There is no internal loopback unless you activated that manually. So it sounds more like a software based feedback.

A relay click would be the output muted, not the input. That could have a different reason - what does the unit show on the display in that moment?

Also, in case you use higher sample rates, the unit might falsely be triggered into DoP (DSD) mode by random signals. In that case disable the automatic DSD detection (in Setup - Device mode / DSD).

And again you did not mention the sample rate...

The analog inputs have no protection that switches them off.

Here's how it looks like for the Right input channel:

Zoom in:

Input device: Analog (1+2) (2- RME ADI-2/4 PRO SE)
Input: LINE_IN (Master Volume)
Channel: Right
Output device: Analog (1+2) (2- RME ADI-2/4 PRO SE)
Output: Default Output
Input RMS target: -12.0 dB
Actual RMS at 1 kHz: -11.8 dB
Sample rate: 192000 Hz
Input volume: 1.000, Sweep level: -12.0 dB
20 Hz .. 20 kHz flatness: +1.5, -119.9 dB
-3 dB points: 7.3 Hz,  14.340 kHz

14.34 kHz is when I hear the relay click and then the signal (output? or input?) turns off.

For example, this is the result for the Left channel under a same test conditions:

Input device: Analog (1+2) (2- RME ADI-2/4 PRO SE)
Input: LINE_IN (Master Volume)
Channel: Left
Output device: Analog (1+2) (2- RME ADI-2/4 PRO SE)
Output: Default Output
Input RMS target: -12.0 dB
Actual RMS at 1 kHz: -11.8 dB
Sample rate: 192000 Hz
Input volume: 1.000, Sweep level: -12.0 dB
20 Hz .. 20 kHz flatness: +0.0, -0.6 dB
-3 dB points: 7.3 Hz,  91.971 kHz

There are no relay clicks or signal cuts on this channel.

And at this moment there are no messages on the display.  In fact, when the outputs are muted, the graphical spectrum analyzer still shows the audio levels, but nothing comes out during this time.

Is that a TRS to 2 x TS split cable? Does the mute happen when it is only unplugged at the input side? Does the same output channel mute when you plug TS L in R input and TS R in L input? Did you disable the DSD detection?

Yes, it is.

Thank you, MC for the help.
I don't fully understand what you mean. Please draw a diagram/flowchart of the proposed test.

I haven't tried it. So far, I have tested Output 3 (Left?) - Analog Input 1, Output 4 (Right?) - Analog Input 2.
I don't have a TS-TS cable, so I use a TRS (output) to 2xTS (input) splitter cable.
But I can flip the input side to simulate the Right output channel to the Left input channel, etc.

I did it now, but haven't tested the system after that yet.

This is called "Soundcard Calibration" to account for any device non-linearity when measuring other devices.

Is there a possibility that the Op Amp at Output 4 has a defect causing ringing at high frequencies?
Is this the case for RMA?

I also tested it with a different signal source (USB device) and found that there were no problems with either Input at any signal level.

DSD Detection is disabled. No difference.

All other tests were conducted from the own headphone outputs (1/2 and 3/4) of the same ADI-2/4 Pro SE.
Loopback mode: Headphone Output 1 -> Analog Input 1 or 2; Headphone Output 2 -> Analog Input 1 or 2; Headphone Output 3 -> Analog Input 1 or 2; Headphone Output 4 -> Analog Input 1 or 2.

The only time I switched the output device to another was today, one time, to figure out the problem.

All tests were conducted in REW software (Soundcard Calibration, Sweep signal mode), if we were talking about a software signal source.

Latest available. FPGA 209, DSP 40.

Due to the silence on this particular case of mine, I decided to investigate further and switch the 3/4 outputs to the rear TRS outputs and perform the same loopback test with sweep tone. This will not be the same test since the linear drivers will be different, but still close.

I've thought about it, and it's likely something special about the load pattern is causing some ringing in the 4th channel of a headphone amplifier that was designed for a low impedance load (16-600 Ohms) connected to high impedance analog inputs (48 kOhms unbalanced).  I've never had any problems with real headphones connected to the PH3/4 and Pentaconn balanced outputs. So I decided to switch to line outputs, redirecting the same DAC 3/4 output signals there. In this case, the output is initially designed for a higher impedance load, and is possibly more stable under such operating conditions.

Now I need to make a 1/4'' TS to 1/4'' TS cable, and then try this idea.

I tend to think your unit has a hardware fault.
Specifically as the problem is limited to the right channel of 3/4 signal path.

