Modifying an SMSL IR remote to reduce directivity and improve reliability

[RandomEar]

I have three SMSL devices in a near field setup, which is just out of range of my arms. The remotes sit beside me, but are at an odd angle relative to the DAC/amp/HP amp (40 cm below and to the left) and inputs have a tendency to not be recognized, unless I pick the remotes up and point upwards, directly at the IR receivers. The new remote, delivered with the SU-9n and other recent releases is especially dumb, because the top of the IR diode if blocked by plastic for design reasons. A "looks over function" kind of choice. The older remote delivered with an AO200 worked most of the time, but about 20% of inputs were not recognized.

As it looked easier to mod, I took a closer look at the old remote. It comes apart easily by inserting a spudger at the back:

Two of the tiny clips broke, but it later fit together again fine. The IR diode is soldered in from the top of the board in the front and the designers were nice enough to mark the cathode contact:

Taking the board out, I noticed that one leg of the LED was actually loose (bend out for the photo):

That was not the cause of the 20% missed inputs, though. I have another one of the old style remotes and it works no better. I measured the diameter (3 mm) and got a replacement diode with the highest power and rad angle I could find: A Kingbright L-934F3BT 940 nm IR LED with 70 mW power and 50° angle. By chance, it happened to be blue instead of clear, but that's just what was in stock. Two minutes of soldering later, it was installed and the remote looked as good as new:

A quick test showed that everything worked as expected. I therefore assume the 940 nm wavelength and 1.2 V forward voltage are close enough to the original. For a more scientific approach, I compared the remote with and without the LED mod at a distance of 1 m directly in front of the IR receiver of my SU-9n (same height) mounted on a tripod with rotating top plate. I rotated the plate in 5° angles starting from straight on, without moving the tripod or the SU-9n. The results are as follow:

• The original LED works at 1 m distance up to 90° rotation (remote angled directly orthogonal to the receiving device)
• The modded LED works at 1 m distance up to 95° rotation (remote angled 5° away from the receiving device)

When moving away from the SU-9n in a straight line with LOS to the IR receiver, both remotes worked until I hit the farthest possible point in my apartment, which equals around 9 m of distance. When moving on a circular path around the IR receiver at 1 m distance at the same height as the reciever, both remotes also worked up until the same angle of about 85°.

After using it for a couple of weeks, the remote with the modded LED has significantly less "dropped" commands than the unmodded one but it still isn't perfect. I'm mostly happy with the results. But my tests also show, that the remote isn't the limiting factor in the chain: The IR sensors on all of my SMSL devices (SU-9n, SH-9, AO200) have a problem with steep angles, especially if the remote is positioned lower than the receiving device. I suspect that the receiving photodiode is either positioned too far back from the front glass, or the hole in the casing is just too small, which leads to an increased directivity. I don't know if the problem is similar on Topping devices, but the remote mod might be worth a shot on those.