New run of orbshields

[VDX]

... for 'virtual XP-modes' on Win7 - I've successfully used the serial SpaceOrb on a WinXP in a VirtualBox on a Win7 Ultimate with the standard drivers.

But too, want to use it in Win7 native for CAD applicarions or space-gaming

Viktor

[rayz007]

Good work vputz, question?Will the smd version still be able to run both wasd and normal space ball modes?

getting excited to try it out

Viktor I don't have a serial port on my comp.so I had to use
a serial to usb adapter. So no fun for me.

Rayz007

[vputz]

The SMD version will be reprogrammable and should support all the features of the original orbshield (WASD, mouse, joystick etc). It's basically the same thing as an Arduino Leonardo with an OrbShield except on one board. It will come flashed with the "basic joystick" program as that will let me do basic testing here and will cover most use cases (a program like X360CE should allow it to emulate other controls like keyboard etc without reflashing, but the Arduino environment will let you reprogram it at the basic level). The only annoyance is that on Windows 8 you will have to authorize an unsigned driver, which is a bit of a pain (forces a reboot) but there's no executable "driver" as such--it's just going to use the Leonardo bootloader and USB/Serial driver with a VID/PID change so that it shows up as an "Orbotron 9000" in your hardware list.

In fact, there's really no reason this hardware wouldn't let you use other old serial devices like joysticks and racing wheels with the appropriate program; I just don't have anything like that to work on at the moment.

[vputz]

And on the subject of "this is safe and normal behaviour, and nothing can possibly go wrong"...

My homemade "smd reflow hot plate" controller basically works, although it needs some tuning since right now there's enough lag in the hot plate that it overshoots significantly. It also needs an enclosure, since it's exposing 240V to open air, which isn't awesome. But fundamentally it works, complete with my ultra-classy "infrared sensor on a microphone stand using ethernet cable" technology.

Sigh.

Prototypes are on order, stencils are on order, and with a touch more tweaking I think my process is going to be pretty stable. Not sure if it's this weekend or next, but I will get the next round of test boards out soon!

[jaycrowe]

ultra-classy "infrared sensor on a microphone stand using ethernet cable" technology

Keep up the great work!

[rayz007]

Sounds great vputz. Any chance I can get a "test" board? I'm dying to try it out. I'm
pretty competent with computers and will familiarize myself with the Arduino programming
environment. But it looks simple enough. Thanks again for all the hard work.

rayz007

[rayz007]

Btw, I am playing Retrovirus and it's pretty good. Has anyone tried their space orb
with that game and how does it perform? Using mouse and keyboard is very difficult
for me. Even if I lower mouse sensitivity to 100 dpi and set in game to 0, it's very
jerky. I resorted to using a Saitek P990, and it works well, but the space orb should
be a better controller. It must be better in joystick mode than wasd mode, if the above
description of wasd and mouse still holds true. Looking foward to hearing how it works
for anyone as well as trying it out myself.

rayz007

[kenyee]

Vic:

If you haven't fixed the size already, is it too late to ask for 0805 SMD components? Much easier to solder by hand, though I suspect most people would have problems w/ the ATMega you'll use for the arduino portion
And any schematics of the board?
The prototype photos on this thread don't look as good as the ones I've gotten back from Oshpark.

Are you using the Dangerous Prototypes sick of beige plexi case? If so, which size? (I'm a bit partial to the DP6037 ones )
I've found other fully enclosed cases, but anything w/ external connectors is a PITA because you then need cutouts which cost an arm and a leg

Thanks for all the work you've done on it...I was just telling a friend about how I used to use a SpaceOrb way back for FPS games and the wikipedia page led me to the OrbShield that I didn't even know about

p.s., if you're doing SMD reflow, it's easier to just buy a $30 toaster oven. Lay the paste down, place the components, stick it in the oven at 240F or whatever the reflow temp is and it works decently well..

[vputz]

Rayz: We can probably get you a fully-assembled test board. My "hot plate in the kitchen" technology is relatively stable and the stencils came in (hooray!) so I should be able to test the SMD version this weekend.

