StickOS

[Bruce]

OK. I'm closing the offer now so I don't have to build 10K of these. If they get popular down the road we can make & sell them to anyone not already in on the group buy.

I have about 1 more week or so to finish a job, then I'll get busy on these. I have received my 3 samples, some from Dave and Walter.

[ScaleRobotics]

Thanks Bruce, I think these are going to be fun!

Here is something that exists, only hard to find, and almost as expensive as the others:



http://mysite.verizon.net/reswwg6r/id9.html

And I found this at the Mikroe site. Lots of bells and whistles (not claimed to be compatible with StickOS). And the graphic lcd would certainly be very difficult, if not impossible to interface with using StickOS. But interesting....

http://www.mikroe.com/eng/products/v...timedia-board/



Walter

[ScaleRobotics]

Hey Bruce,

No rush, especially since you are doing all the work. Just wondering if you had any updates. No worries though.

Thanks,

Walter

[Bruce]

Almost ready. Have a few parts to lookup for BOM, and some hand-routing to finish, but I'll post here when it's ready.

[Bruce]

In the process of selecting parts. Anyone have a problem with tiny surface mount switches that may require the end of a pen or something similar to press, or would you prefer a raised button you can press with a finger?

Last part to select - so please let me know ASAP.

[mackrackit]

What ever you think is good with me.
StickOS
StickPress Buttons

[ScaleRobotics]

Anyone have a problem with tiny surface mount switches that may require the end of a pen or something similar to press

No problem with small switches here.

[tenaja]

In the process of selecting parts. Anyone have a problem with tiny surface mount switches that may require the end of a pen or something similar to press, or would you prefer a raised button you can press with a finger?

Last part to select - so please let me know ASAP.

I'll likely swap it out if it's not tool-less... or never use it.

[ScaleRobotics]

Hey Bruce,
I know you are a busy man. But any updates on the StickOs project?

I noticed that someone is making a zigflea wireless board for the SparkFun StickOS product. Here is the link: http://overtone-labs.ning.com/profil...oard-for-cui32

Thanks,

Walter

[Bruce]

Sorry for the delay Walter. I'll see if I can't get this ready by Monday.

[Bruce]

Walter,

Here's a couple of screen captures of the board so far. I've attached the DipTrace board file also if anyone wants to have a go at making it better, smaller, etc. It would be nice if someone else verified all connections also.

It's similar to the SparkFun board, but I dropped the USB host option, rearranged things a lot, and attempted to make it Stamp-Like, but it's tough to route if you keep the headers close enough together to insert in a small breadboard, and I did NOT want to go any smaller than 0805 resistors/caps.

Board size is 3.2" x 1.8". The smaller capture is 100%. The large is blown-up so you can see everything with connections.

I can have a BOM ready in a day or two, but will wait for feedback, changes, yada-yada...;o)

[Darrel Taylor]

Bruce,

What kind of breadboard will they fit into?

If JP6 is 0.1" spacing, then JP5 and JP7 are 1.5" apart.
But all my solderless breadboards are 1.1" max.

The headers would end up in one of the power rails.

[Darrel Taylor]

Duh! Nevermind.

It would go across the power rails between two breadboard strips.

Sorry, I'm just excited about it.

Carry On.

[Bruce]

Hi Darrel,

I suppose it could be done, but the board would be pretty long, and there wouldn't be any room for labels on the connection headers. The 64-pin PIC32 TQFP package is pin-to-pin/side-to-side 12mm wide, so you have little to no room.

[Darrel Taylor]

No problem.

I'm sure it will work just like this ...



In fact, it fit so well I figure you must have designed it that way.

Works for me.

[ScaleRobotics]

Looks great Bruce. Yeah, I kind of hoped it would fit on ONE breadboard, but that sure is a lot of I/O! I have two or three breadboards to put it on if I need to.

One other thing. What kind of USB connector is that? It looks like one of the printer styles, but I could be wrong. Not sure why I think it matters, but it would be cleaner to keep the low profile over the whole thing and do a smt mini usb connector instead. Who knows, someone might want to send this thing up in a model airplane or rocket or something. (Maybe that's just me )

Thanks a bunch Bruce.

