• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Possible Defect With Ryzen 5700G APUs

Associate
Joined
2 Dec 2022
Posts
690
Location
-
Motherboard: B550 Aorus Elite V2
RAM: 2 x 16GB Crucial Ballistix Sport LT (Dual Rank)
CPU: Ryzen 7 5700G
GPU: Nvidia GTX 1650
DAC: Topping E30

I've been using a 5700G APU with the integrated graphics for a few months with no issues, but the graphics performance wasn't quite enough so I installed a GTX 1650. As soon as I upgraded I noticed major audio crackling issues and occasional screen stuttering/freezing, especially when the GPU is under some load. One way to easily replicate the issue is to play a video with madvr NGU upscaling (1080p to 4K), during which data is constantly copied from the GPU to system memory. This issue also occurs when using the integrated realtek 3.5mm headphone jack.

At first I thought I might have a faulty GPU or motherboard so I tried installing the 5700G into another system with a B450 Tomahawk Max and GTX 1060, but the issue persisted. I've tried many things to fix it including upgrading/downgrading the BIOS, resetting BIOS to defaults, clean install of both Windows 11 and 10, upgrading/downgrading AMD chipset drivers, using a different DAC (Apple 3.5mm dongle), changing the DAC sampling rate. Connecting the DAC to a different USB port did significantly improve the issue but didn't fix it. The only thing that completely eliminated the audio crackling was setting PCIe bifurcation in the BIOS to 2x8, but it's not really a solution if you are limited to half PCIe speed with a performance penalty.

I was sure my 5700G must be faulty so I returned it and purchased another one, but the new one had the exact same issue. I tried a different CPU (5600x) and the issue magically went away. Is there anything else I could try to fix it? Is it possible the 5700G APU has a major defect which isn't getting attention because so few people are using it with a discrete graphics card?
 
Last edited:
What happens if you plug the USB DAC into the end of a USB extension cable? That way you're distancing it from electrical noise.
 
Is it possible to use an onboard USB header instead (usually they're routed through the case)?

The GPU having 8 lanes will be largely irrelevant to performance, particularly for a 1650. I think you'd need a way faster GPU to find a statistically noticeable bottleneck. In TPU's testing, even a 4090 only lost 2% for losing 8 lanes (albeit, PCI-E 4.0 lanes), so I'd be surprised if you even lost 1%.

The 5600X is PCI-E 4.0, whereas the 5700G is PCI-E 3.0, I don't know if that might make a difference to their electrical behaviour.
 
Last edited:
Is it possible to use an onboard USB header instead (usually they're routed through the case)?
The front panel USB ports are connected to the onboard USB header, I already tried them and the same issue occurs.

The GPU having 8 lanes will be largely irrelevant to performance, particularly for a 1650. I think you'd need a way faster GPU to find a statistically noticeable bottleneck. In TPU's testing, even a 4090 only lost 2% for losing 8 lanes (albeit, PCI-E 4.0 lanes), so I'd be surprised if you even lost 1%.
I don't really want to accept any performance loss as a solution even if it's small. A 5700G is limited to PCI-E 3.0 so the 4090 would likely lose more than 2%.

The 5600X is PCI-E 4.0, whereas the 5700G is PCI-E 3.0, I don't know if that might make a difference to their electrical behaviour.
The 5600x also worked fine with the B450 Tomahawk Max which is limited to PCI-E 3.0.
 
I don't really want to accept any performance loss as a solution even if it's small. A 5700G is limited to PCI-E 3.0 so the 4090 would likely lose more than 2%.

Yeah, but you don't have a 4090 :p With respect, the 1650 is hardly a speed demon, so you will never notice outside of 3D Mark, I doubt it can even saturate a PCI-E 2.0 bus. Obviously a proper solution is ideal, but this would concern me about as much (from a performance perspective) as having a red case instead of a blue one.
 
The onboard soundcard is usually directly under the gpu so the odds are you're getting interference from the 1650 rather than the 5700G being dodgy...

Stuttering - check the bios to see if you have an updated bios that fixes the ftpm bug, or even better turn ftpm off if you're not using it (I had stutter on a 5950x due to it)
Not sure if this is possible, disable the onboard gpu in the bios, it's basically the only real difference between the 5700g and a 5600x... why did you change back, the 5600x is slightly faster...

What speed are the usb sockets you're using, usb 3 can cause interference with wifi peripherals such as keyboard/mice (2.4ghz frequency range) so it wouldn't be impossible that they're introducing noise into the 'audio setup' as well.

I've also read about realtek software causing audio issues and removing it helped.


Honestly though if PCIe bifurcation fixed it did you actually do any benchmarks/testing to see if actually cost you any performance... but like Tetras said, the 1650 is hardly a speed demon in the first place.
 
Last edited:
The onboard soundcard is usually directly under the gpu so the odds are you're getting interference from the 1650 rather than the 5700G being dodgy...
The DAC is about 1m away from the PC so electrical interference shouldn't be an issue.

Stuttering - check the bios to see if you have an updated bios that fixes the ftpm bug, or even better turn ftpm off if you're not using it (I had stutter on a 5950x due to it)
Not sure if this is possible, disable the onboard gpu in the bios, it's basically the only real difference between the 5700g and a 5600x... why did you change back, the 5600x is slightly faster...
Already tried disabling FTPM, updating the BIOS and disabling the iGPU. Made no difference. The 5600x was taken from another system for testing purposes only, but it isn't faster than a 5700G.

What speed are the usb sockets you're using, usb 3 can cause interference with wifi peripherals such as keyboard/mice (2.4ghz frequency range) so it wouldn't be impossible that they're introducing noise into the 'audio setup' as well.
I've tried every USB port, 2 of them are USB 2.0 and the rest are 3.0. I'm not sure how interference could be an issue because the USB connection is digital not analogue.
 
