Page 1 of 2 12 LastLast
Results 1 to 10 of 14

Thread: Poor's man Ilda signal Extender via Ethernet cable

  1. #1
    Join Date
    Sep 2008
    Location
    Turkey
    Posts
    16

    Default Poor's man Ilda signal Extender via Ethernet cable

    Hello to all
    I ve been on project abouth Ilda signal extender via standart ethernet cable for a while. I complated both ADC and DAC my boards at %90 but i am stucked at fluctuation issue. However Complicated logos do not project what i expected. I suspect reflesh rate speed but maybe i overlook something else.
    My main issue is image do not project constantly, images a little bit fluctuate like a flag.
    Technical details are;
    - Actual Reflesh rate is 65us
    - X/Y 10bit resolution, RGB 8 bit at both ADC and DAC Side
    - Dataspeed 6Mbitps
    Any comments i appriciated.

    Click image for larger version. 

Name:	IMG-20190423-200543.png 
Views:	49 
Size:	537.1 KB 
ID:	55359
    Click image for larger version. 

Name:	IMG-20190423-200655.png 
Views:	46 
Size:	401.3 KB 
ID:	55360

    Click image for larger version. 

Name:	IMG-20190423-200714.png 
Views:	46 
Size:	372.5 KB 
ID:	55361

    Click image for larger version. 

Name:	IMG-20190423-200835.png 
Views:	48 
Size:	557.5 KB 
ID:	55362
    Click image for larger version. 

Name:	IMG-20190424-132512.png 
Views:	46 
Size:	581.1 KB 
ID:	55363
    Click image for larger version. 

Name:	IMG-20190424-132552.png 
Views:	47 
Size:	380.3 KB 
ID:	55364
    Note: Scope Blue line Source, Yellow line extender range DAC Out
    Last edited by pyrodigy; 04-24-2019 at 04:32.

  2. #2
    Join Date
    Jun 2009
    Location
    St. Louis, MO
    Posts
    708

    Default

    65 uSec is about 15k points sec - which is a problem with faster scanners, more problematic would be if you don't sample all of the inputs at the same time, and trigger all of the outputs at the same time.
    "There are painters who transform the sun into a yellow spot, but there are others who, with the help of their art and their intelligence, transform a yellow spot into the sun." Pablo Picasso

  3. #3
    Join Date
    Sep 2008
    Location
    Turkey
    Posts
    16

    Default

    Quote Originally Posted by laserist View Post
    65 uSec is about 15k points sec - which is a problem with faster scanners, more problematic would be if you don't sample all of the inputs at the same time, and trigger all of the outputs at the same time.
    Laserist thank you for your response. I agree with you, this looks obviously reflesh speed issue. I need to upgrade MCU with faster one.
    At your suggestion If 65us reflesh rate takes 15k pps then can i say 30k pps needs 30us sequence data frames ? And also 60k pps 15us as well ?
    By the way Would it be ignore a couple of uS delay while sample input and trigger outputs ?
    Because significant sequence data both input and output serial so depends on spi speed, there would be 2-3 us delay while sample and trigger.
    Best regards

  4. #4
    mixedgas's Avatar
    mixedgas is offline Creaky Old Award Winning Bastard Technologist
    Infinitus Excellentia Ion Laser Dominatus
    Join Date
    May 2007
    Location
    A lab with some dripping water on the floor.
    Posts
    9,200

    Default

    I think your undersampling.

    Nyquist's law says you must sample at a minimum of twice the highest frequency in the signal. Preferably much faster then that...

    Couple that with a ADC or DAC that has modest linearity issues and I think that would answer for what I see on the scope.

    An analog reconstruction filter is also called for.. A tiny bit of RC lowpass in your case... But that brings up phasing issues.

    1/30,000 = .00003333 seconds...

    Do all your ADCs and DACs sample or update exactly at the same time, ie double buffering?

    Steve
    Last edited by mixedgas; 04-24-2019 at 12:29.
    Qui habet Christos, habet Vitam!
    I should have rented the space under my name for advertising.
    When I still could have...

  5. #5
    Join Date
    May 2009
    Location
    UCSB
    Posts
    685

    Default

    Without a way to synchronize the A/D sampling to the output of your DAC it is going to be quite challenging to get good image quality especially at higher frame rates, in particular with the R/G/B channels which do not have the inherent filtering response of the galvos due to their limited scan speed. I think that you will need at least 100kHz simultaneous sampling (at least for R/G/B to get clean color transitions) ideally 1MHz or more if you will be running at faster than the usual 30kpps

    Is it possible for you to directly output data over ethernet from your laser show software? The hardware you have put together looks very similar to the etherdream, and it would probably be possible to implement the etherdream network protocol on your hardware. Depending on what hardware you used (etherdream is based on a NXP ARM) it should be possible to port the existing etherdream codebase over. Somewhat annoyingly the etherdream2 seems to have switched to a closed source product, but the etherdream 1 and its dll to interface to laser show software is on github under GPL v3 https://github.com/j4cbo/j4cDAC

    There is also the IDN protocol (https://www.ilda.com/resources/Stand...eam_rev001.pdf) but it seems to be very hard to implement (AFAIK the only existing implementations are FPGA based) and not well supported by the laser show software.

