Learned code doesn't work?

seanrich · Post by **seanrich** » Mon Dec 01, 2003 6:33 pm

I am new to this JP-1 thing - I have an OFA 8810 I am trying to use with a Zenith TV. The 'Quit' button to exit the menu doesn't work with the setup code I'm using (all other buttons work), so I tried to learn the code to the button. Using the 'L1' button in the learning process, I checked to make sure it works - it does (by the way I checked this site and previously people had problems because they didn't hold down the button to be learned - I did hold it down until getting confirmation from the OFA that learning was complete). Then I download into IR, and the code is listed as Zenith protocol, Device 7, Sub-device 0, EFC 193, OBC 126, Hex 7E, but there is a second listing that says protocol: Async15:554-588:FF.FD..FF.FD, Device 255, Subdevice 253, EFC blank, Hex blank, OBC 255 and a long string of hex in the misc column. I tried using the EFC 193 code in KM and loading that into the remote, but it doesn't work properly. I also looked up the EFC code in the advanced codes page on this site - all codes worked except for this button which behaved wrong in the same way that the 193 code misbehaved (acted as 'cursor down' vice 'exit'). Any ideas why this EFC that I learned from the OEM remote doesn't work and why there are two commands listed under IR? Thanks for any advice, and sorry for the lengthy explanation.[/quote]

johnsfine · Post by **johnsfine** » Mon Dec 01, 2003 7:34 pm

The "Async" decode is there because my decoder routines aren't perfect. Sorry. When you get multiple unrelated decodes, almost always only one of them is right and the other(s) should be ignored. Usually it's obvious (to a human, not to a computer program) which is right and which should be ignored.

The Zenith protocol doesn't really fit the basic structure enforced by the connection between the decoder and the several programs that use the decoder. That connection allows for a "device" and a "subdevice" which Zenith doesn't have. That connection doesn't allow for a "length" (which the 7 really is) nor for the thing that 0 really is. So we call them "device" and "subdevice".

I forget what KM calls the "length" and that other thing, but if you don't put the 7 and the 0 in the right places it won't work.

Why did you need KM at all? If you have a setup code that nearly works, you can just do KeyMoves within that setup code. What do 193 and the other problem EFC do in that setup code? I don't remember any issues in EFC's for Zenith so I can't imagine why they wouldn't work, but I also don't understand why a KM upgrade would be close enough to send a command at all but not close enough to send the right command.

If there is an EFC issue I'm forgetting then it's worth trying KM in OBC mode rather than EFC mode and try the decoded OBC.

seanrich · Post by **seanrich** » Mon Dec 01, 2003 8:31 pm

I figured the second code was superfluous, which is why I didn't worry about it.

I noticed you mentioned that 7 was the length, but under KM, the Zenith protocol lists a default length of 6. I tried changing it to 7, but then nothing works.

I don't particularly NEED to use KM for this - I could just send the learned code to the 'Quit' button and everything would work. It's more that I'm trying to understand the process of learning.

I tried learning some other buttons that I knew the EFC's for, but for some reason the EFC's that IR lists from the learned buttons are all exactly 4 less than the correct one ('0' button is 016, but IR lists it as 012). I tried adding 4 to the learned 'Quit' button EFC, but that didn't work either. Also, when I typed the plus-4-value into KM, it listed a value in the 'ov' column when all the other functions were all 0. Not sure if this means something.

seanrich · Post by **seanrich** » Mon Dec 01, 2003 8:35 pm

I also didn't mention that I tried switching to OBC values - no buttons worked. Also, you asked what the incorrect EFC's did - they act the same as the menu button - I've found three EFC's that will all activate the menu. Not sure why the 'Quit' button is so feisty.

Mark Pierson · Post by **Mark Pierson** » Mon Dec 01, 2003 9:08 pm

seanrich wrote:I also didn't mention that I tried switching to OBC values - no buttons worked.

