non jp1 SA 3250hd question

hyper · Post by **hyper** » Sat Jun 02, 2007 2:37 pm

I came across this site when looking for a solution to a problem I am having so here goes.

I assist a man who uses a speech computer that has IR capabilities to control tv's stereo's vcr's and such. The device he uses is a dynavox 3100. Company website is here. http://www.dynavoxtech.com/

I have programmed his LG LCD tv, Sirius Radio, DVD player, Sony 300 disc changer into it so it works with those Items.

The dynavox is supposed to capture the ir signal from the factory remote and store each signal.

He just received a Scientific Atlanta 3250 HD cable box and it does not work with it. I am at a loss as to why. the cable company cannot help because they have no clue why it doesnt function.

this device has the ability to import .ir control files, but i wouldnt know how to create one to control the cable box

any suggestions on how to get this box to work with his dynavox? or what is needed to write a .ir file. like I said it works with every other ir signal except that cable box.

thanks

johnsfine · Post by **johnsfine** » Sat Jun 02, 2007 3:56 pm

I expect the modulation frequency is the problem. The Scientific Atlanta uses a higher modulation frequency than the other devices you mentioned.

Hopefully the dynavox can send higher frequency even though it apparently can't learn higher frequency.

Do you have info or examples of the format it can import?

Post by **The Robman** » Sat Jun 02, 2007 4:21 pm

If you can get ahold of the IR files that the dynavox uses, you should zip them together and post the zip file here in the Diagnosis Area folder of the File Section. If the names of the files doesn't clearly indicate the devices that they control, please tell us what device each file controls. With this info, we might be to reverse engineer the format and therefore, we might be able to generate a new file for the SA box.

hyper · Post by **hyper** » Sat Jun 02, 2007 5:29 pm

Thank you for the response.

I can export his current .ir files for you and upload them tomorrow, when i go back to his house. that would be deeply appreciated if you could take a look at them.

Thanks again!

hyper · Post by **hyper** » Sun Jun 03, 2007 12:05 pm

I have uploaded the current ir files to the diagnosis area. file name is dynavox.zip

https://www.hifi-remote.com/forums/dload ... le_id=4622

thanks

the tv.ir commands are the same type of commands i would need for the cable box. ie: power on, power off, vol up, vol down, mute, channels 0-9 etc. if possible

Post by **The Robman** » Sun Jun 03, 2007 12:52 pm

Just FYI, there are 4 "control" files in the zip...

tv.ir - controls an LG lcd tv - file -NEC1, device 4 (38.1 kHz)
stereo.ir - controls a bose waveradio - file - custom exec (38.3 kHz)
cd.ir - controls a sony 300 disc cd jukebox - file - Sony12, device 17 (40 kHz)
sirius.ir - controls an audiovox sirius radio - file - NEC1, device 16 (38.1 kHz)

Just as a reminder, the OP is looking for a new file for a Scientific Atlanta Explorer 3250HD set top box, which probably uses this upgrade - Panasonic (old), device 27 (57.1 kHz)

johnsfine · Post by **johnsfine** » Sun Jun 03, 2007 12:57 pm

The format isn't obvious. Reverse engineering it might be quite hard.

Can you ask dynavoxtech about the modulation frequency of the ir signal?

Post by **The Robman** » Sun Jun 03, 2007 2:36 pm

I've extractedthe hex code for each of the TV buttons (NEC1, dev 4) and dropped it into a spreadsheet, just in case this helps...

https://www.hifi-remote.com/forums/dload ... le_id=4623

johnsfine · Post by **johnsfine** » Sun Jun 03, 2007 3:52 pm

Let me be a little clearer:

The signal structural in those files is pretty obvious. If the objective were just to recognize signals, I could do that. But given the guess that modulation frequency is the problem, the part I understand doesn't help. It doesn't tell me how to fix the frequency or whether that is even possible.

Here is the beginning of the Channel+ signal from that tv.ir file:

01 99 FF FF 03 80 20 00 00 00 00 00 00 00 00 00 00 C0 FF FF FF FF FF 81 0F 7C E0 FF 1F F8 C0 07 3E F0 C1 0F FE FF 83 FF FF E0 07 FF FF C1 FF 7F F0 FF 1F FC FF 07 FF FF C1 0F 7E F0 83 1F FC E0 07 3F F8 C1 FF 7F F0 FF 1F FC FF 07 FF FF C1 FF 7F F0 FF 1F FC FF 07 FF FF C1 FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF

I don't know what the "01 99 FF FF 03 80 20" part means. If I did, that might be what is needed to fix the frequency.

