On July 8, 2007 at 13:11, sekar said...
both streamzap.irp and thomson.irp not at all creating
any responces from these (Zenega and Thomsan) recievers.
none of the key is responding. tried to adding line Define
T=0 and 1 in thomsan.irp without any luck. believe the
codes are not matching??
In post 7 of this thread, you gave some ZENEGA signals. The digit 1 looked cleanest. I decoded it as Streamzap device 34, function 34, with T=0.
0000 0049 0000 000A 0030 0030 0060 0060 0060 0030 0030 0030 0030 0060 0060 0060 0060 0030 0030 0030 0030 0060 0060 05E0
I used MakeHex to create the full set for Device 34, with T=0 and pulled out function 34, which is:
0000 0046 0000 000A 0035 0034 0069 006A 0069 0035 0035 0034 0035 0069 0069 006A 0069 0035 0035 0034 0035 0069 0069 11B0
That may look to you very different from your learned signal, but it is really nearly identical. If we decide, it needs to be closer, we can tweak some values in the .irp file (slightly lower frequency and/or slightly shorter timebase and/or much shorter message time). But I'm not yet convinced we need to.
Please recheck that your learned signal above really works and check to see whether the MakeHex version I made works and whether the MakeHex version you made (device 34, function 34, T=0) matches the one I made.
If you want to experiment with values in a .irp file, lowering the frequency will raise the second number in the Pronto Hex string (and change other numbers, but ignore that). Your cleaner learned signals have 0048 or 0049 in that position. The correct value does NOT vary by command. It just isn't learned very accurately and doesn't need to be exactly correct (46 vs. 49 should be close enough. Even the 42 in your bad learn of power toggle should be close enough).
BEFORE you can compare the rest of the numbers (between to Pronto Hex strings), the second number must match exactly, so even though a perfect match on that won't matter to the final device, you'll need that perfect match before you will know how to tweak other values.
The second largest value in each of your learns is 0060. After the second value matches, you can compare that to the second largest value in the MakeHex signals. Again, the correct value doesn't vary by command and an exact match doesn't matter. If you want it closer, lower the TimeBase to lower that value.
The largest value is last in any reasonably clean learn (your power toggle is not a good learn). The largest value shouldn't even need to be close. But if you want it closer, lowering the Message Time will lower that largest value.
For Thomson, can you learn a signal from the original remote?
Last edited by johnsfine
on July 9, 2007 14:55.
