Another important aspect of my project is to create a clever strategy for connecting the electronics in paper to FTDI and USB port. For that I spent some time designing a connector following Neil's recommendatioin. This connector uses the Molex flat cable connector and by calculating the size of traces and pitch I can insert the paper directly into the molex.I experimented with WM1387TR-ND, WM10938TR-ND, WM10994CT-ND, WM4880TR-ND connectors. I settled on WM4880TR-ND connector (pictured above).
I first inspeted the health of the traces by checking if the printed electrodes were sintering correctly in the paper substrate. The two images below show the sintered contacts.
The hardest challenge was to get the correct alignment on the pitch of the connector and the contact pads on the Molex connector. I resolved that by trial and error and basing myself on the measurements provided in the Molex connector datasheet.
This week I spent some time working on my final project inpyut devices and circuit. For updates on that refer to the project page. To keep working on my technique of inkjet printing circuits I also attemped to modify one of Neil's design for the hal sensor to make it compatible with printed connectors.
I choose the HAL Sesnor board as it has a relatively small number of components and the footprint of the HAL sensor was friendly to the inkjet printing technique. Taking Neil's board design I simply added an adapted interfacce to the FTDI so I could directly conenect the board to Molex connector.
I also made a mask (parts in red) for where I would apply conductive epoxy. Note that the 6-pin header is blocked from the mask and the reason is that I leraned that I could directly program it.
My strategy was to extend the electrodes at the end of the circuit and by aligning the trace width and pitch size with the one of the connetor to make them match. So I took Neil's design and in Photoshop made the necessary changes to make it compatbile with the connector.
I also took the opportunity to experimeent with a different substrate a Mitsubishi Coated PET for circuit printing. The process was somehow simialr to the ordinary paper substrate but in this version I used a paper sheet to protect the substrate from handling.
I then laser cut the mask with the same settings as before and applied to the PET substrate using tape. I found an issue that if the katpon mask is not 100% flat it buldges when the conductive epoxy is applied and in that way the pads are smudged which isn't great. I had to repeat the process 3-4 times until I could get a good mask.
I also develped a technique for applying the conductive epoxy. Instead of using tape to fix the mask to the substrate, with one hand I hold the mask in place. With the other hand I apply the paste. In that way I can do a quick release of the mask without worrying about removing tape from the substrate or smudging the the conductive epoxy.
Another learning is that you want to have enough paste for the leg of the component to 'sink' int he paste. If the leg is just touching the conductive paste at the bottom this makes for a bad connection. So when applying the mask make sure you have enough paste that the leg of the component can be fully immersed.
Anothr learning was on how to carefully place the components. For that I developed a technique with three steps:
(1) using one pair of tweezers land the component on top of the epoxy applied to the pad.
(2) With a second pair of tweezers hold the piece in place.
(3) than release the first pair of tweezers. This avoid that any shaking whilst removing the first pair moves the component in the pad.
During the development process I ran out of conductive ink in the printer, so I couldn't do a second run of printing. In inspecting my tracesI found that the ink had interruptions on ths surface, so in an effort to still save the board I had printed I connected the FTDI pads using the copper tape and soldered them to the FTDI cconnector. I am leaving the piece for the epoxy to cure overnight and will try to program it tomorrow.