Question Fixing a PSU fan

[snatu]

Sooo I have had this problem for at least a year now (maybe even two) and I've had enough. The fan in the PSU is making this really loud rattling noise when I boot my PC, occasionally after booting as well. I believe its the bearing in the fan (correct me if I am wrong) and due to life circumstances cant really afford a new PSU at the moment. My PC is rather old, maybe 7-8 years but I hope I can make it last for another two or so. I have done my research about the topic (precautions, equipment needed etc.) but I want to be absolutely sure I know what I am doing since I dont want to hurt myself or the PC. We are talking about Corsair VS650

So here is my plan:
1. Opening the side panel, disconnecting all of the cables attached to the back of the PC.
2. Discharging the capacitors in the PSU by holding the power button for 20sec, repeating this 3 more times and waiting overnight.
3. Taking the PSU out of the case and opening it (I am planning to leave the wires for components connected to the motherboard because I dont want to mess up the cable management, let me know if this is bad for the PC)
4. Taking the fan out of the PSU
5. Applying sewing machine oil for the bearing and putting everything back in place.

If I am planning to do something wrong, let me know. Help is well appreciated!

 

[kanewolf]

Sooo I have had this problem for at least a year now (maybe even two) and I've had enough. The fan in the PSU is making this really loud rattling noise when I boot my PC, occasionally after booting as well. I believe its the bearing in the fan (correct me if I am wrong) and due to life circumstances cant really afford a new PSU at the moment. My PC is rather old, maybe 7-8 years but I hope I can make it last for another two or so. I have done my research about the topic (precautions, equipment needed etc.) but I want to be absolutely sure I know what I am doing since I dont want to hurt myself or the PC. We are talking about Corsair VS650

So here is my plan:
1. Opening the side panel, disconnecting all of the cables attached to the back of the PC.
2. Discharging the capacitors in the PSU by holding the power button for 20sec, repeating this 3 more times and waiting overnight.
3. Taking the PSU out of the case and opening it (I am planning to leave the wires for components connected to the motherboard because I dont want to mess up the cable management, let me know if this is bad for the PC)
4. Taking the fan out of the PSU
5. Applying sewing machine oil for the bearing and putting everything back in place.

If I am planning to do something wrong, let me know. Help is well appreciated!

Oil on a fan might (STRESS MIGHT) make a difference, but may not. Your best option is to replace the power supply. They are not intended to be user serviced.

 

[Zerk2012]

Taking the PSU out of the case and opening it

That is what your doing wrong buy a new one.

 

[snatu]

Oil on a fan might (STRESS MIGHT) make a difference, but may not. Your best option is to replace the power supply. They are not intended to be user serviced.

But oil is pretty cheap, so its worth trying, right?

Lets assume it doesn’t work. Next cheapest option is to replace the fan itself. What kind of fan do I need? It says on Corsairs website that the stock fan has some kind of ”thermal control”. Is that something that I specifically have to look for when buying a new one, or does every fan have this feature? I have a spare fan for my CPU that should fit in the PSU but it has 3-pin connector and my PSU requires a 2-pin. Is it possible to plug it in despite the cable not being a 2-pin? And if not do I need an adapter or something?

 

[snatu]

Taking the PSU out of the case and opening it

That is what your doing wrong buy a new one.

But how am I doing it wrong?

 

[USAFRet]

If the PSU is old and/or crappy enough for the fan to be rattling, the PSU needs to be replaced.

Oil, even if you can gt to the tiny space where the oil would go, is at best a temp fix.

The rattling noise is coming from the fan shaft wobbling in the solid bronze bearing. Oil doesn't fix that.
In fact, may make it worse.

That is almost certainly oil impregnated bronze bearing. Use the wrong oil may dilute what is currently in the bronze. Or gum it up.

 

[kanewolf]

But oil is pretty cheap, so its worth trying, right?

Lets assume it doesn’t work. Next cheapest option is to replace the fan itself. What kind of fan do I need? It says on Corsairs website that the stock fan has some kind of ”thermal control”. Is that something that I specifically have to look for when buying a new one, or does every fan have this feature? I have a spare fan for my CPU that should fit in the PSU but it has 3-pin connector and my PSU requires a 2-pin. Is it possible to plug it in despite the cable not being a 2-pin? And if not do I need an adapter or something?