The rest is easy to understand: Starting with the long run of 00, you can construct the IR signal by:
1) Convert each hex byte to binary
2) Reverse the bit sequence in each binary byte (example C0 hex is 11000000 binary, so it reverses 00000011).
3) String all that binary together
4) Count runs of 1's and runs of 0's.
The 1's represent signal OFF and the 0's represent signal ON. The time scale is something around 100 microseconds per bit.

So the "00 00 00 00 00 00 00 00 00 00 C0" has 86 0 bits in a row, representing 8600 microseconds ON.

The "C0 FF FF FF FF FF 81" has 43 1 bits in a row, representing 4300 microseconds OFF.

Continuing from there, you can see the bit pattern of the NEC1 signal.

Constructing that aspect of the Scientific Atlanta signal wouldn't be too hard. But if that would have solved the problem, simply learning the signal would have worked.

I don't know how to encode the modulation frequency or if you even could.

Post by **The Robman** » Sun Jun 03, 2007 4:48 pm

johnsfine wrote:I don't know what the "01 99 FF FF 03 80 20" part means.

The "01 99" is the length. The "FF FF 03 80 20" part is there for every button in every file, so I don't think the carrier freq is in there either (as the Sony signals are 40k versus the NEC signals which are 38k).

johnsfine · Post by **johnsfine** » Sun Jun 03, 2007 7:57 pm

The Robman wrote:The "01 99" is the length.

OK.

The Robman wrote:The "FF FF 03 80 20" part is there for every button in every file, so I don't think the carrier freq is in there either (as the Sony signals are 40k versus the NEC signals which are 38k).

Maybe. But if the learning system could get the frequency right, what else could keep the Scientific Atlanta signals from working? Except for frequency, the Panasonic Old protocol is about as generic and easy to learn as a protocol could get.

But if the transmit hardware can handle multiple frequencies while the learning hardware can't, you'd expect a generic frequency to be set on every learned signal.

Sony signals should be 40k, but they'll work just file at 38k. Panasonic Old won't work at 38k.

All that is layering guess upon guess. But what else do we have to go on?

Hyper, if you can't get better info from dynavoxtech, is there any way to use a learning remote or captureir to get the signals it sends? We might want to try tweaking the unknown bytes in a known file and if that doesn't entirely break the file (some checksum etc. or some other kind of count making a tweaked file invalid) we could see how it affects the generated signal and thus deduce the meaning and maybe find the way to control frequency.

binky123 · Post by **binky123** » Sun Jun 03, 2007 8:10 pm

Maybe 03 80 20 is BCD format for 38,020?

Post by **The Robman** » Sun Jun 03, 2007 9:37 pm

binky123 wrote:Maybe 03 80 20 is BCD format for 38,020?

I was thinking that too, until I saw that they used it for Sony as well, but if John's theory about it using the same freq for everything is right, that could well be the answer.

hyper · Post by **hyper** » Mon Jun 04, 2007 2:36 pm

I called Dynavox Tech support, they do not have the frequency information,

However, they will contact their engineering dept.

I have forwarded this thread to dynavox, so maybe the engineering dept. can give you the information needed.

johnsfine wrote:
The Robman wrote:The "01 99" is the length.
Hyper, if you can't get better info from dynavoxtech, is there any way to use a learning remote or captureir to get the signals it sends? We might want to try tweaking the unknown bytes in a known file and if that doesn't entirely break the file (some checksum etc. or some other kind of count making a tweaked file invalid) we could see how it affects the generated signal and thus deduce the meaning and maybe find the way to control frequency.

hyper · Post by **hyper** » Tue Jun 05, 2007 11:47 pm

I havent heard from them, but I think I found my own answer to the problem.

I believe the dynavox is set on 40k band

40K Standard Band
The industry standard carrier frequency of IR remote controls is between 32 and 44 kHz. We call this 40K---the standard band

and the cable box is set on 56k standard band

56K Standard Band
Standard Band designed to work with RCA-DSS, General Instrument, Scientific Atlanta and ProScan equipment, as well as many new Digital Cable and Satellite Receiver boxes.

With my limited knowledge, I would imagine there wouldnt be anyway around this.

Thanks again for trying