Reason most of you guys have input lag
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
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
-
amorou
- Posts: 242
- Joined: 29 Aug 2022, 18:46
Reason most of you guys have input lag
Please don’t move this thread to the EMI section. The root cause of most people’s input lag discussed here is this issue, and the EMI section gets very little visibility
Edit
Do not take action before reading 2nd post written by ChristophSmaul1337 , I was refering to system earthing type in my country +some eu countries only . I didnt checked other countries.
Still problem is the same solution will vary depending on earthing system your distrubtor uses , and you cant consume/use those hf , or you cant store it , it gotta go to earth or you gotta block via xformer
High frequency fields in neutral wont drain to earth , due to neutral ground bond is high resistance or absent . Also if you try to do bonding somewhere else other than panel vs , you will create a loop and make it worse.
What happens
Neutral carries high-frequency switching noise from other loads
This noise appears as common-mode voltage between line ,neutral and earth
PSU does not see a perfectly stable input reference anymore
Input emi filter capacitors couple this noise into the PSU primary
Some of that energy passes through parasitic capacitance of the transformer
Small high-frequency currents reach the secondary side
DC rails gain wideband ripple and phase noise
Why this creates lag / micro-jitter
Cpu gpu ram and chipset depend on stable clock timing obviously.
Clock generators use PLLs referenced to supply rails
Supply noise => phase noise => clock edge timing variation
Timing variation changes when operations complete
Frame delivery intervals become inconsistent
Input sampling intervals become inconsistent
We get , uneven frametime spacing , micro-stutter without FPS drop , inconsistent input response , unstable motion smoothness
How to fix it
Neutral to ground bond must have low impedance , obviously ground there must have low impedance too.
or
Isolation transformer with low interwinding capacitiy , but this is hard to do and some xformers sold as low cap are not low cap or they got leakage at core vs. , but emi approving labs vs must use them so you can ask them where they buy it and ask them to make one for you. First solution is easier but they will produce same result ( actually xformer will be better but neglibile kinda)
Edit
Do not take action before reading 2nd post written by ChristophSmaul1337 , I was refering to system earthing type in my country +some eu countries only . I didnt checked other countries.
Still problem is the same solution will vary depending on earthing system your distrubtor uses , and you cant consume/use those hf , or you cant store it , it gotta go to earth or you gotta block via xformer
High frequency fields in neutral wont drain to earth , due to neutral ground bond is high resistance or absent . Also if you try to do bonding somewhere else other than panel vs , you will create a loop and make it worse.
What happens
Neutral carries high-frequency switching noise from other loads
This noise appears as common-mode voltage between line ,neutral and earth
PSU does not see a perfectly stable input reference anymore
Input emi filter capacitors couple this noise into the PSU primary
Some of that energy passes through parasitic capacitance of the transformer
Small high-frequency currents reach the secondary side
DC rails gain wideband ripple and phase noise
Why this creates lag / micro-jitter
Cpu gpu ram and chipset depend on stable clock timing obviously.
Clock generators use PLLs referenced to supply rails
Supply noise => phase noise => clock edge timing variation
Timing variation changes when operations complete
Frame delivery intervals become inconsistent
Input sampling intervals become inconsistent
We get , uneven frametime spacing , micro-stutter without FPS drop , inconsistent input response , unstable motion smoothness
How to fix it
Neutral to ground bond must have low impedance , obviously ground there must have low impedance too.
or
Isolation transformer with low interwinding capacitiy , but this is hard to do and some xformers sold as low cap are not low cap or they got leakage at core vs. , but emi approving labs vs must use them so you can ask them where they buy it and ask them to make one for you. First solution is easier but they will produce same result ( actually xformer will be better but neglibile kinda)
Last edited by amorou on 26 Feb 2026, 18:38, edited 2 times in total.
- ChristophSmaul1337
- Posts: 114
- Joined: 11 Feb 2024, 21:01
Re: Reason most of you guys have input lag
Just adding some information you might want to be aware of before going to inspect your electrical panel. There's electrical installations that don't allow for a low-impedance neutral to earth bond, for various reasons.
