Dell XPS 15 7590 audio troubleshooting

[Sharx112]

Hello guys. After googling for 3 days and being unable to resolve the audio issue for some time I was looking for the schematic and I found this wonderful website.
It all started from sound just disappearing at some point that I even forgot when. Why did I forget? Because even though speakers and 3.5mm give no reaction whatsoever it works properly with type-c earplugs so I kinda kept using it like that. Now that I was doing an overhaul with repasting, changing pads and battery it struck me, hell, why not look for the sound issue? So here I am, after trying every possible software solution (drivers, bios, etc.)

As for the board, there are no apparent burns, the speakers show 3-4ohms each, however when I was checking the 3.5mm jack with multimeter and schematic I noticed too many grounds (shorts). I am far from a laptop repairman but I got the basic skills to read the schematic and use a multimeter so don't throw your slippers at me if I write newbie questions

Okay, so while in diode mode the speaker lines were okay but the mic lines seem to be shorted (Sleeve/Ring2). I am unsure, is it really a dead realtek here? Because it even has bidirectional TVS diodes on each jack line so I would think it's not that easy to burn it. Then again, if it did burn, why does the sound work from type-c?

In windows it doesn't show any input/output audio devices and doesn't react to headphones plug. I would think there's 3 options to this:
1) bad 3.5mm port;
2) bidirectional TVS diodes on mic burned and now make the mic shorted;
3) realtek chip died.

Would be very grateful for some help with this just so I know if its worth to find the reason or just use earphones and forget it. While there is also an option to just resolder the speakers to a cheap BT receiver PCB and connect it to a 5vdc line inside.

I've added the parts of the schematic in question below. Thanks in advance.

[Sharx112]

Since no one replied yet, I thought I'd update what I found. There was a short on mic line, but I didn't have the time to check it all. Today I tried the easiest thing which was to test the diodes with multimeter and indeed, DA5 is in short. The ones on speaker lines have infinite resistance since they are okay. Now I'll just need to find time to disassemble it for the 5th time to also check DA4 since it's on the hidden part of the motherboard and pray the realtek chip is alive.

[piernov]

If there's not even an HD Audio device showing up, it's unlikely the codec IC is working at all. So need to take some voltage and resistance measurements as usual. One of the ferrite beads on the power rails or the IC itself could be damaged.

[Sharx112]

Hi and thanks for your reply. There is an HD audio device in device manager and the windows utility doesn't find any faults just that no input/output devices since mic line is shorted. I checked the power rails for shorts with multimeter and they were clean. Didn't check voltages as I was almost sure the IC itself is dead. I'll try to change the diodes next weekend and see if the short disappears. If it doesn't then I'll check the voltages and post again. Thanks

[mcplslg123]

Since type-c earplugs are working,i beleive audio codec is ok. Try de-soldering the headphone jack and see if you get any output on speakers.

[Sharx112]

Thanks for replying. This was actually the essential question that bothered me since I'm not entirely sure how the audio circuit is supposed to work. If I assume audio from type-c port also comes from the realtek chip digital part then the chip is alive. If not then I don't see any point to find out the fault since I'm unable to replace the chip with the tools I have at home and even after googling parts most sellers sell the chip in a different form factor than the one I have. And after thinking some more a short on mic line shouldn't matter anyway since if you plug in 3pin headphones, mic is grounded anyhow. I'll try to unsolder the port and then check the diodes again, probably on weekend. Maybe the plug sensor diode is also shorted, then it would make sense. Thank you guys for suggestions.

[piernov]

I'm still not sure what "no input/output devices" mean. The audio codec would typically show up as a "Realtek Audio" in the Device Manager under the "Sound, video and game controllers" section, provided the Realtek drivers are installed (with older drivers it would say "Realtek High Definition Audio"). It should have the Realtek hardware vendor ID 0x10EC. Without the Realtek drivers and using Windows generic drivers it may show up as "High Definition Audio Device", and in that case you should probably install the Realtek drivers.
You may see another device by Intel related to the HDMI audio output but that's irrelevant here.
The "High Definition Audio Controller" under "System devices" is the part inside Intel's chipset and should have hardware vendor ID 0x8086.
If all that is good and of course there are not errors, you should at least see some audio inputs and outputs in the "Playback" and "Recording" tabs of the "Sound" dialog (better look at the old one, not the Windows 11 settings panel, if you're in there click "More sound settings"). They may show up as "Not plugged in", which would at least make sense for the "Headphones" output if the jack detection is damaged.

USB headphones should have their own embedded audio codec, they should show up as a completely separate audio device.