Oil is cheap. Oil is NOT worth "trying". Generally you can't get oil to the bearings. You could create a fire hazard and you might injure yourself.
What kind of fan? You would have to get part number off the original fan to determine RPM, CFM and input voltage. That fan might be 5VDC, might be 12VDC, might be AC. Without a search based on model number there is no way to know. The fan could be soldered to the board rather than socketed.

But how am I doing it wrong?

You are "doing wrong" by being unqualified to attempt it, IMO.

 

[DSzymborski]

If you cannot afford to replace the PSU, then I would guess that you're not in a financial position to replace the entire PC, which is what you risk when you tinker with a PSU. You risk a *lot* by playing around with sewing machine oil. These are not meant to be user serviceable for more than just profit reasons.

If the primary problem is it rattles when you boot it, then it's really not worth the risk of doing any of this. Leave it alone. Just how often are you rebooting it? A well-maintained PC ought to be able to stay running indefinitely except for the occasional update that requires a reboot.

 

[snatu]

Oil is cheap. Oil is NOT worth "trying". Generally you can't get oil to the bearings. You could create a fire hazard and you might injure yourself.
What kind of fan? You would have to get part number off the original fan to determine RPM, CFM and input voltage. That fan might be 5VDC, might be 12VDC, might be AC. Without a search based on model number there is no way to know. The fan could be soldered to the board rather than socketed.

You are "doing wrong" by being unqualified to attempt it, IMO.

Ok, lets forget about the oil then and focus on the fan. Is it possible that all of the info you mentioned can be seen on the fan or figured out by looking up the fan manufacturer etc.? And if I open the power supply, I will also be able to see if its socketed or not.

And yes, I might be a little underqualified, but I dont need a degree in electrical engineering to know that its the capacitors being the safety hazard, and therefore its not generally a good idea to open the PSU. But if they are completely drained, there is no safety hazard anymore, right? The reason why I mentioned the way im going to drain the capacitors is because I want to make sure it is the right way to do it.

 

[snatu]

If you cannot afford to replace the PSU, then I would guess that you're not in a financial position to replace the entire PC.

If the primary problem is it rattles when you boot it, then it's really not worth the risk of doing any of this. Leave it alone. Just how often are you rebooting it? A well-maintained PC ought to be able to stay running indefinitely except for the occasional update that requires a reboot.

Ok, no oil. Fan suggestions for the PSU?

I could deal with the noise if it was only for the first 15min, but it starts when I boot the PC and can last hours after that. I never reboot the PC, only because of updates.

 

[kanewolf]

Ok, lets forget about the oil then and focus on the fan. Is it possible that all of the info you mentioned can be seen on the fan or figured out by looking up the fan manufacturer etc.? And if I open the power supply, I will also be able to see if its socketed or not.

And yes, I might be a little underqualified, but I dont need a degree in electrical engineering to know that its the capacitors being the safety hazard, and therefore its not generally a good idea to open the PSU. But if they are completely drained, there is no safety hazard anymore, right? The reason why I mentioned the way im going to drain the capacitors is because I want to make sure it is the right way to do it.

If you care to replace the fan, I would start by checking youtube for a how-to on the EXACT power supply you have. Since there might have been multiple versions of the VS650.
But my recommendation is to replace the entire power supply. But you don't seem to want to heed that advice. Good luck.

 

[snatu]

If you care to replace the fan, I would start by checking youtube for a how-to on the EXACT power supply you have. Since there might have been multiple versions of the VS650.
But my recommendation is to replace the entire power supply. But you don't seem to want to heed that advice. Good luck.

Thanks for advice and sorry for my stubborness. I will sleep over the night and reconsider tomorrow.

 

[DSzymborski]

I mean, it's not like you'd be salvaging an actually good power supply or anything (unless this is an office PC or a very simple gaming one with a GPU that doesn't require supplementary power). An investment in something decent, if this is unsuitable, would be worth the cost.

 

[sitehostplus]

Dude, those fans are sealed bearings. No way you are going to be able to spray WD-40 in there and lubricate them.

You might be able to replace it, if the housing isn't sealed and you can't get it open. But even then, you would have to know exactly what you're doing, and from what I've read, you don't and you are more likely to get electrocuted than replace the fan.

Those power supplies sell on Newegg for $80. Is that worth risking your life?

