Post by thread:[PIC]: universal IR receiver / decoder (concept or

i'm gonna build an universal infrared control center. i.e. the box has to receive ir signals from different ir controls, decode them and send appropriate control data to different interfaces, to control other devices. as there are different ir transmission protocols (rc5, sony, pana, some korean,...) i need to be able to learn and decode them all (or at least some of them) to be really universal. as the received ir code has to be checked with a manually written look up table, it should really be decoded, as i do not like to enter 20 characters for each ir command. i.e. the result of decoding should really be the value, the ir control has sent. does anyone have a concept or code for universally decoding ir signals? thanx for your help! tino

Well, the input is binary (IR signal on and off), and if you could convert it to a binary number somehow it would be easy to use a lookup table/table read. The IR signals probably go on and off, representing 1 and 0, and a longer on or off probably represents consequtive 0s or 1s. If you can find/make a clock source that matches the time for a single binary digit, then you just read the status of the IR reciever at each clock. The only problem with this is I think that a lot of different controls very likely use different speeds of clocks - you would either have to generate your own dynamic clock source, or have a different chip for each controller. Also, if you have a dynamic clock source, how would it know which type of controller is being used, and select the correct speed? On 6/2/05, Buehler, Martin <spam_OUTMartin.BuehlerTakeThisOuTkeymile.com> wrote: {Quote hidden}> i'm gonna build an universal infrared control center. > i.e. the box has to receive ir signals from different ir controls, > decode them and send appropriate control data to different interfaces, > to control other devices. > > as there are different ir transmission protocols (rc5, sony, pana, some > korean,...) i need to be able to learn and decode them all (or at least > some of them) to be really universal. > > as the received ir code has to be checked with a manually written look > up table, it should really be decoded, as i do not like to enter 20 > characters for each ir command. i.e. the result of decoding should > really be the value, the ir control has sent. > > does anyone have a concept or code for universally decoding ir signals? > > thanx for your help! > tino > > -

Hello, AN657 is your friend. Imre On Thu, 2 Jun 2005 .....dave.w.turnerKILLspam@spam@gmail.com wrote: {Quote hidden}> Well, the input is binary (IR signal on and off), and if you could > convert it to a binary number somehow it would be easy to use a lookup > table/table read. > > The IR signals probably go on and off, representing 1 and 0, and a > longer on or off probably represents consequtive 0s or 1s. If you can > find/make a clock source that matches the time for a single binary > digit, then you just read the status of the IR reciever at each clock. > > The only problem with this is I think that a lot of different controls > very likely use different speeds of clocks - you would either have to > generate your own dynamic clock source, or have a different chip for > each controller. Also, if you have a dynamic clock source, how would > it know which type of controller is being used, and select the correct > speed? > > On 6/2/05, Buehler, Martin <Martin.BuehlerKILLspamkeymile.com> wrote: >> i'm gonna build an universal infrared control center. >> i.e. the box has to receive ir signals from different ir controls, >> decode them and send appropriate control data to different interfaces, >> to control other devices. >> >> as there are different ir transmission protocols (rc5, sony, pana, some >> korean,...) i need to be able to learn and decode them all (or at least >> some of them) to be really universal. >> >> as the received ir code has to be checked with a manually written look >> up table, it should really be decoded, as i do not like to enter 20 >> characters for each ir command. i.e. the result of decoding should >> really be the value, the ir control has sent. >> >> does anyone have a concept or code for universally decoding ir signals? >> >> thanx for your help! >> tino >> >> --

dave.w.turner@gmail.com wrote: > Well, the input is binary (IR signal on and off), and if you could > convert it to a binary number somehow it would be easy to use a lookup > table/table read. > > The IR signals probably go on and off, representing 1 and 0, and a > longer on or off probably represents consequtive 0s or 1s. This is waaaaaay too simplified. There are many different encoding schemes. First, IR is almost always sent as timed pulses of a fixed frequency carrier. This makes it much easier to distiguish from the ambient IR level. The carrier frequencies vary typically in the range from 30-55KHz. The choice of carrier frequency is pretty much an arbitrary pick by the manufacturer. Receiver modules that detect the presence of IR carrier are amazingly sensitive, work much better than I would have thought, and are readily available and surprisingly cheap. However, they are each tuned to a narrow carrier range. One of these module can not be used for a universal receiver. Then there are lots of different encoding formats. If there are any standards, they are widely ignored and most manufacturers cook up their own protocols. There may be compatibility within a manufacturer or product line, but I would assume not beyond that unless specifically stated. I've seen protocols where data was sent by the relative timing between carrier bursts, lengths of bursts, and just about anything else you can imagine. I've personally created one system that used manchester encoding. I've created another system that decoded transmissions from a commercial keyboard. That used at least three different spacings between carrier bursts for signalling if I remember right. Then there is interpreting the bits even if you can decode them from the demodulated carrier signal. Just a sequence of 1s and 0s won't tell you much without knowing the next level up protocol. There are preambles, signatures, and checksums, all mixed in the same stream with some occasional payload bits. And that's not even talking about the deliberately encoded ones. ***************************************************************** Embed Inc, embedded system specialists in Littleton Massachusetts (978) 742-9014, http://www.embedinc.com

> The carrier frequencies vary typically in the range from > 30-55KHz. The > choice of carrier frequency is pretty much an arbitrary pick by the > manufacturer. Receiver modules that detect the presence of > IR carrier are > amazingly sensitive, work much better than I would have > thought, and are > readily available and surprisingly cheap. However, they are > each tuned to a > narrow carrier range. One of these module can not be used > for a universal > receiver. But 36 and 38 kHz are certainly the most common (at least here in Europe), and a 36kHz receiver will work quite well for 38 and vice versa. (maybe half the distance compared to the optimal frequency). And 37 kHz versions exists as a compromise. Wouter van Ooijen -- ------------------------------------------- Van Ooijen Technische Informatica: http://www.voti.nl consultancy, development, PICmicro products docent Hogeschool van Utrecht: http://www.voti.nl/hvu

> i'm gonna build an universal infrared control center. > i.e. the box has to receive ir signals from different ir controls, > decode them and send appropriate control data to different interfaces, > to control other devices. This very project has been on my "to-do" list for about 3 years. My fancy universal remote won't control my cable box, and my cable box remote won't control my fancy stereo receiver. However, since we move in less than a year, I'll probably just write it off...unless I get REALLY bored. > as there are different ir transmission protocols (rc5, sony, pana, some > korean,...) i need to be able to learn and decode them all (or at least > some of them) to be really universal. This would be much easier if they LOOKED the same. But the pulse length modulated ones must be decoded differently than the NRZ ones. AFAIK, most (all?) IR signals are either NRZ or PLM, so if you can figure out a method to detect which type of signal the device is receiving, and the timebase of that (i.e., length of bit in NRZ, or high pulse time in PLM), then extract the binary data from that info, your job would be MUCH easier. I guess an alternative (probably a good one) is to expect that the codes from the remote are all Sony protocol (since you'd probably be using a universal remote for that end anyhow, and you can choose the code you want), then record pulse high/low times for the code to send out and play those back without concern for what the data actually means. I think the best thing to be would to have the user tap the button as fast as possible, then compute a checksum, then have the user do it again until the checksum comes up twice. > as the received ir code has to be checked with a manually written look > up table, it should really be decoded, as i do not like to enter 20 > characters for each ir command. i.e. the result of decoding should > really be the value, the ir control has sent. Why do anything manually? Mike H.

>My fancy universal remote won't control my cable box, and my cable >box remote won't control my fancy stereo receiver. You aren't alone..............the universal remotes don't have all the features and functions for all variations. We bought 3 of them, and finally gave up. However, just recently, the new style universal remotes are on the market. Most of them start with basic programming, just like the conventional universal remotes do. BUT...........they also have a 'learning' mode function. You lay both remotes on a table top facing each other. The learning remote has extra unassigned buttons, which you can custom program by having it read the output of the original remote control unit. In this way, you do not lose any functionality. Some of the units can even be 100 percent programmed from scratch in learning mode...so a non-existant or custom code can be learned. They are a little pricey, but well worth the money. Some even have EL backlighted screens and USB computer interfaces built in for downloading updated code from the internet. To sample what's on the market, do an ebay search for learn* remote. GL to all. Art

> However, just recently, the new style universal remotes are on the market. > Most of them start with basic programming, just like the conventional > universal remotes do. BUT...........they also have a 'learning' mode > function. You lay both remotes on a table top facing each other. The > learning remote has extra unassigned buttons, which you can custom program > by having it read the output of the original remote control unit. Of course, then you won't have done it yourself... ;-). Actually, there are three uber-huge advantages to doing this yourself: 1. This device could be the only visible device (besides your monitor) to an entertainment center. Everything else can be hidden inside a cabinet. 2. You could include a PC interface on this thing, to allow you to control a media player PC with it. 3. You'll learn a whole heck of a lot. Number 3 may be the best argument- factor in your time and the costs involved and there's almost certainly an off-the-shelf solution that'll take care of 1 and 2. Mike H.

4. You could let the PC control the remote - you could make it turn off the radio/TV at certain times, or set a timer on some VCRs with no/bad timers. 5. It's fun 6. Sell it and get $$$ (probably $ though). On 6/2/05, Mike Hord <.....mike.hordKILLspam.....gmail.com> wrote: {Quote hidden}> > However, just recently, the new style universal remotes are on the market. > > Most of them start with basic programming, just like the conventional > > universal remotes do. BUT...........they also have a 'learning' mode > > function. You lay both remotes on a table top facing each other. The > > learning remote has extra unassigned buttons, which you can custom program > > by having it read the output of the original remote control unit. > > Of course, then you won't have done it yourself... ;-). > > Actually, there are three uber-huge advantages to doing this yourself: > > 1. This device could be the only visible device (besides your monitor) > to an entertainment center. Everything else can be hidden inside a > cabinet. > > 2. You could include a PC interface on this thing, to allow you to control > a media player PC with it. > > 3. You'll learn a whole heck of a lot. > > Number 3 may be the best argument- factor in your time and the > costs involved and there's almost certainly an off-the-shelf solution > that'll take care of 1 and 2. > > Mike H. > > -

There is also a type of inexpensive remote control usually sold under the Radio Shack or All-In-One brands that has been reverse engineered and can be completely reprogrammed using a cable (dubbed a JP1 Cable) hooked up to the parallel port of a PC. And I don't mean the relatively small "learning" memory that many remotes have. I mean that you can completely reprogram virtually every button on the remotes, even if it didn't originally even support your device. There is a pretty good overview of these remotes and reprogramming them here (http://www.hifi-remote.com/ofa/nuts-volts.shtml), but there is a lot more information here: (http://www.hifi-remote.com/jp1/index.shtml) and there is a Yahoo Group on the subject here: (http://groups.yahoo.com/group/jp1/) The short story is that there is a little header tucked away in the battery compartment of some remotes that is apparently used by the manufacturer to program the remotes. Somebody reverse engineered the interface and now you can custom design The Perfect Remote for your entertainment center using a remote that only costs $10 to $30 USD. All you have to do is put the codes in an MS Excel spreadsheet and download a freeware program called IR.exe and update the remote using your JP1 cable. I imagine that this JP1 interface might be fertile ground for some interesting PIC hacks. :-)

Bob Blick wrote: >>does anyone have a concept or code for universally decoding ir signals? >> >> > >Have you looked at IRMAN? It does just that. It is a commercial product, >but there is a free compatible called UIR for the 16F84 and 12C509. It >works with MythTV, XMMS, winamp, etc. > >It outputs an 8 character hex for each keypress. > >Cheers, > >Bob > > > > You should check out http://www.cesko.host.sk/ , too. IR to RS232, USB, Ethernet! David

Alan B. Pearce wrote: >>i'm gonna build an universal infrared control center. >>i.e. the box has to receive ir signals from different ir controls, >>decode them and send appropriate control data to different interfaces, >>to control other devices. > > > 'Course once you get this lot done, then you'll want to do the same with the > IR car alarm systems, and copy the thieves that go around car parks ... That's why any decent car alarm system uses 'rolling' or random access codes. R

Hello Martin, Have you been able to get the app going? I recently came across an article "IR Code analyzer". It identifies the IR code transmitted to the device and displays on the LCD. http://www.elektor-electronics.co.uk/Default.aspx?tabid=27&year=2001&month=-1&art=50166&PN=On Regards, -- Chetan Bhargava Web: http://www.bhargavaz.net Blog: http://microz.blogspot.com On 6/1/05, Buehler, Martin <EraseMEMartin.Buehlerspam_OUTTakeThisOuTkeymile.com> wrote: {Quote hidden}> i'm gonna build an universal infrared control center. > i.e. the box has to receive ir signals from different ir controls, > decode them and send appropriate control data to different interfaces, > to control other devices. > > as there are different ir transmission protocols (rc5, sony, pana, some > korean,...) i need to be able to learn and decode them all (or at least > some of them) to be really universal. > > as the received ir code has to be checked with a manually written look > up table, it should really be decoded, as i do not like to enter 20 > characters for each ir command. i.e. the result of decoding should > really be the value, the ir control has sent. > > does anyone have a concept or code for universally decoding ir signals? > > thanx for your help! > tino > > -