[SOLVED] Mouse cursor disappears after the fluorescent lamp in the next room is switched on

[xenthia]

Hello to all the wonderful computer hardware community,
I am facing something really strange that I have never seen or experienced before.
* NOTE: In this situation all the following conditions must be true or else this weird strange incident doesn't happen:
There is a 40w fluorescent lamp in the next room, when it is turned on and if I am holding down the left button of my mouse while moving a window containing graphics like a movie player or a Sega Fusion game emulator, and if while the mouse left button is down moving the window the fluorescent lamp is turned on, I here these clicking scratching noises on my headphone for a few moments (which is usually less for other lamps in the house) and after the lamp is turned on, my mouse cursor disappears on the window that was being moved, and the game or the video freezes for about like a few seconds (it doesn't actually freezes it mostly just slows down). Now I have never experienced such a strange thing before.

The setup:
Input voltage is 230V
PC: Gigabyte motherboard with Core i7-4790k , no graphic cards, only the onboard Intel graphics that comes with the CPU
I have a DVD, and Two hard disks,
A mouse, and a keyboard, no peripherals, a tiny fan for my hard disks, (it gets hot where I live)
just one 8GB ram stick
I have a ATX 330W PSU

Now I have never seen such a strange thing happen before, at first I thought it might be because I have so many 'appearance' settings enabled in the windows performance settings, I disabled all of them, but I was still facing the same issue, then I decided to change my mouse cursor to 'none' thinking there might be something wrong with that, that didn't help the situation either, then I tried to lower the Intel graphics cards settings, the problem didn't go away.

I am using AIDA64 and it is reporting the following voltages as follow, and to be sure I ran the stress test for 10 minutes here is the results:
----------------------Current--------Minimum---Maximum-----Average
CPU core:---------- 0.792-----------0.756--------1.080----------0.954
CPU VRM-----------1.788-----------1.752-----------1.8----------1.785
CPU VID-------------1.062-----------1.007--------1.068---------1.064
+3.3V-----------------3.384----------- 3.384------- 3.404---------3.389
+5V-------------------5.040-----------5.040--------5.070--------- 5.058
+12V----------------12.168---------12.024------- 12.168--------12.154
+3.3V standby------ 3.384-----------3.384---------3.384---------3.384
VBAT Battery------ 2.976------------2.976------- 2.976----------2.976
DIMM--------------- 1.524------------1.524 ---------1.524---------1.524
iGPU----------------- 0.828---------- 0.828----------- 0.876---------0.828

Is there a hidden problem with the PSU that like for a few milliseconds a capacitor discharges due to the noise on the 230V input, because the PSU is damaged?
Or maybe a capacitor on the motherboard not capable to keep up with the voltage need like if due to the noise on the 230V because the lamp's transformer turns on, the PSU tries to maybe protect against the noise so some voltages cut off?

Please help me understand what is going on and how can I fix this? should I change the lamp? Cause if the lamp can cause this problem, then probably any other sources of noise can also induce such problems.
Why is my mouse cursor disappearing?

Thank you.

 

[jonnyguru]

I have a ATX 330W PSU

You don't know what kind?

Most decent PSUs have EMI and RFI filtering.

 

[xenthia]

You don't know what kind?

Most decent PSUs have EMI and RFI filtering.

In its guide it says it has "Input EMI filtering = Double EMI filtering" , now considering this is happening does that mean its EMI filtering is not working?

Nothing about RFI is mentioned though.

 

[jonnyguru]

In its guide it says it has "Input EMI filtering = Double EMI filtering" , now considering this is happening does that mean its EMI filtering is not working?

Nothing about RFI is mentioned though.

So you don't know what kind of PSU you're using?

 

[xenthia]

So you don't know what kind of PSU you're using?

The model on the guide says: GP330A-EUD

 

[xenthia]

I have noticed that sometimes by switching on and off other lights in the house the same thing still happens. Can someone maybe kind enough to help me understand what is really going on?
I have opened up the PSU and there seems to be a big capacitor on the inlet helping with the EMF effects I guess.
Also this didn't use to be a problem before.
Thank you.

 

[Juular]

I have opened up the PSU

Can you provide some photos of it then ?

 

[xenthia]

Can you provide some photos of it then ?

Oh damn, I wish you would have told me a tiny bit sooner, I have closed the whole casing. I would be happy to, but I don't feel like opening everything up again! 🙁 What should I do?

