Actions
Forum Thread Details
  • Replies 4 replies
  • Subscribers 308 subscribers
  • Views 622 views
  • Users 0 members are here
Related

What technology or tool hasnt been made yet in the open source community?

Former Member

Hey everyone,

 

What are some technologies that haven't been attempted, or been poorly attempted to do in open source that you think should be? This could be anything from magnetic core memory to vector signal analyzers. I just want to get a feel for areas that need more development that may be overlooked.

 

Thanks,

 

Austin

  • One of the primary electronics diagnostic tools - the oscilloscope - has not been done well. You would think a 2 GHz ARM chip could manage a pretty decent sample rate at low cost.

  • in reply to dougw

    The issues in making  a decent oscilloscope go way beyond a micro. The Red Pitaya thing is an attempt at this but compared with a real scope it's just a joke.

    First you need the front end, input impedance 1M, variable (switcahble) gain from perhaps 100 to 0.01 in 1,2,5 steps with a flat frequency response over the working bandwidth (DC - 50 or 100MHz to compete with an entry level Rigol).

    Then you need an ADC, 8 bits, 500M samples/second  for openers.

    Then real time digital analysis of the data (for triggering) - this is where your 2Ghz processor can't hack it - it takes an FPGA and some trickery to manage that kind of data rate.

    For big(ger) scope rates (my Lecroy does 4 channels at 10Gsamples/sec per channel) it takes custom ICs and FPGAs.

    Even when you have the data it has to be merged into display memory (so that the effective display update rate is fast (ideally >10e6 times per second) so you have  a chance of seeing infrequent events.

     

    All this stuff takes real cash to develop - Open Source has had all it's big successes with software where the up front input can be all time rather than money.

     

    When you reflect that you can buy a Rigol DS1102E for £270 (2 channels, 1Gsample per channel) I don't think the chances of Open Source hardware are very high.

     

    MK

  • FPGAs: currently, except for an obscure Atmel part that never caught on, you cannot program FPGAs except using the vendor's tools.  IMO this lock-in has prevented FPGAs from enjoying the exponential growth of microprocessors, where you have many choices of tools and for which you can write your own compiler if you wish.

     

    The problem is that vendors do not provide bit-level documentation for how their FPGAs are configured.  While some architectures can be reverse-engineered using vendor tools, this is usually a violation of the EULA and anyone who publishes the results can expect to be sued.  GPUs were in this same situation until recently.  Some GPU vendors have seen the wisdom of opening their architectures and received overwhelming approval from the Open Source community.

  • what i see in the open hardware ecosystem is a trend of most electronics tools to be gimmicky and not of professional quality, and with very few being an exception to this.

     

    the oscilloscope conversation is a good example, ive watched these guys sell out at a makerfaire XMEGA Xprotolab | Development Boards | Gabotronics to be honest your better off with a 40 year old scope without triggering.

     

    Is the open hardware community destined to be flooded with crappy/dangerous tools? i think so unless one of the big players(Tek, HP or Lecroy) develop an open source oscilloscope.