Stumped! - attempting to deploy new mythtv linux PVR

[wagner]

I'm a longtime JP1 user and have a good grasp on the steps required to get my remotes to do what I want (in no small part to the fine guiduence from Rob and others 'round here). Recently I built a new linux based mythtv PVR system to replace my aging linux PVR. The system is built in a Silverstone HTPC case and I am attempting to utililize the built-in imon IR receiver.

My remote is a URC-8811 (purchased from Rob) and I am transferring data to/from it via a parallel cable (also purchased from Rob) attached to a Windows 2000 workstation. I am using the latest version of IR, RemoteMaster and the accompanying dependencies (RDFs, Image/Maps for RM, etc.) I just recently downloaded from this site. I also have the latest Windows patchs and the latest Java from the Sun site.

I tried using IR 6.2 as one test and that made no difference to my situation over using IR 7.0.

I'm effectively attempting to do what is described in this post. I want to create an upgrade where all the keys are available to me to assign to mythtv control. When I saw Indulis's post in this thread, I thought, "perfect! - everything done for me!" No dice. I tried saving all the bits in the thread and later found his updated versions zipped up in one package in the download area. These certainly should work for me, but something else is amiss in my setup.

So, I followed Rob's steps to generate my own upgrade from scratch from this this thread. I did it within RM and it looked quite similar to Indulis's upgrade when I was done.

Here is where things are funky:

When I use "irrecord" in linux to record the codes coming out of my programmed remote (using Indulis's upgrade or my own), I see this as a resulting lircd.conf file:

Code: Select all

  name  lircd.conf2
  bits            8
  eps            30
  aeps          100

  one             0     0
  zero            0     0
  pre_data_bits   16
  pre_data       0x1F
  post_data_bits  8
  post_data      0xDE
  gap          211967
  toggle_bit_mask 0x6EF20012

      begin codes
          1                        0xEC
          2                        0x6C
          3                        0xEC
      end codes

Where it should look something like this from Indulis's lircd.conf:

Code: Select all

  name  urc_8811_mythtv 
  bits           16 
  flags SPACE_ENC|CONST_LENGTH 
  eps            30 
  aeps          100 

  header       9045  4427 
  one           603  1641 
  zero          603   518 
  ptrail        606 
  repeat       9050  2196 
  pre_data_bits   16 
  pre_data       0x9F6 
  gap          108317 
  toggle_bit      0 


      begin codes 
          1                        0x9966 
          2                        0xF906 
          3                        0xD926 

For reference, as another test, I tried this routine with the Sony upgrade assigned to my 8811 DVD key and saw this output:

Code: Select all

  name  lircd.conf-new
  bits           24
  eps            30
  aeps          100

  one             0     0
  zero            0     0
  pre_data_bits   8
  pre_data       0x7E
  gap          203974
  toggle_bit_mask 0x38BE7000

      begin codes
          1                        0xE839FF
          2                        0xC03AFF
          3                        0xF868FB
      end codes

end remote

I tried applying both of our upgrades to both my AUX and VCR keys on the 8811 and they both produce the same output. I tried first on the AUX (changing it's type to "VCR" to match the upgrade designator of the same), but my punch through volume controls didn't work and I thought that may be telling of a problem with my remote. I then tried my upgrade on the VCR button of the 8811 (the punch through DID work there), but the problem persisted. For kicks, I did a MFR RESET, and reuploaded all my codes with IR, but that didn't affect things.

Using "mode2" to display the output in linux of either upgrades (actually, any upgrades I've tried so far) returns output like this:

Code: Select all

code: 0x12931558
code: 0x57ffffff
code: 0x14931558
code: 0x57fffe44
code: 0x6d152b1f
code: 0x14935558
code: 0x57ffffff
code: 0x14931518
code: 0x57fffe44

Rather than what most sites show as typical mode2 output:

Code: Select all

space 2681403
pulse 9058
space 4415
pulse 633
space 494
pulse 622
space 510
pulse 608
space 499
pulse 629
space 1633

When I press and hold a key down on my remote from one of the "all keys enabled" upgrades, mode2 outputs this:

Code: Select all

code: 0x001f6cde
code: 0x2eef6dcc
code: 0x2f6defff
code: 0x001dffff
code: 0x001dffff
code: 0x001fffff
code: 0x001dffff
code: 0x001dffff
code: 0x001dffff
code: 0x001dffff
code: 0x001dffff
code: 0x001dffff
code: 0x001dffff
code: 0x001dffff

Does anyone see anything in the above mess that is indicative of a particular problem? I'm at a loss at this point as to what to try.

I have the homebrew serial IR receiver that I was using on my old linux based PVR system that I plan to test as a final stab (I'd rather keep those drivers off the new system if possible), but I guess I think there is something else going based on the fact that the mode2 output is possibly odd and the lircd.conf generated by using the Sony DVD upgrade is actually usable where the lircd.conf generated from the "all keys enabled" upgrades is bunk.

I used to use KM rather than RM, but don't have Excel on my PC currently, so I went with RM. Is there potentially something amiss with RM or my use of it -- generating bad "paste" data for IR? I could try to track down a copy of Excel to test if that sounds worthwhile.

