Saturday, December 31, 2011

PSoC3 Custom Board - What next?

While development on this is underway,I've been thinking of the future of this project.
The initial goal I set out with,was to provide easy access to the PSoC3 platform to students and hobbyists,at a low price.
Though this sounds good,I'm not so sure now.There are a number of issues to do with the design(?),fabrication and parts sourcing that I'm concerned about.

The Design:
-I've missed 1uF storage caps at 2 places,but I'm hoping that isnt a deal breaker.
-Theres going to be a change in the USB connector,towards a miniusb type instead of the current one..(for sure in the future revisions)
-I guess this list shall extend as I test this further..

Fabrication:
-The current deal I have is just too costly.The way ahead on this front would mostly be engaging the services of either SeeedStudio's Fusion Service or ITead's OpenPCB,both giving me 10 5cmX5cm boards for a total of around USD 14(incl. shipping),excluding the import duty I might have to pay.
-Theres another option here,the SeeedStudio Propagate service,which is board fabrication+assembly,with a minimum order of 100 boards.

Parts Sourcing:
-Currently,I'm using chips I got as free samples from Cypress.The price listed on the chip's page is almost USD 20,with an MOQ of 60,which screws up everything to do with "low cost".
-Other parts(SMD passives mainly)can be taken care of by ordering in larger quantities (MOQ is 100 on most smd parts at locally),and again not a problem if I'm using the Propagate service.

Now the real question and answer to the above doubts and concerns is,"Does anyone even want to buy this?".Since,if there is a market out there for these,then things move to a different scale and the above issues can be resolved.
If not,then the project ends as a very enriching academic pursuit.

13 comments:

  1. Try talking to the people at SparkFun. They only have 1 PSoC board now, so they might like another.

    My understanding is they do all the work to manufacture and sell the boards. Then pay you a royalty.

    ReplyDelete
  2. Also I've made no attempt to see if it's feasible, but using the Arduino header footprint is appealing in that it makes the wealth of arduino shields available for PSoC.

    The combination of PSoC's flexibility with using exiting shields for higher level i/o (like ethernet or various radios) seems like a real winner.

    ReplyDelete
  3. hey davidk,thanks for the interest in my work! :-)
    Well,yes I did plan to contact the guys at sparkfun initially,but shied away from it.

    PSoC on an Arduino Layout is a great idea,but I've exhausted almost all of my hobby budget on this one,and I'm out of chips too.That said,designing it costs nothing :-)

    ReplyDelete
  4. What was your reluctance to approach SparkFun?

    I could probably get you some more chips ... cypress in the US used to be pretty good about giving chips to developers.

    ReplyDelete
  5. Well,I didnt they would look at my design,since theres a large startup cost involved with the PSoC3,since the chip itself is ~$15-20 and you need a $100 programmer just to get going.

    I'll try and sample a few more chips,yes.If I do go into this properly,I'll probably scrape together some funds and use Seeedstudio's relatively cheap PCB fab service.

    ReplyDelete
  6. I'm not sure those startup costs would be considered "large" to a company like SparkFun. They are in the business of making products and commercializing other people's designs, if they can sell a reasonable volume the startup costs don't matter much.

    For them I think think bigger issue would be a lack of tutorial material ... after all, what good is the board if no one knows how to use it. Again, the driver here is for sparkfun to be able build a batch of 100 boards and know they will sell in reasonable time.

    I only glanced through your blog, did you have to use the miniprog3 or did you include a USB bootloader on the board.

    If a bootloader is included, then the end purchaser doesn't need to buy a miniprog3, which is generally a plus. A company like sparkfun will have their own programmers and can probably program the initial firmware directly through JTAG or the like.

    Also, if you can hit a price of <$40USD should be 'low cost' enough for sparkfun.

    ReplyDelete
  7. Yes,the design does involve a USB Bootloader,and the MiniProg3 only to burn it in once.

    I'll work on the PSoC-Arduino compatible layout,and probably submit that to the people at SparkFun.

    ReplyDelete
  8. Personally I think a PSoC-Arduino would be pretty cool. I haven't tried the PSoC Create v2, so I'm up on what all it does [I only vaguely recall trying out v1 a few years ago].

    If you wrote a blog entry or two on using the PSoC with an Arduino shield that would be quite impressive (ex. Danger Shield)

    ReplyDelete
  9. Another idea for you is to go to http://www.kickstarter.com/ and raise some money for a basic PSoC3 board. Think of it as a way to take orders in advance.

    ReplyDelete
  10. well yeah,but then things like fabrication and buying parts etc and supplying tested,working boards to a large audience is not something I think I could give time to,with my current academic workload..But maybe I'm just being a skeptic again.

    ReplyDelete
  11. That said,I have started some work towards a PSoC-Arduino board,and do plan to talk to the guys at Sparkfun when I'm done with the design.

    ReplyDelete
  12. No worries, just pointing out some of the resources available so you can:

    a) Not have to pay for everything yourself.

    b) Get some experience with 'customer' or 'user' interactions. Dealing with real end-users and making decisions among their desires and requirements is one of the most important soft-skills that isn't really taught in school [well ... lets say wasn't taught when I was in school].

    I'm quite interested in what your efforts. I'v subscribed to the comments on this blog entry, so please add a comment when you have news.

    PS. You've inspired me to go back and look at PSoC creator more. I have a PSoC5 dev kit and lots of PSoC1 boards. Looks like only 3 & 5 are supported by PSoC Creator.

    ReplyDelete