Page 1 of 3 123 LastLast
Results 1 to 10 of 23

Thread: Dirty DNA

  1. #1

    Dirty DNA

    Sadly it doesn't look like I will ever be able to buy a DNA board. A while back I contacted Curt to see if he would entertain the thought of sharing the board layout so I could try my hand a making my own. I was actually surprised that I got a response but it contained words such as "..commercial venture.." "..build it for cheap.." "..undercut our market..". Fair enough.

    With that door closed I decided that this time was probably as good as any to learn how to use Cadsoft Eagle. I've always had an interest in electronics even thought I am a Mechanical Engineer by trade. I reckoned that Eagle couldn't be much different to the CAD/CAM packages I am familiar with.

    Cadsoft offers a freeware version of Eagle for boards under 3"x4". This is plenty of room for most hobbyist. There is a bit of a learning curve to the software but there are lots of online tutorials as well as the invaluable YouTube resource.

    I gathered together all of the info pertaining to the DNA venture and set to work on the schematic for the main board. Curt had posted a "napkin engineering" version of it that got me started. I listed out all the components that were mentioned in the blog and gathered all of their data sheets. The puzzle was starting to come together but I was still missing sections

    The saving grace was the pictures of the board layout. This held the key to the parts I was missing. I separated the layer colors into individual images and was able to complete the sections of the schematic that were missing.



    With the schematic sorted, I moved on to board layout. Curt had obviously spent a good amount of time laying out the board as efficiently as he could. There are some neat tricks and what I consider strokes of genius that he used. Again the pictures of the layout were vital in reproducing the board. I imported the images into Eagle's layout editor and adjusted the scale and position until the top 2mm header and bottom .1" header matched exactly. From there on it was a downhill battle as I only had to "color within the lines" with the trace routing. The only issue I ran into really was how Curt filled and enlarged main traces on the board. It seem he used the "Rectangle" tool to block in larger traces and fill areas. This works but is a really hack way to go about it. Eagle's "Design Rule Check" errors out when you have over lapping sections. As a newbie Eagle user I didn't have the skill set at the time to use the polygon tool too correctly to fill the areas. Instead I stuck to copying the board as closely as possible. My final layout was roughly 98% accurate to the real thing. There are a couple differences in the silk screen but the traces and pad layouts are pretty much identical.



    With the DNA board layout finished it was time to get a prototype made. OSH Park has a sweet PCB prototyping service which offers 3 copies of your board @ $5 per square inch. I didn't fully realize the low cost of the DNA until I uploaded my Eagle file.



    I couldn't believe it. $1.70 for 3 boards shipped to my door?!? I just spent hours and hours copying a board that costs $.57 each. It blew my mind.

    Next in line, was reverse engineering the button and OLED board. With my Eagle skills improving it wasn't long before I had a completed reproduction of the button board. Again I couldn't get over how cheap it was to get these boards manufactured. $3.50 for 3.



    I had most of the OLED board layout finished when the first order from OSHPark arrived. Yup, these are small boards.





    Curt mentioned in his email that he doubted I could assemble a DNA board. I've soldered some pretty small stuff by hand. 0603 components and Attiny13's in QFN packages. Do I intend on hand soldering A DNA board? Not really as I can make my own solder paste stencil and reflow it with either a hot air system or modified toaster oven. I do plan on trying to hand solder one just to prove I can do it.

    I finished up the OLED board and put together a parts order from Mouser. Even in small quantities the cost of the components for the main board are only $10. As you can see in the picture below I finally figured out how to “tent” the via’s on the board with solder mask. This should help when assembling the boards. The OLED board being the biggest of the lot came in at $5.15 per batch of three.



    This brings us up to right about the time that Curt reemerged on the forum and it looked like the project was moving forward again. Excellent! Maybe I won't have to attempt to load the DNA alpha code release. It would be great if I could load a polished version of the DNA software.

    Unfortunately that thought was short lived. Curt is once again MIA. Guess I'm going to have put on a programing hat and start learning about that...

    So to where does that leave us now? Can we call the DNA project officially dead as a commercial venture or do Curt and Simon intend on releasing a board to the general public in the near future (by the new year)? Inquiring minds need to know as I would have no problem releasing all the work I have done reverse (dirty) engineering the project. In my opinion this is the only way this project will ever move forward in any capacity.

    The DNA project was presented as an open source project but seems to have transitioned to a commercial venture. The development utilized the open source model and used the community for its collective knowledge. Unfortunately this is where it stopped. The board layout was only shared as images and software has not been updated since the initial release. I feel the project has failed because the community was not allowed to actively participate. There definitely is community interest. Sadly we are relegated to the side lines, offering advice but only allowed to watch painfully as a very potential project dies because the founder loses interest. I'm sure we all have projects that have fallen by the way side. I just hope that the Curt is willing to hand the torch off as there are many of us that would like to see this board come to fruition.

  2. #2
    Insider
    Join Date
    May 2012
    Posts
    1,182
    Really impressive stuff, thanks for sharing.

  3. #3
    Wow. That is some work. Much respect for the undertaking.

  4. #4
    Junior Member
    Join Date
    Apr 2013
    Posts
    28
    Uh, wow. You were totally the guy who, when dared, said "Just watch me" weren't you?

    I'd love to see Curt finish his project. I'd love to see you finish this one.

