Behringer Eurorack UB502 / Xenyx 502 modifications

This (much simplified) block diagram shows how the UB502 is wired internally. You can see that the headphone amplifier is controlled by both the Phones fader and the Main fader. The mods I wanted to do are shown in red: Basically bypass the Main fader from the headphone amplifier's input. That would provide two independent volume controls: one for controlling my stereo amplifier, one for the headphones. I found the schematic for the Behringer Xenyx 502 online, and it seems to be identical to what I found inside my mixer—all the component values and their reference numbers were the same. If you look at the schematic's sheet 2, you see the headphone amplifier circuit gets is signal from switch SW1A/SW1B (the "Tape to phones" switch). It selects either the tape input directly (with the switches in the down-position), or the mixed signal, post-fader (with the switches in the up-position, as drawn on the schematic). I decided to take the signal for the SW1A/SW1B up-position from elsewhere in the circuit, before the "Main mix" fader. I decided a good place to take the signal would be between C42 and R77 (for the left channel, to SW1A) and between C43 and R90 (for the right channel, to SW1B). But first I must isolate those terminals of SW1A and SW1B from the rest of the circuit.

To isolate the switch, you must cut four traces in the circuit board. To access the proper places, you need to completely remove the circuit board from the chassis. First pull up all the various potentiometer knobs (they are only friction-fit and will come off by pulling only). Remove the six screws on the bottom of the chassis (remember which ones go where!) and pull off the bottom panel. Remove the four screws on the sides of the mixer and remove the plastic side panels. Remove the two screws around the XLR connector, the one in the middle of the four tape in/out RCA connectors, and the two near the front of the mixer. Finally remove the nuts and washers holding the eight 6.3 mm phone jacks in place, and remove the circuit board.

	On the top side of the board, cut two traces as shown (see the full image here). The ones you want to cut are the top and bottom one in that bunch of four. Try not to damage the adjascent traces (as I almost did). When cut and cleaned up, I put a drop of cyanoacrylate superglue on the area for no special reason. (Well, it might protect the traces I scratched by accident from corrosion, but the traces I intentionally cut obviously don't need any protection.) I cut the traces with a sharp snap-off knife, but I've later learned that a Dremel-type mini drill with a small milling bit is a lot easier and more controllable.
	On the bottom side of the board, cut two more traces as shown (see the full image here). The ones you want to cut are the ones going to the two rightmost pins of the switch. I likewise covered these areas with cyanoacrylate.
	This has isolated the required pins of the switch, but has also cut the circuit it was connected to in half. Two jumper wires must therefore be soldered in place on the bottom side of the board (the green ones in the image, see the full image here). One wire goes from the positive terminal of C40 to the right-hand end of R40 (when viewing the board as shown in the picture). The other wire goes from the positive terminal of C41 to the right-hand end of R43. Both R40 and R43 are SMD resistors, but it's entirely possible (if somewhat tedious) to solder wires to them. With this done, no signal at all goes to the headphone amplifier when "Tape to phones" is not selected, but the "Main out" works as before, as do the "Tape to phones" and "Tape to mix" switches. Finally, to connect the isolated terminals of the switch to the pre-fader main mix signal, solder another two wires (the red ones). One goes from the negative terminal of C42 to the lower-right pin of the switch. The other goes from the negative terminal of C43 to the upper-right pin of the switch. These are all through-hole components and easier to solder to. This essentially parallels the headphone amplifier section (when "Tape to phones" is not selected) with the main mix fader and Main out section, completing the modification.

When soldering the jumper wires, make sure you don't short-circuit them to any other nearby components or board traces. And while you're inside the mixer, you might consider replacing the power supply filter capacitors (C1 and C2, both are 470 μF 35 V electrolytics. Use 105°C ones, preferably low-ESR type if you have them in-shelf, but ordinary ones are fine too). One of them was visibly swollen on my unit. This is the first time I have ever seen swollen capacitors on a linear power supply! On SMPSs they are an everyday nuisance, but to start failing on a linear supply they really have to be cheap!

Reassemble the mixer and test it. If all has gone well, you can now control the main and tape outputs with the "Main mix" volume control, and the headphone output with the headphone control, independently! The "Tape to phones" and "Tape to mix" switches work exactly as before. The level indicator LEDs, however, are now pre-fader also, so they keep blinking along with the mixed signal even if the "Main mix" fader is turned down to zero. But blinking LEDs are nice—why should they turn off just because the music is silenced? Ok, maybe they're not useful anymore for adjusting levels when mastering a musical recording, but I'm not using this mixer for mastering anyway.

