Slowly dimming radio LCD backlight

[pentode]

I need to tear into this today but I'm just curious if anyone's seen this before - when I can't find a schematic I always like to look for pattern failures before I start. I have a single disc factory radio out of a 2007 (I think) which I swapped in to my '05 to get the aux input. It works fine except for one fault. When I first turn on the day backlighting, the LCD is backlit the way you'd expect and is perfectly legible. Over a period of 2-3 minutes, that one backlight slooowly fades out to the point where it's impossible to read. All other lights in the radio are fine and normally I'd just suspect a bulb (if it even is a bulb)- is just the slow fade out that has me wondering. I expect it'll be something like failing LED and/or something heat related.

Really just wondering if anyone's seen this or if it's totally random.

 

[Curious Hound]

That's a strange one. I woukd expect it to be a problem within the radio itself. If it was anythung else, all the lights on that circuit should act the same way. A led usually just flat out fails - works one moment and then quits permanently. So I wouldn't thjnk it's a failing LED. Without knowing the internal cuecuitry of the radio unit, it will be extremely difficult to fix.

 

[pentode]

Agreed, it's a puzzler. And to be clear, I would't really suspect a LED , but more some other component that drives it. Would Ford be so clever as to use incandescent bulbs for all the other backlights but a PWM circuit to drive a LED for dimming? I really hope not but I've learned to expect surprises from engineers.

It's the slow, steady linear fade out that makes me think either heat or power supply. Should be fun.

 

[pentode]

Following up with more info, although not much useful for the time being.
I must have seen this before as I've had Ford radios apart before but it's been quite a while. Turns out the faceplate comes off as a modular unit, which is kind of cool. And the backlight situation is almost exactly the opposite of what I expected. The backlights for all the buttons are LEDs. The backlights for the LCD look like surface mount LEDs at first glance but they're actually little surface mount bulbs wearing green plastic hats. So the backlights for the buttons and LCD are indeed different circuits, but exactly backwards from what I figured. The button lights will likely have a complicated PWM setup for dimming but the LCD lamps are probably fed by the same circuit coming from the dimmer for the gauge cluster.

This will continue to mildly annoy me so I'll probably check back within the next 6-12 months with a solution as I can't make this a huge priority. However, the local U-pull jacked up the price of a radio from $35 to $55 which is a bridge too far for me, so I will fix it. At least I can read my fuel and temp gauges at night now.

 

[pentode]

These are the 2 lamps that go dim, for the record. If all else fails, I'll just wire it to permanently run the lamps at a voltage that makes the display visible. That's 5V and in a dark room is probably too bright.

This makes me think it's NOT running from the same circuit as the gauge cluster but rather a third, separate supply.

 

[BobSacamano]

Maybe something on this video will help

 

[pentode]

Maybe something on this video will help

Oh, thanks for that, great username BTW. . I won't give it away in case others want to guess.

Yeah, I came across similar info when I first searched. That's a different (later?) design that I think is using a vacuum fluorescent display, which uses higher voltages and is self illuminating.
I'm still suspicious of bad solder joints, although the steadiness of the dimming makes me still think it's a computer thing (microcontroller thing). I'd have to dig deeper into the radio though as everything in the faceplate is apparently fine.

One last detail - the backlight is brighter _before_ I turn the lights on. I don't think I noticed this before, I rarely drive the truck at night and I don't suppose I ever checked the backlight on the radio like that.