H120 remote control

[RandomFactor]

I recently extended the wire of my H120 remote (to around 4m) so I can control it from the sofa.  It works, but I have hit a couple of drawbacks. 

1) The remote screen will light up but no information is displayed.  All functions of the remote work, but I can't see any info on screen.  Does anyone know if this would be expected?  Would extending the wire to that length affect current/resistance to the point that the screen would no longer work?  Or is it more likely a loose connection?  I couldn't see anything loose during testing and have since bit the bullet and bound the connections in tape to protect them.  It works, after all...

2) While my WPS text has always been large (so it can be read from a distance), because of the above problem, I also made the general font size very large so I can browse menus/playlists from a distance too.  However, the screen that allows me to toggle shuffle, repeat etc (accessed by holding the A-B button on the player or the bottom right scroll wheel on the remote) now crashes the player - some kind of display problem on the main player screen.  Is there any way I can change the functionality of those buttons so holding them simply toggles shuffle?  I never use repeat or change the scope of the playlists from that screen.  I suspect changing the font size of that screen alone is not possible.

Thanks...

[bluebrother]

I recently extended the wire of my H120 remote (to around 4m) so I can control it from the sofa.  It works, but I have hit a couple of drawbacks.

I was working on an infrared remote for the h100 series some time back. I haven't finished that project ever but if there is enough interest I could look into it again.
 

2) While my WPS text has always been large (so it can be read from a distance), because of the above problem, I also made the general font size very large so I can browse menus/playlists from a distance too.  However, the screen that allows me to toggle shuffle, repeat etc (accessed by holding the A-B button on the player or the bottom right scroll wheel on the remote) now crashes the player - some kind of display problem on the main player screen.

This shouldn't happen. Can you file a bug report on it? Showing garbage or being unusable is one thing, but the player definitely shouldn't crash.

Is there any way I can change the functionality of those buttons so holding them simply toggles shuffle?  I never use repeat or change the scope of the playlists from that screen.  I suspect changing the font size of that screen alone is not possible.

No. You can customize the items on the quich screen but you can't change button mappings.

[bonapardo]

Thanks for the reply.  I've been thinking about how to create an infra-red control myself but my electronics knowledge is limited to say the least.  In theory, shouldn't it be simple enough to measure then generate the appropriate voltage or modify resistance?  Ignore me.  I know nothing.  But would LOVE an IR remote - the H120 is a permanant fixture in the living room, being connected to a 3.5" drive.  Is one vote enough for you to ressurect the project?!

I'm happy to upload a bug report, where should I look to get instructions?  I'm using a custom font if that makes a difference.

I'll try editing the options on the quickscreen and see if that allows the screen to load.

(remember... +1 for the IR!)

[[Saint]]

Thanks for the reply.  I've been thinking about how to create an infra-red control myself but my electronics knowledge is limited to say the least.  In theory, shouldn't it be simple enough to measure then generate the appropriate voltage or modify resistance?  Ignore me.  I know nothing.  But would LOVE an IR remote - the H120 is a permanant fixture in the living room, being connected to a 3.5" drive.  Is one vote enough for you to ressurect the project?!

I'm happy to upload a bug report, where should I look to get instructions?  I'm using a custom font if that makes a difference.

I'll try editing the options on the quickscreen and see if that allows the screen to load.

(remember... +1 for the IR!)

The quickscreen *really* shouldn't be crashing the player...that's, interesting.
Has this always been the case, or, is it since your remote modification?


[St.]

[bonapardo]

(Just noticed I used a different ID to log in when I started the thread!  We are one and the same user - I'll close the older account).

The problem is not related to the remote, rather the text size.  The text I use is very large - I use three rows of text on the WPS (see photo) and four rows displayed on playlist/menus.  Rockbox just can't seem to display the quickscreen - loads of overlapping and the unit hangs.

[Chronon]

Rockbox shouldn't crash in this case.  It would be worthwhile to post a bug report.

[bluebrother]

In theory, shouldn't it be simple enough to measure then generate the appropriate voltage or modify resistance?  Ignore me.  I know nothing.  But would LOVE an IR remote - the H120 is a permanant fixture in the living room, being connected to a 3.5" drive.  Is one vote enough for you to ressurect the project?!