 

[Juular]

Oh damn, I wish you would have told me a tiny bit sooner, I have closed the whole casing. I would be happy to, but I don't feel like opening everything up again! 🙁 What should I do?

Well, it isn't that important, i just wanted to be sure that this PSU is indeed what them make it appear to be.

 

[Ralston18]

This PSU?

https://green.ir/products/power-supply/gp330a-eud

Verify: your home has 230 v service - correct?

Overall, it appears that there are voltage sags/brownouts occurring. Especially if the problems correspond with lights and (other?) devices being turned on and off.

Does the problem stop when there are fewer devices and/or lights on?

 

[xenthia]

This PSU?

https://green.ir/products/power-supply/gp330a-eud

Yes that is the product exactly, I wonder how you've found it. But yes you got it 😀

Verify: your home has 230 v service - correct?

Yes this is a 230v service, you are totally correct.

Overall, it appears that there are voltage sags/brownouts occurring. Especially if the problems correspond with lights and (other?) devices being turned on and off.

I had a digital voltmeter connected to the outlet watching over the voltage for a couple of days it was between 225 to 234-235 no less than 220 was seen.

Then I decided maybe a digital multimeter is not able to show me some changes, now I have an analog multimeter connected to the outlet monitoring the voltage, while not as exact as a digital multimeter, but it is hopefully more sensitive to big changes (while missing very little and very fast changes) it 'seems' to be rock solid on 230(228) not moving when turning the lights on or off,

Does the problem stop when there are fewer devices and/or lights on?

I have tried that but the problem seems to be going on still. Like during the day time when essentially no lights are on.

Is it possible that perhaps a device in the house is causing really really fast shorts that the fuse(s) are missing it? like maybe an electric short that happens in a nanosecond or in a millisecond?

What are the consequences of very rapid short circuits for computer PSUs and other devices?

How can I know if this is actually the case?

 

[Ralston18]

Short circuits are bad. Electricity moves very fast (e.g. lightning) so even a rapid movement of electrons can be damaging. Generally a matter of both voltage and current (amps).

Since you have a multimeter and have opened the PSU up (not recommended by the way) my suggesting is that you test the PSU per the following link:

https://www.lifewire.com/how-to-manually-test-a-power-supply-with-a-multimeter-2626158

Not a full test per se as the PSU is not underload. However if any actually measured voltages are out of spec that could be a clue as to what is happening.

PSUs are made to not be repairable and doing so or attempting to do so can end badly.

Are you able to obtain another PSU to install and test? My other thought is that that 330 watt PSU may not be up to satisfying the computer's peak power demands and the situation is made worse by the state of the house's electrical circuits.

 

[Karadjgne]

Sounds more like a grounding issue in the house electrical system. Florescent uses a ballast that changes the power from either voltages to 277v single phase, but it can and does leave behind dirty power (transformers work entirely due to magnetic fields) that's supposed to be well grounded. If the lighting circuit isn't well grounded at the house panel, it'll 'jump ship' in order to find a direct route that is.

Easiest fix is a decent UPS, that'll seperate input voltages from output voltages, basically isolating the pc from house electric system. A good UPS will also protect from brown outs, voltage spikes etc.

But I would have someone look at your electric in house, make sure everything is as it's supposed to be.

 

[xenthia]

Short circuits are bad. Electricity moves very fast (e.g. lightning) so even a rapid movement of electrons can be damaging. Generally a matter of both voltage and current (amps).

Since you have a multimeter and have opened the PSU up (not recommended by the way) my suggesting is that you test the PSU per the following link:

https://www.lifewire.com/how-to-manually-test-a-power-supply-with-a-multimeter-2626158

Not a full test per se as the PSU is not underload. However if any actually measured voltages are out of spec that could be a clue as to what is happening.

I actually did do that using the multimeters that I have seemed to be in range, but like you said I thought that maybe because this PSU is not under heavy load it might not be acting up. So Today I decided to put some pressure on it and use CPUID HWMonitor rather than AIDA64. HWMonitor is faster can grabs the voltage even under heavy system load, and then I ran a few benchmarks the voltage seemed to keep up.

PSUs are made to not be repairable and doing so or attempting to do so can end badly.

