I've been kind of wondering where DP is headed lately and Ian's post about his adventures in China have given us some clue. I thought I would reflect a bit on what this community is to me, and I suspect that everyone sees it a bit differently.
For me I really enjoy when we all get to participate in the process of developing some new widget. I like to watch things go from idea, to schematic, to PCB layout to soldering of the first prototype. The ultimate blog post is a video of Ian soldering the first board of something new. I really enjoy those. I come here several times a day looking to see if Matts or someone has posted their latest widget. To me that stuff is gold. Posts about some tangle of wires on a breadboard don't get me that excited, I can go to hack a day for that :)
I think this community has taught us that PCB design can be easy, affordable, and a lot of fun and that surface mount parts make that cheaper and easier. You just need a good soldering iron, good magnification and LOTS of FLUX.
Unlike adafruit or sparkfun where their business is built around selling breakout board, breadboards and the all important bundle of wires, DP is more focused on design and prototyping through PCBs. I can see how Ian is struggling a little with monetizing this community.
Monetizing the PCB based prototyping, seems to be the questions at hand.
If you don't design it, or manufacture it, how do you make money off it? Ian is focused on the tools part of this, that's probably a good idea, but does that move away a bit from the core mission of Dangerous Prototypes? Perhaps, sourcing tools reflow ovens and such is a low overhead, high margin business that can fund other projects and keep DP designing widgets.
Is there a way for DP to add value to the prototype PCB business. One thing that comes to mind is leveraging the Sick Of Beige format. The 60x37 format is a really good size, but is doesn't fit in the 50x50 cheap board size, why not have a DP offer of 60x37 boards (they are smaller than 50x50)! Maybe the order of 10 boards could come with 1 SOB case included.
Another problem to be solved is what to do with the other 9 boards :) Some folks here seems to be trading them, or selling them, but I keep wondering how Ian might broker the trade of boards and pickup a few bucks in the process (without actually having to touch the boards).
Like a lot of businesses today, its all about the experience. DP should make the whole experience of prototyping with PCBs easy. 1. ideation through the community and DP's own open projects 2. design using parts libraries that are know to match DP part kits 3. physical layout using standard SOB shapes and guidelines 4. manufacture with partners who cater to SOB standards and price accordingly 5. source parts from standard parts kits 6. build with DP sourced tools and FLUX (BTW, I would probably buy the Soldering Iron Board, if it came with a handle, so I didn't have to figure out which ones on ebay were any good) 7. debug with DP designed tools, BP, Logic Sniffer, ATX breakout, smoke tester, etc..... 8. show off to all of us on the forum 9. unload the other 9 boards on the swap board.
I think Ian can take a little money on lots of those steps. I realize I am probably just stating what Ian has already figured out, but I'm curious if other see things very differently and see DP very differently?
People have been showing their lipo chargers lately, so I thought I would show mine. Actually mine is designed to charge LiFeP04 batteries, LiPO's safer cousins, and was built from the parts on hand using a pic12510 to implement a rudimentary pulse charger. it implements a constant current/constant voltage charge profile by measuring the battery voltage and charge current and pulsing a pass FET to limit voltage/current. It was built on a home-etched board with a built in USB connector and a paper "silk screen"
LiFePo4 charge to 3.6 volts rather than 4.2 as for most LiPo's, so I added a switch to choose the voltage, but truth to be told I don't use it for LiPo's because I don't trust it. I only use it to charge LiFePO4 batteries from a garden light and I use the battery in various battery powered projects.
I confess, I only have crappy meters. But I often have the need to calibrate ADC measurement, and I would like to calibrate my meters so they are all in the same ballpark. So I have built myself a simple, cheap, and reasonably accurate reference. It is based on a ADR510 voltage reference. According to the specs it is 1V +/- 3.5mV, 70ppm/C.
And I had some boards made from OSHpark [attachment=0] [attachment=1]
Checking it with my crappy meters, it appears to work.
(edit: I replaced images of board with renders due to embarrassing spelling errors)
I am thinking about implementing an microcontroller based I2C slave and I was wondering about the i2c slave address? Are there concerns similar to USB vendor ids? How does one get an official address? And if I don't want to pay for that are there accepted work-arounds?
I wanted to make something tonight, so I threw together a little something. It is a solar garden light turned power supply. It runs off a single AAA and boosts to 5v with a NCP1402 breakout board. Everything was from stuff I had laying around. I broke all the rules -- I soldered directly to the battery and I'm recharging an alkaline. I'm surely going to be sent to prison for this. The picture shows it charging a little bluetooth keyboard. It won't charge my phone, the thing only produced 200mA or something like that. I'll keep it in a sunny window. It might be handy occasionally.
I'm starting a little project. I'm calling it Donatello (because project need stupid names :)
My intention is to make a combination part ninja and super probe. The idea is a USB device and a couple resistors to to measure transistor, caps, etc., generate signals, count frequency, etc, etc, etc. whatever I can jam into the code. My intention is that it will be a HID device and a PC program to display the results and post process raw data, so no display on the device and only raw data collection on the device. I have designed it based on the basic part tester design that has been discussed here, with one small change. I have used a 2x3 header for the probe to bring out the 3 pins for part testing plus power and ground to allow me to create "active probes", like thermocouple amplifier for example.
