I am attempting to build my own universal remote. Does anyone know where i can find some good literature on this, preferably online. Basicly want to make a multi-function ( 8 device ) remote, upgradeable via USB. Any Help would be greatly apreciated.

Can you tell me why you would want to build your own?

Do you mean build your own with actual plastic and tools, or build your own program for your remote on a PC?

To Build your own...

You'll need a good handle on electrical engineering, as this is really a microcontroller, with sophisticated firmware. You will need to find hardware to receive and transmit IR. There are a number of IR transceivers available in hardware. You'll need to be aware that most remotes increase the power level pretty radically from what the typical IR transmitter does. When you find data sheets on the transceivers online, you can find if they have any application notes on increasing the power output levels. You'll need to design some circuit boards, choose mechanical components, layout the PCB, and I'd suggest designing a plastic enclosure.

After picking a micro (there are a number available which support USB, and will have application notes available), designing hardware for the IR transceiver, and choosing a layout for a keypad, you will need to start programming. If you are up on programming micros in assembly language, you have nothing needed but to write a bunch of code.

You can probably get the first unit working, if you have a lab available, and the micro support tools, for a layout of only several thousand dollars, for parts, proto pcb's, proto enclosure, etc.

Or, you could just go buy what you want for about $50.

If you don't understand most of what I've laid out, and you want to ask questions like what micro should I choose....

I do Consulting Engineering work, at a rate of $160 an hour. We can draw up a scope of work, and agree on a rough estimate of time. You would of course be responsible for all materials. If you are serious, post a short message here, and I'll be glad to send you a copy of my resume.

Ok so i never thought it would be quite so expensive to work out the micro circuitry.

what is the best 8 device remote that can be updated via usb? Or some sort of pc link. I want to make my own interface, the guts can be of any model universal. Preferably something that has functions for todays technology. ie dvdrw, dvd, sat, cable, tv, vcr, surround sound systems...

any ideas?

All I can say is....WOW!!

If you provide a bit more info on what you're really trying to accomplish, I probably can provide much better info.

I don't think you need much electrical engineering expertise to design a remote control. Almost everything happens inside a single chip micro. Connecting an IR LED to the micro is easy. Connecting buttons is easy. Providing battery power isn't very hard. The programming interface is a bit more, but there are lots of online examples of programming interfaces for various micros.

Of course, most of the parts you'll want are normally sold and used in bulk. Once you pay extra for shipping and for low quantity, you have probably paid as much for parts as for a fully constructed remote, and you still have a difficult job to solder it all together. Depending on why you want to do all this, you're probably better off starting with a remote and modifying it.

The firmware to run it all is probably even harder (It wouldn't be harder for me, because I'm expert in both IR protocols, and assembler programming, and I'm a klutz at soldering, so I would find programming a remote from scratch much easier than building it. But for most people the programming would be much harder).

How much do you need to customize the remote?

A JP1 remote, plus JP1 cable can be significantly customized using free JP1 software without needing to write any asm code nor even understand the IR signals it will send. The best JP1 choice may be the urc-8910 from Rob at: [Link: hifi-remote.com]
It says they are sold out at the moment, but I think he gets more fairly often.

A JP1.2 remote plus JP1.2 cable gives similar capability for only a few dollars more. The best JP1.2 choice is the 8820. Rob has a good price on those as well, but (unlike an 8910) if you shop around you might do better than Rob's price for the 8820.

If you want to rewrite the firmware of a remote (if you need to customize more than JP1 or JP1.2 will allow), that is impossible with JP1 remotes, but very practical with JP1.2 remotes.

There is a BDM (background debug mode) interface that lets a PC program, control and debug the Freescale micro inside a JP1.2 through a USB interface. The firmware in a JP1.2 remote is protected so the BDM interface is disabled for all commands other than erasing the firmware. Once the firmware is erased, the BDM interface works to load any new programming you like. But you can't make small changes. Either use it as is (via JP1.2 cable, not BDM) or start over from scratch via BDM.

There are several cheap online designs for BDM USB interface. I couldn't figure them out, so I bought a slightly more expensive one included in a Freescale evaluation board from mouser.com.

The Freescale evaluation board itself may be a good starting point, instead of an 8820 if you want to make a remote control almost from scratch. But it includes an inferior processor to the one in the 8820 and it doesn't have a keyboard, so I think the 8820 is a better choice. I bought the evaluation board only because the evaluation board including USB BDM interface cost less than any pre-built USB BDM interface I could find.