Thanks for your consideration and thoughts.

Mike

[johnsfine]

I am attempting to utililize the built-in imon IR receiver.

I don't know anything about that IR receiver, but based on the info in your post, I would conclude it is very different from ordinary LIRC IR receivers.

Probably it supports only a very restricted class of IR signals. If we know what those signals were, we could make your JP1 remote generate a full set of them. But you haven't given any info in that direction.

Did the system come with an original remote?

I am using the latest version of IR, RemoteMaster and ...

There is no reason to believe the problem is in the JP1 remote or in any of the programs or files related to it. Probably it is generating exactly the intended set of signals. The problem is more likely in not knowing what set of signals you want to generate.

When I use "irrecord" in linux to record the codes coming out of my programmed remote

This is an aspect of LIRC where I and most other JP1 experts have very little knowledge. I don't know the characteristics of irrecord. I'm guessing the problem is incompatibility between the signals and the IR receiver. But it could be misuse of irrecord or incompatibility between the IR receiver and irrecord, or some similar issue. In other words it is possible that IR receiver could receive those signals acceptably if the related parts of LIRC were used or configured differently.

I have the homebrew serial IR receiver that I was using on my old linux based PVR system that I plan to test as a final stab (I'd rather keep those drivers off the new system if possible),

I expect that would work. I expect it wouldn't be necessary if we had a slightly better understanding of the built in IR receiver.

the lircd.conf generated by using the Sony DVD upgrade is actually usable where the lircd.conf generated from the "all keys enabled" upgrades is bunk.

That is a further hint that better understanding of that IR receiver would solve the problem.

But why can't you solve it by building a new Sony upgrade with all keys enabled?

[The Robman]

Your first step should be to determine if the problem is with your upgrades or with the way your PC is trying to record the signals. Therefore, you should re-program the remote to use one of the built in setup codes and try learning those buttons. I would recommend TV/0156, which is the main Toshiba TV code. This code is present in just about every remote and it uses the standard NEC1 protocol, which is what you're trying to use anyway.

If the resulting learns looks very similar to the ones from Indulis's lircd.conf, it means the problem is with your upgrades and we can probably help you with that.

If the resulting learns still look like the strange ones that you got from your previous attempts, the problem is with your PC and/or software, and we won't be able to help you with that.

Of course, there's always the idiot question, do you have good batteries in the remote?

[wagner]

I am attempting to utililize the built-in imon IR receiver.

I don't know anything about that IR receiver, but based on the info in your post, I would conclude it is very different from ordinary LIRC IR receivers.

Probably it supports only a very restricted class of IR signals. If we know what those signals were, we could make your JP1 remote generate a full set of them. But you haven't given any info in that direction.

Did the system come with an original remote?

It seems to be in widespread use and is supported by the stock lirc distribution and there is widespread mention/discussion of it being used in mythtv/PVR systems. Your reply has given me more ideas on what to search for. I hadn't considered it may not even be capable of receiving the same signals as my former serial IR receiver! Duh.

When I began all this, I transfered my old systems lircd.conf straight over and expected to use the same upgrade in the remote I had been using with no problems. If the new receiver can't pick up the same range as the old one, that would explain a lot!

Further research turned up the fact that it is a "iMON Soundgraph" IR receiver and I've found a lircd.conf file created by using irrecord and the remote that shipped with the receiver.

Yup! It did come with a remote and I will experiment with that if it would be helpful.

I am using the latest version of IR, RemoteMaster and ...

There is no reason to believe the problem is in the JP1 remote or in any of the programs or files related to it. Probably it is generating exactly the intended set of signals. The problem is more likely in not knowing what set of signals you want to generate.

When I use "irrecord" in linux to record the codes coming out of my programmed remote

This is an aspect of LIRC where I and most other JP1 experts have very little knowledge. I don't know the characteristics of irrecord. I'm guessing the problem is incompatibility between the signals and the IR receiver. But it could be misuse of irrecord or incompatibility between the IR receiver and irrecord, or some similar issue. In other words it is possible that IR receiver could receive those signals acceptably if the related parts of LIRC were used or configured differently.

I tend to believe that it has nothing to do with the software related to programming the remote too, but I wanted to point out what I used just in case someone more in the know may spot something.

It sounds like you're dead on in that I just have to discover what sort of signal range it expects to receive and create an upgrade based on that info.

I have the homebrew serial IR receiver that I was using on my old linux based PVR system that I plan to test as a final stab (I'd rather keep those drivers off the new system if possible),

I expect that would work. I expect it wouldn't be necessary if we had a slightly better understanding of the built in IR receiver.

I'm almost positive that would work, but I'm attempting to utilize the built-in hardware so as not to have extra complexity in the kernel/drivers and/or extra hardware hanging off the back of the box. I have considered caving in and using the old IR receiver, but I haven't quite given up yet.

the lircd.conf generated by using the Sony DVD upgrade is actually usable where the lircd.conf generated from the "all keys enabled" upgrades is bunk.

That is a further hint that better understanding of that IR receiver would solve the problem.