  6. #6
    Join Date
    Nov 2008
    Location
    Cleveland Ohio
    Posts
    2,179

    Default

    Why convert to digital at all. Iíve not tried this but it seems to me you could go a very long way on fiberoptics using one fiber per channel so a bundle of 5. Then all you do is drive the emitter on a fiber @nd put some receiver on the other side. Should be totally immune to any noise along the way and the delay as short as possible without a vacuum.

    if you use five different frequencies you could probable use a single fiber and use dichro on the receivers.

    I had had a similar issue and solved it by changing cat5 cables coming from my laptop to an fb4. Used cat6e instead.

  7. #7
    Join Date
    Sep 2008
    Location
    Turkey
    Posts
    16

    Default

    Thank you guys for your responses with valued informations.

    @mixedgas
    I can only able to reduce reflesh rate as 50uS, not less because of MCU ability.
    Here is the results;
    Scope view X Coordinat only Yellow one Source, Blue one x+ DAC Out.


    fluctuation still occur at the Image



    @krazer
    Krazer first of all i would like to state a point is that this my own design for Extend Ilda signals to long range. Sorry for i didnt introduce clearly in my first topic.
    My main aim was in the project, would read from the source and carry with cable and outs from the DACs as same value as read one. In my design i inspired Open ilda board,
    There is no need software for in my design, i just would like to attach master board to DAC source like FB3, Lumax,..ext and at the laser system side i would like to attach slave board to laser system's ilda in. Signals will be transfer via standart ethernet cable. Not ethernet protocol. I used 8 pin Ethernet cable for carry 4 pin data signals and 4 pin Power with GND to distrubute Sİgnals and Powers to the slaves.
    I just used RS485 standarts and i used my own data protocol. RS485 can able to carry like over 10mbitps speed withouth any problem, this is why I didnt use TCP-IP protocol.
    At 6Mbps i can able to reflesh at 50us minimum, in the very next release i am thinking to upgrade 10mbit or more.
    As you stated in your message reflesh rate importatnt at this point. i can only able to reduce reflesh rate as 50uS but this is demo board anyway. new design will be better i hope.

    @kecked
    I will consider Fiber optic method why not. For very long range results it would be best. For now rs485 can able to handle abouth 100-200 meters long for demo results.

    More Test Videos;
    30kpps QS speed,
    50uS Reflesh rate of my boards;
    Left Laser one directly from FB3, Right one from demo board which signal source from Laser one's ILDA Out > CAT6 > Second laser's ILDA In











    Beam effects and not complicated graphics do not fluctuate. At least i can not see
    Last edited by pyrodigy; 04-25-2019 at 12:46.

  8. #8
    Join Date
    Mar 2007
    Location
    UK
    Posts
    230

    Default

    Quote Originally Posted by krazer View Post

    There is also the IDN protocol (https://www.ilda.com/resources/Stand...eam_rev001.pdf) but it seems to be very hard to implement (AFAIK the only existing implementations are FPGA based) and not well supported by the laser show software.
    Not quite so difficult. There are several implementations of receivers and transmitters. There are open source examples on github. And there is no reason at all that a receiver/DAC couldnít be implemented on a micro-controller.

    A while back, I needed a convenient way of capturing an ILDA output signal, so rather than reinvent the wheel, I just wrote some code to receive the IDN data from UDP datagrams, following the open IDN protocol.

    James
    Laser Safety
    http://www.lvroptical.com
    http://www.facebook.com/LaserSafety

    - Laser Show Safety Training, Assessment, and Realtime MPE Measurement
    - Pangolin PASS System Integrator

  9. #9
    mixedgas's Avatar
    mixedgas is offline Creaky Old Award Winning Bastard Technologist
    Infinitus Excellentia Ion Laser Dominatus
    Join Date
    May 2007
    Location
    A lab with some dripping water on the floor.
    Posts
    9,200

    Default

    OP's A to D appears to not be double buffered... Hence the distortions.. EACH A to D and Each D to A must input and output in a synchronized manner on the X-Y signals or you will get that kind of distortion. I'd second the recommendation for using IDN, I have a set of the IDN dongle hardware and it is wonderful.

    Steve
    Qui habet Christos, habet Vitam!
    I should have rented the space under my name for advertising.
    When I still could have...

  10. #10
    Join Date
    May 2009
    Location
    UCSB
    Posts
    685

    Default

    @pyrodigy
    I see what you are trying to achieve now. I think you are on the right track that the solution is a more powerful microcontroller and use a higher sample rate. I think that finding a way to simultaneously sample the inputs will also help a lot in reducing the flutter, possibly just adding a sample/hold chip between the ILDA lines and the microprocessor will be enough. Of course it would be better to have a dedicated A/D converter for each channel but that would increase the complexity a lot.

    @JStewart
    That is very interesting to hear, I will have to look into this. Do you know if the easylase (or EzAudDac.dll) compatible driver was ever released? And if there is an existing open project like the etherdream that supports IDN? The etherdream is great but it would be nice to migrate over to IDN as support for it grows.

Posting Permissions

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