Hi, sorry I haven’t been to communicative in the past month. This is no way to get people interested in contributing, but I have been working on several things and nothing really to show yet from the efforts.
I have selected different chips sort of. The freescale chip is still quite good for wired USB peripheral lighting design. However, ST Micro has a new line of STM32F4 chips called the 401C that has a minimal amount of features and a tiny UQFN48 footprint. I have gotten some samples from the local rep and the 401C’s discovery board. This week I ported Espruino partially over to it (it only talks over serial, not USB yet). I will probably do some more work to get Espruino building against the newer ST Micro SDK as well as figure out what setting is wrong for USB detection.
I have been working for a few weeks now on a new design of Social Light. I have been struggling to shrink board size and cost. The idea of a big power brick and what was essentially a larger version of the wireless board but with a difficult power design and additional certification costs was a concern.
The new concept moves charging out of the board and so will require USB based chargers as its source of USB power. The board will utilize a watch battery when independent and will run off of a 5V regulator when a USB charger is plugged in. To power peripheral lights, the board will pass through the 5V line and power the 3.3V serial lines through its regulator. This bends USB standards, but it should work I think. The design will likely be 2 micro B or 1 micro B and 1 mini B USB depending on whether the design is specific in input direction or not. There are a few tiny high speed mux switches sold specifically for USB selection with ESD protection.
I finally put all the schematics for Social Light together and the pin count is just north of the 300pin limit on Diptrace. The design will need to be at least a 4 layer board, so I have no EDA solution that is free. Diptrace Light is sort of cheap, so I went ahead and bought for myself. Keeping a separate copy of each schematic will help, but a license will be required for any PCB layouts. There are still a few design decisions left such as USB ESD and mux as hinted above as well as whether to place a 1.8V regulator and if possible to drive the STM32F4 at different voltages if USB is not connected. I assume that the 1.8V regulator will be figured out in a much later revision.
I worked on several board designs trying to test the 32QFN version of the TLC5941. However my trace widths were too small for negative etching and I spent some effort repairing the traces. Along with trace issues, the QFN format requires a well developed solder mask, so it just isn’t possible to do a home etching. I have some dip samples from TI and will make an LED only breakout board. I’ll also send some board variations out to OSHPark and OSHStencils.
I have learned more about hardware/product development. I attended a hardware workshop last week and put a small review up on my blog
http://blog.ambiguous.net . Cindy and I have also gone to some local hacker meetups and met some local people that might be interested in helping. I have a local mechanical engineer and a couple of mobile/javascript developers that might be joining Fyber Labs in some capacity.
I have finished what is hopefully enough documentation for the massager provisional patent. I have been concerned about discussing it further openly without securing IP, so hopefully now I can start showing off prototypes and discuss it more openly.
I’ll also be working on the
fyberlabs.github.io site to update development information and maybe finally make a fyberlabs company site. We still need to make decent design drawings/concept art to explain the products better. The few drawings I have are no longer ‘correct’.
If anyone has time and interest with any of these efforts I would love some help.
Thank you for you patience, but I promise things are still moving forward.