Last edited:
The 5600x also worked fine with the B450 Tomahawk Max which is limited to PCI-E 3.0.

Ahh, good point. It might be worth emailing the motherboard makers and asking them for their thoughts. There are dozens of BIOS options that never get touched * and it's possible one of them changes the electrical behaviour, in a similar way to bifurcation. I guess it could just be an interference issue that is resolved by pushing the bus less hard, but it might be something that one of these 'hidden' options could address. (* I mean things like spread spectrum)
 
Last edited:
I've tried every USB port, 2 of them are USB 2.0 and the rest are 3.0. I'm not sure how interference could be an issue because the USB connection is digital not analogue.
https://www.usb.org/sites/default/files/327216.pdf
It's a pdf but the title is USB 3.0* Radio Frequency
Interference Impact on 2.4 GHz Wireless Devices

Yes your dac is digital but usb 3 is able to interfere with certain frequencies and seeing as we've all heard our mobile phones in our speakers it's not outside the realms of possibility of being part of the issue.... and I wouldn't discount it being a compatibility issue on the dac end based on a quick google.

Just looked into the dac and motherboard... can you not connect via optical, it would still be digital.
 
https://www.usb.org/sites/default/files/327216.pdf
It's a pdf but the title is USB 3.0* Radio Frequency
Interference Impact on 2.4 GHz Wireless Devices

Yes your dac is digital but usb 3 is able to interfere with certain frequencies and seeing as we've all heard our mobile phones in our speakers it's not outside the realms of possibility of being part of the issue.... and I wouldn't discount it being a compatibility issue on the dac end based on a quick google.

Just looked into the dac and motherboard... can you not connect via optical, it would still be digital.
USB interference wouldn't explain why the 5600x works fine, why the 5700G works in PCI-E x8 mode, or the occasional screen stutters. As mentioned in the original post I tried a different DAC (Apple 3.5mm dongle) which had the exact same issue. The optical port is used with a PS5 so it needs to connect over USB, but I will test it to see if it has the same issue.
 
Last edited:
Upon further testing I was mistaken about one thing in my original post, using the onboard realtek 3.5mm jack does not fix the crackling, it's just less frequent than the USB port I was using. I tested it for longer this time and the crackling came back eventually.
 
Last edited:
If you google the audio crackling with Ryzen, there are so many results. It is blamed on different things: drivers, XMP being overly optimistic, voltages (too low or too high, particularly for the I/O). Have you gone through these kind of search results?
 
If you google the audio crackling with Ryzen, there are so many results. It is blamed on different things: drivers, XMP being overly optimistic, voltages (too low or too high, particularly for the I/O). Have you gone through these kind of search results?
Yep, already tried increasing VSOC voltage to 1.1v, VDDP to 0.95v, the G series APUs don't have a VDDG voltage because there's no IO die, XMP was disabled after loading BIOS default settings. The only thing that 100% fixes it is running in PCI-E x8 mode. Considering I've tried 2 5700G's in 2 different computers I think there's a serious problem with PCI-E on these G series APUs, but not many people are running them with a discrete GPU under the right load conditions to notice.
 
Considering it's crackling with 3.5mm, optical (which shouldn't be affected by any of the other issues) and usb I'd say there might be a compatibility issue with the dac, not the pc, and you were just lucky or it's intermittent when you tested it with the bits where it didn't appear in your testing window. I've also seen plenty of comments online about the Topping E30 and crackling so it's not a new thing for the device.

How are you powering the E30, it seems the power supply is basically a usb cable without the data going to a circular pin, right next to the usb socket so I wouldn't be shocked if they're going through a single controller chip. Can you try a different power adapter?

USB interference wouldn't explain why the 5600x works fine, why the 5700G works in PCI-E x8 mode, or the occasional screen stutters. As mentioned in the original post I tried a different DAC (Apple 3.5mm dongle) which had the exact same issue. The optical port is used with a PS5 so it needs to connect over USB, but I will test it to see if it has the same issue.
Slight difference in the usb controller revisions, differences in the silicon so it draws different power levels etc... there's a lot of different variables that could account for some of it. I wouldn't exactly rate the apple dac as a good benchmark in all honesty, especially on something non apple.
 
Last edited:
Considering I've tried 2 5700G's in 2 different computers I think there's a serious problem with PCI-E on these G series APUs, but not many people are running them with a discrete GPU under the right load conditions to notice.

It's curious that the previous gen APUs couldn't even run 16 lanes.
 
Considering it's crackling with 3.5mm, optical (which shouldn't be affected by any of the other issues) and usb I'd say there might be a compatibility issue with the dac, not the pc, and you were just lucky or it's intermittent when you tested it with the bits where it didn't appear in your testing window. I've also seen plenty of comments online about the Topping E30 and crackling so it's not a new thing for the device.

How are you powering the E30, it seems the power supply is basically a usb cable without the data going to a circular pin, right next to the usb socket so I wouldn't be shocked if they're going through a single controller chip. Can you try a different power adapter?


Slight difference in the usb controller revisions, differences in the silicon so it draws different power levels etc... there's a lot of different variables that could account for some of it. I wouldn't exactly rate the apple dac as a good benchmark in all honesty, especially on something non apple.
It's also crackling with the onboard audio and DAC unplugged, it's just less frequent than the initial USB port I was using which had constant crackling. Same problem with the apple dongle. It's not an issue with the topping DAC but to answer your question it's powered with a regular 3 pin plug.
 
Last edited:
I have a similar audio crackling issue on my laptop with 3300U, and haven't managed to find a solution yet.

However, it only started after I replaced the laptop screen, so I put it down to that. Initially I thought i'd damaged the speakers, but it happens over headphones as well.
 
Back
Top Bottom