chappie fan this head is for you! I can't bring myself to take it off :)

Download the files here 6193 or on thingiverse (http://www.thingiverse.com/thing:1055100).

6189

6190

6191 6192

KurtEck
10-06-2015, 04:50 PM
Looks great!

Quick question: Suppose there still exists a few of us who don't have a 3d printer :lol:

Is there a service that you can simply up here that you can simply either send these files to and/or point them to this listing and have them print them?

Thanks
Kurt

tician
10-06-2015, 05:11 PM
The recently quiet birdman cured my ignorance of the rather awesome 3D hubs (https://www.3dhubs.com/) and then there is the ever popular shapeways (http://www.shapeways.com/). Just upload the stl to either and get quotes for the print.

drachels
10-19-2015, 08:04 PM
Hi r3n33,
I recently printed a Chappie head using your file. Thanks. I could send you an orange antenna, if you like, as I'm just three states East of you. I have been trying to learn to use 123D Design and have made some "Jimmy" like hands. If anyone prints them out, I don't think I would have them installed if I was playing around and experimenting with poses or anything else where I was expecting my robot to fall a lot. I printed mine with support and 15% infill.
6206
6207

r3n33
10-20-2015, 01:00 PM
I'm really happy you were able to print a Chappie head. Thanks for offering an orange antenna :)

Those are neat Jimmy style hands. Do you have any thoughts about making more jimmy like parts? I think that might be fun.

drachels
10-20-2015, 05:40 PM
I've been working on a head today. About 5 false starts so far as I'm trying to make it so the back comes off but has a rim so it fits tightly into position. Slow going because I'm having to learn 123D Design at the same time and it is NOT very intuitive. I'm also working on a way that lets the head tilt a lot further than the stock setup. Got one right now that looks up 95 degrees and down 150 degrees. Don't know if it would get that much tilt with a head on the mount though.

Send me your address in a private email and the antenna is yours. :D

LloydF
10-21-2015, 11:27 AM
Your "battery catch" is a most wonderful addition to my HROS1. You ROCK! :o

drachels
10-24-2015, 04:12 PM
I finally got a printable version of a "Jimmy" type head. The bottom of the head replaces the top part of the two round pieces of the standard head tilt. In the included picture, the camera is mounted behind his right eye but has mounts so it could be placed behind the left. The picture on the computer monitor is from that camera. I need to play with the two mounting screws to eliminate a bit of the eye hole that is just visible where the red arrow is pointing.
6215

If that were the only problem, I would just live with it.....but. It wasn't until I was finalizing the design that I imported the servo and tilt links into the project file to see how much of a hole I was going to need for the tilt linkage arms. I was hoping to get away with just a couple of holes in the bottom of the head. Since all the nuts, bolts, screws, etc. are exploded in the HRSO1 3D Models master file, I forgot to account for them and, as you can see in the second picture, I need to do a "little trimming" to use this head.
6216

I am working on a new version as I think this one actually looks a little too small. I plan on increasing the diameter from 120mm to about 140mm, or whatever it takes to make the back cutout unnecessary. It will also allow me to try another type "latch" for the access hatch. With the bigger size, I also plan to include mounts for a couple of speakers. Radio shack has some that are just 29mm in diameter. I might even try using a couple of ancient computer chassis speakers I have lying around. At 57mm, they might give more sound. I might even see about scavenging speakers from old phones I have squirreled away somewhere.

A redesign will also let me add 3mm to the camera mounting posts. In this version, I had to add 3mm spacers so the right edge of the camera circuit board would fit and clear the back side of his cheek.

Since this version has "problems" I'm not going to bother posting the two .stl files. My wife has plans for me over the next couple of days, but I hope to have the new version ready by next weekend and will make the new files available. The real downside to this, is that because of all the support needed, these two pieces took 6:34 and 2:41 to print and a bigger version will just take longer. :mad:

r3n33
10-25-2015, 12:12 PM
That looks great!! And having the camera inside is wicked cool too.

Waiting for 10-12+ hour prints seems to be some form of mild torture. Add to that making multiple revisions to get the desired result... I feel your pain. That said I'm totally willing to wait for this print and am really excited for when you finish the final adjustments.

Looks like you are getting along with your new software pretty well ;)

sonel
10-27-2015, 04:01 AM
I have been looking for some time for a Jimmy head model. I am interested in the model and happy to help with the updates. What 3D modelling program are you using?

I am not a big fan of the current design of the head. It's too complicated and as you pointed out it occupies too much space. Also, as some others have noticed the tilt range is very limited. I'm thinking about changing the design by moving the pan servo in the chest between the two shoulder servos, build a long "neck" and put on top the tilt servo that can be fixed to the head shell.

I've seen you also have the Jimmy hands printed. I think they look nicer.

Alex.

guoshihui
10-27-2015, 04:06 AM
Also, as some others have noticed the tilt range is very limited. I'm thinking about changing the design by moving the pan servo in the chest between the two shoulder servos, build a long "neck" and put on top the tilt servo that can be fixed to the head shell.
Alex.
This is definitely something I am interested! Having a head capable of looking up and down in a big range!

Shihui

sonel
10-27-2015, 05:52 AM
This is definitely something I am interested! Having a head capable of looking up and down in a big range!

Shihui

Shihui, frankly providing better tilt range can be solved with the current design too. I've posted an updated design here:

https://github.com/sonelu/HROS1-Customisation/tree/master/Head%20Tilt

I haven't yet printed and tested. I might try to do this in the evening.

Alex.

guoshihui
10-27-2015, 07:03 AM
Shihui, frankly to provide better tilt range can be solved with the current design too. I've posted an updated design here:

https://github.com/sonelu/HROS1-Customisation/tree/master/Head%20Tilt

I haven't yet printed and tested. I might try to do this in the evening.

Alex.

Thanks Alex. This is fantastic. I have downloaded the file, but can't print until late this week, because we just bought our printer, and it will be assembled this Thursday.

Shihui

drachels
10-27-2015, 06:03 PM
I haven't completely finished the redesigned "Jimmy" head as I needed to incorporate some sort of mounting points for speakers. I bought some Radio Shack 29mm speakers yesterday afternoon and tested them. I can get adequate sound from them using my Emic 2 text to sound card. I just finished printing a set of mounts that can be mounted on a flat horizontal or vertical surface. I'm hoping that mounting them firmly to the inside of the head will give some more amplification like the sound box of stringed musical instruments. I should be able to print the head tomorrow and see if it works.

Edit the next paragraph because I redid the files and made a smaller base version.

Anyway, for r3n33, the side mount only version does fit in the Chappie head a couple of ways, even in the round "ear" areas on the side of the head, if you want them there. My first version that was too thin to slice correctly, fits there just right and this revised one should, too. Anyway, there is another file that has a larger base and has holes for bottom or sidemounting. The included pictures are a back and front view with a speaker mounted in it. The retainer was deliberately made to be flexible because the way the wires are attached is slightly different for each of the four speakers I have.

