Hi folks -
I have an NEC TV from the Truman administration that always gives me problems with learning remotes. I am replacing a ProntoNEO with a TSU3000 and can't get it to work with the TV codes. With the NEO, I vaguely remember that I had to
a) learn the codes because the codeset wasn't covered in the database
b) not use the original remote (which I have), but instead learn from an equally old OneForAll remote that DOES have the codeset in its database.

These tricks aren't working this time, and in addition I have the third option of learning from the NEO. Obviously no happiness there either.

But wait, I exaggerated - I successfully got it to learn the "mute" button, so that should provide some clues as to the protocol involved, yes? Because this relic (which actually has a pretty good picture) is just a monitor, I only need power, mute, and volume controls. Here are the codes that I have loaded, although each time it comes out slightly differently:

Mute (which works):
0000 0065 0026 0000 0170 00BB 0016 0016 0016 0016 0016 0046 0016 0016 0016 0016 0016 0016 0016 0016 0016 0016 0016 0046 0016 0046 0016 0016 0016 0046 0016 0046 0016 0046 0016 0046 0016 0046 0016 0016 0016 0046 0016 0016 0016 0046 0016 0016 0016 0016 0016 0016 0016 0016 0016 0046 0016 0016 0016 0046 0016 0016 0016 0046 0016 0046 0016 0046 0016 0046 0016 0678 0170 005E 0016 0F75 0170 005E 0016 00BB

Power (which doesn't):
0000 0066 0026 0000 016C 00B5 0017 0017 0017 0017 0017 0044 0017 0017 0017 0017 0017 0017 0017 0017 0017 0017 0017 0044 0017 0044 0017 0017 0017 0044 0017 0044 0017 0044 0017 0044 0017 0044 0017 0017 0017 0017 0017 0017 0017 0044 0017 0017 0017 0017 0017 0017 0017 0017 0017 0044 0017 0044 0017 0044 0017 0017 0017 0044 0017 0044 0017 0044 0017 0044 0017 064F 016C 005B 0017 0F49 016C 005B 0017 00B5

Volume Up (ditto):
0000 0067 0026 0000 0168 00B6 0016 0016 0016 0016 0016 0045 0016 0016 0016 0016 0016 0016 0016 0016 0016 0016 0016 0045 0016 0045 0016 0016 0016 0045 0016 0045 0016 0045 0016 0045 0016 0045 0016 0016 0016 0045 0016 0016 0016 0016 0016 0016 0016 0016 0016 0016 0016 0016 0016 0045 0016 0016 0016 0045 0016 0045 0016 0045 0016 0045 0016 0045 0016 0045 0016 0654 0168 005C 0016 0F1E 0168 005C 0016 00B6

Volume Down - nearly the same as above:
0000 0067 0026 0000 0168 00B6 0016 0016 0016 0016 0016 0045 0016 0016 0016 0016 0016 0016 0016 0016 0016 0016 0016 0045 0016 0045 0016 0016 0016 0045 0016 0045 0016 0045 0016 0045 0016 0045 0016 0045 0016 0045 0016 0016 0016 0016 0016 0016 0016 0016 0016 0016 0016 0016 0016 0016 0016 0016 0016 0045 0016 0045 0016 0045 0016 0045 0016 0045 0016 0045 0016 0654 0168 005C 0016 0F1E 0168 005C 0016 00B6

As another exciting data point (whee!), if I open up my old ProntoNEO configuration file with something I have lying around called NeoHacker, I can see codes in there. Note that these all work, although they may be dirty:
Power, I think:
0 f 6c 4 5b ab 16 44 5b b5 3f 5f fc 42 c 40 0 0 40 4 44 44 4 4 4 0 40 40 40 44 4 bf 47 fc 42 8 bf ab bc f2 b 21 7c 3

Volume Up:
0 f 6c 4 5b ab 16 44 5b b5 3f 5f fc 42 c 40 0 0 40 4 44 44 4 4 0 0 40 40 44 44 4 bf 47 fc 42 8 bf ab bc f2 b 21 7c 3

Volume Down - interesting that these are also nearly identical:
0 f 6c 4 5b ab 16 44 5b b5 3f 5f fc 42 c 40 0 0 40 4 44 44 44 4 0 0 0 40 44 44 4 bf 47 fc 42 8 bf ab bc f2 b 21 7c 3

I don't seem to have mute in this file, so perhaps it is a button that works from one of the standard codesets. I wasn't testing with that one on the TSU3000.

