New run of orbshields

[kenyee]

[quote="vputz"]New cases designed. Plus bonus chance to learn more about Blender and Inkscape. At least I can do something semiproductive with the delays...[/quote]

NICE! Wayyyyy better than the previous case IMHO. Very compact and svelte :-)

[vputz]

[quote="kenyee"]Wayyyyy better than the previous case IMHO. Very compact and svelte[/quote]

Well, it's largely your fault, so take some credit; I was pretty complacent with the big fat case :) It might be possible to get it smaller, but I worry it would start getting annoying to build (and I haven't tested this in hardware yet, so I may be missing something, but it sure looks like it'll work). The board's actually taking one more tiny addition, some through-hole pads for the USB lines so I can build a testing pogo-pin jig and test boards without having to plug and unplug the orb and computer. But that shouldn't affect anything electrically and it looks like there's room.


[quote="orb-a-new"]I had a thought about the "belly button". Could it be used as a chording button? [/quote]

Good question! I'm honestly not sure. The short answer is "there's probably a way"--the long answer is trickier because I don't know how to do it off the top of my head; you'd have to not only figure out how to capture that input, but also disable the "auto-zeroing" feature of that button. I'll have a think at some point... I'm still trying to figure out how to feel better about aiming--I'm still awful with fine control, and while the "sniper mode" change-gain-on-the-fly solution technically works, it's not perfect (and my "absolute zone/relative zone" mouse control solution didn't turn out as nicely as I wanted either). So far the best way to get precision with an orblike device is to have a desk-mounted SpaceBall in your left hand and a mouse in your right... (which is awesome, but you run out of keys pretty fast).

[vputz]

Okay, brain is full for a week or so. I have more SMD parts on order and sent off for a batch of SMD boards, more stencils, and case parts. The 10 through-hole kit boards should be here this week and I think I can get those out for cost to testers. I have three prototype SMD boards coming this week that I can probably get out after my guest leaves, and the initial "tiny production batch" of 20 might be here by the time she leaves as well, so I'll get a message out to existing testers soon.



Some "bounty" ideas that I'm considering:



1) I am really interested in adding more devices to this; I mean, basically it should handle any input device that communicated via serial port and turn it into a HID device. There were a bunch of old joysticks, gamepads, wheels, digital tablets, etc that would work.



2) For the Orb, I'd like to see scripts that would interface it with 3d CAD and such: Google sketchup and Blender come to mind since I use those.



As for today, to my considerable shock the sun is shining so I think I will emerge from my cave for a bit and brave the shock of the outside world ;)

[VDX]

... "digital tablet" sounds really intersting too - have an old professional one from the Win98/NT-times with pen and crosshair-mouse ... would be great to reactivate it again :mrgreen:



Viktor

[JimH]

vic,

I saw on the Orbitron beta forum, mention of the new boards being for sale on the webstore. what is the address of the webstore? can't find it her or there. i missed the saturday date for the boards to go on sale.

thanks,

jim

[vputz]

Jim:



Ha, hadn't meant for the store to really be active yet--it's still in a trial mode! ;) You'll get a kit sent out hopefully tomorrow... with my guest in town I haven't had time to assemble things although she'd probably find it funny. The kits are in and the first 3 SMD prototypes are in, so tomorrow should be some good fun.

[vputz]

OK, 9 kits made with 2 accounted for going to testers. The three prototype SMD boards... not so successful. I didn't have quite enough parts so only one is fully populated and while it accepted programming OK, it's not functioning and I probably can't test more while my guest is here (bless her heart, she even helped stuff bags, but I can tell it's not her favorite activity). Production is pretty much on hold for a few weeks, but I can ship existing kits out.

[pshros]

I am very interested in acquiring/building one of these shields. Where is it most convenient to track progress on this? Here?

[vputz]

For now, yes... but this site is really more about the orb itself and it would be rude to coopt it, so I am setting up a webstore and forums for the "Orbotron" itself (silly name for the new device). It's not really a shield any more but its own built-in Arduino-compatible board so that the final SMD board SHOULD be a "buy it and plug it in and it works" device but the kit is still available for those who want to do it themselves and understand it a bit better.



Sorry for the delays, but it really is finally getting to its "done" form. I'm feeling good about the kits, but the SMD process will take more work (right now it's a little labor-intensive for me to make a lot of these, but maybe I can either get it down or, if demand ends up high enough, contract out construction). I just keep learning by making mistakes, and I'm learning a LOT--which means loads of mistakes ;)

[kenyee]

[quote="vputz"]prototype SMD boards... not so successful. I didn't have quite enough parts so only one is fully populated and while it accepted programming OK, it's not functioning[/quote]

Did the solder mask help? And did you solder by hand or did you use your heating plate?

[vputz]

Irritatingly, the pitch between the pins on the QFP package was fine enough there wasn't any solder mask there! However, when I was able to put the chips down straight, I didn't get many bridges; when my hand slipped and smeared the paste, I got them. I'm going to do more work on my "vacuum pen" jig to make it more stable.



On the "kit" boards, only the 32u4 chip is SMD, and I had a stencil, so those came out OK (I think of 8 boards 2 required significant rework, one because I was stupid and touched the chip with tweezers after the solder was molten). The three SMD boards I tried were the prototypes so I didn't have a stencil and had to apply the paste with a wire, which explains the bridges (hard to put down "little enough" solder for the fine pins when you're daubing it on...) I think one would work but the USB connector had bridges, and those are really hard to fix with the way the connector is made. I just got the remaining parts for the SMD boards so I'll try to fix them. The upcoming shipment of SMD boards has a good stencil, which will mostly fix the problem.