6228
6229

6221
6222

drachels
10-27-2015, 06:20 PM
Here is what the Jimmy head looks like in 123D Design, right at the moment. The object in the left eye is a mock-up of the camera just to make sure position and mounting holes are in the right places to clear the inside surfaces of the head. The tilt mechanism parts are just there for the same reason. The servo mount in the bottom center is actually part of the head and replaces the top plate of the two in the tilt mechanism. The two holes near the bottom edge on the side of the head are access holes that let you mount the tilt servo. I probably will replace them with vertical slits for the speakers with the slits being wide enough for the 1.5mm Allen wrench to get through to mount the servo.
6223

drachels
10-27-2015, 06:59 PM
sonel,
I hope I'm wrong, but I think you'll find that your design does give slightly more tilt up, but not more forward tilt. The limit for tilting up is the rear edge of the two circular disks hitting the top of the pan servo so making the pivot posts taller should help there. However, the forward tilt is limited by the crossbar between the two Head Tilt Cams hitting the back of the tilt servo and the way the lower end of the right Head Tilt Swing Arm connects to the Pan Servo Bracket causes it to bind just above the lower right hand bolt.

Anyway, I am currently printing layer 18 of 63 of your new tilt plate bracket to try it out.

r3n33
10-28-2015, 10:35 AM
6221
6222

This looks nice and clean. Have you tired these little speakers yet? I played with one last night and it had a pretty decent volume when you put your hand or an object behind it. I was thinking I could get away with using two to give a little extra volume. So far I've only tested with this 5V amplifier found on amazon (http://www.amazon.com/DROK-Amplifier-Amplifiers-Portable-Headphones/dp/B00C4N410G) but I'm definitely satisfied for $6. We might start a new thread for robot audio when you start testing your TTS module?


Here is what the Jimmy head looks like in 123D Design, right at the moment. The object in the left eye is a mock-up of the camera just to make sure position and mounting holes are in the right places to clear the inside surfaces of the head. The tilt mechanism parts are just there for the same reason. The servo mount in the bottom center is actually part of the head and replaces the top plate of the two in the tilt mechanism. The two holes near the bottom edge on the side of the head are access holes that let you mount the tilt servo. I probably will replace them with vertical slits for the speakers with the slits being wide enough for the 1.5mm Allen wrench to get through to mount the servo.
6223

Still doing a little dance for this! I can't wait to give it a try.

The simplified camera model is a good idea to get a good fit.. One thing that helped me was to also draw the field of view coming from the center of the lens. Keep in mind I've only measured mine but I observed a 45 degree horizontal angle and a 32 degree vertical angle from the supplied camera. Also I noticed my camera looks a little to the right as the components were not in perfect alignment from manufacturing.

sonel
10-28-2015, 10:48 AM
Guys, thanks for the feedback. I had the chance to do some more testing and have come up with the following design:
6227

It includes changes to tilt plate bracket, tilt servo bracket, arms, etc.
The parts are in the github as well as the assembly pictures.
https://github.com/sonelu/HROS1-Customisation/tree/master/Head%20Tilt

With this setup the tilt servo works from 260 (aprox 75º down) to 730 (aprox. 63º up).

drachels
10-28-2015, 06:06 PM
This looks nice and clean. Have you tired these little speakers yet? I played with one last night and it had a pretty decent volume when you put your hand or an object behind it. I was thinking I could get away with using two to give a little extra volume. So far I've only tested with this 5V amplifier found on amazon (http://www.amazon.com/DROK-Amplifier-Amplifiers-Portable-Headphones/dp/B00C4N410G) but I'm definitely satisfied for $6. We might start a new thread for robot audio when you start testing your TTS module?



Still doing a little dance for this! I can't wait to give it a try.

The simplified camera model is a good idea to get a good fit.. One thing that helped me was to also draw the field of view coming from the center of the lens. Keep in mind I've only measured mine but I observed a 45 degree horizontal angle and a 32 degree vertical angle from the supplied camera. Also I noticed my camera looks a little to the right as the components were not in perfect alignment from manufacturing.

Yes, I tried the speakers out last night. At the full volume of the TTS module, one speaker in the open was fairly decent. Two was much better. I have provision in the head to mount up to four. We'll see. A LOT depends on which voice is selected for the output.

Your angles on the camera are what I think I found somewhere last week. I hate to say I didn't think to do the field of view trick. Just went with what I got from the first head I'm not using, plus I made the eye holes 2mm larger in diameter. I will probably go back and add it into the working files.

I am 45 minutes into a trial print of the new head design and still on the second layer...and it is a print I'm fairly certain will fail near the end. I'm using my Printrbot Simple Metal that I just upgraded to a heated bed. I was working on getting the Z stop dialed in, got in a rush, and forgot to switch to a profile with support, so I think the top of the head is going to cave in. I am letting it go as it is an old spool of gray filament, and it will a least give me a base plate that I can use to verify clearances, hole positions, and hole sizes before I use my new white filament.

sonel
10-29-2015, 03:00 AM
We might start a new thread for robot audio when you start testing your TTS module?

I think this is a good idea. I'm also working on audio, using instead the 2x2.5W amplifier from Adafruit (http://www.adafruit.com/product/1552). It is very small has seems to have very good efficiency.

The speakers I found on eBay are very thin and 40mm diameter:
62306231

They are rated 3W at 4ohm, which I'm sure is a gross exaggeration. But I will test them soon. I plan to place them in the front cover just in front of the shoulder servos, with the amplifier between them.

The issue is that powering the amplifier from the regulator that is on the ArbotixPro might not be a very good idea. That is rated at 1.5A and the Pi takes about 700mA. In my design I also have a 2.2'' HAT TFT display in the front, and with a camera on the USB things get easily to about 1A. I'm afraid that powering the amplifier from the same source will not be a good idea.

So I'm started preparing a "power" board that includes 2 Murata DC/DC regulators (I'm having a second one because in the end I want to run another RPi in the head dedicated to stereo vision - that is why I'm interested in the Jimmy head from darachels: it' the most suitable head that might include a Pi and two cameras). Since I'm there I want to include also two XT male connectors for the battery and 2 Schotky diodes to allow hot swap of the batteries.

For the time being I've changed the lower part of the body as follows to make space for the new board:
6232

I have removed the 2 standoffs at the back and placed instead a U shaped support that has 4 M2 heat inserts - these will hold the board. I've also modified the plate P4 by adding 2mm standoffs for the servos so that now the distance between the P4 and P5 plate is 34mm, enough to mount the battery upright, thus leaving me more space at the back for the power board.

As I progress with this I will post it on the git hub.

Alex.

drachels
10-30-2015, 11:15 AM
Hi Alex,
I did a trial print of the head yesterday and I'm printing the back cover right now. r3n33 looked at a copy of the files for me yesterday, and between us we found the need for three modifications I'm working on today. Should have the files ready later today.