 

[snatu]

UPDATE:

I decided to listen to you guys and didn’t open the PSU. Instead I tried poking the fan inside with a small wooden stick trying to see if it has any effect on the noise. I put the PSU back in place and booted the PC and another error occured;

My PC wont let me log in. In the start screen it says something along ”Preparing Windows for automatic repair” and ”diagnosing computer” (my language is not set in english, so translations may have flaws)

It continues this loop when I try to restart it, and I have 0 idea what to do. Any possible fixes?

 

[Misgar]

It looks like Windows has failed to boot up normally 3 times and has triggered the automatic repair. You can let it attempt to fix Windows automatically. Sometimes it works, sometimes it doesn't.

Check on line for "fixes" when automatic repair doesn't work. If all else fails, I'd boot up the machine from a Windows installation USB memory stick with the same version as Windows as that on your hard disk, and choose the Update option. This will refresh Windows, keeping all your data and programs intact (if you're lucky).

You would be wise to pull the defective drive out of the computer, connect it to another computer and backup all the data, before trying to Update Windows. If you perform a normal "clean" Windows install, you'll lose all your data.

Back to the fan problem. It's a relatively simple task for a competent person to replace the fan. If you have a history of breaking everything you attempt to mend, don't try this at home. Ask a professional instead.

Yes, it's true lethal voltages above the SELV limit exist inside an ATX PSU when it's switched on and people do get electrocuted, but if you unplug the mains lead and hold your finger on the power button for 30 seconds, the 420V bulk capacitor usually discharges to a safe level. If you leave the PSU for a few hours and take care not to touch any bare wires or the uninsulated ends of component leads, your chances of electrocution are minimal.

If you're paranoid, wear a pair of thick gloves. I remember working on valve (tube) TVs where the chassis was sometimes at mains potential (240V AC), before I became a qualified engineer with a university degree. We all have to start somewhere. I suspect more people die in automobile accidents and gun shot incidents than get electrocuted by ATX PSUs. Even one death by electrocution is too many, but life is full of risks.

The two wires from the fan may be terminated in a small white 2-pin connector, plugged into the main PCB. The plug/socket connection might be covered in a blob of glue, to stop the lead from falling out, which makes unplugging the fan more difficult. Alternatively, the fan wires may be soldered directly into the board. Make a note of where they are connected before unsoldering. Check which side of the fan has a label in the middle of the hub and note it down. This determines the direction in which the air leaves the fan.

When the fan has been removed from the PSU, test it on the 5V supply of another ATX PSU Molex connector (red +5V and black 0V wires). If the fan spins at full speed, it's a 5V fan. If the fan spins slowly, it's probably a 12V fan. In this case, connect the fan to the 12V supply on a Molex connector (yellow +12V and black 0V). If the fan spins at full speed, it's a 12V fan. The PSU PCB will feed a variable voltage to the fan to change its speed.

I remove the fans from dead PSUs before recycling them, for use as spares or case fans. They're normally 12V operation.

You can buy fans with the appropriate small white connector already attached to the two wires. If you aready have a fan with the correct voltage rating, e.g. 12V, cut off the 3 or 4-way plug and solder the appropriate wires to the 2-way plug removed from the old fan.
https://superuser.com/questions/1126136/three-wire-and-four-wire-cpu-fans-whats-the-difference

Pop the fan back into the PSU with the label on the hub in the same orientation. Tap the fan blades to see if they spin freely. Fit the two halves of the PSU back together, taking care not to trap or crush any wires. Cross your fingers and switch on the PSU.

Some PSUs don't start the fan until the PSU is supplying at least 250W. Others spin up the fan immediately.

TLDR. PSU fans can be replaced if you're competent enough, but dangerous voltages might still be present inside the PSU, after disconnecting the mains. If you lack confidence or ability working with electronics, don't open the PSU.

 

[Zerk2012]

It looks like Windows has failed to boot up normally 3 times and has triggered the automatic repair. You can let it attempt to fix Windows automatically. Sometimes it works, sometimes it doesn't.

Check on line for "fixes" when automatic repair doesn't work. If all else fails, I'd boot up the machine from a Windows installation USB memory stick with the same version as Windows as that on your hard disk, and choose the Update option. This will refresh Windows, keeping all your data and programs intact (if you're lucky).