I've looked after the hardware yesterday evening and still found it lying around. Haven't figured if it's still working to any degree, and haven't searched for the firmware. It will definitely include an Atmel controller so for building it you'd need to flash it. If I finish it, no promises (yet ).

I'm happy to upload a bug report, where should I look to get instructions?  I'm using a custom font if that makes a difference.

See http://www.rockbox.org/wiki/FlySprayHowto

Edit:

(remember... +1 for the IR!)

ok, after some hours of work I have a working prototype. It tries to fake a h300 non-lcd remote. As a result, it can use the buttons Play/Pause, Forward, Rewind, Volume Up and Volume Down (i.e. the buttons that remote has). I'm using an old Hauppauge remote for that which sends RC5 codes. There's also a strange problem: when the receiver is connected the device won't power up. I haven't figured what's going on there.

Anyway, the current schematic is here: http://www.alice-dsl.net/dominik.riebeling/rockbox/h100remote.png

[bonapardo]

Wow, I didn't even get a chance to respond to your last post!

Thanks for the schematic.  I have an old Hauppauge remote myself somewhere so seems like a great starting point.

When you mention the device won't power up are you referring to the player itself?  If so, do you mean it won't power up from just the remote or from any of the buttons on the player too?

You're a star, I'll let you know how I get on.  With my limited electronics knowledge, I may be some time...

I love the internet!

[bluebrother]

When you mention the device won't power up are you referring to the player itself?

Yes, the player itself. No idea why this happens, but it seems to be related to the circuit -- if I erase the memory of the controller the same thing happens, so I think it's safe to say it's not caused by some wrong initialization in the controller doing weird things. Let's see if I find some time to look into that the next weekend.

You can't power up the player using the remote in any case: first I thought the remote isn't powered when the player is off. Later I found that there is some power supply, but it's not the normal 3.3V and therefore won't be sufficient to make sure the receiver is in a usable state.

If there is enough interest I might consider doing a PCB. So ... are there others around that are interested?

Edit:
I've pushed the current state (code for the controller will follow later, most likely this weekend) to https://github.com/bluebrother/ir-remote

Edit2:
Found the problem that prevented the player from powering on, it's working fine for me now. An updated schematic will follow in the github repository later.

[bonapardo]

Keep up the good work!

btw, I'm having problems viewing the updated schematic - do I need a special viewer?

Once I've cobbled everything together, I'm hoping to attempt navigation control (as I can view my large text on the iRiver from the sofa).  What was your approach to working out what the controller needed to output (i.e. what the iRiver was expecting to be sent to it)?  What/how did you measure?  Do you think, with a spare LCD remote, that it should be pretty simple for me to program the extra commands into the controller - or would I need to redesign the circuit?  The latter is pretty much out of the question for me!

Thanks again...

[bluebrother]

btw, I'm having problems viewing the updated schematic - do I need a special viewer?

It's an Eagle file -- there's a (limited) free version available which I was using for drawing it. I've put a png export of it online: http://tinyurl.com/bluebrother/ir-remote.png

What was your approach to working out what the controller needed to output (i.e. what the iRiver was expecting to be sent to it)?  What/how did you measure?  Do you think, with a spare LCD remote, that it should be pretty simple for me to program the extra commands into the controller - or would I need to redesign the circuit?

The Rockbox wiki has some information on how the remote works (see the links in the Readme file in the github repository). If you want to add navigation you'll face two problems:

• you need to tell the player it's an LCD remote. This is done via pulling th Hold line around (I've figured this out in the meanwhile but haven't documented it yet. The change is simple).
• when you're faking the H100 LCD remote you need completely different resistor values. That is, the direct connection of RMC to a pin is useless right now -- it generates a value of 0, and that value is Stop for the LCD remote but not handled for the non-LCD remote. All other values are different as well, and you need more buttons (i.e. GPIO pins). So you basically need to redesign parts of the circuit.

However, depending on what exactly you want to achieve there isn't much need to change anything: I'm usually playing music from the file browser. Once the playlist ends Rockbox drops me back into the file browser, and I can navigate and select a new folder / playlist that way. The only "problem" is that I can't switch out of the WPS (by pressing Select or Stop since those don't exist) with the current solution.

[bonapardo]

Brilliant, I'll download the software.  Thanks for the PNG.  