A common example is the TT earthing system which not only doesn't require this bond, but explicitly disallows it. There's a good chance you do NOT want to install any low-impedance bond between neutral and earth on a TT system, even though you theoretically could. In case of a ground fault, without the bond fault current will go through the PE and the ground rod, through the literal earth soil back to the transformer. This creates small currents which aren't enough to trip a common 10A / 16A breaker; a RCD is therefore used to break the circuit in case of a ground fault. If you connect a low-impedance bond between neutral and earth, the fault current will go down the PE conductor until it reaches your electrical panel and then use the now-"PEN" conductor all the way back to the transformer as its return path instead of the earth's soil. If your installation is TT, there's almost always a reason why it's TT: Because the pseudo-"PEN" conductor you just created is too small in diameter to handle the fault current of a fully-fledged ground fault. In the worst-case, the momentary fault-current overloads the N conductor so much it causes a fire to happen. Don't fancy your chances.
Another old installation type, not commonly used today anymore, is TN-C. You get a 2-wire conductor to the receptacle and the neutral is tied to the ground prongs at the receptacle. This installation type does not have a ground rod, and there is also no way to bond the "ground" to neutral in the panel, because there won't be a ground to bond to. This installation requires bonding of "ground" and neutral at the receptacle.
There's more installation types like IT or TN-S, but that's usually not used at residential or apartment buildings, so it's unlikely you'll have one of these.
The only installation types where you would expect a ground to neutral bond are TN-C-S or the American split-phase power systems. If you're on one of these, it might not be a bad idea to check for that bond to be present and installed correctly, but if you happen to not find it in your panel, don't automatically assume something's wrong and it needs to be added. Especially if you're in Europe, there's a good chance it's not required and your installation code even prohibits the use of one.
You can check for the grounding system yourself if you know what you're looking at or give your power company a call. They'll always be able to tell you what grounding system you have.
Considering I've personally played on computers connected to TT, TN-C and even on entirely ungrounded systems altogether, all of them without problems of input lag and the like, I'd like to ask how that might be possible. What's the explanation for millions of players playing on either TT or TN-C systems (lacking the neutral to ground bond by design) without problems? I'm not questioning your findings, I'd just like to understand more. Thanks!
A common example is the TT earthing system which not only doesn't require this bond, but explicitly disallows it. There's a good chance you do NOT want to install any low-impedance bond between neutral and earth on a TT system, even though you theoretically could. In case of a ground fault, without the bond fault current will go through the PE and the ground rod, through the literal earth soil back to the transformer. This creates small currents which aren't enough to trip a common 10A / 16A breaker; a RCD is therefore used to break the circuit in case of a ground fault. If you connect a low-impedance bond between neutral and earth, the fault current will go down the PE conductor until it reaches your electrical panel and then use the now-"PEN" conductor all the way back to the transformer as its return path instead of the earth's soil. If your installation is TT, there's almost always a reason why it's TT: Because the pseudo-"PEN" conductor you just created is too small in diameter to handle the fault current of a fully-fledged ground fault. In the worst-case, the momentary fault-current overloads the N conductor so much it causes a fire to happen. Don't fancy your chances.
Another old installation type, not commonly used today anymore, is TN-C. You get a 2-wire conductor to the receptacle and the neutral is tied to the ground prongs at the receptacle. This installation type does not have a ground rod, and there is also no way to bond the "ground" to neutral in the panel, because there won't be a ground to bond to. This installation requires bonding of "ground" and neutral at the receptacle.
There's more installation types like IT or TN-S, but that's usually not used at residential or apartment buildings, so it's unlikely you'll have one of these.
The only installation types where you would expect a ground to neutral bond are TN-C-S or the American split-phase power systems. If you're on one of these, it might not be a bad idea to check for that bond to be present and installed correctly, but if you happen to not find it in your panel, don't automatically assume something's wrong and it needs to be added. Especially if you're in Europe, there's a good chance it's not required and your installation code even prohibits the use of one.