Enough detail? Thanks for reading this far, and any help you can give.

It may be informative to see the Pronto Hex from learning from the original remote, to see how that compares to learning from the OFA.

I assume the OFA correctly controls the TV.

It would help to know which model of old OFA, and especially which setup code.


All those samples are fairly clean examples of NEC protocol, device number 4. They are all at higher than the nominal frequency for NEC protocol. I'm not sure whether they're supposed to be at that higher frequency. But mute is slightly higher frequency than the others.

They also all look like they were learned as the same length of short press of the original remote. Most protocols learn better as a long press, so the Pronto can detect the repeat pattern. But if the Pronto does detect the repeat pattern for these, it would probably recognise the whole signal as NEC and store it in the condensed form. Then it would ignore the stored frequency and transmit at the nominal frequency for NEC, which is probably wrong.

But the condensed form is worth trying anyway. Here are your four signals in condensed Pronto Hex:

900A 0065 0000 0001 04FB 0AF5

900A 0065 0000 0001 04FB 08F7

900A 0065 0000 0001 04FB 02FD

900A 0065 0000 0001 04FB 03FC

Regarding the short press issue I mentioned before, what do the vol+ and vol- keys on the original remote and/or the OFA remote do if you hold them pressed a long time? Do they continuosly change the volume? Or do you need a new press for each volume step? If they continuously change the volume, then the Pronto would need the right repeat pattern in order to duplicate that behavior. After we figure out what detail of the difference between mute and the others you posted is reponsible for mute working, we still may need to fix the repeat pattern and disguise the result so ProntoEdit won't recognise and condense it (assuming the condensed version doesn't work).

This message was edited by johnsfine on 04/20/05 08:01 ET.

Thanks, John. I tried the condensed codes that you posted, but no luck. However, I have more answers to your questions...


Power: 900A 0068 0000 0001 04FB 08F7
Mute: 900A 006D 0000 0001 04FB 0AF5
Vol Up: 900A 006D 0000 0001 04FB 02FD
Vol Down: 900A 0068 0000 0001 04FB 03FC

Using a longer button press on the original remote, the Pronto appears to have recognized the pattern and condensed it. As you stated later in your message, the frequency may then be wrong, and in fact these still don't work.


URC-4063, setup code 030


Yes, they do on both the original and the OFA.

Thanks so much for your help.

This is quite strange.

Setup code 030 in an OFA remote is NEC protocol (device 4 with same command numbers as the Pronto Hex in this thread) and it is at the nominal NEC frequency.

The condensed form Mute and Vol Up commands you just posted perfectly match the Mute and Vol Up from setup code 030. I'm having trouble even guessing a reason what those condensed Pronto Hex values might work and the Pronto fail.

As I said before, I think the Pronto ignores the frequency value stored in the condensed form, so the Power and Vol Down you just posted look like they were learned with a higher frequency, but I think they would transmit the same as setup code 030.

An NG Pronto isn't usually very good at learning the right frequency. But it isn't usually this inconsistent. I'm pretty sure your original remote and OFA remote are transmitting a fairly stable frequencies. But the learns you've done are a widely different frequencies. I'm not sure what makes an NG Pronto act even worse than typical. Maybe low batteries?

OFA's setup code 030 is one of the ones included in the 3000's universal database. The codeset numbers in the 3000's database are 27 higher than the OFA numbers. So setup code 030 is codeset 0057 in the 3000. But I don't expect using that would give different results from using the condensed codes.