As far as sticking a Pi in the head, I think it might be doable as it looks like there is just enough room above the tilt servo, but there is a possibility the cam arms would impinge into the required space. I won't know for sure for a while. Got four Pi's in the house and not a single one is outside of a case. I'll have to try removing one to see.

There is definitely room for the supplied Edison and it's breakout board to go in the head. In fact, more than one. It might be feasible to run the robot off an Edison in the head with two cameras running off the Pi mounted in the torso.

I'll try to post the updated files this afternoon,

AL

drachels
10-30-2015, 02:41 PM
Well, I think he looks pretty good, even though this is the trial print from yesterday with today's back cover.
62376238

The "ear" slots have been changed to a pattern of circles because these wound up having trouble printing. The ones on the other side of the head do not look as good as this side. If anyone prints the files, I had to rotate the back piece forward 90 degrees as the normal upright way was not sticking to the support. Of course that makes the lines of the print pattern match, but I think I can live with that....for now. I'll try printing upright again, but if I have problems I'll probably reverse the edge lips later so that I can make it print it upright without problems.

@Alex, as you can see lying there beside him in the first picture, I took one of my Pi's out of a case and tried it out in the head. If I lay my speakers down and I'm very careful about how I do it, I can squeeze a Pi into the head. I put it in upside down with the USB connectors over by the ear holes. You can have them to the rear, but the tilt cam might hit the them. An interior redesign to have the servo standing up, (would need longer tilt arms,) would make it to where you had room to stand the Pi on it's edge right behind the cameras. Would probably also want to make the whole top of the head removable, in that case. Actually, that's the way I initially thought about trying to make this thing, but I got side tracked and went with the present design version.

There is ALMOST enough room for a computer on a stick in there. If I opened up a slot in the back cover, one would fit. I haven't tried to see yet if I can replace Windows 10 with Linux on one yet, though. It would be nice to have a quad core 1.33 GHz processor in there. I don't have one, but I bet an Odroid-XU4 would also fit. At 2GHz and an Octa core, that would probably be able to run everything. Hmm, just place it where the normal Pi goes.

Anyway, here are the files if anyone wants to try them out.
62406239

drachels
10-30-2015, 06:47 PM
I have been looking for some time for a Jimmy head model. I am interested in the model and happy to help with the updates. What 3D modelling program are you using?

I am not a big fan of the current design of the head. It's too complicated and as you pointed out it occupies too much space. Also, as some others have noticed the tilt range is very limited. I'm thinking about changing the design by moving the pan servo in the chest between the two shoulder servos, build a long "neck" and put on top the tilt servo that can be fixed to the head shell.

I've seen you also have the Jimmy hands printed. I think they look nicer.

Alex.

I forgot to mention this when I saw your post the other day, but I do have a .stl file for the P1 plate that lets you place the pan servo in the chest between the shoulder servos. It will also let you keep the servo on top, but flipped over so the servo horn is on the top side of the servo instead of on the bottom, which is standard. It is made by just adding a cutout and screw holes from the P1 plate in the master file.

6241

r3n33
10-31-2015, 11:39 AM
Thanks AL!!

The JimmyV5 model looks great! I tried to print one last night and while it came out OK the supports that Slic3r generated were pretty excessive in the middle. It was late last night so I tried to start the print anyway. Long story short I ended up breaking (er ripping apart) the print while trying to remove them and then hot gluing the print back together. Hah. I'll see what I can do to adjust my support settings or create a flat ceiling in the model so it can bridge the top and avoid the need for the support pillar in the middle. The very worst case is it's about $62 at shapeways for a really nice copy ;)

62436244

62456246

KurtEck
10-31-2015, 11:56 AM
Looks like I may need to order some parts (Head and Hands)... Probably to add Neopixels to all :D

drachels
10-31-2015, 01:08 PM
Thanks AL!!

The JimmyV5 model looks great! I tried to print one last night and while it came out OK the supports that Slic3r generated were pretty excessive in the middle. It was late last night so I tried to start the print anyway. Long story short I ended up breaking (er ripping apart) the print while trying to remove them and then hot gluing the print back together. Hah. I'll see what I can do to adjust my support settings or create a flat ceiling in the model so it can bridge the top and avoid the need for the support pillar in the middle. The very worst case is it's about $62 at shapeways for a really nice copy ;)

62436244

62456246

I haven't tried a Slic3r generated print. So far the two I've done were with the latest Cura and whatever it has as internal slicer and the support was parallel lines which were very easy to get out pretty cleanly. In some spots I had to take out just one vertical line at a time around the servo and camera mounts. The only problem I had with cracking/breaking is that this thing sticks to my heated bed so well, that I cracked it getting it loose. Gotta make me a wedge to slide between the print and the tape to see it that makes it easier. Maybe I can find a real thin and flexible putty knife. As far as the price, wow, didn't realize we are saving that much!

drachels
10-31-2015, 01:16 PM
Looks like I may need to order some parts (Head and Hands)... Probably to add Neopixels to all :D

The other night, when our dog kept me from going to sleep for an hour or two, I got to thinking about r3n33's lighted hands, when I suddenly realized that a pair of Adafruit NeoPixel rings (http://www.adafruit.com/products/1463) right behind the eyes/cameras, would allow for some pretty neat lighting effects.

LloydF
11-02-2015, 03:56 PM
My chair is almost finished printing, Thank you. Every bot needs his spot.:cool: (Sheldon Cooper quote i think)

LloydF
11-05-2015, 07:17 AM
I noticed that you are running pressure sensors in your feet. How is that working out?

r3n33
11-05-2015, 02:04 PM
I noticed that you are running pressure sensors in your feet. How is that working out?

They work great! The HR-OS1 FSR feet are just a first rough draft and I'd like to try adjusting the design to get the robot's weight back to the center of the foot. I cheated and moved the bracket up a few millimeters to get them assembled, programmed and tested. This moved the ankle pitch rotation axis forward on the foot. I can compensate with tuning but I'm wanting to get back to AX12 knees and get rid of the cooling fans anyway so Chap is mostly in a state of development.

The FSR firmware is working great and a demo in the framework has been adapted which streams the data to a web page hosted by the robot. I would like to update the design slightly and then post all the information needed to build your own. The total cost was just under $100 I believe.

Eventually what I'd like to see would be some body position compensation for the readings obtained by the feet and possibly fused with some inertial readings. Even if it was just a demo and not necessarily used in the walk gait.

sonel
11-05-2015, 04:10 PM
All, I'm in the process of finishing a slightly more complex change for the head:

6261

The P1 is modified so that the pan servo is between the shoulders. It has an added "collar" that protects the neck in case of a fall so that not all the impact is taken by the servo horn. This also helps guiding the cable from the pan servo to the tilt servo - you will see more detail when I'll finish.

The tilt servo is in the neck and the cable goes nicely through that skin.