Oh thank you dear sir for the warning, I might know a thing or two about electronics but personally I would never dear attempting repairing a sensitive and highly specialized unit like a computer power supply unit. But thank you for the warning and a teachable note on not using a repaired PSU for sensitive environment. Highly appreciated

Are you able to obtain another PSU to install and test? My other thought is that that 330 watt PSU may not be up to satisfying the computer's peak power demands and the situation is made worse by the state of the house's electrical circuits.

Well I have a very dear friend in the business who lends me stuff to test, but due to the corona situation I am not sure if things are going to work out as easily as they used to before. Also I have another unit but it is a cheap one which I am afraid to test it.
Although I have put the system on benchmarking and heavy duty tasks (like running neural net learning on all cores and such) and monitored for drop of voltage but it seemed to handle even a peak of 144W some moments without the voltages dropping, only the CPU VID dropped to 0.92 instead of the normal 0.98 momentarily which was totally new and never seen before.

So do you think that the PSU should be replaced?
Thank you for trying to help me.

 

[xenthia]

Sounds more like a grounding issue in the house electrical system. Florescent uses a ballast that changes the power from either voltages to 277v single phase, but it can and does leave behind dirty power (transformers work entirely due to magnetic fields) that's supposed to be well grounded. If the lighting circuit isn't well grounded at the house panel, it'll 'jump ship' in order to find a direct route that is.

Easiest fix is a decent UPS, that'll seperate input voltages from output voltages, basically isolating the pc from house electric system. A good UPS will also protect from brown outs, voltage spikes etc.

But I would have someone look at your electric in house, make sure everything is as it's supposed to be.

Yes I think that also needs to be checked out. Also I decided to use an isolating transformer like a 230V->230V transformer but a friend of mine told me it would only cause the harmonics to be even worse. Now I don't know what harmonics are or what a 'bad harmonic phase' is in city AC, but I decided not to buy an isolating transformer, but if you think a 230V->230V AC transformer used for isolating the system will be better, I will order it online.

 

[Karadjgne]

If you know what a sine wave looks like, that's alternating current - AC. Your household probably has 50Hz AC, which means you'll get 50 sinewaves per second electricity. That's single phase power. You have 240v, which is 2 sets of 120v single phase power, out of phase. A phase is positive when B phase is negative, and then they switch.

The way electricity moves is the electrons get charged and move to the next atom in line, constantly, so as 1 electron goes in the front of the wire, an electron drops out the other end. This movement creates a magnetic field that moves with the electrons. When wrapped around an iron core with a very, very long wire, you'll create a very strong, moving magnetic field traveling through the wire. Being magnetic, it'll pick up the electrons from a second wire that's wrapped around that, creating a voltage flow. That's a transformer.

What you describe as an isolator transformer is a set of coils that when 230v is applied to the center coil, you get 230v 'created' in the outer coil. A harmonic is an additional wave created by a multiplication of the frequency when using non-linear loads, in saturated magnetic devices (transformers). A non-linear load is when the amperage changes.

With as much amperage change as that transformer would see when hooked to a psu, you'll get some funky harmonics, in effect making the original 'dirty' power, dirtier. It'll cure any RFI backfeed from lousy grounds, but won't exactly be healthy for the psu.

A good UPS (uninterruptable power supply) is like a conveyer belt dishwasher, dirty power goes in one end, clean power comes out the other. If there's a power fluctuation, (spike) , it gets capped so only supplies the correct upper range, if the power is low (brown-out), it gets supplemented by battery power, so will supply the correct minimum range. EFI/RFI effectively gets neutralized by getting stuffed into the battery.

In many older homes, they originally only used 2 prong receptacles for regular circuits, grounds were treated as being useless. Even when moving to 3 prong, grounds were still considered by many electricians as unnecessary waste of time, didn't get corrected, hooked up, or even tied into the full circuit. Why would a light bulb need a ground. Or a toaster. Unfortunately, they had no idea about the impending electronic explosion that totally relies on good grounds.

 

[xenthia]

If you know what a sine wave looks like, that's alternating current - AC. Your household probably has 50Hz AC, which means you'll get 50 sinewaves per second electricity. That's single phase power. You have 240v, which is 2 sets of 120v single phase power, out of phase. A phase is positive when B phase is negative, and then they switch.

