Actions
Forum Thread Details
  • Replies 100 replies
  • Subscribers 2563 subscribers
  • Views 8343 views
  • Users 0 members are here
Related

What's Stopping You Building Your Next Project with an FPGA? (Please, Don't Blame the Cost!)

rscasny

The community does a ton of projects with MCUs. If they do the job for you, that's great. But the element14 community is about learning, experimenting and roadtesting. In this spirit, I am proposing that you should try building your next project with an FPGA (SoC), if it suits the application, of course. What's stopping you? Oh, some members have voiced the cost issue. Granted, some FPGAs do cost a lot. But chip manufacturers are rolling out economical chipsets that most makers, pro-makers or experienced hobbyists can afford. (If you can't, then apply to a RoadTest and if you win you can get a dev board for FREE.) I think FPGA / SoCs will move into a more important place for electronic designers, especially for IoT applications. Perhaps it's time to experiement with one. What's stopping you?

 

Here's a link to a current roadtest:Digilent ARTY S7 Dev Board (Xilinx Spartan 7)   Apply today!

 

Randall Scasny

RoadTest Program Manager

  • in reply to rsc

    Here's the kind of thing you can do with them, drive 9216 WS2812B smart pixels

     

    image

     

    http://makerlogic.com/ezpixel/

  • in reply to Workshopshed

    I hope you are powering the LED's externally.

     

     

     

    DAB

  • in reply to DAB

    When using lots of WS2812 LEDS it's very, very, very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very,very, very important to provide them ample clean power and control signal.

    Ensure that you don't get any excessive under or overshoot of the control signal.

     

  • As much as I would love to road test it, I must concentrate on road-tests that I am currently involved in my current projects, work and studies

  • in reply to mcb1

    Mark Subtly Hinted:

    >You do recall the Digispark bootloader can be changed to start instantly ...?

    ...

    >Besides some clever person sent you the code, so all you need to do is arrange some diodes to power it.

    ...

    >Digispark has an decent sized 7805 regulator, should handle 12v okay.

     

    You're absolutely right, Mark. After thinking about this more the Digispark looks like a much easier option, so I've dug one out of my drawer and I'll give that a go. I'll leave that FPGA learning curve for later, and thus unenrolled from that RT - it's just the wrong time for me take that on right now.

     

    Looks like I'll still need two more 7805's for the 2 inputs, right? Or do you have a much more clever way of doing that?

     

    ps, The 4000-series cmos chip method is also still in the works. So close. sooooo close. can't give up now!

     

    -Nico

  • in reply to ntewinkel

    Any signal overshoot and insufficient power can blow up WS2812B LEDs.

  • Like there are other obstacles than price for FPGA.

    Evaluation Board, Spartan-7 50 FPGA, 256MB RAM, 4 x Pmod Headers, Vivado

    Price:£103 (Price is before tax)

     

    Like there are other obstacles than price for FPGA.

    Think about it for a second, OK?

    You can build a PC for that kind of money.

    You can buy 50+ arduinos for that kind of money.

    You can buy month worth of food for that kind of money.

    You can buy 4 raspberry PI's for that kind of money + get some spare change for some small fun parts.

    I think even watching a movies worth of that money is better investment than buying this board.

    How delusional are this manufacturers, i bet no one in their right mind buys this garbage in a first place other than on impulse, no wonder they give this crap away for free.

  • in reply to recurzive

      wrote:

     

    You can build a PC for that kind of money.

    You can buy 50+ arduinos for that kind of money.

    You can buy month worth of food for that kind of money.

    You can buy 4 raspberry PI's for that kind of money + get some spare change for some small fun parts.

     

    These are all irrelevant comparisons.

     

      wrote:

     

    I think even watching a movies worth of that money is better investment than buying this board.

    How delusional are this manufacturers, i bet no one in their right mind buys this garbage in a first place other than on impulse, no wonder they give this crap away for free.

     

    Why are you being so judgemental? There are plenty of uses for FPGA's which enable you to do things which otherwise wouldn't be possible. Yes, they aren't for everybody, a lot of applications can be covered by cheap microcontroller based boards and in the hobby space this is exceedingly likely to be the case. But there are plenty of times in a development lab where you need to test a proof of concept design and need an FPGA to be part of that. This is where development boards like this come into play and at £103 this is really cheap compared to how much FPGA development boards have cost in the past. The cost of this board is small compared to the cost of one engineer for a day so if it saves a days effort it's paid for itself several times already. If it saves a board spin because something important is discovered early on it'll pay for itself tens or hundreds of times over depending on the complexity of the board under development.

     

    If FPGA's aren't your thing and you have no uses for them then that's fine but I fail to see why you feel the need to rant about how expensive and useless you feel they are. Just ignore them and move on.

  • in reply to recurzive

    There are things a single FPGA can do that are not possible with 5000 Arduino's, or Pi's. Specifically, throughput related or timing (accuracy) related. When you sent that message through the Internet, it probably went through infrastructure containing dozens (a hundred maybe) FPGA's across the world. The infrastructure manufacturers used FPGA's to produce the best product for a given cost. That's just one example of FPGA use. Those who can use FPGA's will earn that months' worth of food within an hour. The major obstacle is not the cost of the development board, it is the very different skill-set than for an Arduino or Raspberry Pi software programmer. Learning that skill-set ideally requires courses, or books that cost more than the development board. Some resources are free online however, and such a development board can be used to get those skills. But digital knowledge is also required (in contrast for software development on Arduino/Pi), so it is worth studying that too (there are some online resources for that too).