The head is slightly modified - is not round but oval so that the cameras fit better as the curvature is less strong. Also the head is made of two parts: the bottom (which is in the picture) and the top (which is printing now...another 6 hours to go) making it much easier to work on the parts fitted in the head.

Now there is a lot of space in the head and you can see I have already included the spacers for the Pi:
6262

The cable you are seeing is the normal Dynamixel bus coming from the tilt servo that will be used to power the Pi and the audio amplifier. I will use a prototyping Pi Hat where there will be 2 Murata regulators (one for PI and one for the amplifier, the amplifier itself and some MOLEX connectors).

Also in the head there will be 2 cameras (VGA) and the speakers. The cameras come with microphones - those will be unsoldered and wired on the sides. Hopefully that will help doing to audio triangulation.

I will post more pictures and the 3D files when I'm done with it.

Alex.

sonel
11-05-2015, 04:23 PM
They work great! The HR-OS1 FSR feet are just a first rough draft and I'd like to try adjusting the design to get the robot's weight back to the center of the foot.

r3n33, this is very interesting. I had it on my list to do. Do you use a controller board with an STM32F and feed the info through the Dynamixel chain, or are you using another approach?

Interestingly enough the original Darwin project is open source except the FSR implementation. My idea was to build a small board with a STM32F that would read the analog values from the 4 FSRs and implement the normal Dynamixel protocol to respond to commands as any other Dynamixel device. You can see the Robotis specification for their FSR controller here (http://support.robotis.com/en/product/darwin-op/references/reference/hardware_specifications/electronics/optional_components/fsr.htm).

Alex.

drachels
11-05-2015, 06:43 PM
Wow! that new head base neck combination is looking great. I've been working on basically the same thing today while printing some of the armor. I'm about 90% finished, but I think I'll either modify it like yours, or scrap it and wait for yours. Either way, I'll need to modify it some as I got a couple of Odroid-XU4's delivered today and while it is basically about the same size as a Pi, the hole patterns are different. Gotta make at least one head with one it.

sonel
11-06-2015, 10:28 AM
I've finished printing and it looks decent, but I do have to make some changes, the most important being that I want to place the Pi fixing standoffs in the top of the head, instead of the bottom as it is now. That will help with the cabling as all the stuff (speakers, cameras, microphones) will be in the top part of the head. Moving the Pi there (upside down) will make it simple to open and close the head - there is just the power cable coming from the neck.

@drachels: do you have any idea where I can find some dimensions for the Odroid-XU4 mounting holes and board layout? I can already include standoff for it too so there should be no need to change the model (it's actually a pain to change other peoples' models...).

Alex.

KurtEck
11-06-2015, 10:59 AM
I am not drachels: but I do have Xu4 and probably plan to use on HROS1... Currently using C1...

But you can find a lot of information at: http://www.hardkernel.com/main/products/prdt_info.php?g_code=G143452239825&tab_idx=2

drachels
11-06-2015, 11:19 AM
I've finished printing and it looks decent, but I do have to make some changes, the most important being that I want to place the Pi fixing standoffs in the top of the head, instead of the bottom as it is now. That will help with the cabling as all the stuff (speakers, cameras, microphones) will be in the top part of the head. Moving the Pi there (upside down) will make it simple to open and close the head - there is just the power cable coming from the neck.

@drachels: do you have any idea where I can find some dimensions for the Odroid-XU4 mounting holes and board layout? I can already include standoff for it too so there should be no need to change the model (it's actually a pain to change other peoples' models...).

Alex.

Hi Alex, the Odroid is a rectangle 59mm X 83mm and the holes are in each corner with the center-to-center of mounting holes 52mm and 76mm apart. More details are at http://www.hardkernel.com/main/products/prdt_info.php?g_code=G143452239825&tab_idx=2. I can't find it on the drawing, but the hole diameters appear to be 3.5mm.

Oh, and the height of the board is basically the same as a PI due to the stacked dual USB connector.

r3n33
11-07-2015, 12:05 PM
r3n33, this is very interesting. I had it on my list to do. Do you use a controller board with an STM32F and feed the info through the Dynamixel chain, or are you using another approach?

Interestingly enough the original Darwin project is open source except the FSR implementation. My idea was to build a small board with a STM32F that would read the analog values from the 4 FSRs and implement the normal Dynamixel protocol to respond to commands as any other Dynamixel device. You can see the Robotis specification for their FSR controller here (http://support.robotis.com/en/product/darwin-op/references/reference/hardware_specifications/electronics/optional_components/fsr.htm).

Alex.

Actually I have about 75% of a write up completed for creating your own FSR feet that I hope to finish soon. What I used was a Trinket Pro 5V and wrote a Dynamixel device driver. The full register table from Robotis was implemented as well as some other improvements.

LloydF
11-08-2015, 02:30 PM
With FSR feet this bot could literally stand on one foot.:robotsurprised:

sonel
11-10-2015, 06:37 AM
All,

I have finished the head updates. Sorry for taking so long - but there are some long prints...

I have posted the parts in my github (https://github.com/sonelu/HROS1-Customisation/tree/master/Jimmy%20Head). I hope by tonight I will also include an assembly guide with pictures.

Alex.

sonel
11-10-2015, 06:43 AM
Actually I have about 75% of a write up completed for creating your own FSR feet that I hope to finish soon. What I used was a Trinket Pro 5V and wrote a Dynamixel device driver. The full register table from Robotis was implemented as well as some other improvements.

Nice, is the trinket small enough to fit in the feet (between the F5 wide bracket and the edge of the foot)?

Also I'm curious if you're using any buffering between the UART and the Dynamixel bus - the original Arbotix-M was simply joining TX and RX from UART and straight to D on Dynamixel bus. All Bioloid controllers (at least the ones that have schematics published) use a combination of LS125/LS126 tri-state buffers controlled through one I/O pin. The Arbotix-PRO is also similar.

Alex.

tician
11-10-2015, 07:54 AM
Buffering is safer since the TTL dynamixel buss is supposed to be pulled up to 5V by servos and/or controllers, but with 5V AVR microcontrollers like those used in the arbotix-m and CM-5/510/700, it is not actually necessary since the AVR permits quick enabling/disabling of TX or RX functionality of UART pins. On 3.3V microcontrollers like most ARM variants (as used in CM-530, CM-900, CM-730/Arbotix-Pro, Arduino Due), it is absolutely necessary to add buffering unless the pins are known to be 5V logic compatible and permit quick enable/disable of TX and RX functions.

KurtEck
11-10-2015, 08:39 AM
Note: after seeing all the stuff r3n33 has been doing with the Adafruit pro trinket (5v), especially with the color LED, I have started to play along :D The Trinket Pros are Atmega 328 processors like the Arduino Uno and the like, but unlike those boards they don't have a built in USB to serial adapter like an FTDI chip... However it does have a funky USB connector on board, which is connected up to two IO pins, that when you reboot the processor it does startup a boot loader which allows you to program the board using USB...