You can check for the grounding system yourself if you know what you're looking at or give your power company a call. They'll always be able to tell you what grounding system you have.
Considering I've personally played on computers connected to TT, TN-C and even on entirely ungrounded systems altogether, all of them without problems of input lag and the like, I'd like to ask how that might be possible. What's the explanation for millions of players playing on either TT or TN-C systems (lacking the neutral to ground bond by design) without problems? I'm not questioning your findings, I'd just like to understand more. Thanks!
-
vnb
- Posts: 70
- Joined: 04 Nov 2025, 14:44
Re: Reason most of you guys have input lag
very interesting topic, will follow closely.
Since you seem very knowledgeable about this, i am wondering what about countries like : Belgium and Norway?ChristophSmaul1337 wrote: ↑20 Feb 2026, 18:31Just adding some information you might want to be aware of before going to inspect your electrical panel. There's electrical installations that don't allow for a low-impedance neutral to earth bond, for various reasons. (...)
In fact both of those countries do not have neutral. May i ask you where are you from?
I happen to be in Belgium right due to work and i have tested my setup in 3 different locations.
location 1 had old 220V single phase + bad TT earthing System (like 99% of habitations in Belgium)
location 2 had 230V single phase - no neutral, no ground.
location 3 had 3x230V (three-phase), no neutral, no ground.
So far third location with no neutral + no grounding and a very well balanced three phase is the best. On my days off everything is perfect all day long except from around 19:00 to midnight (peak-hours) where input lag is crazy high (not internet related because even on desktop i can feel it). Easiest way to describe it is like you are dragging your cursor in mud. Everything that is moving on the screen (video, cursor, etc) is choppy and not smooth. Outside of peak hours it's perfect and you can really feel the difference.
Also a bit more general but what about voltage stability? Nobody seems to talk about it, I noticed that compared to back home; here in Belgium the voltage varies a lot constantly with wild swings like 218V to 239V in every locations.
- Slender
- Posts: 1812
- Joined: 25 Jan 2020, 17:55
Re: Reason most of you guys have input lag
take a simple system TN-C.
in that systen we have working neutral and shield neutral (mean, both PEN comming from one bus from main panel).
We want add good isolated ground and connect it to neutral bus in panel.
What happens?
Let's assume the transformer's neutral conductor is damaged. Does this mean the neutral's operating current will choose a different path (i.e., our new ground) to drain current, and the current will bypass the electricity meter?
- ChristophSmaul1337
- Posts: 114
- Joined: 11 Feb 2024, 21:01
Re: Reason most of you guys have input lag
Hey from your neighboring country Germany! As far as I'm aware and read online, Belgium mostly uses 3-phase systems just like we do over here. Every home installation requires a neutral conductor, maybe you didn't find it in your panel because it's hidden out of sight, but rest assured, your electrical installation would not work without a neutral. There are applications where one could omit the neutral wire, like an industrial motor which runs on 3-phase, but these only work without a neutral because the current draw on each phase is exactly equal and therefore cancel each other out perfectly. In any home installation you'll never be able to achieve a perfect balance between all 3 phases, so a neutral is always required in order for your devices to work. Additionally, your outlets in Belgium are of Type E, the same as in France for example. They're very similar to the Type F "Schuko" receptacles we have over here. One hole holds live, the other the neutral. You don't get two hots in the receptacle, just one hot and one neutral. Without the neutral, there wouldn't be a return path for the current, so anything you plug into it wouldn't work. You HAVE to have a neutral.
How do you know the earthing system is bad? Do you have access to the official documentation from the electrician who did the installation? Usually they measure the earth impedance (don't know what it's called, in Germany we say "Schleifenimpedanzmessung", Google Translate calls it "loop impedance measurement") and the effectiveness of the installed RCD. The earthing system would only be considered "bad" if the earth impedance is so high it wouldn't be able to conduct enough current to trip the RCD in case of a ground fault. Depending on the installed RCD, it's possible to test this yourself: Most RCD's have a "test" button (which you should press every half a year or so anyways) and some of them apply a current between live and PE, which should trip the RCD if the earthing system is working correctly.
