magore@watdcsu.UUCP (10/08/87)
Hello Peter, In article <2152@ihlpe.ATT.COM> psfales@ihlpe.ATT.COM (Pete Fales) writes: [munch...] >Do you really need to know the signal format? I recently built a >mult-device IR controller (similar to the one recently described in Byte >magazine). No need to decode the format, just digitally "record" a >sample of the signal and play it back. The Byte article describes >the tricks to reducing this to a manageable volume of data. [munch...] >Peter Fales UUCP: ...ihnp4!ihlpe!psfales I have a few comments about the Byte article. The Byte article used a PC via serial interface to the 8031 for initial set-up of menus and such. If you drop the serial interface and have the controler do all the work you can use the internal shift register to sample at up to 1.25MHZ [ with 15Mhz version ] - thus dropping the need for an external shifter as in the article. I find that the 8031 has no problem generating the codes for my SONY VCR _without_ the external hardware timers used... This is obvious, the timings can only be as good as the original sample rate and the 8031 can surely put out at the same rate it samples in the first place! Using the internal 8 bit shifter mode is not really needed for output though. Most instructions take 12 clocks and this includes port bit invert instructions so an inline section of code can actually 'bit-bang' at 1MHZ or better depending on clock. As Peter implied if you want to save memory space by say storing just the time between transitions. Durring play back you can reconstruct the original bit stream and then play it back -or you can reconstruct the timings on the fly. In the latter case to avoid hard coding delay loops all one has to do is generate the delay code in place in RAM and branch to it ! [I know - ugh] This requires connecting (PSEN & RD) ] - So you then have self generating code which is a timing win over branch delays which are 24 clocks. With regards to Peters comment about just needing to sample store: The Byte article 'cheated' by _assuming_ a carrier coded data stream to save more space then the methods suggested here. Still I think if someone just samples and then use edge transition time compression alone I think they could get quite enough starage with only a few 32K CMOS RAMS for several remotes. Add to edge transition run-length compression of the edge data itself and you have something just as good as the Byte version but with far less wired in assumptions or complexity - which of course is what Peter is saying... Another point - the 8031 is also made in CMOS... Best Regards, # Mike Gore # Institute for Computer Research. ( watmath!mgvax!root - at home ) # These ideas/concepts do not imply views held by the University of Waterloo.
psfales@ihlpe.ATT.COM (Pete Fales) (10/10/87)
Hello Mike, In article <3935@watdcsu.waterloo.edu>, magore@watdcsu.waterloo.edu (Mike Gore, Institute Computer Research - ICR) writes: > Hello Peter, > > In article <2152@ihlpe.ATT.COM> psfales@ihlpe.ATT.COM (Pete Fales) writes: > [munch...] > >Do you really need to know the signal format? I recently built a > >mult-device IR controller (similar to the one recently described in Byte > >magazine). > > I have a few comments about the Byte article. The Byte article > used a PC via serial interface to the 8031 for initial set-up of menus > and such. If you drop the serial interface and have the controler > do all the work you can use the internal shift register to sample > at up to 1.25MHZ [ with 15Mhz version ] - thus dropping the need > for an external shifter as in the article. > [munch] Well, now that at least one person has read my note, I feel justified in providing a few more details. In my case, I used the Hitachi 63B03X (CMOS) microcontroller which includes standard asynchronous serial port (for the PC interfac), as well as a fairly sophisticated timer interface. The timer includes an edge-capturing feature that allowed me to sample the signals with .5 microsecond resolution without any external hardware such as the serial shift register. For playback, I felt the large reduction in data volume as well as eliminating the need for contorted assembly coding justified the additional hardware for the 82C54. The other innovation I feel quite proud of is realizing the the modulated 40 KHz carrier is the same method used by the ultrasonic remote for my BSR X-10 home controller (which has been previously discussed on this net). So, after connecting an ultrasonic transducer in parallel with the infrared LEDs, my controller can handle all the IR devices: TV, VCR, CD player, etc. as well as control lights and appliances all over the house. It's a fun toy. I would be happy to provide more details for anyone who might be interested. -- Peter Fales UUCP: ...ihnp4!ihlpe!psfales work: (312) 979-7784 AT&T Information Systems, IW 1Z-243 1100 E. Warrenville Rd., IL 60566