So the same tricks used for the Arbotix-M work here as well. I have a breadboard setup right now. Where I have the pro trinket hooked up to the AX Buss, with RX and TX connected to each other, I have an inline resistor to give the processor a little protection. I then have an Adafruit Neopixel hooked up (I thought r3n33 was using these, but...). So I updated her sketch (and library) and I believe have code that can emulate what she has in her hands, but then I swapped in an Adafruit Jewel (7 Neopixels) and updated my version of the code, such that the the center pixel works like before, but the AX code also monitors other messages on the AX buss and currently I have it monitoring SYNC-Write messages, and have it setup for my PhantomX hexapod where one pixel is associated with each leg. So if the leg is moving, it sets a color depending on what is moving... (Code also works with Teensy 3.2s...)

But again I don't want to hijack this thread! Will post more later when we are happy with the code (and likewise circuit boards... ;) ) But again this all deserves a separate thread.

LloydF
11-11-2015, 04:08 PM
62776278
I got the Jimmy Head Finished and it is really nice, Thank you so much.

drachels
11-12-2015, 10:15 AM
62776278
I got the Jimmy Head Finished and it is really nice, Thank you so much.

Hi LloydF,
That shade of black looks pretty good. I still have to print out a front and back cover. I also need to reprint the calf armor. I printed two before I remembered to eliminate the two holes that are incorrectly spaced. For anyone interested, I'll try to post the altered files in a day or two. I guess I need to start adding all these HR-OS1 files on my github account so people can get to them sooner/easier.

@r3n33 and LloydF - Did you have any kind of problems with the head? Any suggestions for improvement? Any problems getting the rear cover to fit and fasten in place?

I'm about done with an improved version that has a bigger rear hatch to make it easier to access whatever a person sticks inside the head. I'm mostly finishing up on hole placements for speaker mounts and standoffs for mounting a Pi or Odroid0XU4. It will look sort of like stylized hair if printed, or painted, in another color.
6279
It is designed to use a modified P1 plate with the pan servo between the shoulder servos and the tilt servo as the neck, leaving the interior of the head available for cameras, speakers, amp, Pi, Odroid, or anything else you desire to stick in there. I have a prototype printed for making final design changes of where to locate holes, fasteners, standoffs, etc.
6280
As you can see, I currently have a camera, four speakers, and an Odroid-XU4 in my testing head. For anyone interested, yes those two speakers right behind the camera and other eye, do make a difference in the perceived audio volume. I am also experimenting with alternate ways of mounting a Pi or Odroid. The Odroid also presents another problem. Without it being in a case, the eMMC storage chip comes loose pretty easily, so I have to come up with some sort of brace or clip to hold it securely in place.

LloydF
11-13-2015, 07:31 AM
I'm thinking smaller is better here, as I seem to remember this head design has issues.
"Jimmy, while cute, has an impractically large head. We had a lot of problems with the head cracking like an egg when it would take a hard fall. Furthermore, we don't own the IP to open source the Jimmy 3d models, and the owner of those chose not to release them."

drachels
11-13-2015, 10:44 AM
I think smaller is better too. I have been planning to go back to the original, smaller version that I started with as with full body armor, hopefully the chest or back plate should hit the ground first in a fall, instead of the rim of the head. The current the larger version sticks out further, especially to the front, and will hit before the chest and thus be subject to cracking more so than the smaller version.

I am also planning to make it so that the top curved portion prints in one piece and the bottom, more or less like a small saucer plate, will print separately. Pretty much the same way Alex has done. The two parts of the current larger version I've been experimenting with just do not match up well enough due to print "squish" expansion and I don't feel like trying to figure out how much smaller to print the back cover so that it fits better.

My main goal is just to have something with enough room to securely mount other controllers for a little experimenting. I would rather not try stuffing experiments into the little space between the shoulder servos. Besides, the things I am trying out right now just won't fit in there. I've even thought of expanding r3n33's Chappie head so I could hang an Odroid or Pi right behind the the tilt servo, and then add a display on the front like the "real" one from the movie has.

Since I am "retired" and have plenty of spare time, I'll probably try that, plus "Elvis" might even enter the building. :D
6281

r3n33
11-16-2015, 10:48 AM
All,

I have finished the head updates. Sorry for taking so long - but there are some long prints...

I have posted the parts in my github (https://github.com/sonelu/HROS1-Customisation/tree/master/Jimmy%20Head). I hope by tonight I will also include an assembly guide with pictures.

Alex.

Very nice stuff! I checked out your assembly guide as well. Thanks for sharing all the hard work!


Nice, is the trinket small enough to fit in the feet (between the F5 wide bracket and the edge of the foot)?

Also I'm curious if you're using any buffering between the UART and the Dynamixel bus - the original Arbotix-M was simply joining TX and RX from UART and straight to D on Dynamixel bus. All Bioloid controllers (at least the ones that have schematics published) use a combination of LS125/LS126 tri-state buffers controlled through one I/O pin. The Arbotix-PRO is also similar.

Alex.

Yes the Trinket fits great between the F5 bracket and the edge of the F12 foot. I started a thread so we could discuss FSR feet as progress is made. http://forums.trossenrobotics.com/showthread.php?7690-DIY-FSR-Feet-for-the-HR-OS1 To answer your question though, TX and RX are tied together. We are very much experimenting at this point. I've had good success using a diode between TX and RX which simplifies the firmware by not needing to switch the pin modes on the MCU. And just as well without the diode by setting registers to enable/disable comms.



62776278
I got the Jimmy Head Finished and it is really nice, Thank you so much.

Your robot looks great!! :) :)



@r3n33 and LloydF - Did you have any kind of problems with the head? Any suggestions for improvement? Any problems getting the rear cover to fit and fasten in place?

I'm about done with an improved version that has a bigger rear hatch to make it easier to access whatever a person sticks inside the head. I'm mostly finishing up on hole placements for speaker mounts and standoffs for mounting a Pi or Odroid0XU4. It will look sort of like stylized hair if printed, or painted, in another color.
6279
It is designed to use a modified P1 plate with the pan servo between the shoulder servos and the tilt servo as the neck, leaving the interior of the head available for cameras, speakers, amp, Pi, Odroid, or anything else you desire to stick in there. I have a prototype printed for making final design changes of where to locate holes, fasteners, standoffs, etc.
6280
As you can see, I currently have a camera, four speakers, and an Odroid-XU4 in my testing head. For anyone interested, yes those two speakers right behind the camera and other eye, do make a difference in the perceived audio volume. I am also experimenting with alternate ways of mounting a Pi or Odroid. The Odroid also presents another problem. Without it being in a case, the eMMC storage chip comes loose pretty easily, so I have to come up with some sort of brace or clip to hold it securely in place.

To be honest I still have your V5 jimmy style head installed but haven't printed the back half. I never did put the time in to improve my support structuring and have been lurking about watching a few of you making awesome progress so I figured I'd hold off and tend to some of my other projects.

