That is a very strange set of learned signals.
Maybe bad batteries in the original remote could cause that. But it isn't the symptom I'd expect.
Jerking the remote as you press the button would cause that. But the failures are SO consistent. You'd need to jerk the remote almost identically each time.
If the OBC numbers of all the failed learns are very high and there is a firmware flaw (that some other brand learning remotes have) in handling the repeat pattern for NEC device 0 with high OBC, that would fit exactly. But I think someone would have reported before if a 9910 had that flaw, and the signals that learned right have low OBC numbers, so I think most of the signals will have low OBC numbers.
So, no decent explanation for those bad learns.
Capn Trips wrote:Do NONE of the learned signals work? Interestingly, of your learned signals, Up, Down, Right, Vol+ and Vol- appear to be legitimate learns. Do any of those work?
Those ARE learned correctly. they SHOULD work.
The ones that learned incorrectly are also NEC protocol, device 0. They just didn't learn well enough to be decoded or used.
Capn Trips wrote:(hopefully one of the experts is) to identify if you are either not holding the SENDING button long enough, or if the signals are just too long
No to both of those theories. But as I said above, I don't have a very good theory myself.
Capn Trips wrote:
Is your learning technique proper?
(a) Both remotes laying flat on a yable facing each other at about 2-4 inches apart?
Failing to do that may fit these symptoms.
Capn Trips wrote:
(b) press the 9910 button TO which you are learning BEFORE pressing the OEM button FROM which you are learning?
It's NEC2 protocol, so getting that wrong would have very minor consequences and not at all like these symptoms.
Capn Trips wrote:
(c) HOLD the OEM button untiil the 9910 LCD indicates "SUCCESS"?
Might help.
Capn Trips wrote:
Also, if none of that helps you get proper learns on the 9910 from the OEM remote, I again suggest attempting learning from the Sony to the 9910, to see if something in the learned signals changes.
That is also a good idea.
You could also build a few NEC2 upgrades with device 0 and try all the OBC numbers (0 to 255) and see what each does.