There's a clear tolerance limit for voltage in the EU electrical network. The allowed deviation from the nominal 230V in continental Europe is standardized to +6% and -10%, or 207V on the low end and 244V on the high end. Anything in-between these values is fine and every appliance designed to operate on 230V is also capable to operate on any of the other voltages within tolerance without any problems. However, the vast majority of devices will also tolerate anything up to +10% deviation, or up to 253V. The values you read are within tolerance and therefore should not pose any problem.
Nothing. Technically, you couldn't even do this, because strictly speaking there isn't a neutral in the panel, it's still a PEN. In a TN-C installation, you get 4 conductors from the utility (assuming 3-phase): 3 phases and a PEN. The difference between TN-C and TN-C-S is the fact that the PEN conductor in a TN-C system isn't being separated (hence no -S) at the panel or the service entrance. The PEN is instead run all the way to a receptacle and only split there. A modern receptacle should have 3 poles: Live, neutral and PE. In a TN-C system, you get 2 conductors to your receptacle. You connect live to one terminal, neutral to the other and bond the PE terminal to N. Only here would the PEN be split into PE an N. In a TN-C-S you would get 3 conductors to the receptacle in the first place, (L,N,PE) because the separation of the PEN was done at the service entrance and both conductors (N and PE) are being kept separate throughout the installation. In a TN-C-S you could theoretically bond a ground rod to the neutral bus bar, but this won't do much:
The PEN conductor you get from the utility is already grounded at the transformer site. In case of normal operation, current will still take the PEN conductor back to the transformer. In case of a ground fault, the fault current still takes the PEN conductor in both scenarios, because the impedance of copper wire is orders of magnitude smaller than that of the earth's soil. Adding a ground rod doesn't do anything in both scenarios. It can't hurt and isn't "bad", but it won't do anything either.
That's really bad. Assuming 3-phase, when the neutral gets damaged, it becomes a "floating neutral", meaning it doesn't hold reference to 0V anymore. This is called a neutral point shift and it will kill any appliance in seconds. Depending on the impedance of your new ground rod, it might return some milliamps of current, but a floating neutral is still going to be catastrophic.
No. Electricity consumption is usually measured at the service entrance and at the live conductors. Any current going into your house will be registered by the meter, no matter which path it takes back to the transformer.
-
MK92
- Posts: 152
- Joined: 06 Oct 2025, 15:11
Re: Reason most of you guys have input lag
Can confirm this, I also have TN-C and the issue is 100% related to high impendance of neutral/ground combined wire when the electricity usage in the entire building is too high, but after 11 pm when almost no one uses electricity anymore, it became stable. When the impendance is too high, the current would literally flow through your PC and monitor case, not enough to shock you, but enough to cause jitter at rails, which would then cause incorrect frame timings (delays, 60hz feeling, choppy anminations etc., and of course the mouse cursor position which updates every milisecond won't be processed correctly = floaty mouse).
And what can we realisticaly do, because re-wiring is obviously not an option (if you re-wire only your apartment you didn't do jackshit, because you would still share noisy neutrals with all neighbours)? Here in EU, they never ever upgrade old buildings to a newer electrical system, too much work and too expensive, they would do this ONLY if the wiring would be so bad that people would get burning smell or would get shocked (PEN failure), otherwise they won't give a crap just because our PCs are working incorrectly.
And what can we realisticaly do, because re-wiring is obviously not an option (if you re-wire only your apartment you didn't do jackshit, because you would still share noisy neutrals with all neighbours)? Here in EU, they never ever upgrade old buildings to a newer electrical system, too much work and too expensive, they would do this ONLY if the wiring would be so bad that people would get burning smell or would get shocked (PEN failure), otherwise they won't give a crap just because our PCs are working incorrectly.