One final test:
Set: I/O / Phones 3/4 / Mono to “ON” and repeat the test from Rear TRS set to output 3/4.

This way you can make sure left and right output carry the exact same signal.

And full factory reset is always a good idea to exclude a software problem:

While powering up, push and hold the VOL button (NOT the VOLUME dial) and Encoder 1 and Encoder 2, until you see the message: “Reset done“.

All user Setups and EQ Presets are deleted.
Take notes before, or store them through the Remote App, if you want to keep some.

I double-checked this with KaiS's suggestions:
- in the 'Mono' mode of the PH3/4 outputs, this never happens for the 'Right' channel.
- this also does not happen when both PH3/4 outputs are turned on in the test ('L+R', not just 'R'), despite the fact that 'L' output is not connected anywhere.
- resetting to factory settings did not change anything.

I have to try this. Thanks for the offer.

What makes you believe so?

I need to ask John Mulcahy(REW Author) what he thinks about my particular case.
Perhaps he also has some thoughts about this.

***************************************************************************************
Measurement Sweep

The Measurement Sweep signal is used by REW when measuring system response. It consists of a logarithmic sweep from half the start frequency to twice the end frequency (with an overall limit of half the interface sample rate) to provide accurate measurement over the selected range. If the start frequency is below 20Hz the signal begins with a linear sweep from DC to 10Hz, followed by a logarithmic sweep from there to the end frequency.

Well nothing is excluded. I recently heard from another owner that he had to return the device under warranty due to a defective op-amp.

Actually yes. I can get it to shut off with a simple sine wave, but the problematic frequency range is >6370-7650< Hz.
Everything below 6370 Hz and above 7650 Hz is stable.

I think this correlates with my post https://forum.rme-audio.de/viewtopic.ph … 07#p219207 where the sweep test is interrupted at about 7 kHz. The problem always gets worse with repeated tests (2nd, 3rd time).

Possibly because I switched to the Rear 3/4 ports, but the sweep test is now interrupted before it reaches 7 kHz.

Does the problem occur on 44.1 or 48 kHz sample rate too?

I lack a sine wave generator, app or online, with 192 kHz sample rate to test this here.
Playing 7 kHz @ 48 kHz sample rate on my ADI-2/4 Pro SE, firmware. V 35, did not reveal any problem.

In the coming days, I will conduct several more tests to double-check this problem with all the tips from this topic. As I said, the problem with the 4th output channel is fairly consistent and easy to reproduce under my test conditions.

In the meantime, if I decide to request a Return Authorization (RA) from my local dealer, what is the best way to describe the problem?

I think incoming units are subject to inspection and testing before replacement is made.
I want to make sure the problem is understandable to the recipient and also reproducible on their end.

Thanks for the idea, KaiS.

I have to try this too.

I think the easiest way to do this is to use the balance adjustment slider moved to the right for the PH34 outputs.

100% a hardware defect.

Just so everyone knows, I returned the device and received a new one in exchange. After a quick assessment, to my surprise, it does the same thing, only on the right channel (4th) and on the same frequencies. I think it's some kind of hardware "feature" and I'm leaving it as is. There's no point in doing anything about it anymore. Just remember to always use the L+R output configuration in all cases, not just R-ch.

Thank you, KayS! Would you like to trade your unit?

Both of my RME ADI-2/4 Pro SE devices were detected by the Flash Update Tool as HW-Rev.7. And both had an older version of firmware (v. 3x, I don't remember the exact one) out of the box, and did not work with Remote 2.0 software until I flashed them to firmware v.40.

I believe this "feature" is independent of level. I did this with the "Low power" settings on the PH34. But I can double check this at the IEM level. More important is the digital playback level, which should be between -6...-0dBFS. And also the frequencies played should be 5~20 kHz.

After seeing this behavior on two brand new devices, I accepted it as a "feature" and not just a faulty device I got from the seller. So less worry now, and I get around this problem by turning on both 3+4 (L+R) output channels at the same time, even if I only need 4 (R).

Thank you, MC!

I am glad to hear that my contribution has led to the root cause of this unusual situation being found.  There is no rush to fix this as it is very specific and technically never happens in real life scenarios.

My faith in RME company and my choice of device has been rewarded with this customer support.  Sincerely!

FPGA firmware 73 and 210 for the ADI-2/4 Pro SE should fix this issue - it does for us, so please try:

Windows:
https://www.rme-audio.de/downloads/fut_ … in_210.zip

Mac:
https://www.rme-audio.de/downloads/fut_ … ac_210.zip