Walter

[Bruce]

3 breadboards! Yikes...

I'm putting female headers on mine for the analog pins.

I could suck-it-in for a version that would drop right into a standard single breadboard, but I'm just not in the mood to sit & hand-assemble 30+ boards under a microscope for a freebie...

And - it's really a drag for someone that gets one to replace any parts < 0805 by hand.

If someone prefers SMD switches, I can change that, but they're normally more expensive. I definitely do not want the SMD USB-A connector. These things tend to rip-off & lift traces with just a little wiggle. I tried really hard to keep it super cheap, but I'm open to feedback if someone wants something changed.

You should be able to edit/view it with the free version of DipTrace. The raw file is there if anyone wants to have a go at making it better or smaller.

Walter - the mini USB connectors are a pain, and not everyone has those teeny cables. I used a standard through-hole A type for stability, and it's easier for me to assemble. And - if we used the SMD mini USB connector, it's way too easy to rip-off the board if you lift up on the USB cable. These need to be installed in an enclosure (like say a PICKit2 or 3) where the enclosure itself provides support.

[mackrackit]

I do not have a problem with three bread boards.

It all looks good to me Bruce!

THANKS!!!

[ScaleRobotics]

I'm putting female headers on mine for the analog pins.

That's a good idea!

I could suck-it-in for a version that would drop right into a standard single breadboard, but I'm just not in the mood to sit & hand-assemble 30+ boards under a microscope for a freebie...

Can't blame you there, totally understand.

If someone prefers SMD switches, I can change that, but they're normally more expensive. I definitely do not want the SMD USB-A connector. These things tend to rip-off & lift traces with just a little wiggle. I tried really hard to keep it super cheap, but I'm open to feedback if someone wants something changed.

I don't see a need for smt switches, and I can get by with whatever USB connector, or pinout you put on it.

You should be able to edit/view it with the free version of DipTrace. The raw file is there if anyone wants to have a go at making it better or smaller.

Looks like a nice layout, and I know I will be happy with the silkscreen labeling, so I don't have to refer back and forth to a pdf file. Not too big, and with your idea of the female header for the analog, I'll be able to fit it on one of my larger breadboards (not three).

Walter - the mini USB connectors are a pain, and not everyone has those teeny cables. I used a standard through-hole A type for stability, and it's easier for me to assemble. And - if we used the SMD mini USB connector, it's way too easy to rip-off the board if you lift up on the USB cable. These need to be installed in an enclosure (like say a PICKit2 or 3) where the enclosure itself provides support.

Bruce, I totally agree with you keeping the cost down, and you have done an excellent job. I know that this kind of project takes time, even for someone as professional as you, so thank you again for jumping in and offering to do this!

As for the mini-usb, I can see your point of not wanting to install them, and I can live with that just fine. But I have purchased plenty of cheap boards from Sparkfun, Microchip, Olimex, and other board manufacturers who include a mini USB connector on their board, but offer no case. Don't take me wrong, I am just saying that in my opinion, I don't think they "need" to be installed in an inclosure. But I agree that would make them stand up to more use/abuse.

Your pinout would certainly be one of the most durable, compact, and cheap. So by all means, I agree it is a good way to go. Just think a few people will have to make an easy cable to connect it with.

[Bruce]

OK here's one that fits on a breadboard. Headers from side-to-side are 0.9" so you'll have 1 hole left over on the outside edges for wire. Board edges will hang over a tad, but it now fits. Overal size is now 3.1" x 1.475".

Changed the 1x16 analog header to a 2x8 so that fits, and added Walters Mini USB connector.

If someone else with Diptrace could verify all connections it would be a big help. It's passed DRC at Freedfm.com, but a second set of eyes might catch something I missed.

The .ZIP file contains the Diptrace board design.

[ScaleRobotics]

Wow! You shouldn't have (but I love it!). Thanks again. I can't wait.

I will try to look over the board with Diptrace tonight, but doubt I will be able to find anything.