Sounds like the only additional functionality I need are just the select and stop buttons in order for me to back out to a menu while the music is playing and go up a folder level respectively. I wonder if we even need to fool it into thinking it's an LCD remote or whether we can get away with fooling it into thinking it's a non-lcd remote with additional buttons (!).  I suspect that's wishful thinking.

How did you finalise the resistor values for each control, was it just trial and error, did you connect everything up and measure values while using the remote or were there hidden truths within the non-lcd remote itself?

EDIT:  I just looked at the section above the remote section in the wiki (which I was familiar with from the extension cable) and can see some resistor values.  When you mention completely different resistors being necessary for faking the LCD remote, is this due to the fact that we're faking LCD (i.e. does this even affect resistors for your existing non-lcd buttons) or do you simply mean we need different resistors for the additional two buttons?  Also, would I need a different controller (with more generic I/Os) for the additional buttons?

Sorry to keep pestering you with questions!  Navigation is one of my requirements and just wanted to check whether I'd need to know this info before I actually put things together - either way, I'm going to give it a shot for fun.  I really appreciate your advice and don't want you to regret kick starting the project again!

Cheers,

Ben.

[bluebrother]

I wonder if we even need to fool it into thinking it's an LCD remote or whether we can get away with fooling it into thinking it's a non-lcd remote with additional buttons (!).  I suspect that's wishful thinking.

Yes, you need to.

It's basically this way: Rockbox checks for the type of the remote. Depending on the type of the remote it checks the input ADC for various value ranges. The H100 LCD remote produces completely different values than the H300 non-LCD remote due to different resistors in the setup. As an example take the connection between PB4 and RMC: for the H100 LCD remote this is Stop, for the H300 non-LCD remote this does nothing. It is wired for the simple reason that I had this connection in my initial circuit (since I started with the H100 LCD remote) and didn't remove it. Since the controller software doesn't use PB4 it doesn't hurt.

You could of course also adjust the Rockbox source code to work with different resistor values. I would try to avoid that to simplify updating (and adjusting the values really isn't that much of an issue).

How did you finalise the resistor values for each control, was it just trial and error, did you connect everything up and measure values while using the remote or were there hidden truths within the non-lcd remote itself?

I measured the values directly on the remote.

When you mention completely different resistors being necessary for faking the LCD remote, is this due to the fact that we're faking LCD (i.e. does this even affect resistors for your existing non-lcd buttons) or do you simply mean we need different resistors for the additional two buttons?  Also, would I need a different controller (with more generic I/Os) for the additional buttons?

First, the Play button (which is a special case for the buttons) is the same for both remotes, so no changes needed here. Second, all other buttons (except Hold, but that isn't of interest anyway) are handled via the resistor network you see both in the Wiki and the circuit. However, H100 LCD remote and H300 non-LCD remote use completely different values.

For two additional buttons you won't need a different controller -- the used one still has various GPIOs left.  For adding those two buttons you even need only one additional line due to th PB4 connection already been made. It's basically an additional element in the line of resistors, connect that to PD6 (which has even a quite good location on the controller pinout; furthermore you could even add more buttons by dropping the UART, LED and unused crystal lines). Then you need to adjust all resistors to produce the needed values and adjust the controller firmware. Last, you need to tell the player that you use a different type of remote. This is done by pulling the Hold line to a certain level. I haven't tried this but a resistor network like the following should do

Code: [Select]

Vcc---[2k9]---HOLD---[2k1]---GND

I might even update the circuit with that some time later -- I've kept the UART lines open to be able of interacting with an Ipod later (since that uses a simple UART). Unfortunately I haven't managed to get that working with my current prototype, and I fear the UART might have been damaged during my repeated (un-)soldering work, so I might build up a new prototype. This won't happen too soon though since this weekend DevCon is coming up.

I really appreciate your advice and don't want you to regret kick starting the project again!

As long as there is someone interested I'm unlikely to do

[bonapardo]

Brilliant, that's exactly the kind of info I was after.  Doesn't seem like too great an obstacle for me to add those buttons (he says with trepidation).  I'll order the controller and get started over the coming weeks.

Thanks again for all your help, enjoy DevCon!

[bonapardo]

(I had a question, but worked it out)