-
darkbebe
- Posts: 24
- Joined: 10 Apr 2025, 07:11
Re: Reason most of you guys have input lag
hello but why a external battery doesn't fix the problem? do we have to charge the battery when the electricity / area usage is low?MK92 wrote: ↑21 Feb 2026, 13:17Can confirm this, I also have TN-C and the issue is 100% related to high impendance of neutral/ground combined wire when the electricity usage in the entire building is too high, but after 11 pm when almost no one uses electricity anymore, it became stable. When the impendance is too high, the current would literally flow through your PC and monitor case, not enough to shock you, but enough to cause jitter at rails, which would then cause incorrect frame timings (delays, 60hz feeling, choppy anminations etc., and of course the mouse cursor position which updates every milisecond won't be processed correctly = floaty mouse).
And what can we realisticaly do, because re-wiring is obviously not an option (if you re-wire only your apartment you didn't do jackshit, because you would still share noisy neutrals with all neighbours)? Here in EU, they never ever upgrade old buildings to a newer electrical system, too much work and too expensive, they would do this ONLY if the wiring would be so bad that people would get burning smell or would get shocked (PEN failure), otherwise they won't give a crap just because our PCs are working incorrectly.
-
amorou
- Posts: 242
- Joined: 29 Aug 2022, 18:46
Re: Reason most of you guys have input lag
I'd put my money on people using cheap inverters , i would guess decent ones should be 1000 euro for inverter onlydarkbebe wrote: ↑22 Feb 2026, 14:13hello but why a external battery doesn't fix the problem? do we have to charge the battery when the electricity / area usage is low?MK92 wrote: ↑21 Feb 2026, 13:17Can confirm this, I also have TN-C and the issue is 100% related to high impendance of neutral/ground combined wire when the electricity usage in the entire building is too high, but after 11 pm when almost no one uses electricity anymore, it became stable. When the impendance is too high, the current would literally flow through your PC and monitor case, not enough to shock you, but enough to cause jitter at rails, which would then cause incorrect frame timings (delays, 60hz feeling, choppy anminations etc., and of course the mouse cursor position which updates every milisecond won't be processed correctly = floaty mouse).
And what can we realisticaly do, because re-wiring is obviously not an option (if you re-wire only your apartment you didn't do jackshit, because you would still share noisy neutrals with all neighbours)? Here in EU, they never ever upgrade old buildings to a newer electrical system, too much work and too expensive, they would do this ONLY if the wiring would be so bad that people would get burning smell or would get shocked (PEN failure), otherwise they won't give a crap just because our PCs are working incorrectly.
-
ZINZIRIO
- Posts: 93
- Joined: 25 Oct 2021, 10:29
Re: Reason most of you guys have input lag
That's not true, a neutral isn't absolutely necessary. I have a two-phase installation: 130V between A and ground, and B and ground, and 220V between A and B. Only two wires enter my house, and there's no neutral. Therefore, it's not necessary because either phase can function as both phase and neutral simultaneously, since they are 180° out of phase. So, the current comes in through one phase and returns through the other; it doesn't matter which way you connect the polarity, either one performs the same function. This is called an IT system, and it's what I have in my country, Uruguay.ChristophSmaul1337 wrote: ↑21 Feb 2026, 12:39Hey from your neighboring country Germany! As far as I'm aware and read online, Belgium mostly uses 3-phase systems just like we do over here. Every home installation requires a neutral conductor, maybe you didn't find it in your panel because it's hidden out of sight, but rest assured, your electrical installation would not work without a neutral. There are applications where one could omit the neutral wire, like an industrial motor which runs on 3-phase, but these only work without a neutral because the current draw on each phase is exactly equal and therefore cancel each other out perfectly. In any home installation you'll never be able to achieve a perfect balance between all 3 phases, so a neutral is always required in order for your devices to work. Additionally, your outlets in Belgium are of Type E, the same as in France for example. They're very similar to the Type F "Schuko" receptacles we have over here. One hole holds live, the other the neutral. You don't get two hots in the receptacle, just one hot and one neutral. Without the neutral, there wouldn't be a return path for the current, so anything you plug into it wouldn't work. You HAVE to have a neutral.