This message was edited by johnsfine on 04/20/05 12:21 ET.

Well, glad to see I didn't just lob up another easy problem for you. Given your information that OFA code 030 (which worked fine on the old OFA itself) is represented in the Pronto database, I focused on that one. I found that if I stand way to the side of the TV's IR receptor, I can get the buttons to work ~50% of the time using that built-in code. My suspicion is that the frequency is in fact the problem, and that by increasing the duration of each pulse (with the angle change) I got results.

Is there some code generation application you can point me to that I can use to try out different frequency variations on the 0057 codes? If not, it would be great if you could dump out some samples for me to try.

Another approach I could try is to launder the IR signals through an IR repeater, which I was considering doing anyway for other reasons. Might that have an effect, or are they too good at mimicking the original signal?

Thanks.

What sort of batteries does the Pronto have? I still think that might be the problem.

How does that Pronto behave with other devices? It sounds like the Pronto has some problem generating IR signals (too weak or coming out at a strange angle). But that should have symptoms with other devices.

Does the room have a white ceiling? Lots of aiming problems are fixed by bouncing the signal off a white ceiling. Aim at the ceiling around halfway between you and the TV. There are many cases in which that works surprisingly better than aiming at the device itself.

Do you have a camera or camcorder with an LCD viewfinder? Those generally "see" IR signals quite well. Try looking at the IR signal through the viewfinder, and see if something is visible wrong with its angle, or if it is significantly weaker than some remote that works better.

My MakeHex program is available at
[Link: john.fine.home.comcast.net]

That includes a file nec1_40.irp that can be used to generate the higher frequency signals that some of your samples learned as. It can also be used to experiment with other frequencies and timing values.

Read the readme file included in that .zip.
Edit the device number to 4.
Your four commands are number 8, 10, 2 and 3.

The "Form" part of that .irp file is disguised enough that ProntoEdit should be unable to recognise and condense the signal.

I don't know which meaning of "pulse" you were using above, nor why you think a different angle would change the duration of either meaning of pulse. But editing the .irp file lets you change the frequency or the timing.

I can't believe the TV is this picky about frequency or timing. But the one sample of Pronto Hex you said worked had a frequency of 41000 and a TimeBase of 560.

This message was edited by johnsfine on 04/21/05 11:29 ET.

You're the man! A frequency of 40000 and timebase of 560 worked flawlessly. If you're ever in the San Francisco area, I owe you a beer.

Not sure if this is interesting, but ProntoEdit would alter the codes generated by makehex after I pasted them in. For instance, Power (function 8) would be pasted in as
0000 0068 0024 0002 0165 00B3 0016 0017 0016 0017 0016 0043 0016 0017 0016 0017 0016 0017 0016 0017 0016 0017 0016 0043 0016 0043 0016 0017 0016 0043 0016 0043 0016 0043 0016 0043 0016 0043 0016 0017 0016 0017 0016 0017 0016 0043 0016 0017 0016 0017 0016 0017 0016 0017 0016 0043 0016 0043 0016 0043 0016 0017 0016 0043 0016 0043 0016 0043 0016 0043 0016 06CD 0165 0059 0016 0F2C 0165 0059 0016 0F2C

but end up (working nicely) as
0000 0067 0024 0002 0168 00B4 0016 0017 0016 0017 0016 0043 0016 0017 0016 0017 0016 0017 0016 0017 0016 0017 0016 0043 0016 0043 0016 0017 0016 0043 0016 0043 0016 0043 0016 0043 0016 0043 0016 0017 0016 0017 0016 0017 0016 0043 0016 0017 0016 0017 0016 0017 0016 0017 0016 0043 0016 0043 0016 0043 0016 0017 0016 0043 0016 0043 0016 0043 0016 0043 0016 06DB 0168 0059 0016 0F1E 0168 0059 0016 0F1E

This message was edited by Thomas Smith on 04/22/05 11:48 ET.