Thank you so much for this very educational post. But going through it I have a few questions.
Yes the city has 50 Hz alternating current, I think, if not mistaken, the majority of the world has 50 Hz, while some have 60Hz. Going to https://en.wikipedia.org/wiki/Mains_electricity_by_country actually seems like it is sort of half and half 😀 Actually I am from Iran, in here on paper it is said you will get "220V", but essentially it is 230V when I use a multimeter. I have seen places with 220V, and 235V some other places. Too bad I don't have a decent oscilloscope but I am sure I would have been able to make what you say visible in a more tangible way. 😉

The way electricity moves is the electrons get charged and move to the next atom in line, constantly, so as 1 electron goes in the front of the wire, an electron drops out the other end. This movement creates a magnetic field that moves with the electrons. When wrapped around an iron core with a very, very long wire, you'll create a very strong, moving magnetic field traveling through the wire. Being magnetic, it'll pick up the electrons from a second wire that's wrapped around that, creating a voltage flow. That's a transformer.

I always thought that the electromagnetic field around the first wire cause the rearrangement of electrons (and positively charged atoms as a result too) in the second wire, and inducing electric current in the second wire, I didn't know that electrons jump from the first wire to the second wire too. That is a new thing learnt thank you.

What you describe as an isolator transformer is a set of coils that when 230v is applied to the center coil, you get 230v 'created' in the outer coil. A harmonic is an additional wave created by a multiplication of the frequency when using non-linear loads, in saturated magnetic devices (transformers). A non-linear load is when the amperage changes.

With as much amperage change as that transformer would see when hooked to a psu, you'll get some funky harmonics, in effect making the original 'dirty' power, dirtier. It'll cure any RFI backfeed from lousy grounds, but won't exactly be healthy for the psu.

So at first hearing the term "harmonic" was a bit strange for me and I didn't really know what to make of it, you see because "harmonic" always carried a positive meaning in my head like in music without harmonics it wouldn't sound nice, or when we speak we have https://www.britannica.com/topic/speech-language/Harmonic-structure in our voice.

So correct me if I am wrong here about electronics: based on what you said harmonics are like different sine forms that are formed on the wire's voltage which was supposed to be 50Hz good and pure, but due to 'noise' this 50Hz has gotten some 'harmonics' and now looks weird and out of phase and even not a 'classical' sine wave but a strange looking one.

So in essence if I have a device connected to the electrical outlet that either:
1: eats, or grabs a lot of current, in sudden and random forms that would mess up the sine form of the electricity for other devices in the house. Like 4 consecutive sharp current usages or grabs would do it.
and or two:
2: have a device that very rapidly causes shorts in the network, which the fuses would not be able to detect and that would essentially mess up the original wave form of the 50Hz current.

right?
Which is very dangerous for electric devices and could zap them. because the nice good old 50Hz sine wave form is essentially destroyed:
https://en.wikipedia.org/wiki/Aurora_Generator_Test

Can this perhaps be used like as way like fuzzing the electric devices for these kind of vulnerabilities?
Like how you fuzz a software with random stuff so that it shows vulnerabilities, instead this time you would feed the electric component with 'fuzzed' currents and out of phase power supply to detect its weaknesses.
Because this is not my field feel free to slap me, hit me, shoot me, and jump i to correct me any time you feel like it, 😀 I could use a good teacher
seems like a good red teaming project

A good UPS (uninterruptable power supply) is like a conveyer belt dishwasher, dirty power goes in one end, clean power comes out the other. If there's a power fluctuation, (spike) , it gets capped so only supplies the correct upper range, if the power is low (brown-out), it gets supplemented by battery power, so will supply the correct minimum range. EFI/RFI effectively gets neutralized by getting stuffed into the battery.

Yes I noticed that UPSes are relatively cheaper. I have seen a couple of them online, but I am somewhat classic and I would like to go to the bazaar😀 to buy my electronics instead of ordering them online. haha
But due to corona restrictions I am somewhat limited.

In many older homes, they originally only used 2 prong receptacles for regular circuits, grounds were treated as being useless. Even when moving to 3 prong, grounds were still considered by many electricians as unnecessary waste of time, didn't get corrected, hooked up, or even tied into the full circuit. Why would a light bulb need a ground. Or a toaster. Unfortunately, they had no idea about the impending electronic explosion that totally relies on good grounds.

Are you saying that by good grounding I can get rid of all these noise?