You would be wise to pull the defective drive out of the computer, connect it to another computer and backup all the data, before trying to Update Windows. If you perform a normal "clean" Windows install, you'll lose all your data.

Back to the fan problem. It's a relatively simple task for a competent person to replace the fan. If you have a history of breaking everything you attempt to mend, don't try this at home. Ask a professional instead.

Yes, it's true lethal voltages above the SELV limit exist inside an ATX PSU when it's switched on and people do get electrocuted, but if you unplug the mains lead and hold your finger on the power button for 30 seconds, the 420V bulk capacitor usually discharges to a safe level. If you leave the PSU for a few hours and take care not to touch any bare wires or the uninsulated ends of component leads, your chances of electrocution are minimal.

If you're paranoid, wear a pair of thick gloves. I remember working on valve (tube) TVs where the chassis was sometimes at mains potential (240V AC), before I became a qualified engineer with a university degree. We all have to start somewhere. I suspect more people die in automobile accidents and gun shot incidents than get electrocuted by ATX PSUs. Even one death by electrocution is too many, but life is full of risks.

The two wires from the fan may be terminated in a small white 2-pin connector, plugged into the main PCB. The plug/socket connection might be covered in a blob of glue, to stop the lead from falling out, which makes unplugging the fan more difficult. Alternatively, the fan wires may be soldered directly into the board. Make a note of where they are connected before unsoldering. Check which side of the fan has a label in the middle of the hub and note it down. This determines the direction in which the air leaves the fan.

When the fan has been removed from the PSU, test it on the 5V supply of another ATX PSU Molex connector (red +5V and black 0V wires). If the fan spins at full speed, it's a 5V fan. If the fan spins slowly, it's probably a 12V fan. In this case, connect the fan to the 12V supply on a Molex connector (yellow +12V and black 0V). If the fan spins at full speed, it's a 12V fan. The PSU PCB will feed a variable voltage to the fan to change its speed.

I remove the fans from dead PSUs before recycling them, for use as spares or case fans. They're normally 12V operation.

You can buy fans with the appropriate small white connector already attached to the two wires. If you aready have a fan with the correct voltage rating, e.g. 12V, cut off the 3 or 4-way plug and solder the appropriate wires to the 2-way plug removed from the old fan.
https://superuser.com/questions/1126136/three-wire-and-four-wire-cpu-fans-whats-the-difference

Pop the fan back into the PSU with the label on the hub in the same orientation. Tap the fan blades to see if they spin freely. Fit the two halves of the PSU back together, taking care not to trap or crush any wires. Cross your fingers and switch on the PSU.

Some PSUs don't start the fan until the PSU is supplying at least 250W. Others spin up the fan immediately.

TLDR. PSU fans can be replaced if you're competent enough, but dangerous voltages might still be present inside the PSU, after disconnecting the mains. If you lack confidence or ability working with electronics, don't open the PSU.

We don't tell people to open a PSU you have no clue about their skill level!!!!!!!!

 

[Misgar]

I appreciate we shouldn't encourage all and sundry to open their PSUs, but snatu's plan of action (see below) shows a good grasp of what needs to be done and the risks involved (see step 2.) and it was obvious he was intending to open the PSU without being prompted by forum members.

So here is my plan:
1. Opening the side panel, disconnecting all of the cables attached to the back of the PC.
2. Discharging the capacitors in the PSU by holding the power button for 20sec, repeating this 3 more times and waiting overnight.
3. Taking the PSU out of the case and opening it (I am planning to leave the wires for components connected to the motherboard because I dont want to mess up the cable management, let me know if this is bad for the PC)
4. Taking the fan out of the PSU
5. Applying sewing machine oil for the bearing and putting everything back in place.

Here is another statement by snatu which gives me confidence he has an excellent grasp of the hazards involved.

And yes, I might be a little underqualified, but I dont need a degree in electrical engineering to know that its the capacitors being the safety hazard, and therefore its not generally a good idea to open the PSU. But if they are completely drained, there is no safety hazard anymore, right?

I would not have described in detail how to proceed, had snatu not produced exactly the same method I would have used, apart from oiling the fan. Mind you, I've just oiled the bearings of the ailing fan motor in my fridge/freezer, so perhaps it wasn't such a bad idea.