The JP1 cables exist in three versions:
Parallel port: About $15 or easy low cost build-it-yourself.
USB: About $35
Serial port: Hard to find but a build-it-yourself design isn't too hard. It works only direct to an ordinary serial port, NOT through a USB to serial adapter.

The JP1.2 cable is only serial, about $20. It works fine through a USB to serial adapter (but that is another cost).

Connecting is easy, just a couple easy solders and you are there. Just remember that the micro you've chosen won't be able to drive that LED directly! Even if you've chosen a micro that provides 5V drive at 5mA, it's not going to get the job done. At the very least the micro will drive a transistor, and the transistor will drive the LED.

Note that this gets you only IR output. How do you learn the IR codes? That would require an IR receiver. I'm not aware of any micro that supports IR receive, and the USB interface in a single IC. So that will be a seperate IR receiver. The IR receiver will include AGC circuitry so that the level of IR in the room does not cause the remote to be unable to learn. There goes your statement that almost everything happens inside a single chip micro.


This is an easy task for a EE, but how about the DIY'er. Probably want 60 buttons, so are they going to need a micro with 60 I/O pins? Only if they want to pay a whole lot for a micro. It will require using keyboard matrix techniques. Using a keyboard matrix, I can provide a 64 button remote using only 16 I/O. Depending on the requirements (does N-key rollover need to be recognized) it could be done in less.


Huh! I guess my job where I provide circuitry for just such devices to regulate voltage isn't really needed? I guess that's why we have so many customers lining up for my help. Or perhaps they have bosses like you, "Ahhh, that's easy". Note that most micro's won't be able to run on unregulated voltage, and even if they would, the noise injection from when the IR LED's are turned on will likely cause only a few bits to flip every now and then!


It won't probably be more expensive... It will be more expensive! I'd make a very large wager on that any day!

If you have the electronics skills to design the circuitry, the soldering will be the least of your worries! I don't think anyone with the skills to accomplish this would put difficulty soldering as anywhere among this risks in this DIY project!

Edit: I originally posted some fairly stupid comments about directly connecting the minus side of the IR LED to the micro, because the I/O pin can pull low more current than high. Generally it is better to switch the minus side of the LED and use constant connection to power on the plus side, rather than switch the plus side. But after giving it more thought, I think leaving out that transistor is a lame idea. I expect the designs that do so limit the range seriously.


Don't. Who said all universal remotes need learning. Get the codes online or through some other IR capture device and program them into the remote via the PC interface.



Yes. Trivial electrical design. The rest is mechanical and soldering.


Now you're just trying to make it hard. Who puts N-key rollover in a remote.


But most low cost universal remotes on the market run the micro and the IR LED directly off batteries with just a cap. It works. Just select one of the same micros they use, such as the Freescale one I mentioned.


I was assuming a lot of cut corners, especially on the mechanicals. I didn't mean to say you could duplicate a low cost universal remote for parts cost less than that remote. I meant you could duplicate the basic electronic parts (micro, LED, etc. ) assuming you kludge the mechanicals.


I am quite sure that for ME the soldering would be the hard part and the major failure risk. Neither electrical design nor firmware would be that hard for me. I didn't say that soldering would be the hardest part for a more typical DIYer.

But anyway, I don't disagree on the main conclusion from your post that is modifying an existing remote makes much more sense than building from scratch.

You'll find that good designs use more than a single transistor. They use multiple transistors, configured to provide a constant current source, that will drive the IR LED at the full rated power over the entire voltage range of the cells.


A proprietary universal remote, like what is being proposed with a need for an external capture device doesn't make the remote any easier to design, because now a capture device must be designed. And a proprietary remote design that requires finding codes online will be a real pain. I couldn't find codes for at least 10% of the items I have, and suspect many people would find even higher incidence.


Trivial, just like hooking up the IR LED. Wait a second, you didn't do too well on that one...


Which Freescale micro is that? (You don't actually give a part number for the micro, you give the part number for the remote.) The Freescale micros that I know of, fall into two categories. (Limiting this to Flash Based, as ROM based is worthless for this purpose.) The first category includes several micros that include a voltage regulator on the IC. This is not running the micro directly off the battery. The second category consists of a number of micros with narrow input voltage ranges, or varying voltages. Any of which that I'm aware of would require an external voltage regulator IC.