[Bruce]

No problemo. I dorked-up the 1st one with the wrong 64-pin TQFP package type anyhow, so I had to change it. The PIC32 we're using is 10x10x1 with 0.5 lead pitch. That just made it even harder to route, but did give up a teeny bit more realestate.

It needs to fit on a breadboard anyhow since it's not dev-board size, so I spent a few hours changing it around. A 2x8 female header for the analog pins should make it nicer than the 1x16.

And I went by the datasheet VS the Sparkfun schematic for the VUSB connection (pin #35) to the cap. It's not crystal-clear what they're doing with this pin. It shows ---> with 3.3V just below it. Not sure if they didn't connect this pin or if they ran it to the other 3.3V connections with --->3.3V?

It could be a LOT smaller with < 0805 components, but it wouldn't be anything I would even consider assembling by hand unless it was just 1 or 2 units. And the price per board would definitely go up, so I'm leaving it at this size.

I'm going to get in touch with Rich on a clone connector. If I have room, I'll try to squeeze that in too. Then you could connect 2 boards & clone them, but that's a big maybe. If it turns out to be a hot item, I'll make it really small, and probably even ad the wireless, but I'll have someone else assemble them. I just want 1 to mess with to see how it works with the BASIC OS for starters.

[Nicmus]

I believe that having the board wireless ready (even if the involved parts are not populated) will be a big plus.

I am also interested on USB host feature (if possible) which might solve the well known problem of depending on FTDI chips and modules as the only source of USB host capability for hobbyists. Grouping these pins and making them available to the user in a USB connector fashion arrangement or any other way will add value to the board at no expense.

In any case, I just can’t wait to try this new device.

Regards,

Nick

[Bruce]

Walter,

Use this latest version. It's been optimized for a better ground plane, and clearance.

Nick,

I may include wireless & USB host in a later version, but not this 1st prototype. But you're welcome to use the existing board file if you like to ad new options. I'm just short on time at present.

[Ioannis]

Hi Bruce. Very impressive job! Well done!

To ease things for you, maybe just assemble the smd parts on your pick and place machine, then let us assemble the rest, through hole parts, like a kit.

I believe this will save you a lot of time.

Ioannis

[ScaleRobotics]

And I went by the datasheet VS the Sparkfun schematic for the VUSB connection (pin #35) to the cap. It's not crystal-clear what they're doing with this pin. It shows ---> with 3.3V just below it. Not sure if they didn't connect this pin or if they ran it to the other 3.3V connections with --->3.3V?

It wasn't clear at all to me. I located the source files of the CUI32. Here is a view of how they really have it connected... Following the datasheet sounds like a good idea.

I looked over the obvious on your pcb, and I can't find any mistakes. Pin6 of the programming header doesn't go anywhere, but you probably knew that.

Thanks,

Walter

[Bruce]

Thanks Walter. I'm guessing they have the internal Vreg disable, so I changed it.

This is the final version unless someone notices something wrong. It has a lot of spacing changes, copper pour islands removed, and has passed DRC.

I'm attaching the BOM, Gerber files, and the final CAD file so anyone can get quotes on boards & component prices if they like, or modify the original design as needed.

I have a bunch of parts on-hand like resistors, caps, /MCLR diode, etc, and some female headers if people want to use them for the analog pins - so it will save everyone a few bucks. See the notes in the BOM for whatever I have a ton of in stock already.

For a few parts I've included part #s for Mouser & DigiKey, but everyone's welcome to lookup cheaper parts if they can find them.

Ioannis,

Thanks, but my pick & place machine is a pair of SMD tweezers.. That's why I didn't want to go with any parts smaller than 0805 since I'll be placing them all by hand.

As for the through-hole parts, I don't mind installing the 2 switches, but the headers I'm going to leave to everyone else to install themselves.

Note: A few manufacturers from China have contacted me asking to quote on PCBs and full assembly of these, so I have attached pretty much all files they'll need to do so. You can contact me directly if you're interested.