I've met many unqualified engineers and mechanics in third world countries who have far more experience than I, but who knows how many of their associates were injured learning their trade?

 

[Zerk2012]

I appreciate we shouldn't encourage all and sundry to open their PSUs, but snatu's plan of action (see below) shows a good grasp of what needs to be done and the risks involved (see step 2.) and it was obvious he was intending to open the PSU without being prompted by forum members.

So here is my plan:
1. Opening the side panel, disconnecting all of the cables attached to the back of the PC.
2. Discharging the capacitors in the PSU by holding the power button for 20sec, repeating this 3 more times and waiting overnight.
3. Taking the PSU out of the case and opening it (I am planning to leave the wires for components connected to the motherboard because I dont want to mess up the cable management, let me know if this is bad for the PC)
4. Taking the fan out of the PSU
5. Applying sewing machine oil for the bearing and putting everything back in place.

Here is another statement by snatu which gives me confidence he has an excellent grasp of the hazards involved.

And yes, I might be a little underqualified, but I dont need a degree in electrical engineering to know that its the capacitors being the safety hazard, and therefore its not generally a good idea to open the PSU. But if they are completely drained, there is no safety hazard anymore, right?

I would not have described in detail how to proceed, had snatu not produced exactly the same method I would have used, apart from oiling the fan. Mind you, I've just oiled the bearings of the ailing fan motor in my fridge/freezer, so perhaps it wasn't such a bad idea.

I've met many unqualified engineers and mechanics in third world countries who have far more experience than I, but who knows how many of their associates were injured learning their trade?

No never if he was sure about what he/she was doing then they would of never posted here.
To me and many others here with years and years of experience that is the one line to never cross.

We don't help people pirate, bypass passwords, or tell them it's OK to do something that has a risk of death.

 

[Misgar]

I have known only one person (a fellow engineer back in the 1970s) who died after disabling the safety interlocks on a 10,000V 2.5A DC power supply in a 25kW power supply for a MIL-SPEC 10kW RF Linear Amplifier, so I appreciate the dangers of working with high voltages.

The force of the shock threw him across the laboratory and he could not be resuscitated. He should have known better than to tamper with the interlocks.

In future, I shall not be offering unsolicited advice to anyone on this forum involving work on high voltages above the SELV (Safety Extra Low Voltage) limit.

 

[USAFRet]

In future, I shall not be offering unsolicited advice to anyone on this forum involving work on high voltages above the SELV (Safety Extra Low Voltage) limit.

People who come here asking about how to fix the internals of a PSU are at this level:

If they knew more, they would have just done it, or known enough to not do it.

 

[Misgar]

Been there, done that. When I was 10, I poked both fingers into an open standard lamp socket after removing the bulb. After receiving a 240V AC electric shock, I didn't do it again.

It was the first of a number of mains shocks. Back in the 1960s, there were no mitigations against electric shock once you removed the cover on an appliance.

A friend of mine poked a knife into a vacuum cleaner motor. Shards from the knife took out one eye. He still qualified for his private pilot license and took us up for a flight over the Everglades before the first Space Shuttle launch.

At work I learned to specify the Safety Critical components in mains power supplies I designed.

 

[snatu]

Back here again, haven't found time to fix the Windows yet. I will get into it propably this evening.

People who come here asking about how to fix the internals of a PSU are at this level:

If they knew more, they would have just done it, or known enough to not do it.

I was asking if I am planning to do something wrong, not HOW to do it. Big difference imo

It looks like Windows has failed to boot up normally 3 times and has triggered the automatic repair. You can let it attempt to fix Windows automatically. Sometimes it works, sometimes it doesn't.

Check on line for "fixes" when automatic repair doesn't work. If all else fails, I'd boot up the machine from a Windows installation USB memory stick with the same version as Windows as that on your hard disk, and choose the Update option. This will refresh Windows, keeping all your data and programs intact (if you're lucky).

You would be wise to pull the defective drive out of the computer, connect it to another computer and backup all the data, before trying to Update Windows. If you perform a normal "clean" Windows install, you'll lose all your data.

Back to the fan problem. It's a relatively simple task for a competent person to replace the fan. If you have a history of breaking everything you attempt to mend, don't try this at home. Ask a professional instead.