I like your improvements to the V5 design and it's impressive how much you can fit inside the head.

LloydF
11-16-2015, 05:01 PM
I used some long self taping servo screws I had laying around to attach the rear hatch. I used shapeways as I had the same issues as
r3n33 had with my 3d printer but shapeways apparently does not have.:cool:

guoshihui
11-19-2015, 09:27 PM
Did anyone notice the following problem with the Onion armor design?
6292
For the front torso, I understand that the four screw holes can be used to attach your logo or something, is it correct?

But I don't understand the other four screw holes, what these for? The orientation doesn't match with the robot design, so can't attach to the robot with these screw holes. In this way, the armor is only attached to the robot by the two screws on the shoulder.

LloydF
11-20-2015, 07:54 AM
Well the two on top and the four in the middle of the chest are to screw the chest down to the frame but the extra four
on the outside of the chest are used the same way, I just ran out of L bracket's (note this is were you use them little L brackets to
fasten the chest to the Frame plates 2 and 3, it only goes one way so.)
62936294

drachels
11-20-2015, 04:11 PM
Sorry guys, that I did not attach these sooner, but I've had some dental surgery recently and have not been feeling like doing much of anything. These two files have had the mounting holes removed that do not line up with the bracket holes correctly. I felt that eight screws were plenty for attachment, so I just removed them, instead of moving them.
6296
6295

tician
11-20-2015, 05:01 PM
Sorry guys, that I did not attach these sooner, but I've had some dental surgery recently and have not been feeling like doing much of anything.
That's never fun. First or second year at uni, I had my entire upper jaw cut away from my skull, slightly reshaped at the back (removed wedges of bone from both sides), moved forward slightly, then reattached with 17 screws and 4 plates all so that the teeth in my upper jaw would actually line up with the teeth in my lower jaw. Thankfully had a great surgeon with no complications and no painkillers required, but haven't worn my retainer very often the last several years because there is not much need. My sister also had a great surgeon for her upper jaw surgery, but my brother was not so lucky. He had a not very good surgeon who cut and reshaped both upper (cut into three pieces) and lower jaws, accidentally severed nerves in his chin during chin shortening/reshaping (leaving him unable to feel lips; later fixed by another great surgeon so no longer requiring a mirror to eat), and added unwanted cheek implants. Now his second round of braces (invisalign) with some rubber bands - after religiously wearing his retainer from the first round - has left him with a tooth that will probably need to be pulled because the root appears to be in the process of being re-absorbed.

Robots are so lucky. Anything breaks, just buy/build/install a replacement.

LloydF
11-22-2015, 12:05 PM
Started printing speaker frames:cool:. I noticed how terribly powerful sound is with a robot and also how terribly cpu intensive.
With this Head design a MP3 Trigger can be really effective and fit in the head just fine. I'll post some pictures as I only have a
robertsonics trigger at the moment. (and I think there are much smaller one's out there)

drachels
11-22-2015, 04:00 PM
The robertsonics trigger sounds like it would be the perfect item to use in something like an R2D2 type robot, but with 256 tracks available, you could still cover an awful lot of possible pre-programmed voice responses.

I am making VERY slow progress with my own sound idea. I am trying to use something like Jasper or voicecommand. The idea is that since they can can be set to silent mode with just text replies, I can pipe the text to my Emic2 text-to-speech module and let it handle the audio output processing. My biggest problem with the two voice recognition programs is getting them to run reliably for more than a few minutes. For instance, a few minutes ago I had voicecommand working, stopped to look at the various commands listed in the manual, and haven't been able to get it to record or speak since then.

Now I understand why so many guys like to build robots that use smart phones and tablets to do all the voice/audio stuff.

drachels
12-10-2015, 01:23 PM
6347
@r3n33, here is the current version 8 of the Jimmy head. After the root canal 2 days ago, I have been feeling too poorly to try and convert the extra board mounting holes to a universal mount like I was hoping to do. So, this one will handle a Pi2 or Odroid-XU4 orientated front to back, or left to right.

The zip contains three files. One for the top of his head, one for the bottom, and one for pivot arms. I have tried to design this so that in a fall, he should land on the seam between the top and bottom part of the head. I have purposely made the pivot arms pretty thin, so that if something does break, it will be them first. This is an advantage as they only take a few minutes to print compared to about 3 hours for the bottom and 11 hours for the top, on my printers.

I just realized that I forgot to include the file for the Radio Shack speakers. I'll add them in a couple of hours as I have to go pick up some medicine right now.

Edit: Been feeling poorly so it took longer than I meant to, to add these speaker mounts for this head. 6347

KurtEck
12-10-2015, 07:59 PM
Hope you feel better!

Question: I know that a few people are now using an XU4 on their HRSO1. I still have a C1 in mine as it has the same footprint as the RPI2, so fits without any mods.

However I would like to move to using the XU4. I understand that you have it setup such that it can fit in the head, which probably would work fine, but I was wondering if anyone has adapted it to mount in the position that the RP2 or Edison, if so did you setup to print a new plate for the HROS1? Or did you mount it a different way?

I may try laying out the board and print one and see if it works, but it would be nice to find out that someone has already done so and has a design file to share.

Thanks

drachels
12-11-2015, 03:56 PM
Hi KurtEck,

I'm still feeling miserable due to the shingles, flu, tetanus, diphtheria, and pertussis shots I got in the afternoon after the root canal.

If totally switching to an Odroid, my first inclination would be to print a new P2 plate with the holes spaced for an Odroid-XU4 and mount it in place of the orignal. The only reason I am making room for one in the head is for trying out extra video and audio stuff without having to mess with my normal Pi2 and the code I have on it.

If you want to use the existing P2 plate, I would just try adding a couple of new holes and mount it using only three screws. I would only use 3 of the Odroid holes as the width between them do not match the Pi2 and on the chest side of the plate I don't think there is enough room to drill two new holes, so one screw at the front and two at the back where there is plenty of room to add two holes.

One of the unforeseen problems with using the Odroid-XU4 is that I will probably switch from using the eMMC storage to SDC storage even though it is noticeably slower. The reason is that it is so easy to bump and jar the eMMC loose from the bottom side of the board. I just did it again a few minutes ago. One time, I had to completely restore the operating system to the eMMC as it got trashed when it fell off. There is no provision for adding a retainer to the Odroid board itself, so I am going to see about adding one inside the head.

KurtEck
12-12-2015, 09:52 AM
Sorry to hear that! Hopefully you are starting to feel better now.

Thanks, I am thinking of doing a new P2 plate, but always wondered about instead mounting to p3 plate. But will probably stick with the main method, although different hole pattern.

So far I have not had the emmc fall off, although I have not had it on the hominoid where it can fall. was wondering about two possibilities. The first is to mount it upside down, such that the emmc is on top. The 2nd is to maybe add a retainer as you mentioned to the plate, or a simple plate that goes between two of the mounting holes of the Xu4...

Looks like time for some more disassembly.