[ScaleRobotics]

Here are some possibilities for the LED's:
Digikey 160-1414-1-ND Green 805 LED $0.12 each
Mouser EL-17-21/BHC-AN1P2/3T Blue 805 LED $0.20 each

[Ioannis]

Thanks, but my pick & place machine is a pair of SMD tweezers..

Ooh, I see. I thought you had the magic machine...

In that case why not send the parts and PCB to assemble it? It will save your valuable time. I suppose all we that ordered the StickOS can do this.

You have done a lot of work on this project and thank you very much for this.

As for the Chinese, what is their MOQ?

Ioannis

[mackrackit]

Bruce,
I agree with Ioannis. The reason I sent you my four chips is because I thought you had the magic machine. I can do the assembly with my tweezers.

So if it is any help just bag the stuff for me and let me know what I owe you.

[Nicmus]

I completely agree with Ioannis and Mackrakit that Bruce had done a lot for this project so if Bruce decides to send out kits I’m supporting it 100%.

One thing I’m not too clear is if we all should have send Bruce the blank chips (I didn’t) or he will be ordering them. Also I'm not sure if Bruce can do the initial programming of the chips before installing them.

I have a couple suggestions related with this project:

By now Bruce should have a fairly good idea what the cost per kit (assembled or not) should be but he has no financial backing from us. So to avoid this financial burden on him my suggestion is to have each of the participants in the program send their financial support. I personally suggest estimating it on the plus side just to make sure that Bruce doesn’t take a hit on this aspect also.

Even if there is a dedicated forum for StickOS users I suggest opening a separate section in this forum for this new project. It will make it much easier for the members of this forum to exchange information about this particular project.

Regards,

Nick

[ScaleRobotics]

I agree with everyone as well.......... Bruce has over done his generosity.

However, it sounds like my toaster oven skill level is not up to par with everyone else. I am at the "never done it" level. Speaking of off topic, can I get some feedback on how everyone else is soldering their 64 pin devices?

I have watched a few videos on the toaster oven technique (sparkfun). I do not have a ramp controller on the toaster I purchassed to try this. I do have some solder paste, but obviously will not have a stencil. I have seen it done a few ways, from a syringe of solder paste, to a solder iron applying solder in a pre-application to "toasting".

Can anyone elaborate on an "easy" way to do this. I am leaning toward offering $ for a working device, but obviously in the minority. Would love to be a member of the majority though!

[Nicmus]

Few tips and tricks I’m using:

• Keep your soldering iron tip clean all the time (use wet sponge) and re-tin the tip periodically (nothing helps better than a clean tip)
• Use a reliable soldering iron (I’m using Metcal) that can maintain the tip temperature when touching the pads
• Use a pointy flat at the very end tip not necessarily the thinnest. This shape works for me better than pointy round tips when dealing with multiple pin devices
• Add extra flux (fresh brownish resin type) it makes the solder flow better and avoids bridges
• Use lower temperature solder (leaded one is much easier to wok with)
• Keep your soldering iron temperature few degrees higher than normal but not too high so it oxidates too quickly
• Find solder as thin as possible (I’m personally using 0.3mm which I’m not sure is readily available)
• Make sure your pads do not have extra solder on them before you install the device (use solder wick to remove the excess)
• When soldering two pins components add solder on one pad. While holding the soldering iron on the pad hold the component with the tweezers in a slight angle so the pin you working with lands first on the pad and slide it in place. Finish by soldering the other pad. When soldering multiple pins chips add solder on one pad, slide the component in position while holding your soldering iron on the pad. Make sure all the pins are aligned with the pads. Unsolder and repeat this step until you have perfect alignment. It makes next steps much easier. When done aligning solder one more pin on opposite corner of the chip.
• Apply solder only where needed not on the tip first and then to the board.
• To finish a multiple pin device work one whole side at the time not pin by pin
• Add flux (on the side of the device over the pins and pads) with a fine tip dispenser if your solder has too little or none
• Hold your soldering iron on the first pin of the line you working on (if the tip has the small flat I mentioned above hold the flat on the pin). The axis of the soldering iron should be held at about 45 - 60 degree in reference to the board. Pull few inches on solder and using the other hand to hold it try to place its end right at the point of contact between the tip and board. While gently pushing down the tip slowly start sliding it over the pins on that side. In the same time make sure you feed fresh solder at the contact point and the tip does not starve.
• I prefer to kind of over feed solder which usually sticks with the iron and does not create bridges and get rid of it by quickly shaking the iron
• If any bridges use flux and solder wick.
• For uniformity and professional look, especially if you had to fix bridges, run the soldering iron once more over the affected sides and add solder only if necessary
• Last but not least it is important to use nonmagnetic tweezers especially when dealing with smaller sizes components
• Use Head Cleaner to clean when finished soldering. Flux remover has the tendency to leave a whitish film and it is much harsher on certain plastics. For better results when cleaning the flux I use acid brushes. I cut their bristles short for more stiffness