Let me be even clearer, I'll take a bet, that you can't produce a working electrical design (without mechanicals) for less than the cost of a low cost universal remote. I think I'll win this handily, as any design will need a printed circuit board, and even if you etch your own, you'll spend more on the chemicals and photoresist, than the money for the JP-1's you point out.

Then you won't find many "good designs" in the most commonly sold universal remotes. Zero transistors or one doing that job in every one I've seen the inside of.


My capture device is just an IR sensor (I have experimented with several different ones, costing at most a few dollars. All have worked easily and well) connected to a PC printer port. It needed zero of what I would call "design" and even the soldering was easy enough for me (though for most of them, I wimped out and used solderless breadboard). But I'm still not convinced eclipse_storm needs one at all (he hasn't told us).


I believe the 8820 uses the MC9S08RC60FG. If that is not correct, I'm at least sure it is one of the MC9S08R?60?? versions.


Why?

Anyway, I'm not planning to take the time to shop around for those parts nor further argue arbitrary rules with you. You might as well assume you won your bet, because you clearly would assume that regardless.


I didn't pretend to understand what kind of universal remote eclipse_storm was considering designing/building. I asked. You, made a bunch of assumptions about requirements and design constraints without knowing whether those fit the purpose of the project. I don't assume learning is required. I'm not even sure a significant button matrix is required. I expect there should be some significant DIFFERENCE from a typical universal remote, or he wouldn't want to build his own. If enough of those unknowns fall in the direction that starting from parts makes sense and if you value labor at zero dollars (as DIYers sometimes do), I still think the parts cost from scratch could come in under the cost of something like an 8820 (also assuming that some kind of BDM interface is required either way, rather than a JP1.2 cable being enough for the 8820).

If eclipse_storm is still reading this thread and still interested, I'd be glad to provide more details about many of the things I said above, after he provides more info about what he is trying to accomplish.

This is a funny thread, it's too bad the original poster hasn't chipped in with what he really wanted to do ... but nevertheless I thought I'd contribute.

As for driving the IR led, of course it's better to use an explicit drive circuit, but if you had to kludge keep in mind that if you're bit banging you don't have to use just a single output from the CPU. You could certainly gang together multiple outputs to increase the current sinking/sourcing ability.

For what it's worth I think John already won the bet with his proposal of just reprogramming a JP-1 remote using the BDM interface.

I guess its worth reading the spec sheet of the chip to see if that would work. I don't have time to dig up details, but I recall reading some of those spec sheets in which the limit of total current for all the I/O pins in the chip wasn't much higher than the limit for just one. So with those spec's ganging together wouldn't help much.


Thanks for the support, but I don't think I actually made a bet, and I think the proposed bet was strictly regarding parts cost if one DIDN'T start with a pre built remote.

Making reasonable guesses about the desired project, my best guess is that the best plan would be to use an 8820 and a BDM interface. So the discussion of parts cost and related design details was probably a meaningless sidetrack.

I'm still curious about what the desired project really is.

Well... the bet diverged quite a bit from you must be an EE and if you'd like to hire me at $160/hr I'll draw up a plan, but the tooling costs alone will be thousands...

The brilliance of DIYers is coming up with unconventional solutions that can be done on a shoestring. Once upon a time all you needed to start building your own homebrew computer was a bag of parts, a circuit board, an assembly drawing, and willingness to learn. Or even a perf board and some wire-wrap if you wanted to make your own design.

Surface mount parts make everything trickier but some people are better at soldering then others. If you just had to hook up power, an IR sensor, and a handful of discretes - I don't see why that couldn't be hand soldered.

The OP may very well be in over his head, but if he has the will, by all means there's always a way.

I can't even solder wires onto a DIP at ten pins per inch. But I've seen someone double sided tape a very tiny 40 pin surface mount package onto a perf board then quickly and perfectly solder nearly 40 tiny wires onto that package and the other ends onto thicker wire and components mounted in the perf board. It was done with a magnifying system so powerful even I could see what was happening, but with a simple hand held soldering iron with a tip so much bigger than the pin spacing that it seemed imposible that he could control which pin he was soldering, much less do each pin well.


Probably.


I don't think will makes up for lack of skill. But of course I have no idea what skills he has.

Other than asm programming skill needed (depending on how complex a remote you want) the path of 8820 plus BDM doesn't need much skill (electrical, mechanical, etc.)