KurtEck
12-12-2015, 01:22 PM
The HROS1 is in more parts again as I now have the P2 plate out. I have now relaid it back out using Fusion 360 and I think it is pretty close. I used my scanner to make a picture of the current board and then had fusion read that in as a canvas, which I then dimensioned it. I then drew it out, added all of the existing holes and some new holes for XU4. While laying out where the Xu4 would go, some of the holes were real close to where the Edison holes are currently, so I then moved the edison slightly...

Maybe later today will try to print it out:

6348

But first I am going to lay out a strap that goes accross the board, with hopefully built in washers, which when screwed in with the mounting holes will hopefully hold the eMMC in...

Maybe something like:
6349

Assuming I measured correctly (and also looked at diagram on Hardkernel), I have the main part of strap something like 1.5mm thick, and the offsets at both ends are 3mm high, to hold it up to about same height as top of emmc... Hopefully will get some time to print out and see if it works...

Note, could also potentially only go from the closest mounting hole, it might not be quite a secure, but would make it easier to maybe loosen the screw and rotate out of the way, if/when you need to reflash the emmc.

drachels
12-12-2015, 04:36 PM
Sorry to hear that! Hopefully you are starting to feel better now.

Thanks, I am thinking of doing a new P2 plate, but always wondered about instead mounting to p3 plate. But will probably stick with the main method, although different hole pattern.

So far I have not had the emmc fall off, although I have not had it on the hominoid where it can fall. was wondering about two possibilities. The first is to mount it upside down, such that the emmc is on top. The 2nd is to maybe add a retainer as you mentioned to the plate, or a simple plate that goes between two of the mounting holes of the Xu4...

Looks like time for some more disassembly.

Installing on the P3 plate would probably work okay. I think if you did use the Odroid-XU4 in the chest area, either on P2 or P3, it might be a good idea to rotate it 90 degrees to make the access to the USB ports a little easier. It would also make access to the Ethernet, HDMI, power plug, eMMC and SDC easier.

I don't think it would really make a difference if the eMMC is on top or on the bottom as long as you use your idea of adding the extra clamping bar. Your design looks okay except I think you would want to add a wide spot in the bar to cover more of the eMMC. Right now your design would only clamp the forward edge making it possible for the socket connection to possibly come loose anyway.

drachels
12-12-2015, 04:53 PM
.... I have now relaid it back out using Fusion 360 and I think it is pretty close. I used my scanner to make a picture of the current board and then had fusion read that in as a canvas, which I then dimensioned it. I then drew it out, added all of the existing holes and some new holes for XU4. While laying out where the Xu4 would go, some of the holes were real close to where the Edison holes are currently, so I then moved the edison slightly...

Have you tried importing the step file into Fusion 360 so that you don't have to do all the layout work over again? https://github.com/Interbotix/HROS1-3D-Models/tree/master/Assembly_STEP

It took a while to open the first time in 123D Design, and it was a bit of a pain to save out all the different groups that I have needed, but in reality, it has saved me from having to redesign anything from scratch except the Jimmy head.

KurtEck
12-12-2015, 08:49 PM
Sounds good. I have opened up the step file in to the Fusion 360 which works. I have not tried anything with exporting stuff or 3d printing from it yet, but should try it.

Right now I am using this as a learning experience.

LloydF
12-13-2015, 06:53 AM
If you get a good working design, holler as I am looking to use the xu-4 as well, I just don't have the the plates to mount it, and with out a spare I don't know. There Ubuntu 15 has all the drivers I need, hum.

KurtEck
12-13-2015, 11:27 AM
Hi, still playing around some with the strap.

The current one looks like:
6350

Which I just printed out, and you can see it laying on the bottom of XU4:
6351
Looks like a pretty good fit. Won't work with the Hardkernel risers, but for this case does not matter. I was playing with a second version
6352
But my first attempts at printing it did not work too well, tried with scaffolding, but since now very thick trying to clean it up, I broke one end off. Also would need to play more with the edges as parts of the bottom then were hitting some bumps on bottom of ODroid so would not go in float. Would need to skin off some of the inner parts... Not sure if I will continue on that direction, until I need to. Currently only have one XU4 which is headed for this one and an XU3 lite, which is currently in PhantomX.

Not sure if anyone will find it useful, but here is an STL frile for the strap I just printed. Note: I have not actually printed this file yet, as I had Fusion 360 talk directly with Cura to print it.

Will try to print main board later today.

drachels
12-14-2015, 11:09 AM
Hi KurtEck,
Thanks for the strap. Just finished printing and installing it on my Odroid. Used four left over Hex Socket Head M3x8mm screws and two left over 3mm spacers from the HR-OS1 kit. Using the chrome part of the supplied standoffs should allow for easy mounting either in the head or on the P2 or P3 plate. Can also replace the M3x8mm with M3x10mm screws and mount the board from the bottom side without using the chrome standoffs.
63546355

KurtEck
12-14-2015, 06:58 PM
Yesterday I went ahead and printed my updated plate:
6357

It printed pretty good, ran into a few issues where the holes still are not big enough. The ones that I said that were 2.5mm were almost closed off, and the ones that I said to print 3.5mm, were pretty close to fitting the 3mm screws. I used a couple of drill bits and cleaned them out some.

I started to try to reassemble the robot and find things are a bit tight! Not really any room to mount the strap.
6358
It actually fits in pretty nicely, but I am concerned that the screws on the F3 board that hold the standoffs for the Arbotix-pro ride up into the ODroid, which I don't like. May not be hitting anything critical, but. So trying to decide what I want to do. May look through my parts and see about using shorter standoffs. Might try using some short standoffs and mount upside down. Also from looking how far back the Odroid fits into the space, from a back end picture:
6359
I could probably move the Odroid back a few mm and maybe miss those standoffs. Don't want to move too far or will hit some in the back. Will play some more and see what works.

Update: I switched to using the metal standoffs that came with the ODroid (or one of the boards with Odroid), and it is about a mm shorter and now the board is not on the lower screw :D. Tomorrow, will look at what to do about emmc strap. The current stuff wont fit in the space. but can probably fit something that is held in by friction or the like.

drachels
12-15-2015, 09:46 AM
While you are trying various mounting options, you might want to see about inverting the Odroid and just using your strap and two 3mm spacers. From looking at mine, that configuration should give you 3 or 4 mm of clearance below the Odroid and the P3 plate.

KurtEck
12-15-2015, 10:56 AM
Thanks

I might try that. Right now I am updating my plate slightly, to move the ODroid back about 6mm, such that there is more room for the
power plug and still fit stuff in front, plus enlarged the holes for screwing the plate into the upper servos. Once I print out a new version, will try it again and decide which way to go.

Thanks again

Kurt

drachels
12-16-2015, 05:32 PM
I got tired of waiting for my new filament shipment (which actually came about an hour ago) and went ahead and finished printing a V8 Jimmy head in orange yesterday. From the pictures you can see that it contains 1 camera, 4 Radio Shack speakers, and an Odroid-XU4. I plan on adding another camera and an Emic2 text to speech board in there.