And no, I don't solder the SMD parts by hand except for rework--I use the plate. It's much more reliable. I don't even think I could solder the USB connector by hand because the contacts are mostly under the connector.



The "pogo pin" fixture for programming works like a champ, though, and I may be able to work out a way to use the JTAG TAP pins to do an automated "boundary scan" and check for solder bridges, which would be awesome (right now I use a USB microscope which is only accurate visually).

[vputz]

Had a little time this AM to try and repair the SMD boards. I'm having a real problem with the USB connectors; they're hard to align, and VERY hard to repair if something goes wrong. So far my success rate with the 3 prototype SMD boards is 3/3 for programming the ATMega, but only 1/3 for plugging into USB and functioning. But it [i]does[/i] function, and I'm sure with the stencils things will get a bit better.



Anyway, here's the current revision, and this does seem to work fine. I'm still not happy about the USB micro socket; I wish it had better strain relief, but this is evidently the way it's done. The "case" could even have smaller standoffs, although this isn't bad at all (the screws are obviously too long, but I can get different screws).



In fact, now that I've worked with this a few times, I'm sure I could get it even smaller (and if I could get it a little smaller than 2" end to end, price on the circuit boards would drop too) but I really think it's about time to just start shipping things rather than faffing with the design. I have 20 boards coming, and stencils here, so in a couple weeks I should be able to cook up a batch; just hope I can get the reliability on the builds higher or it's going to eat me alive on cost of lost parts...



Edit: make that 2/3 working. The last one acts completely dead, but I may get lucky (there's a bad connection somewhere... when I measure VCC to ground I should get 5v but I instead get 1.5, so there's a short or some such somewhere. Even so, getting 2/3 working convinces me the design is basically sound).



Edit: 3/3. Damn micro USB connectors.

[VDX]

Hi Vic,



you can send me a kit with loose SMD-parts -- can attach them on my own with microscope support ... my until now smallest 'soldered' connectors were 30 microns wide and wires with 10 microns diameter - but this is a bit different from 'normal' workflow :wink:



Viktor

[kenyee]

[quote="vputz"]Irritatingly, the pitch between the pins on the QFP package was fine enough there wasn't any solder mask there! [/quote]

Eek. Can you find a place that does have enough accuracy to add a solder mask between the QFP package's pins? You're still going to get bridges even w/ a stencil IMHO :(

The prototypes I've gotten back from Oshpark have soldermask between pitch that small I think...I'm using the MSP430 in 20-pin TFP (.65mm pitch). If your QFP is standard, it's .5mm pitch.



For the microUSB port, what does the back of it look like? Can you see the pins?

I took a quick look for microusb ports and a lot have those pins buried underneath which seems stupid :shock:

And do you have a solder mask between the pins?

These look very slightly longer than the rest that are available:

http://www.ebay.com/itm/5pcs-USB-Micro-Type-B-5pin-Female-Jack-Connector-SMD-SMT-Socket-Surface-Mount-/261201638747?pt=US_Radio_Comm_Coaxial_Cables_Connectors&hash=item3cd0d4b55b

Or maybe this:

http://www.newark.com/fci/10103593-0001lf/micro-usb-b-receptacle-5pos-smt/dp/82R7406



If it's not reliably buildable, you should change the part...even switch it to mini-usb like the launchpad if you need to. Finding issues like this is what prototypes are for.



And yes, you could have shrunk the size down...lots of space between resistors. If you make it two-sided SMD, you should be able to bring it under 2"x2".

[vputz]

VDX: I'll see what I can do... not sure I can come up with a good way to package everything nicely, but I'll think about it. You'd have to flash the chip yourself, though.



Kenyee: Actually as it turns out the footprint I used didn't have solder mask, so I can't blame the PCB folks. But this was the footprint adapted from Adafruit's 32u4 breakout board, so I'm going to ask if they really don't have solder mask on their 32u4 chips.



The microUSB port has the contacts barely visible with my microscope, but it's hard to get the soldering iron under there (I can wedge it between the pins and the casing, so I can heat and reflow the pins, but it's hardly accessible). I haven't found any that don't have the pins pretty much underneath. I asked around and one guy suggested [url=http://www.farnell.com/datasheets/44939.pdf]these[/url], which still have the pins mostly underneath but do have legs which fit in holes on the PCB, which would help with alignment and also provide a little strain relief. If I can find a better one than that, I'll give it a try, but with the first batch of 20 I'm going to just do my best and see what I get.



Since I have boards on order and one set of cases/stencils built, I'll build these as the first small batch since they should function just fine and I don't want the boards/cases to be a total loss, but I'll plan a revision; I'm pretty sure I can drop the length under 5cm, which would be only a mildly tight fit but drop the board cost a little through places like itead/seeed who have "max 50mmx50mm" boards pretty cheap... although on the current revision it does look a little more crowded as I added some thtrough-hole pads to make pogo-testing the USB port easier and possibly giving access to the TAP pins for boundary scan. I could use tag-connect for the boundary scan, though, and I'm not convinced the USB pogos are worth it for testing since they would just test the board, not the USB connector. And 2-sided SMD would be much tricker for me since I can only use the hot plate for one side and this is labor-intensive enough already that I can't afford to do any real hand soldering on each one. Contract assembly would be one option, but I didn't sell all that many of the original OrbShield so I almost certainly wouldn't have the volume to make that worthwhile here.



Edit: Sigh... yeah, I can get it under 50mm. Needs a lot more vias and I'll have to move the USB test pads to just solder pads underneath (and probably doing the tag-connect for TAP since I can't imagine anyone but me using it) but it should work and be very pretty... there's JUST enough room to put USB-ATMega-ADM232-DB9 since it's going from 2.5" to just under 2". I'm still selling the rev1 boards, though!