Data/Feedback required. Please test/try something for me

Separate area for niche lag issues including unexpected causes and/or electromagnetic interference (ECC = retransmits = lag). Interference (EMI, EMF) of all kinds (wired, wireless, external, internal, environment, bad component) can cause error-correction latencies like a bad modem connection. Troubleshooting may require university degree. Your lag issue is likely not EMI. Please read this before entering sub-forum.
Forum rules
IMPORTANT:
This subforum is for advanced users only. This separate area is for niche or unexpected lag issues such as electromagnetic interference (EMI, EMF, electrical, radiofrequency, etc). Interference of all kinds (wired, wireless, external, internal, environment, bad component) can cause error-correction (ECC) latencies like a bad modem connection, except internally in a circuit. ECC = retransmits = lag. Troubleshooting may require university degree. Your lag issue is likely not EMI.
🠚 You Must Read This First Before Submit Post or Submit Reply
User avatar
ChristophSmaul1337
Posts: 111
Joined: 11 Feb 2024, 21:01

Re: Data/Feedback required. Please test/try something for me

Post by ChristophSmaul1337 » 12 Aug 2025, 16:18

Kipperii wrote:
12 Aug 2025, 03:44
Did you keep the previously problematic PSU/PC to test this theory? Would love to hear your results.
Yes, of course. I still have all my hardware, I usually keep it around anyways. I have all the PSU's here to try, except the ones that blew up, or aren't operational otherwise, of course.

I have obtained basically a whole new setup, because I needed a CPU upgrade over my old 10900K (can't handle BF6 with >100 FPS), so I got a super good deal on a used 14700K with an ASUS PRIME Z790-P D4. Also, my 2080Ti died the other week, so I had to replace it with a used 4070 Ti. This setup still won't give me any problems, on any PSU that will fit the required wattage. I ran this setup for testing with a 150W power limit on the CPU with the 550W PSU, without any problems.

For now, everything still works and I didn't encounter problems ever since. If I did, I would've updated this posting already. Again, I'm not sure about what's exactly happening here and how it can be explained, but all it takes to make desync (somewhat) return, is for me to take the PC's power cord out of it's current socket and move it to the "last position" on the strip. However, it's not nearly as severe as it was before I figured this out.

All of this might be unnecessary, and the culprit might have been a faulty device at a neighbour's house, which has been removed exactly when I was testing stuff around. You never know, and I will continue to look into it, try to make sense of it and try to help best I can.
Slender wrote:
12 Aug 2025, 12:31
What's the point? The extension cord is the same circuit as the sockets on the wall in a row.
The point isn't about EMI/RFI/Noise. I might be totally wrong, but what I think about right now is more related to harmonics, and it goes like this: Your power company delivers you a somewhat clean sine wave, right at your service entrance / electrical panel. Depending on where you live and what's around you (industry, residential, etc), your sine wave might be more or less distorted already. However, it will always be within tolerance and within acceptable limits, so any electrical device that's been certified for sale will be able to handle it without problems.

If we "follow" this - acceptable - sine wave down the line in your house, it might get distorted. Switched-Mode Power Supplies (SMPS) will distort a sine wave, because of the way they're drawing power. Usually, power is drawn when the sine wave it at its peak. This will create a flat-topped sine wave, given enough distortion from SMPS'es. You can read more about how that works here: https://powerquality.blog/2021/11/22/wh ... hemselves/

Now let's say your power strip is the "first" thing that's encountered from the breaker box. I think here is where the difference lies within: If you plug your computer into the "first" position, it will still see the (somewhat) clean sine wave from the PoCo, as it hasn't been distorted along the way yet. However, if there's already one monitor, or maybe 3 of them, some audio equipment, routers, switches, anything "in front" of where the PC is plugged in, these devices (their SMPS) will distort the sine wave. By the time the PC gets to "see" the sine wave, it might already be distorted beyond the acceptable limits.

I doubt a single network switch, or a single router, will make a significant difference to the sine wave, but if you're rocking 4 monitors like I am right now, these can draw some substantial amount of current. Especially these gamer monitors with high refresh rates can draw some good power, sometimes even over 100W when you crank the brightness. Now imagine 4 of those, that's 400W of power draw. That kind of current might very well have the ability to distort the sine wave just enough.

There's a lot of "moving parts", I think it depends on the incoming power quality for the most part. If your power quality at the entrance is near-flawless, it's probably irrelevant if there's 4 monitors worth of distortions on the sine wave. However, if your power company already delivers a sine wave that just about scrapes by the limits, anything might tip the scales.

Again, I'm not an electrical engineer and I have no idea if what I'm supposing is actually real. I like this approach though, as at least the flat-topping of sine waves and its adverse effects on electrical equipment is a real phenomenon backed by science. I would love to be corrected on this by an actual EE and learn something new.