Yes, it's true lethal voltages above the SELV limit exist inside an ATX PSU when it's switched on and people do get electrocuted, but if you unplug the mains lead and hold your finger on the power button for 30 seconds, the 420V bulk capacitor usually discharges to a safe level. If you leave the PSU for a few hours and take care not to touch any bare wires or the uninsulated ends of component leads, your chances of electrocution are minimal.

If you're paranoid, wear a pair of thick gloves. I remember working on valve (tube) TVs where the chassis was sometimes at mains potential (240V AC), before I became a qualified engineer with a university degree. We all have to start somewhere. I suspect more people die in automobile accidents and gun shot incidents than get electrocuted by ATX PSUs. Even one death by electrocution is too many, but life is full of risks.

The two wires from the fan may be terminated in a small white 2-pin connector, plugged into the main PCB. The plug/socket connection might be covered in a blob of glue, to stop the lead from falling out, which makes unplugging the fan more difficult. Alternatively, the fan wires may be soldered directly into the board. Make a note of where they are connected before unsoldering. Check which side of the fan has a label in the middle of the hub and note it down. This determines the direction in which the air leaves the fan.

When the fan has been removed from the PSU, test it on the 5V supply of another ATX PSU Molex connector (red +5V and black 0V wires). If the fan spins at full speed, it's a 5V fan. If the fan spins slowly, it's probably a 12V fan. In this case, connect the fan to the 12V supply on a Molex connector (yellow +12V and black 0V). If the fan spins at full speed, it's a 12V fan. The PSU PCB will feed a variable voltage to the fan to change its speed.

I remove the fans from dead PSUs before recycling them, for use as spares or case fans. They're normally 12V operation.

You can buy fans with the appropriate small white connector already attached to the two wires. If you aready have a fan with the correct voltage rating, e.g. 12V, cut off the 3 or 4-way plug and solder the appropriate wires to the 2-way plug removed from the old fan.
https://superuser.com/questions/1126136/three-wire-and-four-wire-cpu-fans-whats-the-difference

Pop the fan back into the PSU with the label on the hub in the same orientation. Tap the fan blades to see if they spin freely. Fit the two halves of the PSU back together, taking care not to trap or crush any wires. Cross your fingers and switch on the PSU.

Some PSUs don't start the fan until the PSU is supplying at least 250W. Others spin up the fan immediately.

TLDR. PSU fans can be replaced if you're competent enough, but dangerous voltages might still be present inside the PSU, after disconnecting the mains. If you lack confidence or ability working with electronics, don't open the PSU.

Thank you for the tips and trust. However I decided not to fix the fan after all. Instead I'm gonna bear with the noise and save up money for a new PC (even if the fan was okay, the PC is still getting pretty old and outdated anyway). I just want to ask out of curiosity; is it possible that after draining capacitors that big (with the method you and I mentioned) they can still pack a lethal punch?

 

[Misgar]

I'm very sorry but I seem to have have broken an (undocumented?) rule of this forum, that no help shall be given to anyone considering opening a unit containing lethal voltages and attempting repairs. There is just too much risk involved. If you need to ask for help, you obviously don't have sufficient knowledge. Catch 22.

As a result I cannot offer any more advice on this subject. You could always consider enrolling on a course at a college or university to discover the delights of becoming an electronics engineer. A much simpler option (as you say) is to leave well alone and buy a new computer. Best of luck.

 

[snatu]

I'm very sorry but I seem to have have broken an (undocumented?) rule of this forum, that no help shall be given to anyone considering opening a unit containing lethal voltages and attempting repairs. There is just too much risk involved. If you need to ask for help, you obviously don't have sufficient knowledge. Catch 22.

As a result I cannot offer any more advice on this subject. You could always consider enrolling on a course at a college or university to discover the delights of becoming an electronics engineer. A much simpler option (as you say) is to leave well alone and buy a new computer. Best of luck.

I dont fully get this rule. Isn’t it better to give advice to someone whos about to do something potentially dangerous so that the risks are minimized that way rather than letting people attempt these kind of ”stunts” on their own?

If one is as stubborn as I am, spamming ”dont do it” is more encouraging than anything. If someone told me how to properly do it, and gave me a list of things that could go wrong I would then decide if the risks are worth taking and if I am capable of eliminating those risks.

But I kinda understand that its a double edged sword and in some cases giving full tutorials would in worst case lead to someone getting seriously hurt.