I played through Retrovirus with the orb (partially on the first version of the kit and partially on a mockup of the second). It handled like a dream although I did have to set up the orb to have higher sensitivity. And my usual issues with the orb -- fine-grained aiming. I didn't even use my "sniper mode" fix... but I have another idea I'm going to try to change the mouse mode emulation which I think will make it even better. Retrovirus is great--the ending boss is TERRIBLE (everything after the point at which you beat the "main boss" is pretty unnecessary and even that fight is pretty bad, but everything leading up to it I quite liked).

Kenyee: most of the components are 0805, but unless you're just wanting it I'm not likely to offer an SMD kit (too many fiddly bits); the kit version will be PTH with the ATMega already in place (I can reflow and program it here, so when the kit goes together it should "just work").

The prototype boards look that way because they're seriously prototypes; SpiritCircuits here in the UK offers a fast (within a couple days) and free turnaround with no solder mask or silkscreen (which is why they look so rough); I just have to place a paid order after every three prototypes--and since they let me tile multiple Gerbers (I have to cut them out myself, also why they look rough), this is HUGE since I'm able to cycle through three boards of the SMD version, the PTH kit version, and the Kettletron 9000 (the temp controller) and make progress on all three. Great company. The actual boards are a bit spendy from them, but the free/fast prototyping service makes it worthwhile.

No schematics yet, but the whole project will be open source once everything's stable.

The idea for the case came from DP, but as is my usual contrary way, I just learned how to write a program to generate them in OpenSCAD, and this way I can change two numbers in a file for dimensions and get the case to fit whatever board I'm making. The same file with a few tweaks generates enough to fill a page for the laser cutters, and I can add whatever holes I need (I guessed wrong first time round, but I did get what I asked for). Much easier and cheap as chips in the long run because the laser cutter guy charges by the sheet and this just fills up a sheet. I added another routine in the file which "draws" the case with a circuit board and screws/spacers for the label... still needs some tweaking in Blender for a nice "drawn" look but it works pretty well.

case_xplod.png (33.44 KiB) Viewed 103658 times

I debated the "toaster oven vs hot plate" item for quite a while, and went for the hot plate on the basis that the Sparkfun guys liked it a little better than the toaster oven... and I thought my temperature controller would actually work with one, so I wanted to learn something new (usually a bad idea...). But they had a nice cheap Target griddle with relatively low thermal mass... this stupid hot plate has such a lag that my homemade PID controller overshoots pretty badly. I've tweaked the constants and it looks more stable, but measuring the temp is tricky (thus the IR sensor).

HOPEFULLY this weekend will be good for testing, but I think I'm out of ATMega chips and my blasted supplier is having trouble again (billing address in the US, invoice and shipping here in the UK means confusion EVERY TIME I ORDER, but we'll get there).

[rayz007]

Sounds good Vputz. PM me and let me know what you need to
make this happen. Really looking forward to trying this tech out.


Rayz007

[kenyee]

Kenyee: most of the components are 0805, but unless you're just wanting it I'm not likely to offer an SMD kit (too many fiddly bits); the kit version will be PTH with the ATMega already in place (I can reflow and program it here, so when the kit goes together it should "just work").
The idea for the case came from DP, but as is my usual contrary way, I just learned how to write a program to generate them in OpenSCAD, and this way I can change two numbers in a file for dimensions and get the case to fit whatever board I'm making. The same file with a few tweaks generates enough to fill a page for the laser cutters, and I can add whatever holes I need (I guessed wrong first time round, but I did get what I asked for). Much easier and cheap as chips in the long run because the laser cutter guy charges by the sheet and this just fills up a sheet.