	A dynamic microphone can be plugged directly into the mixer's mic input, of course. A typical PC headset's condenser microphone, however, needs a "bias voltage", typically provided as shown here. That's very different from the 48 V "phantom voltage" used in professional microphones, by the way! And this mixer does not provide phantom power on its mic input anyway.
	Since CH1 is a single mono input, I don't think I'll ever be connecting a line source to it. So I decided to convert its line-level 6.3 mm TRS jack into a mic input that mimics an unbalanced PC sound card microphone input, with the necessary bias voltage. Here is an excerpt of the schematic, with my modifications shown in red. Firstly, the "TIP" and "RING" contacts of the jack, and the "TIP" contact's switched terminal as well, are isolated. The switched contact will not be reconnected anywhere, as I want "TIP" of the jack floating when no plug is inserted. Secondly, the zener diode, resistor and capacitor provide a stable 3.9 V bias voltage into the "RING" contact through the 2.2 kΩ resistor. Finally, the "TIP" contact is wired directly to the positive end of C3, which places the mic signal onto the "HOT" line (pin 2) of the XLR connector, while the "COLD" line (pin 3) remains grounded through R11. When nothing is plugged into the TRS jack, this does not affect the XLR input in any way, and a dynamic XLR microphone can be used as usual. Since nearly all electret microphones will connect "TIP" and "RING" together, the DC bias voltage will be present also on the mic input signal, but C3 will conveniently block that from the following amplifier stage. If an unbalanced dynamic microphone is plugged in, its mono plug will short the bias voltage to ground (no matter, the 2.2 kΩ resistor limits current) and its audio signal will be fed into the XLR's "HOT" line as usual.
	To isolate the "TIP" and "RING" contacts, I simply cut the two traces leading to them. Isolating the switched contact, however, was a pain. I first ground away the copper all the way around the solder joint, but that did not isolate it! Apprently there's a ground plane on the front side of the board as well. Since the connector body itself would be in the way, I wouldn't be able to do the same on the front side, so I went ahead and ground all the way through the circuit board from the back side! See a bigger photo of the whole board (not yet ground through) here, and see a picture of the completely ground through solder joint here. Yes, that's your basic circuit modification!  :) I had already tried to isolate the switched contact of "RING" as well (you can see that in the above photo; that cut did not isolate it either), but luckily I realized that that switched contact does not need to be isolated after all. It would have been a lot more difficult to grind, as it's so close to an SMD resistor. (I had already slipped once, and inadvertently cut through one trace while grinding out the "TIP" switched contact, so who knows how much damage I would have caused.) An alternative to grinding the board like that would be to leave a stereo 6.3 mm–3.5 mm adapter permanently in the jack, even when using the XLR input. That disconnects the "TIP" contact itself from its switched contact. But it's not as elegant a solution.
	I then wired the "TIP" terminal (mic signal input) to the positive end of C3 (and soldered a jumper wire to fix the damage I had caused by accident), and connected my circuit (with the two resistors, capacitor and zener, soldered together "ugly style" and covered with heat shring tube) to the "RING" terminal (bias voltage output) and "SLEEVE" terminal (ground) of the jack, and to R1 (+15 V), as shown here. Looking at the schematic, you should be able to deduce which of the red wires is part of the modification, and which just fixes the destroyed board trace when my hand slipped while drilling... Knowing the locations of various components on the front side of the board (from the photo I took during my first modification with this mixer), these latter two modifications were possible without having to remove the board from the chassis.

With all the modifications finished, I can plug a headset into the headphone output and its microphone into the CH1 TRS input with nothing more than 6.3 mm–3.5 mm stereo plug adapters. Then I plug my "TRRS-CTIA" adapter into my laptop, connect its mic input to the "Tape out" RCA connector as described in the previous modification, and its headphone output to the CH2/3 stereo input with a 3.5 mm stereo to 2× 6.3 mm mono cable. Now I adjust the headphone audio as usual (by adjusting the PC's output level, CH2/3 fader and headphones volume. While speaking into the microphone, I adjust CH1 gain and the PC's software mixer so that the volume is correctly set in the PC. Finally, I increase the CH1 fader until I hear just enough of my own voice in the headphones to make speaking feel natural with the headset on.

A word of warning