 

[Karadjgne]

I didn't know that electrons jump from the first wire to the second wire

They don't jump. It's like having a moving magnet under a piece of paper will 'drag' anything magnetic on the other side of the paper. As the field moves through the first wire, it 'grabs and drags' the electrons in the second wire. When electrons travel from atom to atom, you get a charge. V=IR, the 'charge' is amperage, moving through the resistance of the wire, you get voltage.

This is what 240v should look like.


This is what happens with harmonics, it's a single phase 120v, so imagine a mirrored black line like above.

Then multiply that 50x and you get a sawtooth for electricity, not a smooth series of waves.

You are in an Arabic region. That means basically nothing but dry sand for grounds unless you go very deep. Which means mostly relying on the service ground at the electrical hookup to the city. Which I know from living in Saudi is not exactly brilliant. If you have any copper plumbing piping near the service, you could probably 'water-bond' that pipe, increasing the grounding potential multifold. Good grounds are essential to electronics, as that's where all efi/rfi gets dumped to, eliminating as much as possible, which keeps power cleaner.

 

[xenthia]

They don't jump. It's like having a moving magnet under a piece of paper will 'drag' anything magnetic on the other side of the paper. As the field moves through the first wire, it 'grabs and drags' the electrons in the second wire. When electrons travel from atom to atom, you get a charge. V=IR, the 'charge' is amperage, moving through the resistance of the wire, you get voltage.

Thank you. Again very educational and highly informative I respect that.

Yes I thought because you mentioned "it will pick up electrons from a second wire" it was something like electrons jumping wires, and then I thought maybe that's why some transformers on the internet were so pricey when they were using terms like "double shielding wires" and such.

But yes the positivity of the electromagnetic field attracts the negativity of an induced electromagnetic field on the second wire and the negativity of it induces a negative electromagnetic field in the second wire. like because there are a lot of positively charged atoms in the first wire they are missing their electrons, when adjacent to a second wire, it will attract electrons in that region. and this over a long wire causes induction and makes the final transformer. Which keeps me wondering why should this be used for isolation in the first place because to me it seems whatever wave we feed the transformer would essentially be seen in the other side of it too. So noises and "harmonics" would be able to easily jump isolating transformers right?

You are in an Arabic region. That means basically nothing but dry sand for grounds unless you go very deep. Which means mostly relying on the service ground at the electrical hookup to the city. Which I know from living in Saudi is not exactly brilliant. If you have any copper plumbing piping near the service, you could probably 'water-bond' that pipe, increasing the grounding potential multifold. Good grounds are essential to electronics, as that's where all efi/rfi gets dumped to, eliminating as much as possible, which keeps power cleaner.

Yes not Arabic exactly, I mean geographically speaking this is the "Iranian Plateau", or olden days Persia, But it is mostly dry lands in Iran. I have seen many westerners and Americans mistake us with "Arabs" while it is ok and smiled at it should be noted that mostly the people in here prefer to be called "Iranian" even in the south west of the country. Anyways getting off-topic there.

So a good grounding would reduce a lot of noise and bad electromagnetics right? But the city doesn't provide grounding where I live, is there anyways I could make one myself? Like taking a long thick wire and stick it in a wet ground like near a tree in the garden backyard? Or would that do more damage than good?
Also I have seen some devices essentially using like a big plate of aluminum with a large surface as grounding, would that work for high voltages like 240V 50Hz, or that is only good for tiny devices?
I am asking these "apparently" light-minded-ish question because I am trying to explore my options before I go ahead and dig a hole in the street 😀😀

 

[Karadjgne]

Ideally you'd want the ground as close to your house as you can get it. Over here they use an 8' long galvanized or copper rod driven into the ground and attached to the service entrance, inside panel by #4 copper.

RFI is rarely mentioned, if at all, because of the psu casing. It should be grounded to the supply (3 prong connector). Any radio frequencies trying to penetrate the case get shunted straight to ground. As does any external EFI. It's the interference in the supply itself that's supposed to be removed by that cap, but honestly a single cap does very little by itself, just enough to make claims of protection.

A good household ground is essential, but a clean supply is also necessary. Electricity vibrates (for you it 50x per second) and over time any 'not so perfect' connections in the main panel (both line and load sides) can lead to bad supply at the socket.

A water bond from the service to the closest main water inlet should fix most ground issues, and having someone check your electrical panel and make sure all connections are snug (not over-tight), would be a good start.