If you just changed the Button Codes entry on the Setup sheet AFTER entering EFC's on Functions, the resulting hex would be all wrong. When you want to change from EFC to OBC style (or vice versa) after entering EFC's, you must use the [Swap EFC/OBC] button on the Functions sheet.

seanrich · Post by **seanrich** » Mon Dec 01, 2003 9:17 pm

Mark -

Tried that - still got no result from any buttons. I even double checked IR, which even if you tell KM to use OBC's will list the corresponding EFC's under the devices tab, and this value was correct.

jon_armstrong · Post by **jon_armstrong** » Mon Dec 01, 2003 9:48 pm

seanrich,

What John is saying is that the Zenith IR protocol is very different than most other IR protocols. John's decoder, DecodeIR.dll that should be used in conjunction with IR.exe, probably decodes the EFC's correctly. So first try a key move of EFC 193 and see if that works. If so, great! If not, we can troubleshoot from there.

You can use IR.exe to do the key moves. While I think KM Master works correctly, I think it assumes that you understand the Zenith protocol. The Zenith protocol requires fixed data that is 1 PLUS the number of data bits and I don't know if KM compensates for this.

Zenith TV's are either 5 or 7 data bits. You didn't mention what setup code you are using. I am assuming that it is TV_0016 that is 7 data-bits.

usblipitor · Post by **usblipitor** » Mon Dec 01, 2003 10:05 pm

seanrich wrote:under KM, the Zenith protocol lists a default length of 6

jon_armstrong wrote:The Zenith protocol requires fixed data that is 1 PLUS the number of data bits and I don't know if KM compensates for this. Zenith TV's are either 5 or 7 data bits.

Since KM is listing a length of 6, and since the zenith TVs use either 5 or 7, perhaps we are using 5 data bits and KM is indeed compensating for this by adding 1 to get 5+1=6 ? Or am I adding apples to oranges again?

Revelator · Post by **Revelator** » Mon Dec 01, 2003 11:05 pm

I just went though this myself a few days ago.

The "Quit" button is EFC 145 on device 0017.

jon_armstrong · Post by **jon_armstrong** » Tue Dec 02, 2003 7:26 am

Steve,

The reason that I assumed it was the 7-bit variant (and that is less common, I think), was that DecodeIR returned a 7 as the "device" value. That is also why seanrich should tell us what setup code (either TV_0016 or TV_0017) works for most commands.

johnsfine · Post by **johnsfine** » Tue Dec 02, 2003 8:51 am

seanrich wrote: I noticed you mentioned that 7 was the length, but under KM, the Zenith protocol lists a default length of 6. I tried changing it to 7, but then nothing works.

I'd like to see those learned signals (The file generated by IR's file/save command after downloading from the remote with the learned signals in memory). Can you upload that to the diagnosis folder
http://groups.yahoo.com/group/jp1/files ... is%20Area/
or email it to me.

I'd like to figure out why my decoder sees a length of 7, but you claim partial success with KM with a length of 6 and no success with a length of 7.

Also, what setup code gave you partial success?
TV/0016 is Zenith protocol with length 7 ( fixed data=8 ).
TV/0017 is Zenith protocol with length 5 ( fixed data=6 ).

KM knows that the fixed data is one more than the length, so it gives fixed data 7, when you keep the default length of 6 (which Jon says doesn't match Zenith TVs) and fixed data 8 when you select length 7 which matches the decode you reported.

seanrich · Post by **seanrich** » Tue Dec 02, 2003 5:24 pm

Thank you for all your great help! The EFC 145 works perfect. I was using setup code 0017 with KM listing default signal length as 6 and fixed data as 06. IR also listed fixed data as 06. Thanks again for all your help.

Sean

jon_armstrong · Post by **jon_armstrong** » Tue Dec 02, 2003 6:57 pm

Sean,

Could you please upload a copy of the file that John Fine requested that contains the learned commands decoded as "device"=7. I think we are still curious how a 7-bit command was learned unless it is a decoding error and I haven't seen many of those for the Zenith protocol.

seanrich · Post by **seanrich** » Wed Dec 03, 2003 5:29 am

Sorry, I emailed the file to John, but I didn't know others were interested. I uploaded the file now - called 'Zenith TV - 1017'. Thanks again for everyone's help.