A notice for russian speaking people: я написал эту статью на английском, поскольку видел много стенаний на форумах по поводу бракованных этих контроллеров с неработающими наушниками, и мне подумалось, что мой опыт может быть полезен не только русскоговорящей части интернета. Если вы совсем не бум-бум в английском, пишите, переведу 
Many threads on DJ related forums mentioned that Numark Mixtrack Pro 3 has very low headphones out level. I got my hands on one of such units and managed to find and fix the problem. As my experience may help other unlucky owners of this controller all around the world i decided to write this repair log in english. I'm terribly sorry for the language, it's not my native. And please take into account that i have no statistics and my particular case may not be the same as yours so please pay your attention to what you do.
So, this unit was brought to me by my friend, a beginner DJ who was "lucky" enough to buy this unit on ebay. A friend of mine thought his low headphones level issue had something to do with his software settings, but it turned out that the problem is hardware. No matter what settings nor what kind of software used - headphones signal were low and remarkably distorted. So i borrowed my own controller to a friend and took this one for repair. Here's what i found.
In order to open the unit you'll need to unscrew all the screws you may find. There's a dozen of screws on the bottom of the unit, and one more is located right in between two RCA connectors on the back. You'll also want to pull off all master section knobs including faders and navigation knob. Then unscrew three screws that hold connectors board in place and a number of screws holding master section PCB. Disconnect all the ribbon cables and two wire connectors to the right and left top of the board. Then board might be removed.
On the board (just to the right of master section) there is an IC marked "DRV604" This IC is both line output driver and headphone amplifier combined. Pin 23 of this IC is it's 3.3V power input. I checked the pin with multimeter and found that 3.3V line is down. Tracing the tracks first led me to the right side of the board, where the electrolytic capacitors located. Near the capacitors there are two ICs in tiny SOT23-5 packages. These are DC-DC converters that convert 5V input to 3.3V and -3.3V outputs to feed the audio part of the device. No 5V input on these as well. You may check this voltage on big 4K7 resistor. Further tracing led me to the middle left part of the board. Here, just next to USB connector you'll find a number of transistors with a bunch of R and C as well as 3.3V voltage regulator for the digital part. Here's how it work: Q2 is a MOSFET that feeds 5V from USB and down the line while it's drain (pin 3) is pulled to the ground. This is controlled by Q9 - a PNP transistor.
Here's the picture of these transistors:
On this picture Q9 is removed and you may see its place marked with "BCE" letters. B for base, C for collector, E - for emitter. During normal operation it's "C" pin must be down, forcing pin 3 of Q2 down as well. In my case Q9 seems to be dead. Whatever voltage i feed to it's base, collector stays high. It drives Q2 output low and cuts the power to all audio part of the schematic. I just wonder why i heard anything at all.
I found no exchange for Q9 in my stock. But then i took a more precise look to this part of schematic. It seems that Q9 base is just driven to a certain level from 5V bus through a resistor divider R154/R153. I have no idea why it has been done that way, but there is no external control to this parts. You connect USB, 5V bus is up, current flows through R153/R154 right to the base of Q9, it drives Q2 open, boom! It seems that we don't really need this Q9 guy. We just need to drive pin 3 of Q2 down to the ground. I picked up a random (well, almost random, 3.9K) resistor and soldered it between base pin of previously removed Q9 and nearest ground which happened to be voltage regulator pin.
I connected a board to USB cable and checked if i have power line back. Done! 5V line is up, 3.3V and -3.3V lines are here again. After i assembled the whole thing back i checked the sound to find that phone output provides good amount of power and the unit operates great! Everything seem to work perfectly.
Hopefully this article will help you get your controller back to life. Once again, i can't stress it too much: you have to understand what you're doing. If you don't then just show this article to a nearest electronics geek and let him do the job.