It might be that, if it is due to the switching noise, it’s finding its way back to the ribbon cable … in which case the tape over the board won’t help.
You could test it with the amp disassembled with grounded metal foil just manually held between the display board and the amp.
Can try that as well.
I did email Loxjie through the email on their site but I’m not hopeful at all they will reply. 
Subjectively of course the amp sounds fine with less sensitive headphones but as the noise is very audible on the SE846s I have I am sure it is enough to destroy objective measurements. The fact that noise varies based on the volume setting but is independent of the actual volume but related to which sections of the display are lit to show the volume is pretty bad too.
Disheartening to see a great value product arguably crippled by one component. Leaving the LED unplugged for now and just memorized how to change settings blind.
I used mine almost exclusively through the balanced and didn’t have hum…but I do know how infuriating it can be to hear it =( I’ll have to look into plugging mine up again and see if I can hear the noise with my headphones both SE/Balanced. I’ll try to remember to do that this weekend…no promises 
I used SE for testing (15 minutes) and then never again…
It’s very important to note that you need to use rather sensitive headphones to hear it. My SE846s are under 10 ohm of impedance and 114 dB SPL/mW. The noise is very audible on that one. I haven’t tried them in balanced as I don’t have the cable. The noise is also audible on my Sennheiser Momentum 2.0 wireless (in wired mode)
Note that setting the P20 volume between -12 and -10 in balanced mode would blow my eardrums and perhaps blow my headphone drivers! I’ve never listened for more than few seconds above -20, even with my planar magnetics. Per my iPhone db testing, I generally don’t exceed 80 db.
The noise is audible even when the amp is not plugged into a source so you can turn it up without a signal and not blow your eardrums 
. It is also audible on lower volumes. It is entirely dependent on which crystals are lit on the display and not the actual volume level. I used -10 -11 and -12 as an example because -11 seems to give me the lowest noise and the difference between that and -12 is obvious.
I will try it with my CA Andromedas…pretty sure those are sensitive enough…I also have an adapter for Balanced - XLR for them… my bigger problem is finding the time to do said tests (wife and daughter, then relaxing sanity check take up the rest of my day on the weekends).
I would say if you don’t hear the noise on lower volumes you won’t notice it at the -12 -11 -10 levels I mentioned. The noise is not being amplified. Appreciate your testing.
When I was borrowing the Campfire Orions from this forum, and used it with the P20, I did not get any humming or ground noise or any noise which was actually a pleasant surprise, because most other amps I plugged it into had some sort of white noise. I was using it primarily using balanced connection to an XLR adapter.
Worst-case when I do have humming or white noise, I use the iFi IEMatch (balanced) or EarBuddy (unbalanced) and it gets rid of all my issues. If it’s EMI, then it’s probably due to cable not shielded correctly (in my case).
I have tested the SE output with the KZ ATE which is the lowest impedance set I have (16ohms) and also both the balanced and SE with Sony MDR-ZX100 (agains, because it is the lowest impedance set I have) and can’t find any trace of the noise you are referencing. Is it possible that you have an issue with the screen itself? If anyone close to you has the P20 and could lend you the screen that would be a good test.
BTW, I use the single ended output daily with the Custom Studios (because they don’t like balanced for some reason) and have not noticed any noise either, however they are 80ohm so it is probably not relevant.
This info was enough for me to order a replacement off Amazon. I used the advance replacement option so I’ll be able to compare the old with the new one.
This 4-Pin XLR male to 2.5mm balanced female adapter is available again on Massdrop. I already own a similar adapter from Norne Audio but picked up this Massdrop one as well. This is by far the best price I’ve ever seen this type of adapter ($20 USD). I thought about making my own, but this is cheap enough that it’s going to end up a wash anyway.
Interpreting the Loxjie P20 Volume Display - Predicting Settings on other Headphones
Following some questions in the HE-560 thread, I completed some simple math to interpret what the Loxjie P20 is showing with this -60 through 0 digital volume display.