How do you know the earthing system is bad? Do you have access to the official documentation from the electrician who did the installation? Usually they measure the earth impedance (don't know what it's called, in Germany we say "Schleifenimpedanzmessung", Google Translate calls it "loop impedance measurement") and the effectiveness of the installed RCD. The earthing system would only be considered "bad" if the earth impedance is so high it wouldn't be able to conduct enough current to trip the RCD in case of a ground fault. Depending on the installed RCD, it's possible to test this yourself: Most RCD's have a "test" button (which you should press every half a year or so anyways) and some of them apply a current between live and PE, which should trip the RCD if the earthing system is working correctly.
There's a clear tolerance limit for voltage in the EU electrical network. The allowed deviation from the nominal 230V in continental Europe is standardized to +6% and -10%, or 207V on the low end and 244V on the high end. Anything in-between these values is fine and every appliance designed to operate on 230V is also capable to operate on any of the other voltages within tolerance without any problems. However, the vast majority of devices will also tolerate anything up to +10% deviation, or up to 253V. The values you read are within tolerance and therefore should not pose any problem.
Nothing. Technically, you couldn't even do this, because strictly speaking there isn't a neutral in the panel, it's still a PEN. In a TN-C installation, you get 4 conductors from the utility (assuming 3-phase): 3 phases and a PEN. The difference between TN-C and TN-C-S is the fact that the PEN conductor in a TN-C system isn't being separated (hence no -S) at the panel or the service entrance. The PEN is instead run all the way to a receptacle and only split there. A modern receptacle should have 3 poles: Live, neutral and PE. In a TN-C system, you get 2 conductors to your receptacle. You connect live to one terminal, neutral to the other and bond the PE terminal to N. Only here would the PEN be split into PE an N. In a TN-C-S you would get 3 conductors to the receptacle in the first place, (L,N,PE) because the separation of the PEN was done at the service entrance and both conductors (N and PE) are being kept separate throughout the installation. In a TN-C-S you could theoretically bond a ground rod to the neutral bus bar, but this won't do much:
The PEN conductor you get from the utility is already grounded at the transformer site. In case of normal operation, current will still take the PEN conductor back to the transformer. In case of a ground fault, the fault current still takes the PEN conductor in both scenarios, because the impedance of copper wire is orders of magnitude smaller than that of the earth's soil. Adding a ground rod doesn't do anything in both scenarios. It can't hurt and isn't "bad", but it won't do anything either.
That's really bad. Assuming 3-phase, when the neutral gets damaged, it becomes a "floating neutral", meaning it doesn't hold reference to 0V anymore. This is called a neutral point shift and it will kill any appliance in seconds. Depending on the impedance of your new ground rod, it might return some milliamps of current, but a floating neutral is still going to be catastrophic.
No. Electricity consumption is usually measured at the service entrance and at the live conductors. Any current going into your house will be registered by the meter, no matter which path it takes back to the transformer.
- dervu
- Posts: 437
- Joined: 17 Apr 2020, 18:09
Re: Reason most of you guys have input lag
One thing is faulty part of one electrical system and another thing is another system not needing this part.
That does not mean that first system's faulty part does not matter because other one is not using it...
That does not mean that first system's faulty part does not matter because other one is not using it...
Ryzen 7950X3D / MSI GeForce RTX 4090 Gaming X Trio / ASUS TUF GAMING X670E-PLUS / 2x16GB DDR5@6000 G.Skill Trident Z5 RGB / Dell Alienware AW3225QF / Logitech G PRO X2 SUPERSTRIKE / SkyPAD Glass 3.0 / Wooting 60HE / DT 700 PRO X || EMI Input lag issue survivor (source removed)