User avatar
Slender
Posts: 1609
Joined: 25 Jan 2020, 17:55

Re: Data/Feedback required. Please test/try something for me

Post by Slender » 13 Aug 2025, 00:33

ChristophSmaul1337 wrote:
12 Aug 2025, 16:18
you can take your old psu with problem and just connect power cable to it. Do you hear some coil wine sound inside? TRRRR TRRRRR TRRRR TRRRRR TRRRRR, no?
Unfortunately, in 99% of cases, the problem is already at the input of the panel. You can get a good sine wave without your PC, but as soon as the PC turns on, your sine wave breaks. Well, it doesn't break, but when it goes through your PSU, it's already crooked. You mentioned that if the PC is after your 4 monitors, the desynchronization comes back... That's right, your 4 monitors create a high resistance neutral. Think of it as a current flow from the PC that wants to go back into the circuit, it will meet the current from your monitors on its way, and accordingly, the resistance increases.

Kipperii
Posts: 88
Joined: 26 Mar 2021, 03:19

Re: Data/Feedback required. Please test/try something for me

Post by Kipperii » 13 Aug 2025, 03:10

I've been testing the power strip order method for a day, but haven't come to a conclusion yet.

However, I have a question:
Could this be related to a ground loop inside the computer?
Would strengthening the computer's internal grounding significantly help?

User avatar
Slender
Posts: 1609
Joined: 25 Jan 2020, 17:55

Re: Data/Feedback required. Please test/try something for me

Post by Slender » 13 Aug 2025, 09:13

Kipperii wrote:
13 Aug 2025, 03:10
I've been testing the power strip order method for a day, but haven't come to a conclusion yet.

However, I have a question:
Could this be related to a ground loop inside the computer?
Would strengthening the computer's internal grounding significantly help?
internal "grounding" is not help you. External ground connected to pc doest help you too.
only if you make ext. ground and connect it to panel neutral bus.
edit: you can take ext. ground and connect to neutral on pc input, and take phase from socet.
free electricity and perfect pc.

User avatar
ChristophSmaul1337
Posts: 111
Joined: 11 Feb 2024, 21:01

Re: Data/Feedback required. Please test/try something for me

Post by ChristophSmaul1337 » 13 Aug 2025, 09:21

Kipperii wrote:
13 Aug 2025, 03:10
Could this be related to a ground loop inside the computer?
Would strengthening the computer's internal grounding significantly help?
That's EXACTLY what I've thought before, too. Read this: https://www.igorslab.de/en/the-imbalanc ... stigative/

I've thought about possible EMI coming from internal ground loops within the computer before. I still haven't found anything that goes further into detail apart from this one article, but I think there's merit to it. If this can cause a power connector to melt, there's some serious current involved.

In the comment section down below, the arcile's author writes this:
Am PCIe Slot gibt es nur 3 offizielle 12V-Pins (mit Reserve-Pin sind es 4), aber deutlich mehr Masse-Pins. So gesehen passt das schon noch. Ich hatte mit einer MSI RTX4090 SUPRIM X und 516 Watt Leistungsaufnahme knapp 13 Ampere am Slot, bei 600 Watt BIOSen sind es sicher 14 bis 15 Ampere. Das ist zwar schon arg sportlich, aber für satte 34 Gnd-Pins noch locker machbar. Ich mache mir da eher Sorgen ums Mutterbrett und die Interferenzen
The translation is:
There are 3 official 12V-pins (with a reserve pin it's 4), but there are distinctly more ground pins. Seeing it this way, it's still OK. I measured 13 amps at the slot with a MSI RTX 4090 SUPRIM X and 516 watts power draw, and it's surely in the 14 to 15 amps range with a 600W BIOS. That is pretty sportly, but absolutely doable for a whopping 34 ground pins. I'm more worried about the motherboard and the interference
I've read this a while back and though about a temporary workaround. I measured a voltage difference between the back of my GPU and the PSU screws. I tried to create a low-resistance path for some of the current back to ground via thick copper wires running directly between the PSU and GPU mounting plate. I've detailed this in a posting in another thread. It's here: viewtopic.php?f=24&t=13054&start=190#p114932. There's also an image attachment there which shows my, admittedly crude, implementation: https://imgur.com/a/ruXGZWr

This 100% exists and is a problem in the real world. It's most likely responsible for melting 12VHPWR cables on high power draw GPU's. If this can create enough interference, well the article's author certainly thinks so, and that guy definitely knows what he's talking about.

Now, here's the weird part about this. Since I've employed the "technique" in this thread, by plugging in all my devices in the "correct" order, I'm no longer so sure this is actually a thing. With my current setup, it's virtually impossible to break it again. Only when moving the plug it starts to get worse again. The workaround with the wire doesn't have any impact any longer. My system works with and without it. To add to all that, the back of my current GPU and PSU don't show voltage differences any more, and there's no more stray current on the DP cable, even with a grounded monitor, leading me to believe my old 2080Ti had a hardware defect somewhere and it was leaking voltage.

amorou
Posts: 218
Joined: 29 Aug 2022, 18:46

Re: Data/Feedback required. Please test/try something for me

Post by amorou » 13 Aug 2025, 11:50

ChristophSmaul1337 wrote:
13 Aug 2025, 09:21
Kipperii wrote:
13 Aug 2025, 03:10
Could this be related to a ground loop inside the computer?
Would strengthening the computer's internal grounding significantly help?
That's EXACTLY what I've thought before, too. Read this: https://www.igorslab.de/en/the-imbalanc ... stigative/

I've thought about possible EMI coming from internal ground loops within the computer before. I still haven't found anything that goes further into detail apart from this one article, but I think there's merit to it. If this can cause a power connector to melt, there's some serious current involved.

