Behringer DCX2496 Noisy Output Repair

Fixing the "frying eggs" problem


The Behringer DCX2496 is the go-to loudspeaker management system for those who aren’t made of money. Good features, meets the specifications, unbeatable price. Popular in live sound applications, with some appreciation in audiophile communities too. What’s not to like?

Reliability, of course. The Behringer curse. It’s getting better lately though!

Anyway, these units are older designs known for developing various problems, the most well-known of which is probably “frying eggs”. This problem manifests as the outputs developing a raspy background noise with some audio distortion, which sounds a lot like frying eggs. There are various theories as to what causes this - PCBs sagging and shorting against the case, badly made ribbon cables, and so on and so forth. Folk remedies include putting cardboard under the PCB, and cutting the IDC connectors off the ribbon cables in order to solder the wires directly to the board.

Overall though, these units are cheap enough that if used in a revenue generating application, you can pretty much afford to buy a new one if it fails. They’re not taken too seriously. As one forum user puts it: “I don’t remember the fix but I’m pretty sure the sound was frying bacon, not eggs. If yours is frying eggs, then it is probably a whole new problem.”


My DCX2496 developed Rapid Onset Frying Eggs. No problems for years, and then one week the problems appeared and got progressively worse. It started as raspy background noise that would come and go in the first few minutes after power-up. Thermal issues! I thought it sounded a lot like potential ribbon cable problems, but percussive maintenance and reseating the connections made no difference. I put it back in the rack and dealt with the noise, but eventually the noise stopped going away, and there was plainly audible distortion in the output audio. It got to the point that it was unusable.

Scope traces comparing a good channel to a bad channel reveal the pathetic state of the signal on the failed channels, 1 and 5:

Good channel Bad channel

As usual for me, none of the common fixes that work for everyone else were applicable here. They may have worked for those forum users, on whatever revision of the hardware they had, but they weren’t relevant for me. I’m also not in a position where I can drop $300-400 USD to replace something without at least trying to fix it first. So I tried, and was successful.


I tried to start clean and work through the issues methodically. Only two of the six outputs had problems. Since each of six outputs can be mapped to one of (or a sum of) three input channels, I tried remapping to see if it made any difference, which it didn’t. A bad output was always a bad output, and a good one was always good.

In a related incident I had just recently found some undocumented test modes in the DCX2496, one of which digitally generates a sine and sends it to all six outputs. This completely isolated all of the input circuitry, and narrowed down the issue to either the DSP, or the output DACs, or the analog stage after the DACs. The problem could not possibly be in the input stage.

Looking at the PCB and comparing it to the AK4393 DAC pinout, we can see the serial data input for each DAC is run through a 100R resistor right there on top of the board. With three DACs driving two channels each, and one of those DACs having both channels working, I did a bodge job of lifting two resistors and using a wire to swap the data leading to one of the good DACs to the input of one of the bad ones. The output on the bad DAC was still noisy. This eliminated the DSP as a potential cause of the problem.

Hack job to test the DSP

The analog output stage was then eliminated as well. It’s on a separate board and connected over one of the ribbon cables, so I just pulled the cable and verified I could still see the noisy output when probing directly on the connector.

The problem was now confirmed. Two of three DACs had failed.


With the problem isolated to the DACs, I went looking for replacements. The original AK4393 was readily available on AliExpress, but I had seen some notes on audiophile forums about the AK4396 being a pin-compatible replacement with superior sound quality. I’m not really an audiophile and was already happy enough with the sound quality, but I ordered some AK4396s anyway for about $5 each. My reasoning was that since this is such a widespread problem, and two of the three DACs in my unit are confirmed bad, then maybe the AK4393 is a flawed component with a high failure rate. Who knows!

The DACs arrived some time later, and I got to replacing them. Hot air, solder paste, and some manual touch-up work with the iron was all it took. Here’s a pic halfway through the process:

One DAC replaced, working on the next, and another final one remains. Brown residue under the missing chip is old burnt flux.

Might have gotten a bit careless with the hot air and overheated some of the coupling capacitors, but no matter, they’re easily replaced later. I’ve done a fair amount of hand work and reflow for SMD assembly, but hot air rework is something I’m still learning.

After I was done, I repopulated the resistors in their original configuration, and confirmed the problem was solved. The DCX2496 lives on!

Built with Hugo
Theme Stack designed by Jimmy