[Sharx112]

Thats the issue, I do have a device in device manager. At first it was a High Definition Audio Device, later on after I tried a ton of drivers thinking it was a driver issue it became Audio for monitors with ID 8086. In sound settings there are no connected devices (even grey inactive ones). After I plug in 3.5mm headset there's no reaction, but if I use type-c headset with mic it works perfectly, both sound and mic and they are displayed in sound input/output devices with Audio for Monitor device being grayed out. So the realtek device is missing. But its strange to think it just burned without any reasons, right? After reading up for a few days usually when it dies the speaker output should be shorted (left, right or both) but in my case the speaker outputs are okay but the mic outputs are in short. I'll check the voltages as you mentioned in the first reply when I disassemble it again and try to desolder the port to see if thats why mic is in short.
But the first question was why the usb-c works if the realtek chip is busted, so it does use the intel chipset via usb and then decodes it irrelevant of realtek chip? If yes then chances are I'm wasting time 🥲

The reason why I thought the realtek chip is alive was because the sound is working via type-c, but also with these dell xps models, there are a lot of sound related issues like after you plug in an HDMI to a monitor and unplug it, the realtek device dissapears from device manager and takes the sound devices with it. And other strange things. In my case the internal speakers aren't even directly connected to the realtek chip but via a daughter board. The assumption was that the chip couldn't have burned just from the 3.5mm jack being faulty.

[Eniaro]

Anything audio USB-related usually has its own DAC and functions as an additional sound card... (This may not apply to USB C Hubs)
HDMI audio is a part of the video card which operates its own sound card.

If you have audio from your internal speakers the internal DAC is fine, you may have issues closer to the 3.5 jack

Have you double-checked if you have the correct drivers installed, if you have a 3.5 TRRRS jack, software is usually used to prompt the user requesting information as to what was plugged in from there its adjusted accordingly, do not use Windows update but instead straight from Dell

[piernov]

If there is no Realtek device with hardware vendor ID 0x10EC (Realtek) in the Device Manager, there we're back to what I was saying in post #3.
The device with hardware vendor ID 0x8086 (Intel) is unrelated as it is just for the HDMI output.

Also it is much more likely to damage the audio codec through the external audio jack rather than something happening with the internal speakers. You can plug whatever in the audio jack, and despite the TVS diodes, if you plug e.g., a microphone with phantom power you'll cause irreversible damage. This is something we sometimes see on MacBooks used for professional audio production.
For the internal speakers, what can get damaged is the speaker audio amplifiers if they are overdriven all day long. On top of that, looking at the schematics of this machine, the link between the Realtek audio codec and the audio daughter board is digital (I2S) not analog, so the audio daughter board has its own DAC (or it is part of the speaker amplifiers).

[Sharx112]

Thanks, this actually clears up most of the things then. I did check the daughter audio board just to see what it was. It does have a separate IC (which is probably the DAC) but there's no schematic for it so cant really say much besides there not being shorts to the ground there. As you suggested, I'll try starting from the voltages on the realtek IC when I open it up again. I don't think anybody connected anything besides a headset to the 3.5mm port which is why I thought it's strange for the mic output to be shorted. If I saw, for example, the speaker outputs being shorted I wouldn't even be asking these questions lol.
I already started feeling a slight disgust to dell because audio I can understand is a frequent fault on any OEM however I also replaced the old battery and ffs, if it's 40%+ charge left it charges properly. If its below 35% left it refuses to charge blinking a battery error but passing all diagnostic checks. And the old battery with 50% wear works and charges properly...
So thanks for your answers, I'll get back with the voltages when I have time.

[Eniaro]

I don't know if this helps... the daughter audio board does seem to have the DAC on it, if this is true then the part number for the entire board is WD50F and is used in XPS 15 7590, XPS 15 9570, and Precision 5530

[Sharx112]

Hello again. Today got the time to measure the voltages since I was busy with work and then had to replace the back cover of the LCD, so finally got to the insides.

I checked all the voltages on the ALC3266 chip, 1.8v, 3.3v, 5v are present on the pins just as schematic suggests. Only one I didn't check was pin10 because there's no pads there besides the pin and I was afraid to short it with the multimeter.

Piernov, you said there's a possibility to short the IC with a phantom mic voltage. While I was checking also checked the 3.5mm connector for voltages and figured 3v comes into jack for "knowing" when a jack is inserted. Could you also short the IC if those somehow went into mic output of IC?

I guess the next step is unsoldering the 3.5mm slot to see if it changes anything and if it doesn't then also taking out the TVS diodes on mic output?

[Sharx112]

Update:
Desoldered 3.5mm jack port and DA4, DA5 - still got a short between Ring2 and Sleeve (pin36-37 on IC). I guess further actions are wasted time since I doubt any of the small capacitors on output could be causing this so 99% chance the IC is dead. But then how come it gets all the voltages on it, like shouldn't it be disabled by the intel bridge IC?

[rogfanther]

The signals to the jack come from the IC on the daughter board, not from the Realtek chip on the motherboard. If you do not have the necessary equipment, maybe it is easier to replace the audio board. Eniaro posted the model of it in post #12.

[Sharx112]

The daughter board is working, however there's a short circuit on the realtek IC mic output. Since I haven't got the equipment AND the IC can't be found on the internet in the correct formfactor I smacked inside a BT receiver board for $1 and connected it to the speakers. It works, I guess, tho not the solution I was hoping for xD