nlian
DCP_4751
PCB for cc1100 wireless tranceiver (reference design). I needed to prototype a few wireless sensor platforms (aka smartdust, motes) using the brand spanking new cc1100 tranceiver from chipcon/ti. This chip only comes in a qfn package, with a bottom pad (the primary ground) that MUST be soldered to the PCB. There was no way I could do this middle pad by hand, so I tried the griddle reflow method. Passive/supporting components are 0402 size (0.5mm by 1mm). These photos document the PCB fabrication using inkjet transparencies and a test solder reflow using a griddle (thanks spark fun). I'm astounded, tiny prototypes made entirely in my kitchen!
About the PCB making process: the PCBs were make using the photo process with photo-positive pre-coated boards. Artwork was printed on an inkjet transparency using all black ink. The printer was set to its maximum resolution (something like a fake 2400 dpi). Two copies of the art were taped together to give a thicker mask. I think that I could do it in a single layer if I printed using all colors for black (called high quality black on my printer), or using photographic cartridges with uv resistant ink. As of yet, I am to cheap to buy a new cartridge. I'm thinking of asking the inkjet refill shop if they would fill an empty black cartridge with a UV resistant color ink.
About the reflow: I used the sparkfun griddle reflow method. The solder paste was a $5 tube of "silver solder paste" from mouser. I made a small pool of paste on a note card. I dipped the syringe tip in the solder and dotted the pads on the board. I cranked the griddle up (240c) and in five minutes the parts were soldered and aligned.
About the cc1100: the cc1100 is a low power tranceiver for wireless networks. It is a big upgrade from the cc1000 used in many previous generations of MOTES (mica2, mica2dot, etc). I had a really hard time actually getting the cc1100 - ti won't sample and distributors all list them on back order if you don't want a whole reel. The boards shown here will use a draft (E) revision of the silicon. That was the only thing I could find anywhere in the world. I hope to get my hands on the final silicon soon (hint, hint TI).
DCP_4751
PCB for cc1100 wireless tranceiver (reference design). I needed to prototype a few wireless sensor platforms (aka smartdust, motes) using the brand spanking new cc1100 tranceiver from chipcon/ti. This chip only comes in a qfn package, with a bottom pad (the primary ground) that MUST be soldered to the PCB. There was no way I could do this middle pad by hand, so I tried the griddle reflow method. Passive/supporting components are 0402 size (0.5mm by 1mm). These photos document the PCB fabrication using inkjet transparencies and a test solder reflow using a griddle (thanks spark fun). I'm astounded, tiny prototypes made entirely in my kitchen!
About the PCB making process: the PCBs were make using the photo process with photo-positive pre-coated boards. Artwork was printed on an inkjet transparency using all black ink. The printer was set to its maximum resolution (something like a fake 2400 dpi). Two copies of the art were taped together to give a thicker mask. I think that I could do it in a single layer if I printed using all colors for black (called high quality black on my printer), or using photographic cartridges with uv resistant ink. As of yet, I am to cheap to buy a new cartridge. I'm thinking of asking the inkjet refill shop if they would fill an empty black cartridge with a UV resistant color ink.
About the reflow: I used the sparkfun griddle reflow method. The solder paste was a $5 tube of "silver solder paste" from mouser. I made a small pool of paste on a note card. I dipped the syringe tip in the solder and dotted the pads on the board. I cranked the griddle up (240c) and in five minutes the parts were soldered and aligned.
About the cc1100: the cc1100 is a low power tranceiver for wireless networks. It is a big upgrade from the cc1000 used in many previous generations of MOTES (mica2, mica2dot, etc). I had a really hard time actually getting the cc1100 - ti won't sample and distributors all list them on back order if you don't want a whole reel. The boards shown here will use a draft (E) revision of the silicon. That was the only thing I could find anywhere in the world. I hope to get my hands on the final silicon soon (hint, hint TI).