One thing I am not so sure about is the plethora of resistor values that the superprobe uses, My choices will be much more limited I'm not sure if that is going to cause trouble. I haven't really done all my homework on how the superprobe works.
I have designed it to be really, really, cheap. I made a 1x1.6in golden rectangle board in the SOB form, which I should be able to get for $8 for 3 from OSH park, a pic18f14k50 ($2.10 from avnet) so it should come in at around $6-$8 each and maybe I'll spring for an SOB custom made.
I've attached the files here and I would like some feedback, from anyone willing to provide it. I expect that people will suggest adding a display, but I don't want to put one on. I would rather save the code space for more features and I intend to use it with a laptop as the power anyway so why not have a full feature UI on the PC without and display or buttons on the device.
Thanks in advance for your feedback
PS about the name -- think geek ninja and the color purple.
I finally got around to finishing my ATX breakout free build. After a long wait from parts from ebay and a long wait on the back of my work bench I finally finished it.
It wouldn't be a free build if something wasn't bodged together. In my case it was the ATX connector. I got all the parts from ebay -- I couldn't find a right angle ATX connector on ebay, so I got a vertical one. I thought it would fit, but it didn't. So I ended up making it a right angle by bending the bottom row of pins and extending the top row of pins with 18ga finishing nails! It actually worked better than I expected. Also the push button I got from ebay was smaller than the footprint, but I was able to make it fit. The downside of buying from ebay.
You will also notice that the -12v LED is kinda cock-eyed. That's because I had to take it off and flip it around, I didn't really think about the fact that it was -12v when I plopped it down, so it was backwards. A trap for young players as Dave Jones would say.
Thanks so much DP, it's a great little tool and building free PCBs is so much fun!
Now that I have this, and all these power connectors sitting there. I am thinking about building a sister board which is a programmable supply using one of the other connectors. Maybe I'll start a design here.
I have been thinking lately that the BP would be a lot more useful for me if it were an HID device. I typically work on a computer that I can't install drivers on (long story), which is why I am so interested in HID. Clearly not an option with V3, but should be possible with V4. I was thinking that my generic HID code should work. We would also need to work on a client app for the PC, but there are several out there that should be easy to convert.
I don't think that I am the only one who might benefit from this, it would seem that plugging in a BP to any machine and pop up an app from your flash drive could be useful to many people.
I don't have a V4 to try on, but do people see a reason that won't work?
I am starting a small project to see if I can add support to the pickit2 for the pic16f1507 part. I bought a few of them because they were cheap and they have programmable logic cells which I thought might be fun to play with. I looked at them long enough to decide to buy them, but not long enough to realize that the pickit2 does not support them. I am hoping that I can hack in some support for them, with a little luck maybe all I have to to is add an entry to the device file.
The chips I bought were SSOP package, I don't know if that was the only package available or if I was blinded by price, but I didn't even know if I could make a board for an SSOP package. So this is my first task. I'm making a simple breakout board with an ISCP header and the rest of the pins brought out to pins. I make boards with the toner transfer method and I have often had problems with find pitch packages.
My first few attempts didn't transfer well, I tried all the usual things, glossy paper, transparencies etc. and all I got was a big black blob where the chip was supposed to be. So I went in the opposite direction a tried some blank newsprint that I have. The image transferred very well, it took high heat and a very long soak to make it work but the image was very good. However the board didn't come out too well, all of the SSOP pins were shorted. Here is a picture of it.
So I modified the package definition in eagle to make the pads smaller so there would be more space between them. I was hoping that the toner would spread out a bit to make the pad about the right size. It worked pretty well. Here is the results:
this board was much better, there were two minor shorts, but they were easily solved with a razor. I went ahead and soldered the pic onto the board and a ICSP header, soldering went pretty well, much better than I expected. Here is a pic:
My last thing for tonight was to plug it into the pickit2 to see what happened. All that happened was that the pickit2 recognized it as an unsupported device with an ID= 2D00 -- which is the correct number, so I think all the basics are in place.
My secrets to success so far were: 1. stay away for glossy paper, the traces have a tendency to spread out too much when ironed onto the board. 2. used newsprint 3. be patient for a good long soak to get the news print off. I put a little vinegar in the water to help break down the paper --- maybe it helped 4.lots and lots and lots of flux to drag solder the SSOP package.
All the parts came in today, so I had to build it right a way!
A few points: Ordering: 1. The mouser cart was great --- I swapped the header pins for a shrouded header 2. The FTDI chip was out of stock, so I got mine from ebay
Building: 1. I had lots of trouble with the FTDI chip. Solder bridges that wouldn't die and some problems with a poor solder joint. This took me a bit of time to work out and I made quite a mess of it! 2. Put the VR3 in before C10, your life will be much happier. it is hard to do the other way around 3. I had issues with the resistor networks. I don't know why, but I had some a cold solder joint or something. It was causing the self test to fail with wrong voltages, even though the voltages measured fine with a meter. re-soldering the RNs seemed to resolve the issue.