636163626363

Simply because I always like to try other options, there is a good chance that I will do like Kurt and move the Odroid into the chest, replacing the RPi. Later though...I really need to get back into playing with the software and programming side of things. I just realized I haven't had anything but the head powered up for almost 6 weeks.

The color really doesn't matter. My son does miniature painting as a sideline business at the gaming store where he works. I bought him a new airbrush system for Christmas so I'm gonna turn him loose on all the printed parts and see what I get back. I really need to do some more work on the Elvis head so I can get him to paint that too.

KurtEck
12-16-2015, 06:04 PM
Good looking head!

On the Odroid front, I made some more progress yesterday on mounting mine in the chest. I printed up a new one where I moved the Odroid farther in (pretty close to centered). I then decided to go ahead and mount it upside down. I did not have any 3mm screws that were long enough so I used some ones I had from earlier for Lynxmotion robots. Also found my wide part of the strap was off when hanging upside down, so I made another version where it would fit for this orientation. I then used the original one on the other side as the set of standoffs. I can post all of the stl files if anyone would like them.

6364

Next up need to work on wiring. The 90 degree usb adapter to connect to Odroid pro may not work here. May use normal USB cable.
But that maybe more of a topic for another thread. Also since the XU4 only has 3 USB connections, may need to find a place to mount a USB hub (maybe head?).

Actually next up for me is more side projects, like I received my FSR boards back, so need to assemble them and install in feet, likewise need to finish/install new hands... But again probably different topics

Update: Included zip files with the printed parts for XU4...

KurtEck
12-21-2015, 10:22 AM
Quick update: Right now I am in the process of trying to get my HROS1 (first beta) back into one piece with some new parts including FSR feet, ODroid XU4 and printed hands and now a Chappie head.

Yesterday I tried putting on the Chappie head, but the head does not want to go over the neck brackets and I do not want to try to force the opening at the bottom to be wide enough to go over the brackets. Looking at the instructions (http://learn.trossenrobotics.com/projects/155-hr-os1-kit-assembly-guide.html)
Likewise I believe these parts look the same as in the 3d model as it is currently up on github (https://github.com/Interbotix/HROS1-3D-Models)
You can see these parts in this picture:
6374

However when I go up to the shapeways store, the Head mount 3 bracket (https://www.shapeways.com/product/8DBHG53LH/hr-os1-head-mount-3?li=shop-results&optionId=37489698) looks different, so I downloaded the file and sure enough the diameter appears to have changed from about 60mm to 50mm. So I just printed one out. So it looks like I need to disassemble #5 :D and put in the new neck!

I thought I would mention this for the other beta 1 people who wish to update to some of these newer things. Wonder what else changed?

Kurt

LloydF
12-21-2015, 11:00 AM
Oh , my, nice catch. Some new armor is supposed to be coming down the track as well and, "oh long Johnson!" said the cat.

KurtEck
12-22-2015, 04:43 PM
Thanks, still putting my HROS1 back into one piece. But I keep creating diversions for myself, as I am still learning to 3d print and also play with 3d programs. Currently using Fusion 360 and Cura...

Right now playing with adding a front chest piece to the HROS1, which should hopefully allow me to fit a Teensy 3.2 controlling an 8 neopixel bar. So far it is real real basic, but learning:

63786379

The front of it shows some black as I just swapped in the red PLA for the Black one that I was using.

The Neopixel bar fits in from the back. I will make a few enhancements for it, like printing a ridge right around the inside of the hole that is the same height as the resistors on the front of the bar, as to let it lie flat. Also may print a few round knobs to go into the two mounting holes of it to hold it in place.

The Teensy (not the one I show) fits into the inside of the tabs I show (currently a little too tight), I am now thinking I want the front to wrap around the 45 degree edges on both sides a little. Currently I have the screw holes on the outside edges larger than on the inside tabs, as to allow me to tap the inside holes.

Again no great piece of work, but it is an interesting learning project...

tician
12-24-2015, 11:24 PM
Since I've been on a bit of a nostalgia trip recently and there are more plot holes in chappie (https://www.youtube.com/watch?v=OR2TI0hfg7A) than a block of aerogel, thought I might as well add some of the designs I've been working on.

First thought several months ago was to make a Mr. Stay Puft (https://www.youtube.com/watch?v=oPr7cYLq3dk) suit from Instamorph and/or silicone, but the HROS-1 has enough trouble with its default gait and I've still not finished my stochastic gait seekers that might find a decent waddling gait to move around with the suit restricting joint ranges.

Given some thought to making HK-47 from KOTOR, but might be a bit difficult to scale down to an HROS-1 while keeping stereo video. Might work if it were chibi (http://www.vgcats.com/comics/?strip_id=93), although it would still never do it justice without the voice and attitude.

Now thinking something like Glitch from Metal Arms (https://youtu.be/YJeAm0KYL0E) since the head (http://vignette1.wikia.nocookie.net/metalarms/images/6/69/Glitch_.jpg) is literally a braincase with attached hardhat and chin/face guard (http://fotos.trucoteca.com/wallpapers2/5139.jpg). Would also make it possible to have swappable 'helmet'/'persona' options if there were 'standard' dimensions and mounting points on the braincase. Thinking along the lines of 64mm x 64mm x ~110mm braincase (simplest would be 64mm x 64mm tube with endcap on back and pcb on front; 60mm x 60mm internal cavity) with 35mm blower mounted to the face PCB that holds the two RPi cameras (with exhaust at the bottom of the chin), eight WS82812B/neopixels, and a small speaker in front.
6388

Really hoping that the usb_gadget kernel modules will start to be included with Rasbian since the introduction of the RPi-Zero. Not so interested in the RPi-Zero, but having a precompiled g_ether module for RPi-A+ would make a nice USB camera with some pre-processing of images.

LloydF
04-01-2016, 12:50 PM
Are there multiple beta parts from the Trossen folks. I can not remember ever being notified there were changes in the Beta HROS1 model. Nor any instruction on how to mount skins and that just wrong, but they don't offer them so I guess It is all good.

KurtEck
04-01-2016, 05:08 PM
Are there multiple beta parts from the Trossen folks. I can not remember ever being notified there were changes in the Beta HROS1 model. Nor any instruction on how to mount skins and that just wrong, but they don't offer them so I guess It is all good.
Have not seen any new stuff yet. Keep looking. I know they mentioned having a few different sets of printable models back when we first ordered, but have not seen anything posted here, nor any updates to the models project.

For example I remember Andrew talking about the different skins in the video up on 21st Century Robots:

https://www.youtube.com/watch?v=1IcDh4i2ag0

I see some different skins up on those web pages: http://www.21stcenturyrobot.com/
Probably done by some different schools and the like. I have not looked yet to see if any of them are open source or not.

Kurt