Following these simple guidelines, even with my failing eyes, I can still replace (that includes removing a defective chip and installing a new one) 64 – 100 pin chips in about two minutes and toy with 0402 or 0201 components without any problems.

After all this is not rocket surgery .

HTH

Regards,

Nick

[mackrackit]

If you want to do the "oven" method here is a primer
http://www.picbasic.co.uk/forum/cont...MD-Prototyping
and I will add a bit...

Do not worry about the paste bridging, when heated it will flow to the metal. On these small parts the paste will bridge as it is almost impossible to apply the paste to one pin at a time. I end up "smearing" a little paste across the pads and during alignment (sliding the part around) the paste will "bridge". Not to much paste though, it is easier to add more than "wicking" it off.

My oven is unmodified. Get one with a heating element selector and use the bottom element only. Set the temperature to 350 and the timer for 10 minutes, you may need to increase the time with larger or multiple boards. Put the PCB on a thin piece of aluminum plate (1/16 inch) to even out the heat when it sits on the grill.

Put the PCB in a "cold" oven when doing the 10 minute thing, at the end of 10 minutes or just before you will see the paste flow. When you get to doing several boards a day and the oven is pre-heated you will have to work out the time per batch.

Best thing to do is just play with it some like the link to the wiki article shows.
Like Nick said "After all this is not rocket surgery ."

[ScaleRobotics]

Like Nick said "After all this is not rocket surgery ."

Maybe not, but perhaps a little bit of brain science is involved.

Thanks guys, I will give it a go. Appreciate the tips from you two.

Walter

[mark_s]

I also agree, special Thanks to Bruce!

I have had very good results with a Walmart electric skillet and a IR thermometer from Harbor freight. I preheat the boards to around 180 F. Then turn it up to full heat, once the solder paste has flowed you turn it off and let cool. Its like magic when it melts!
The idea is descibed in a Sparkfun tutorial

Dave's advice on less solder paste is really good, which I have learned the hard way.

Regards
Mark

[ScaleRobotics]

Rich has added a few neat features in v1.80, like I2C. Check out 2.2 in http://www.cpustick.com/stickos.htm (and all other improvements in red).

[rmteo]

Any news on this?

[ScaleRobotics]

I tried to upload Bruce's gerber files (from post# 67) to Sparkfun's BatchPCB site, but I received a message saying " Unfortunately, your design "Bruce__s UBW32" violates our design rules of 8 mil traces with 8 mil trace spacing."


I am getting a quote on 30 boards from www.pcb-pool.com for the design v2.6 located here: http://www.schmalzhaus.com/UBW32/doc...easeNotes.html



It looks to be less than $5.00 for each board

Will let you know more when I get an official quote.

Walter

[Bruce]

Hi Walter,

I sent you some updated Gerbers, but I don't think you received the email so I'm posting them here. Just in case someone else wants to try making a prototype.

Sorry I haven't had time to move on this. We sold our big oven before the move, and our desktop unit went ker-flop shortly afterwards, so I'm not doing much fine pitch assembly.

I increased the spacing, trace widths, etc to 8 mil. Note I haven't made a prototype of this yet, but anyone that wants to is welcome to the files. Guess I better start returning a few chips too. Sorry. If it ain't one thing it's another.