    Scenario Dreams (produces of the Universal T-Board for paintball markers) also has graciously shared a bunch of their programming, wiring diagrams, etc. here:
    http://www.scenariodreams.com/produc...iversaldev.htm

    I don't know if that helps you, but I thought I'd post it just in case.

  5. #5
    Thanks for the kind words.

    I've learned a lot about micro controllers, PCB's and electronics in the last couple months. Here are some of my thoughts on the DNA board.

    I feel I have a pretty good understanding of the DNA board design now. There some things I really like. It's small, dual solenoid compatible, has expansion and versatility options. The board layout has some neat layout ideas. On the other hand there are a couple things that I would personally do differently.

    It's small, too small? Seems like it was designed with the smallest form factor as a high priority. I agree compact size is important but anywhere this board is going (in the paintball world) will have .5" width minimum. (Grip frames, trays, battery tubes) It uses 0402 components but has large (in relation) expansion headers thrown on the bottom that negate the space savings.

    Mounting screw. Mounting locations are important but the offset location on the DNA seems impracticable. I would rather a tiny daughter board solders onto the bottom of one the pin headers with a mounting hole in it. This way the end user can adjust for most practical application in projects.


    2mm right angle top header is bulky. Consider using an edge connector as dual sided board is conducive to this.

    Onboard led. It's LARGE (0805) and in my option on the wrong side of the board. I would prefer to see an led with remote mounting locations or soldering on to a user selectable side. Possible a solder bridge for onboard led or pinout for a remote LED.

    ATtiny84 while a good chip is right on the edge of been impracticable. It seems like it was selected early in the projects and as the scope of the project grew (OLED etc) it was not reevaluated. I/O is maxed out and additional features have to use work arounds and additional chips to function. It lacks native USB and expansion is flakey at best. Consider moving to the ATtiny167 found in the new Digispark Pro. More I/O as well as on board usb/I2C and a larger user base to pull support from. I should have one of these next week from the kick starter.



    Attiny84 seem to be the in-between chip. Either people use the smaller Attiny85 or jump to the ATmega328 chipset. The more I think about it, the more I feel this might be the perfect board.




    It has amazing specs. 32 bit ARM Cortex-M4 72 MHz CPU, 256K Flash Memory, 64K RAM, 2K EEPROM. $17 shipped from OSHPark.

    With the current DNA/Switch/OLED there are 3 micro controllers and 2 memory chips. Seems excessive when a single chip can handle all this.

    USB connection is nice but with the advent of wifi/ble is it needed? Wifi chips are now the size of usb connectors (ESP8266).


    Consider Lipo batteries and an on-board charger/boost converter

    Mosfet - it is a beast, the data sheet says it can switch a 9 amp sustained load. While this is great, what in paintball has that kind of load? Emags? That's a pretty small demographic. Could be handled by an additional daughter board. Other issues I see are that the trace width is 1mm to and from the mofset. At a standard 1oz/foot we get can only safely move 2.5 amps. The 2mm headers at the top is only rated for 3 amps. Seems like an overkill. I suggest considering using a dual mosfet rated for 2-3amps (ZXMD63N03X) . This will also allow for a smaller footprint.

    The problem is that I keep coming back to the same issue with the DNA/Attiny84 board. It uses dated hardware to reinvent the wheel. Standalone micro controllers are plentiful and cheap now. $3 will get you one with twice the spec's of the ATtiny84 and it is in a very usable package. An additional $14 will get you a kick ass board (Teensy 3.1). I feel this project needs to move forward but to do so we need to start a simpler approach. A "Paintball shield" board. A tinkering board that has a header for the normal eye/trigger/power/solenoid/etc connections and has a suitable dual mosfet chip onboard as well as button connections and display expansion options. This "shield" would then plug into/backpack on a widely available micro controller.

  6. #6
    The one thing i can contribute, as i have followed this is about the mosfet. The reason for that is the nonpaintball uses. This was spurred on as a true universal board that doesn't have to be used in paintball. I do know that Doc Nickle had wanted a board to run a machine of his that was totally programmable to do what he wanted. And if i believe, the idea of having a universal board that wasn't paintball specific lends itself to be in the grey area, so as to not pay any possible royalities to anyone. Since the board never had programming when it shipped, it was a total user/free space(i know the concept but the name eludes me).

  7. #7
    Oh, that makes sense. Thanks for that background info. I actually used this mosfet for a backpack/shield for a Digispark clone.




    It's in my next shipment from OSHPark.


    I plan on using it to control the under-counter/over-cabinet lighting in my kitchen. That's the nice thing about the arduino environment, it's simple to use and program for projects like that.

  8. #8
    Junior Member
    Join Date
    Apr 2013
    Posts
    28
    Agreed that the board was supposed to be a universal board to skirt the pb patents, wider audience, greater revenue, etc.

    While Emag/Automag users are a minority nowadays, they are a majority (along with cocker users) when it comes to dumping money on upgrades, enhancements, repairs, etc. Not catering to them eliminates a large part of your marketplace.

  9. #9
    Insider PBSteve's Avatar
    Join Date
    May 2012
    Location
    Los Angeles
    Posts
    3,084
    The other reason to use a beefy mosfet is to keep the drain-source resistance low.

  10. #10
    Insider
    Join Date
    May 2012
    Posts
    1,182
    You know, teensy + your shield + Gabriel's Gcode is like 8/10 of the way to DNA anyway.

    "...there might yet be a heaven, but it isn’t going to be perfect. And we’re going to have to build it ourselves."

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •