Post by thread:[EE] IRDA here to stay?

Hi all, I am about to design a product which will be in use for at least 10 years and needs a medium range (about 50cm max) free-space optical data link, low speed (under 9600 bps is fine - even 1200 bps would be acceptable). The product contains both sides of the link so it does not have to communicate with standard products. It needs to be robust to a fair amount of dust (up to a 1cm layer of paper dust) and be able to work in normal ambient room light (i.e., not outdoors). I prototyped a solution using a microcontroller, IR LED, and IR phototransistor and FSK modulation which seems to work well although I have not tested it yet under all of the relevant conditions (e.g. dust). However, a co-worker of mine pointed out that it would be nice to use an off-the-shelf IRDA module, which would have the side benefit of allowing this optical interface to be used also as a debugging port for the product, since IRDA would be easily capable of much higher data rates. I am a bit concerned about designing in IRDA, though, since I do not see it commonly used anymore in consumer devices. I wonder if it is going obsolete and may be difficult to source components for it in, say, 3 or 5 years from now. I searched the web and haven't found anything definite about the future of IRDA - wikipedia seems to indicate that its popularity is being revived by something called "irSimple" which seems to be a variant of IRDA. Also, I found articles which indicate that it is still widely used in consumer products in Japan. I found that Digikey and Mouser both seem to have plenty of stock of IRDA offerings from several manufacturers. What is your opinion? Is IRDA going to be around for at least 5 more years or is it disappearing quickly? Thanks, Sea

On Fri, 2011-06-03 at 11:17 -0400, Sean Breheny wrote: > What is your opinion? Is IRDA going to be around for at least 5 more > years or is it disappearing quickly? It's always hard to predict the future. IMHO IRDA has already mostly disappeared. Very few devices still have an IRDA port. It's been pretty much usurped by bluetooth. It is still used alot for debug/programming ports on things like electrical meters. Personally, I'd avoid designing something with IRDA and instead use wireless. If you are stuck with optical then perhaps you can just buy enough modules so you'll never have to worry about them no longer being available? TTYL

On Fri, 3 Jun 2011 11:17:23 -0400, you wrote: {Quote hidden}>Hi all, > >I am about to design a product which will be in use for at least 10 >years and needs a medium range (about 50cm max) free-space optical >data link, low speed (under 9600 bps is fine - even 1200 bps would be >acceptable). The product contains both sides of the link so it does >not have to communicate with standard products. It needs to be robust >to a fair amount of dust (up to a 1cm layer of paper dust) and be able >to work in normal ambient room light (i.e., not outdoors). > >I prototyped a solution using a microcontroller, IR LED, and IR >phototransistor and FSK modulation which seems to work well although I >have not tested it yet under all of the relevant conditions (e.g. >dust). > >However, a co-worker of mine pointed out that it would be nice to use >an off-the-shelf IRDA module, which would have the side benefit of >allowing this optical interface to be used also as a debugging port >for the product, since IRDA would be easily capable of much higher >data rates. > >I am a bit concerned about designing in IRDA, though, since I do not >see it commonly used anymore in consumer devices. I wonder if it is >going obsolete and may be difficult to source components for it in, >say, 3 or 5 years from now. I searched the web and haven't found >anything definite about the future of IRDA - wikipedia seems to >indicate that its popularity is being revived by something called >"irSimple" which seems to be a variant of IRDA. Also, I found articles >which indicate that it is still widely used in consumer products in >Japan. I found that Digikey and Mouser both seem to have plenty of >stock of IRDA offerings from several manufacturers. > >What is your opinion? Is IRDA going to be around for at least 5 more >years or is it disappearing quickly? > >Thanks, > >Sean I wouldn't be too concerned about IRDA parts disappearing, they're fairly cheap, so assuming you buy from someone that will warn you of imminent obscolesecence, , you should get enough warning to do a lifetime buy. Something else to consider is IR remote receivers - cheap, good for up to about 2-4kbits, tons of range and IR remote controls aren't going away any time soon.

On 6/3/2011 11:17 AM, Sean Breheny wrote: > Hi all, > > I am about to design a product which will be in use for at least 10 > years and needs a medium range (about 50cm max) free-space optical > data link, low speed (under 9600 bps is fine - even 1200 bps would be > acceptable). The product contains both sides of the link so it does > not have to communicate with standard products. It needs to be robust > to a fair amount of dust (up to a 1cm layer of paper dust) and be able > to work in normal ambient room light (i.e., not outdoors). > > Is that 1cm or 1mm of dust? Did my wife send you the specs for my office

On 03/06/2011 16:17, Sean Breheny wrote: > What is your opinion? Is IRDA going to be around for at least 5 more > years or is it disappearing quickly? > It's never had much use and is an overkill for what you need. I thought it was dead now (replaced by Bluetooth) though my phone and laptop have it. The range on most is less than 30cm usually. Some are only 10cm, though 5m is possible. Use an 38KHz IR RX as used in TVs and set-boxes. It can manage 1200bps with range of 5m to 10m You use Manchester encoding and 38kHz clock. The very cheap IR RX gives filtered decoded TTL. No complex SW stack needed as per IRDA. IRDA is also a two way protocol, you only need one direction? (If you need two directions just have IR LED and IR RX at both ends). Virtually all the remote controllers are pre-programmed very basic PICs

On 03/06/2011 16:17, Sean Breheny wrote: > What is your opinion? Is IRDA going to be around for at least 5 more > years or is it disappearing quickly? > I think it has obviously become far less popular, with the advent of bluetooth and stuff. However, I don't think it will disappear altogether for quite some time, there are still plenty of applications that are more suited to an easily implemented, cheap, line of sight solution (for which RF would be overkill) Even if it did disappear, as noted I'm sure you could quite easily conjure up your own. I doubt very much this would be needed though - I know it's not quite the same , but look at e.g. RS232, USB, SPI, I2C etc, has been around for ages but it is still reasonably common and ICs available. Hard to be certain, but I'm betting on it being good for a while (as long as remotes use the technology anyway..) I think I would go for it.

On Fri, 2011-06-03 at 08:49 -0700, William "Chops" Westfield wrote: > On Jun 3, 2011, at 8:36 AM, Herbert Graf wrote: > > > I'd avoid designing something with IRDA and instead use wireless. > > At like 5 to 10x the price and complexity!? 5-10x? Wireless modules can be had for a few bucks each, I don't know what the price of an IRDA module is, but I'll give you that on price. Complexity? It's pretty trivial to drive a wireless module, in fact the algorithms used are pretty similar to what you'd do with an optical connection, so complexity? The same IMHO. TTYL

Can you give me an example of a few $ wireless module? DougM On Fri, Jun 3, 2011 at 10:32 AM, Herbert Graf <spam_OUThkgrafTakeThisOuTgmail.com> wrote: {Quote hidden}> On Fri, 2011-06-03 at 08:49 -0700, William "Chops" Westfield wrote: > > On Jun 3, 2011, at 8:36 AM, Herbert Graf wrote: > > > > > I'd avoid designing something with IRDA and instead use wireless. > > > > At like 5 to 10x the price and complexity!? > > 5-10x? > > Wireless modules can be had for a few bucks each, I don't know what the > price of an IRDA module is, but I'll give you that on price. > > Complexity? It's pretty trivial to drive a wireless module, in fact the > algorithms used are pretty similar to what you'd do with an optical > connection, so complexity? The same IMHO. > > TTYL > >

Thanks to everyone for their replies. I think RS-232 is a valid comparison, although the more troubling thing about IRDA is that it needs not only an IC or two but also the optical components, so it isn't just a matter of a single IC like an RS-232 level converter. My gut feeling therefore is that RS-232 will outlive IRDA. You heard correctly that I must be able to deal with 1 centimeter of dust. This is because the two halves of the link are one above the other so that the TX side is part of a horizontal surface and will collect dust. Yes, I also only need unidirectional comms. This is in an environment where several of these devices could be in close proximity to each other and there cannot be crosstalk between them - which is why I am opting for optical instead of RF. Sean On Fri, Jun 3, 2011 at 12:45 PM, Oli Glaser <.....oli.glaserKILLspam@spam@talktalk.net> wrote: {Quote hidden}> On 03/06/2011 16:17, Sean Breheny wrote: >> What is your opinion? Is IRDA going to be around for at least 5 more >> years or is it disappearing quickly? >> > > I think it has obviously become far less popular, with the advent of > bluetooth and stuff. > However, I don't think it will disappear altogether for quite some time, > there are still plenty of applications that are more suited to an easily > implemented, cheap, line of sight solution (for which RF would be overkill) > Even if it did disappear, as noted I'm sure you could quite easily > conjure up your own. I doubt very much this would be needed though - I > know it's not quite the same , but look at e.g. RS232, USB, SPI, I2C > etc, has been around for ages but it is still reasonably common and ICs > available. > Hard to be certain, but I'm betting on it being good for a while (as > long as remotes use the technology anyway..) > I think I would go for it. > > >

how is light going to get through a cm of dust? Is this pixie dust? DougM On Fri, Jun 3, 2011 at 10:54 AM, Sean Breheny <shb7KILLspamcornell.edu> wrote: {Quote hidden}> Thanks to everyone for their replies. > > I think RS-232 is a valid comparison, although the more troubling > thing about IRDA is that it needs not only an IC or two but also the > optical components, so it isn't just a matter of a single IC like an > RS-232 level converter. My gut feeling therefore is that RS-232 will > outlive IRDA. > > You heard correctly that I must be able to deal with 1 centimeter of > dust. This is because the two halves of the link are one above the > other so that the TX side is part of a horizontal surface and will > collect dust. Yes, I also only need unidirectional comms. > > This is in an environment where several of these devices could be in > close proximity to each other and there cannot be crosstalk between > them - which is why I am opting for optical instead of RF. > > Sean > > > On Fri, Jun 3, 2011 at 12:45 PM, Oli Glaser <.....oli.glaserKILLspam.....talktalk.net> > wrote: > > On 03/06/2011 16:17, Sean Breheny wrote: > >> What is your opinion? Is IRDA going to be around for at least 5 more > >> years or is it disappearing quickly? > >> > > > > I think it has obviously become far less popular, with the advent of > > bluetooth and stuff. > > However, I don't think it will disappear altogether for quite some time, > > there are still plenty of applications that are more suited to an easily > > implemented, cheap, line of sight solution (for which RF would be > overkill) > > Even if it did disappear, as noted I'm sure you could quite easily > > conjure up your own. I doubt very much this would be needed though - I > > know it's not quite the same , but look at e.g. RS232, USB, SPI, I2C > > etc, has been around for ages but it is still reasonably common and ICs > > available. > > Hard to be certain, but I'm betting on it being good for a while (as > > long as remotes use the technology anyway..) > > I think I would go for it. > > > > > > --

This is where you get into the area of error correction, though. very cheap modules just push through bits and errors be damned. bt and other modules do your error correction for you, so what goes in is guaranteed to be what comes out. Thanks, DougM On Fri, Jun 3, 2011 at 10:59 AM, Michael Watterson <mikespam_OUTradioway.org>wrote: > On 03/06/2011 18:46, doug metzler wrote: > > Can you give me an example of a few $ wireless module? > > > > DougM > > > > On Fri, Jun 3, 2011 at 10:32 AM, Herbert Graf<@spam@hkgrafKILLspamgmail.com> wrote: > > > > OOK 315/433/868/915 MHz is very cheap and simple. > 2.4Ghz is less cheap, simple and not so dear if not WiFi or BT. >

On 03/06/2011 19:11, doug metzler wrote: > This is where you get into the area of error correction, though. very cheap > modules just push through bits and errors be damned. bt and other modules > do your error correction for you, so what goes in is guaranteed to be what > comes out. Even some quite expensive 2.4Ghz it's up to you to encrypt and add FEC. Only if you have a module implementing a complete stack that has ECC included in spec gives automatic FEC and/or retrie

On 03/06/2011 18:57, doug metzler wrote: > how is light going to get through a cm of dust? Is this pixie dust? > > DougM > > > On Fri, Jun 3, 2011 at 10:54 AM, Sean Breheny<KILLspamshb7KILLspamcornell.edu> wrote: IRDA is a high speed protocol it likely won't. with basic IR remote control you can use more power on LED (it's only an LED and you add your own driver) and it would only be 1200 bps on 38kHz OOK carrier. Depends how tight packed the dust is, I'd not rule it out without a test

On 6/3/2011 1:32 PM, Herbert Graf wrote: {Quote hidden}> On Fri, 2011-06-03 at 08:49 -0700, William "Chops" Westfield wrote: > >> On Jun 3, 2011, at 8:36 AM, Herbert Graf wrote: >> >> >>> I'd avoid designing something with IRDA and instead use wireless. >>> >> At like 5 to 10x the price and complexity!? >> > 5-10x? > > Wireless modules can be had for a few bucks each, I don't know what the > price of an IRDA module is, but I'll give you that on price. > > Complexity? It's pretty trivial to drive a wireless module, in fact the > algorithms used are pretty similar to what you'd do with an optical > connection, so complexity? The same IMHO. > > TTYL > > I think the complexity with wireless comes with understanding the FCC rules.. The dirt cheap modules probably aren't certified. And the certified modules are suspect depending on how they're packed into the final product

I would not worry about IRDA disappearing any time soon. It's widely used in metering equipment (electricity/gas meters, etc) and other industrial products. I'm working with such equipment and, as a requirement, newly designed units have IRDA estimated lifetime is about 30 years. So , as far as I am concerned, IRDA is here to stay. Kind Regards, Robert Csaba Molnar --- On Fri, 6/3/11, Charles Craft <RemoveMEchuckseaTakeThisOuTmindspring.com> wrote: From: Charles Craft <spamBeGonechuckseaspamBeGonemindspring.com> Subject: Re: [EE] IRDA here to stay? To: "Microcontroller discussion list - Public." <TakeThisOuTpiclistEraseMEspam_OUTmit.edu> Date: Friday, June 3, 2011, 9:58 PM On 6/3/2011 1:32 PM, Herbert Graf wrote: {Quote hidden}> On Fri, 2011-06-03 at 08:49 -0700, William "Chops" Westfield wrote: > >> On Jun 3, 2011, at 8:36 AM, Herbert Graf wrote: >> >> >>> I'd avoid designing something with IRDA and instead use wireless. >>> >> At like 5 to 10x the price and complexity!? >> > 5-10x? > > Wireless modules can be had for a few bucks each, I don't know what the > price of an IRDA module is, but I'll give you that on price. > > Complexity? It's pretty trivial to drive a wireless module, in fact the > algorithms used are pretty similar to what you'd do with an optical > connection, so complexity? The same IMHO. > > TTYL > > I think the complexity with wireless comes with understanding the FCC rules.. The dirt cheap modules probably aren't certified. And the certified modules are suspect depending on how they're packed into the final product

On Fri, Jun 3, 2011 at 2:58 PM, Charles Craft <RemoveMEchuckseaTakeThisOuTmindspring.com> wrote: > I think the complexity with wireless comes with understanding the FCC rules. > The dirt cheap modules probably aren't certified. > And the certified modules are suspect depending on how they're packed > into the final product. You're correct. The MRF24J40MA is one of the cheapest FCC certified modules, at around $8 a piece (last I checked) If you're going to sell 10,000 units you can go much cheaper by rolling your own. Time to market will be longer. -- Martin K

Oh don't be silly - I think the best solution would be to have a a small bird tied to the device along with a photo of an owl. When the link fails (presumably due to dust), a small linear actuator will raise the photo of the own and play a "hoot" sound on a tiny speaker. This will cause the bird to try to fly away - which will blow the dust off. MUCH simpler and more elegant solution. :) Sean On Fri, Jun 3, 2011 at 4:14 PM, John Gardner <goflo3EraseME.....gmail.com> wrote: > Sean - > > Will the budget allow a tiny servo/feather duster? > > Rube >

On 04/06/2011 22:10, Sean Breheny wrote: > Oh don't be silly - I think the best solution would be to have a a > small bird tied to the device along with a photo of an owl. When the > link fails (presumably due to dust), a small linear actuator will > raise the photo of the own and play a "hoot" sound on a tiny speaker. > This will cause the bird to try to fly away - which will blow the dust > off. MUCH simpler and more elegant solution.:) Brilliant! Heath and Rube would be most impressed. You would need to feed the bird of course, but that's a minor detail...

On 04/06/2011 23:20, Michael Watterson wrote: > On 04/06/2011 23:09, Oli Glaser wrote: >> > Brilliant! Heath and Rube would be most impressed. >> > You would need to feed the bird of course, but that's a minor detail.... > Can any avians digest cellulose? > Or keep a guinea pig. They can digest paper made from wood pulp. But they don't fly, which might impair the workings of the proposed device somewhat.. :-) > > Newspapers do them no harm as long as they don't read them before eating. > :-)

The guinea pig could be mounted to the device in such a fashion that it would "outgas in the general direction" of the sensor, thereby flushing it clear of all dust :) On Sat, Jun 4, 2011 at 6:32 PM, Oli Glaser <EraseMEoli.glasertalktalk.net> wrote: {Quote hidden}> On 04/06/2011 23:20, Michael Watterson wrote: >> On 04/06/2011 23:09, Oli Glaser wrote: >>> > Brilliant! Heath and Rube would be most impressed. >>> > You would need to feed the bird of course, but that's a minor detail... >> Can any avians digest cellulose? >> Or keep a guinea pig. They can digest paper made from wood pulp. > > But they don't fly, which might impair the workings of the proposed > device somewhat.. :-) > >> >> Newspapers do them no harm as long as they don't read them before eating.. >> > > :-) > > >

On Sat, Jun 4, 2011 at 6:09 PM, Oli Glaser <RemoveMEoli.glaserEraseMEEraseMEtalktalk.net> wrote: > On 04/06/2011 22:10, Sean Breheny wrote: >> Oh don't be silly - I think the best solution would be to have a a >> small bird tied to the device along with a photo of an owl. When the >> link fails (presumably due to dust), a small linear actuator will >> raise the photo of the own and play a "hoot" sound on a tiny speaker. >> This will cause the bird to try to fly away - which will blow the dust >> off. MUCH simpler and more elegant solution.:) > > Brilliant! Heath and Rube would be most impressed. > You would need to feed the bird of course, but that's a minor detail... And something to clean up the bird droppings as well. A yak hair brush, perhaps? -p

On 6/3/2011 8:56 AM, Charles Craft wrote: > On 6/3/2011 11:17 AM, Sean Breheny wrote: >> Hi all, >> >> I am about to design a product which will be in use for at least 10 >> years and needs a medium range (about 50cm max) free-space optical >> data link, low speed (under 9600 bps is fine - even 1200 bps would be >> acceptable). The product contains both sides of the link so it does >> not have to communicate with standard products. It needs to be robust >> to a fair amount of dust (up to a 1cm layer of paper dust) and be able >> to work in normal ambient room light (i.e., not outdoors). >> >> > Is that 1cm or 1mm of dust? > Did my wife send you the specs for my office? There is an almost-hidden defect in the IRDA module that is often overlooked: loss of IR emitter intensity over time. If you need 10+ years of life, IRDA is the wrong device unless it is used infrequently, like taking a meter reading, or checking status monthly, etc. Every time I tried to use an IRDA or a TV-style IR remote system, I was foiled by one thing or another. Mostly the problem was loss of distance in sunlight. Works to 50m at night but in daylight you are lucky to have one meter of distance. --Bob

On 6/3/2011 10:32 AM, Herbert Graf wrote: > On Fri, 2011-06-03 at 08:49 -0700, William "Chops" Westfield wrote: >> On Jun 3, 2011, at 8:36 AM, Herbert Graf wrote: >> >>> I'd avoid designing something with IRDA and instead use wireless. >> At like 5 to 10x the price and complexity!? > 5-10x? > > Wireless modules can be had for a few bucks each, I don't know what the > price of an IRDA module is, but I'll give you that on price. > > Complexity? It's pretty trivial to drive a wireless module, in fact the > algorithms used are pretty similar to what you'd do with an optical > connection, so complexity? The same IMHO. > > TTYL > Herbert, I agree completely. I found a 2GB-band wireless module controlled by I2C complete with SPI interface as well as a built-in antenna for $18 for a PAIR of modules. It has 154 channels, built-in encryption/decryption and IN/OUT FIFO (32 character) buffers. Range is supposed to be 100m with the built-in antenna. I'll let the list know how it goes when they get here from China. --Bob

Inductive coupling / "Near Field Communication" / NFC sounds close to ideal.. Can be used to power remote too if desired. They say 4 cm range here but range can be metres if desired. http://en.wikipedia.org/wiki/Near_field_communication > several of these devices could be in close proximity to each other Depending on the value of "close", 50 cm is getting a bit far if you have multiple units but if you can get the two closer together and reduce available range then you can use N in closish physical proximity. Dust no problem. Probably no certification issues either (YMMV). Potentially low cost and simple. Capacitive coupling also worth thinking about. Russel

Hi Bob, Interesting. I would guess that this issue of life is due to overdriving the IR emitter with pulses of high current? or are you saying that IR LEDs fail early even when operated at low to moderate current? Do you know where I could find more info on this? I may run into the same problem with the "roll-your-own" solution. I don't have to deal with direct sunlight as this is indoors only and the direction of the transmission is vertical and the detector is facing downward and covering the transmitter. Sean On Sun, Jun 5, 2011 at 6:35 PM, Bob Axtell <RemoveMEengineerspam_OUTKILLspamcotse.net> wrote: {Quote hidden}> There is an almost-hidden defect in the IRDA module that is often > overlooked: loss of > IR emitter intensity over time. If you need 10+ years of life, IRDA is > the wrong device unless it is > used infrequently, like taking a meter reading, or checking status > monthly, etc. > > Every time I tried to use an IRDA or a TV-style IR remote system, I was > foiled by one thing > or another. Mostly the problem was loss of distance in sunlight. Works > to 50m at night but > in daylight you are lucky to have one meter of distance. > > --Bob A >

Interesting idea, Russell, but at this point time is quite tight and I don't think I could switch to NFC unless I were to use an off-the-shelf solution. I already have an IR-based proto working. Sean On Sun, Jun 5, 2011 at 8:59 PM, RussellMc <RemoveMEapptechnzTakeThisOuTspamgmail.com> wrote: {Quote hidden}> Inductive coupling / "Near Field Communication" / NFC sounds close to ideal. > Can be used to power remote too if desired. > They say 4 cm range here but range can be metres if desired. > > http://en.wikipedia.org/wiki/Near_field_communication > >> several of these devices could be in close proximity to each other > > Depending on the value of "close", 50 cm is getting a bit far if you > have multiple units but if you can get the two closer together and > reduce available range then you can use N in closish physical > proximity. Dust no problem. Probably no certification issues either > (YMMV). Potentially low cost and simple. > > Capacitive coupling also worth thinking about. > > > Russell >