But why can't you solve it by building a new Sony upgrade with all keys enabled?

I have considered that, but don't want unecessary clutter (protocols, etc. - though I guess the required Sony protocol will already be there, eh?) consuming space in the remote, but mostly because I want to figure this out. I will try that this evening.

I did just turn up a site that seems to discuss this particular IR receiver at length, so I'll poke around there a while and check out the source of the iMon remote lircd.conf.

Thanks for your repsponse,
Mike

[wagner]

Your first step should be to determine if the problem is with your upgrades or with the way your PC is trying to record the signals. Therefore, you should re-program the remote to use one of the built in setup codes and try learning those buttons. I would recommend TV/0156, which is the main Toshiba TV code. This code is present in just about every remote and it uses the standard NEC1 protocol, which is what you're trying to use anyway.

If the resulting learns looks very similar to the ones from Indulis's lircd.conf, it means the problem is with your upgrades and we can probably help you with that.

If the resulting learns still look like the strange ones that you got from your previous attempts, the problem is with your PC and/or software, and we won't be able to help you with that.

I'll give that a shot and let you know what I find. I'll also see what I get using the supplied Soundgraph iMon remote. If I can make that work, I guess the hardware can be considered functional and I'm not sending the right range of signal with my 8811.

Of course, there's always the idiot question, do you have good batteries in the remote?

Yep! I did ensure I switched the existing batteries out at some point during my testing.

Thanks,
Mike

[johnsfine]

I've found a lircd.conf file created by using irrecord and the remote that shipped with the receiver.

Yup! It did come with a remote and I will experiment with that if it would be helpful.

Your 8811 supports learning. So the most powerful way to understand that IR receiver is to examine both the learned signals and that lircd.conf file, looking at the results from the same original keys of the original remote.

That way we could see what the signal looks like to that IR receiver and what the same signal looks like to JP1.

So learn some signals from the original remote into the 8811 and download.ir.exe and look at the decodes. If the decodes are simple post the decode info (protocol, device, subdevice, obc and frequency). If the decodes aren't simple, put the whole .ir file containing those learned signals in the diagnosis area and post its URL.

Post lircd data: either the header data plus the lines from the same functions you learned, or just put the whole file in the diagnosis area. If it is unclear which function names in that file go to which learned keys in the 8811, give details.

I did just turn up a site that seems to discuss this particular IR receiver at length,

Post a URL. Maybe something in that discussion will be meaningful to us that wouldn't mean anything to you.

[wagner]

I've found this lengthy discussion on the JP1 forums that seems to be the same discussion we're beginning here.

... after reading through this discussion, it sounds like I will be pushing the built-in IR receiver out of the way and squirreling my trusty old DIY serial IR receiver in there next to it.

Thanks for the replies and sorry I hadn't found that post sooner so we didn't have to dredge this topic up again.

Mike

[johnsfine]

I found this document describing an imon IR protocol:
http://krieger.free.fr/ir_data_spec.pdf

I compared that to the .ir file posted to diagnosis in that other thread. It isn't obvious how to match things up between them, but it might be right.

There seem to be a few very different imon IR receivers. I don't know whether they all use the same protocol.

So I'd still like to see a .ir file of your learned signals compared to the lircd.conf file of the same signals.

If you want to give up and use the ordinary IR receiver instead, I understand. That probably would end up being less effort. But if you want to get a JP1 remote to duplicate the imon signals, I think we can figure that out.

[wagner]

I'd love to figure it out (or assist in figuring it out -- you'd be figuring it out I reckon) for myself and for any other folks who end up in my situation and may need it.

I'm not sure I ever generated a .ir file of learned signals in my time dealing with JP1 related activities.

If you would point me toward documentation to do that, I'll do it right away. Do I essentially learn all the buttons on the included iMon remote to my 8811 and then dump that out to a file?

Thanks,
Mike

[The Robman]

If you would point me toward documentation to do that, I'll do it right away. Do I essentially learn all the buttons on the included iMon remote to my 8811 and then dump that out to a file?

Read this...
http://www.hifi-remote.com/jp1/help/

[johnsfine]

If you would point me toward documentation to do that, I'll do it right away.

I don't know which aspect of the task you don't know how to do.

Do I essentially learn all the buttons on the included iMon remote to my 8811 and then dump that out to a file?

Yes.

1) Learn some signals from the iMon remote into the 8811.
2) Download from the 8811 to IR.exe
3) Save that to a .ir file

4) Post that .ir file and the related lircd file to our diagnosis area

5) Write a post giving the URL(s) of those files and explaining whatever may be non obvious about which learned signal buttons in the .ir file correspond to which function names in the lircd file.

If you need documentation on how to do those individual steps, tell us which steps you need help with and someone will give you a more specific link into documentation.

[wagner]

Johnsfine and Rob,

I ran out of time to peck away at this (visitors from out of town and I wanted a functioning system!) and fell back on using my serial based IR receiver and Indulis's upgrade. That worked out well... just as we expected.

Thank you for the suggestions, pointers, and, Johnsfine, your offer to work through the situation to the end.

I'm out of town for a week currently, but wanted to drop in and say thanks.

Mike