In the comment section down below, the arcile's author writes this:
Am PCIe Slot gibt es nur 3 offizielle 12V-Pins (mit Reserve-Pin sind es 4), aber deutlich mehr Masse-Pins. So gesehen passt das schon noch. Ich hatte mit einer MSI RTX4090 SUPRIM X und 516 Watt Leistungsaufnahme knapp 13 Ampere am Slot, bei 600 Watt BIOSen sind es sicher 14 bis 15 Ampere. Das ist zwar schon arg sportlich, aber für satte 34 Gnd-Pins noch locker machbar. Ich mache mir da eher Sorgen ums Mutterbrett und die Interferenzen
The translation is:
There are 3 official 12V-pins (with a reserve pin it's 4), but there are distinctly more ground pins. Seeing it this way, it's still OK. I measured 13 amps at the slot with a MSI RTX 4090 SUPRIM X and 516 watts power draw, and it's surely in the 14 to 15 amps range with a 600W BIOS. That is pretty sportly, but absolutely doable for a whopping 34 ground pins. I'm more worried about the motherboard and the interference
I've read this a while back and though about a temporary workaround. I measured a voltage difference between the back of my GPU and the PSU screws. I tried to create a low-resistance path for some of the current back to ground via thick copper wires running directly between the PSU and GPU mounting plate. I've detailed this in a posting in another thread. It's here: viewtopic.php?f=24&t=13054&start=190#p114932. There's also an image attachment there which shows my, admittedly crude, implementation: https://imgur.com/a/ruXGZWr

This 100% exists and is a problem in the real world. It's most likely responsible for melting 12VHPWR cables on high power draw GPU's. If this can create enough interference, well the article's author certainly thinks so, and that guy definitely knows what he's talking about.

Now, here's the weird part about this. Since I've employed the "technique" in this thread, by plugging in all my devices in the "correct" order, I'm no longer so sure this is actually a thing. With my current setup, it's virtually impossible to break it again. Only when moving the plug it starts to get worse again. The workaround with the wire doesn't have any impact any longer. My system works with and without it. To add to all that, the back of my current GPU and PSU don't show voltage differences any more, and there's no more stray current on the DP cable, even with a grounded monitor, leading me to believe my old 2080Ti had a hardware defect somewhere and it was leaking voltage.
As you can see psu metal case is connected to circuts ground , I wanted to lift them long time ago but got lazy , maybe I will now.
Attachments
643534 - Copy.jpeg
643534 - Copy.jpeg (239.01 KiB) Viewed 2515 times

Kipperii
Posts: 88
Joined: 26 Mar 2021, 03:19

Re: Data/Feedback required. Please test/try something for me

Post by Kipperii » 14 Aug 2025, 08:06

1. Are you on a computer, a laptop or a console? PC Desktop
2. Did the feeling change at all?
The game suddenly experiences a lag of about 0.1 seconds every few dozen seconds. However, in tests with CS2 and Rainbow Six, the lag seems to have improved. Plug in PC->monitor->router, and my FTTH modem is plugged into a different outlet.
3. Assuming it changed, does it feel better or worse?
better
4. Can you pinpoint what exactly feels better or worse?
(2)
5. If the feeling didn't change: Are you in a house or an apartment?
6. If the feeling changed to better: Unplug the computer from the 1st outlet and plug it into the last one. Does that worsen the feeling again?
Not tested yet
7. If the feeling changed to worse: Unplug the computer from the 1st outlet and plug it into the last one. Does that improve the feeling?
Not tested yet
8. Which country / region are you from?
Asia 3 prong
9. With "3-prong" plugs, does your plug/socket enforce polarity? Can you only insert the plug one way, or can you rotate it 180°? (Example: North America's NEMA 5-15 plugs enforce polarity, Germany's SCHUKO plugs don't)
only one way, it cant be rotate 180°

Also: I installed a new 6cm fan on the motherboard today. The small bracket attaches to the motherboard screws (direct contact). The 3-pin connector plugs into the motherboard's amp/fan label for power. My gaming performance instantly deteriorated a bit, which has made me more interested in ground loops inside the computer (on the motherboard). I've ordered ten 10AWG, 30cm thick, pure copper wires (with rubber sheathing). I plan to run some tests in the next few days to optimize the ground loop. (I don't have a deep understanding of electricity, but this is just for fun.)

Post Reply