Wow...you're really hardcore
Thanks for making the components 0805...if the board's smaller, I think I'd like that and I have a bunch of 0805 parts depending on what's in the schematic. Did you see the Sick of Beige V2 boards w/ the extra stacked rings to seal up the sides? They basically added extra rings w/ cutouts for the components you could just stack as ring pieces to close up the sides of the "case". It does seem like something that would be perfect for a program to generate cutouts in the right places.
I'm jealous you guys have SpiritCircuits...we have nothing like that in the US...Oshpark is as close as it gets and that takes weeks to get new prototypes that are cheaper as you get smaller, but still around $19 for four 2x2" boards

ken

[vputz]

I like the "rings" design! I considered that for this but with cutouts on both sides (for the problematic DB9 port and then the USB port) I would have a few "half rings" which wouldn't work so well, and really I just wanted something light and cheap to prevent the board from being knocked around. If I finish the Kettletron into a workable product I'll need a proper enclosure for that since having 240V exposed isn't great for the hobbyist

I'm glad we have SpiritCircuits here myself, because what I did in the states was have my minimill set up to do 1-layer boards... prototype in half an hour! But a) two-sided was tricky, b) didn't have resolution for fine SMD parts. And over here--I just don't have the room at ALL; all this is being done out of my living room and I don't have a place to put the mill up (it's in pieces lying in my server rack). In the US I did use Sparkfun's associated service (can't remember the name) as well as Seeed Studio, but you're right--the delays are painful (I complain about delays with ordering parts here, but it's usually just a few days).

Now I'm actually wondering if the components are really 0805, because you describe them as being easier to SMD solder and I still find them hard as heck! But it's getting there... I've got a vacuum pen I'll try this time and I even rigged up a little gantry to stabilize it.

[vputz]

Okay, the new stencils work pretty well, the vacuum pen works better than tweezers for placing chips, the hot plate and Kettletron work pretty well for getting solder to flow, and I'm pretty sure the production boards with solder mask will help with the solder bridges... so why doesn't the dang thing accept programming? SO frustrating, but every failure gets us a little closer. I'm going to plug away with this one a bit more; maybe I'm missing another solder bridge or something, but all the right pads seem to be getting voltage and such, so I'm mystified. GRR.

[kenyee]

I like the "rings" design! I considered that for this but with cutouts on both sides (for the problematic DB9 port and then the USB port) I would have a few "half rings" which wouldn't work so well
...
Now I'm actually wondering if the components are really 0805, because you describe them as being easier to SMD solder and I still find them hard as heck!

FYI, the "half rings" would still work. You still have the holes to put the screw into on the tips of the rings.
The 0805's should be barely small enough to read lettering on (4 letters). Much easier than 0603 because of the bigger pads. Soldering by hand is fairly easy...tin a little solder on a pad on one side, glue the part to it, then solder the other side...assuming your pads are big enough.

For the bootup issue, are you sure you have all the caps/resistors needed for programming? I've done MSP430 hacking and you need the 330 ohm resistor to program it and the 2.2nf cap for it to boot and be recognized by the programmer. I've never worked w/ atmegas though.

You shouldn't be having solder bridges w/ the right solder paste...the flux should make sure the chip pins land on the pads and not connect each other...of course that only works if you have a solder mask

p.s., you're lucky getting parts too...mouser/digikey take a week to get stuff to me

[vputz]

For the bootup issue, are you sure you have all the caps/resistors needed for programming? I've done MSP430 hacking and you need the 330 ohm resistor to program it and the 2.2nf cap for it to boot and be recognized by the programmer. I've never worked w/ atmegas though.

Well, the 328ps worked well (the first version of this kit, the one that went to testers before I decided I needed the revision) used the AtMega 328p in DIP format, and it was really nice because I could program them with nothing more than a ZIF socket and a 6-pin header--if I could still use the 328p I would, but the 32u4 is just going to be so much nicer with proper USB support. I haven't worked out if the 32u4 needs more to program... I've had some people tell me I just need the 6-pin header hooked up and some say that it needs a crystal as well. At any rate with the two prototypes I've gotten working, both needed a lot of fiddling. The lack of solder mask IS one big downside of these free prototypes; I get lots of bridges on the chip itself. But if I can get the board itself working, that'll be enough to order the first semiproduction batch.

Eh, I'll worry about it tomorrow. Out way too late tonight pretending to have a social life; we'll have no more of that