Sony MDR-IF245R transmitter LED replacement

[lti]

I'm looking at the transmitter from these wireless headphones because the range is too short. One of the LEDs failed many years ago (nine years ago - I found pictures that I took then) and was simply bypassed to make them work again, but it seems like the range has been degrading since then. They're near-IR LEDs (I can see them dimly lit red, so they aren't the standard 940nm), and I haven't been able to find the actual wavelength. Does anyone here happen to know what replacement LEDs I should use? Maybe I can adjust the transmitter was well while I'm in there, or there's another fault. I should look for the classic Sony bad solder joints.

I did find that other Sony wireless audio products used 850nm LEDs, so maybe I'll try that.

[redwire]

Clean up the PCB there is some flux and corrosion on the second IR-string driver transistor Q254, maybe one trace on the base I can see bare copper. I use IPA and q-tips.
You can measure and compare V_F on diode-test, the LEDs should match closely unless damaged.
I would suspect the tuning has drifted out but no idea how it works here because there are two circuits driving the IR string. The OSC coils are labelled 2.3MHz and 2.8MHz, possibly for each R/L channel. A scope might show something.
The headphone end receiver should have a test point for RX signal strength.

I have seen some IR remote control LED's faintly glow red, don't look directly at them as they can damage your eye up close.
You can see they are wide-angle IR LED's, the die not set far back in the lens. Lumex OED-EL-1L2 is 940nm, 60 degrees and good for 1A pulses, and bright 60mW/SR at 50mA seems the usual.

[lti]

I remember that I did a lot of cleaning and resoldering after I took those pictures. The LEDs had a higher forward voltage than standard 940nm LEDs, and I was never able to see light from 940nm LEDs either. It seems intentional to use a shorter wavelength either for show (Sony's marketing pictures also show the dim red glow) or so Sony could save a cent on an indicator LED.

They are using two FM oscillators at 2.3MHz for the left channel and 2.8MHz for the right channel. That appears to be a common design for IR wireless headphones, but older units used a dedicated IC (probably discontinued by the time this model was released). The receiver in the headphones has more adjustments (and selectivity doesn't seem too good - you can hear WWV in the left earpiece if you go just far enough out of range). Unfortunately, I don't have an oscilloscope.

I haven't had a chance to open them up yet because my mom is still using them to watch TV. She keeps complaining that one channel cuts out until she wiggles the plug (using the 3.5mm input and included RCA to 3.5mm cable - dissimilar metals?). A while back, I tried swapping her to an RCA cable, and one channel was completely missing through the RCA input. That should be easy to fix, but I also noticed a lot of static at her normal watching position (within the range specified by Sony) during that visit. Now she tells me that there was always a little static, but I don't remember that from nine years ago. That's why I asked about the LEDs so I can properly fix it when I either upgrade the jacks or fix the RCA input.

She also wore out the power switch on the headphones a few years ago. I found that E-Switch EG1206A is a suitable replacement.

[lti]

I finally got to look at these. The static was only in the right channel, and the problem was my fault. When I did the original LED bypass hack, I messed with those two coils. However, I marked their original positions, and turning the 2.8MHz coil back to its original position got rid of the static. The 2.3MHz actually had slightly shorter range with the coil in its factory setting, so I adjusted it slightly from factory. I guess I don't need new LEDs, but it would be nice to fix it right. There's currently a 100 ohm resistor bridged across the bad LED (which still lights, but it has high internal resistance).

I cleaned the board a little. The base pad for Q254 was damaged (it looks more like a lifted pad from bad soldering than corrosion), but it is still making a connection. The soldering quality is awful in general, but they're holding. I added some newer pictures (but not better - this is my new phone camera), and you can see shiny solder joints where I resoldered the worst joints with leaded solder.I tried to get a macro shot of the area around Q254, but I didn't notice that it was blurry until this was already reassembled. If anyone is curious, the two wires go to a separate PCB that just has the charging LED and its current limiting resistor.

The 3.5mm jack is flaky (and has been cleaned multiple times), but the RCA jacks worked today. They're just cheap connectors with bad plating, but they're oddballs. The 3.5mm has all three pins in a line along the center of the jack with 4.5mm spacing and a height (center of the hole to the top of the PCB) of 6.5mm. The closest I can find is Adam Tech ASJ-125-3, but that has 3.5mm pin spacing.

The dual RCA jack also has an unusual pin layout. There are single RCA jacks that have the correct pin layout (use two individual jacks - one white and one red), but the distance from the center of the jack to the PCB is 1mm higher than the original Sony jack (I measured 7mm). I guess you could shave down the plastic to make them fit if you were desperate.
https://www.cuidevices.com/product/resource/rcj-04.pdf
https://www.kycon.com/Catalog_PDF/KLPX.pdf

[redwire]

What I do for alignment is find a signal test point in the RX (headphones) and scope that signal.
Another approach is to cover the TX with stuff to weaken the IR signal to where the headphones just start to cut out and then align the TX coil for best sensitivity.

You need a plastic/tiny metal screwdriver to adjust these tunable transformers, presence of a metal screwdriver detunes them and you can get fooled it all tunes great until the screwdrier is pulled away.

[lti]

Another approach is to cover the TX with stuff to weaken the IR signal to where the headphones just start to cut out and then align the TX coil for best sensitivity.

That's how I adjusted them this time, and I used a plastic screwdriver both times. This was a different pair of receiving headphones (she impulse bought a replacement when the power switch failed instead of having me replace the switch), so I guess the original headphones were tuned a little off.