Well, if I understand it correctly, the difference between Power and source signals on the RioCar remote is the length of time the signal is broadcast. That is: they're the same button, if you hold it for a long press, the player reads that as Power; a short press is read as Source.
(Consults the button guide
You are correct: Longpress=Standby/power, shortpress=source.
I know that the Harmony programming system is tricky. I own a Harmony One, and although I love the remote, and I love the way that the *results* of the programming behave once you've got it properly programmed, I find that the programming software itself is just downright awful.
I have never tried to use my Harmony One to control my empeg, but I have had to make it do backflips through hoops before to successfully control other pieces of equipment. I have found the following things helpful sometimes. Try some variants of these:
- It is possible to configure an "Activity" to do additional steps when powering on. So you could say that when you switch to the "Listen to empeg" activity, you could have it deliberately change the source back again to the correct source. I forget which menus in the programming software you use to do this stunt, and I know it's a bit byzantine in their software. But you should be able to find it. For instance, on my current A/V system, each one my Activities is followed up by several additional steps, such as re-sending a second "input" command to the TV which doesn't always get the first command.
- I don't recall whether there are additional steps you can program for *ending* an Activity. If there are you could maybe put some steps in there which do the trick to put the player into Standby mode.
- You might be able to dig deep and tell the Harmony software to use a different set of button presses for the "Empeg" *device* for power on and power off. I think that's an even deeper dig than adding commands to Activities. Don't remember how to get to it in the software.
- I don't recall if any of these Harmony devices have "Learning" capability, but if they did, you could "Learn" it the power button all over again. I never needed to "Learn" anything into my Harmony One, so I'm not sure.
- I wonder if Logitech has a system for reporting/accepting updates and corrections to their device profiles. Because that's what this really is at the heart: Their device profile for the empeg is flawed in the way you describe.
Now, a side note... This might not have any bearing on this at all, or this might help you as you go about fixing the problem... Infrared remotes handle long button presses differently than you might imagine that they do. It's not like this:
Infrared remotes can't do that, because there is no guarantee that the device is going to see both the press and the release. For example if you pressed "volume up" and then covered up the IR emitter before releasing the button, the stereo would keep turning up infinitely.
So IR Remotes don't work that way. Instead, they work like this:
(then, if user has held down the button...)
RepeatLast (as long as the user is holding down the button).
The Harmony and the Empeg both know this and should be handling this correctly. I think the Harmony profile for Empeg is just wrong in this regard. Question: Do other repeated buttons work as expected on the Harmony controlling the Empeg? For instance, volume or Fast Forward? That would indicate whether this is a global thing with the "RepeatLast" ir code, or, if it's simply that the profile doesn't contain the longpress version of the Source button.
An interesting side trick you can do with IR remotes for the RepeatLast command. There is only one "RepeatLast" code for any given remote, the repeats are not button-unique. So if you have two remotes for the same device (or the original and a Universal remote), you can hold down a button on one of them (say, volume down) and tap a different button on the other remote (say volume up) while still holding the repeat on the first remote, and you'll see how the repeated command changes. Because what the device saw was this:
... and the device just does what it was told to do. Totally useless, but really interesting from a design perspective.