Recap of Data Collection Process (from the HE-560 thread)
Equipment: FiiO Q5 DAC line out to Loxjie P20 (balanced out)
Summary of Data and Findings
| Headphone Model | Published Resistance (Ohms) | Published Sensitivity dB/mW | Loxjie P20 Setting Required for 75dB | Percent of Loxjie Display Range (-60 to 0) | Conversion Factor (Rated Sensitivity/Loxjie Display Value) |
|---|---|---|---|---|---|
| Audeze LCD-2C | 70 | 101 | -42 | 30.00% | -2.4048 |
| AQ NightHawk Carbon | 25 | 99 | -40 | 33.33% | -2.4750 |
| Focal Elex | 80 | 104 | -40 | 33.33% | -2.6000 |
| HiFiMan HE-560 | 45 | 90 | -32 | 46.67% | -2.8125 |
| MrSpeakers AEON Flow Closed | 13 | 92 | -32 | 46.67% | -2.8750 |
| Sennheiser HD-600 | 300 | 97 | -38 | 36.67% | -2.5526 |
| Published | Published | Estimate | Estimate | HE-560 Value | |
| HiFiMan HE-6se | 50 | 83 | -29.5 | 50.81% | -2.8125 |
Simple Correlations
I ran simple correlations in a spreadsheet between the variables, and found that headphone sensitivity ratings were an extremely strong predictor of the display setting:
Correlations:
Headphone Rated Sensitivity X Loxjie Scale = -0.9238 [VERY HIGH]
Headphone Rated Resistance (Ohms) X Loxjie Scale = -0.2161 [LOW]
For comparison – Headphone Rated Sensitivity X Headphone Resistance = 0.1479 [MINIMAL]
Predicting Volume Settings for other Headphones
The final column of the table is a simple ratio of Headphone Sensitivity divided by the Loxjie display setting. This follows from the high correlation. For example, the Focal Elex is: 104 / -40 = -2.6000.
Regarding the question from @rrwwss52 on the HiFiMan HE-6se, I used the conversion factor for the HE-560 (as the closest technological match). The estimate is shown in the bottom row: to reach 75 decibels would likely require a volume setting between -30 and -29.
Math: 83 / -2.8125 = -29.5
Summary
The Loxjie P20 volume display largely follows from rated headphone sensitivity. The resistance (ohm) ratings have little to do with the volume setting required.
And finally, this proves I have too much time on my hands.
Let us know the outcome.
These are good findings but I don’t can you use an iPhone to really properly measure headphone SPL? You don’t have a seal between the ear cup and the mic so you lose a lot of energy, and the iPhone’s mic is hardly an accurate instrument.
Good question. I take the high correlation as evidence that it’s possible. Seems pretty consistent with the perceived volume too.
It’s maybe a 90% answer. Those with real test equipment could reach 99% accuracy perhaps. But I’ll never own an audio test rig—method included for future improvements.
Unfortunately I do hear the noise on the second unit too. Perhaps a tad quieter on this one. It exhibits the same behavior on the unbalanced output, varying with the display. Keep in mind that the SE846 is incredibly sensitive (9 ohms 114 dB SPL/mW - nearly half the impedance of your KZ ATE and 8 dB SPL/mW more sensitive!). It could very well be that the noise is beyond the limits of human hearing on your IEMs. But the listed 97 dB SNR in unbalanced mode, I don’t believe it, not with the display plugged in.
At this point I would advise anyone who is using more sensitive cans seeking the lowest noise possible to unplug the screen. Maybe someone can find a mod. I just find it a huge shame that the amp is otherwise great - it’s just a non-essential feature that introduces imperfection in the output. If I were designing this amp, I have done without the screen in the design and made buttons to switch outputs.
That is certainly a shame. If I come across any more sensitive headphones/IEMs I will see if I can see the noise but I certainly can’t find it with what I have (using balanced input and both SE and Balanced outputs).
I’m not an electrical engineer by trade but I do like to tinker. For fun, going to see if I can borrow an oscilloscope to characterize the noise, and then maybe somehow put in some filtering caps in between the screen and the motherboard, to filter out the high frequency noise. If that makes sense.