VOGONS


First post, by appiah4

User metadata
Rank l33t++
Rank
l33t++

I have an FSP 300W ATX PSU I want to use with a Slot or Socket A system (it has a pretty decent 5V rail) but the fan is gone and needs replacing. I have to desolder it and solder some jumpers in its place to use a different 80mm fan. I don't want to kill myself in the process though, and this thing has some pretty large capacitors inside. What should I do to safely discharge it all before I start doing soldering work?

Retronautics: A digital gallery of my retro computers, hardware and projects.

Reply 1 of 35, by PCBONEZ

User metadata
Rank Oldbie
Rank
Oldbie

Any PC PSU I've ever torn apart had discharge paths built in so the caps discharge all by themselves in just a few minutes.
This is the section of a schematic for a generic 300w showing the two "Beer Can" (my pet name for them) caps.
The Beer Cans are C3 and C4.

Beer Can Caps_.jpg
Filename
Beer Can Caps_.jpg
File size
41.96 KiB
Views
6764 views
File license
Fair use/fair dealing exception

C3 discharges through R3.
C4 discharges through R4.
.

Last edited by PCBONEZ on 2019-12-25, 10:39. Edited 2 times in total.

GRUMPY OLD FART - On Hiatus, sort'a
Mann-Made Global Warming. - We should be more concerned about the Intellectual Climate.
You can teach a man to fish and feed him for life, but if he can't handle sushi you must also teach him to cook.

Reply 2 of 35, by derSammler

User metadata
Rank l33t
Rank
l33t

Turn the PSU on and then pull the mains power cable.

You can also put the PCB onto a damp towel for some minutes. This will gently discharge the caps.

Most people however just short the caps with a screwdriver. I would not recommend this, as it can damage the caps.

Reply 3 of 35, by mdog69

User metadata
Rank Newbie
Rank
Newbie
appiah4 wrote:

I have an FSP 300W ATX PSU I want to use with a Slot or Socket A system (it has a pretty decent 5V rail) but the fan is gone and needs replacing. I have to desolder it and solder some jumpers in its place to use a different 80mm fan. I don't want to kill myself in the process though, and this thing has some pretty large capacitors inside. What should I do to safely discharge it all before I start doing soldering work?

Leave the PSU disconnected from the main. Wait 10 minutes, then carefully measure the voltage across the input electrolytic capacitors (the ones rated at 400v) to verify that the unit is safe to work on. I would consider 50v or less as safe.

Reply 4 of 35, by appiah4

User metadata
Rank l33t++
Rank
l33t++

It has been disconnected for a couple weeks now (I like playing safe) but I will measure the voltage on the input caps thanks.

Retronautics: A digital gallery of my retro computers, hardware and projects.

Reply 5 of 35, by gdjacobs

User metadata
Rank l33t++
Rank
l33t++
mdog69 wrote:

Leave the PSU disconnected from the main. Wait 10 minutes, then carefully measure the voltage across the input electrolytic capacitors (the ones rated at 400v) to verify that the unit is safe to work on. I would consider 50v or less as safe.

The 20 to 30V range is more widely considered safe, assuming dry skin conditions, no probe stabbing, etc. 50V isn't particularly dangerous, but in certain circumstances it could pose some cardiac threat. However, any PSU worth it's salt will have bleeders in shunt with the bulk caps that will deplete any residual charge very quickly.

All hail the Great Capacitor Brand Finder

Reply 6 of 35, by luckybob

User metadata
Rank l33t
Rank
l33t

If its unplugged, its fine.

If it does bite you when its unplugged, DESTROY IT; a PSU that does not discharge its primary caps is a death trap and shouldn't be used. PERIOD. If they cut cost by the 2 PENNIES those bleeder resistors cost, they cut costs EVERYWHERE and it's a trash unit.

That said, I've been nailed by 240V on multiple occasions, and as long as you aren't grabbing it with both hands, even a cap bite is a non-issue. This level of caution is appropriate for CRT's, and I appreciate being cautious.

It is a mistake to think you can solve any major problems just with potatoes.

Reply 7 of 35, by derSammler

User metadata
Rank l33t
Rank
l33t
gdjacobs wrote:

The 20 to 30V range is more widely considered safe, assuming dry skin conditions, no probe stabbing, etc. 50V isn't particularly dangerous, but in certain circumstances it could pose some cardiac threat.

There's no need to "consider" what is safe. For AC, 50V is the limit. For DC (which is what we deal here with, since caps don't output AC), the limit is 120V. Unless you are a child or have heart issues, that is.

Normally, a PSU that has not been plugged in for hours is safe to work on, especially if you only replace a fan on the secondary side anyway.

Reply 8 of 35, by SirNickity

User metadata
Rank Oldbie
Rank
Oldbie

Handling a PSU is never restricted to the secondary side. I don't think I've ever done anything in a PSU case without bumping up against a heatsink (which may or may not be connected to a pin on a transistor / diode / regulator package, or ground), or a wire connected to the power inlet or switch, or the through-hole bits on the bottom of the PCB.

IMO, it's either safe to handle, or it isn't, and you shouldn't be futzing with it if it isn't. 😀 It only takes a second to probe the reservoir cap pins, or at least cross them with a screwdriver. By the time you get the thing undressed, they should be at zero (or nearly zero) volts. If not, I would also be suspicious about what other corners they cut. Lingering HVDC is just bad design.

Reply 10 of 35, by SirNickity

User metadata
Rank Oldbie
Rank
Oldbie

Oh get over yourself. It's cramped quarters in there, and sooner or later you're going to bump a knuckle into something metal. If your safety practice is "don't do that" without regard to whether the rest of the supply is safe, then you're going to get away with it right until you don't. It's a moot point anyway unless you're not working on a supply with the mains plugged in.

Reply 11 of 35, by retardware

User metadata
Rank Oldbie
Rank
Oldbie

LIve PSU handling is not recommendable unless you have an insulation transformer.
As I don't like getting shocked, I short the primary caps using a wire with two alligator clamps.
This is better than accidentally short and damage some components.
If there is a noticeable zap, then be careful for the reasons the others already mentioned.

Reply 12 of 35, by Tiido

User metadata
Rank l33t
Rank
l33t

I have a 150ohm ceramic block power resistor with two pins on one side for safely discharging the big capacitors without damaging them or causing huge sparks.

T-04YBSC, a new YMF71x based sound card & Official VOGONS thread about it
Newly made 4MB 60ns 30pin SIMMs ~
mida sa loed ? nagunii aru ei saa 😜

Reply 13 of 35, by PCBONEZ

User metadata
Rank Oldbie
Rank
Oldbie

Clearly someone has nil experience working inside a modern (90's forward) PSUs.

gdjacobs wrote:
mdog69 wrote:

Leave the PSU disconnected from the main. Wait 10 minutes, then carefully measure the voltage across the input electrolytic capacitors (the ones rated at 400v) to verify that the unit is safe to work on. I would consider 50v or less as safe.

The 20 to 30V range is more widely considered safe, assuming dry skin conditions, no probe stabbing, etc. 50V isn't particularly dangerous, but in certain circumstances it could pose some cardiac threat. However, any PSU worth it's salt will have bleeders in shunt with the bulk caps that will deplete any residual charge very quickly.

CORRECT

luckybob wrote:

If its unplugged, its fine.

If it does bite you when its unplugged, DESTROY IT; a PSU that does not discharge its primary caps is a death trap and shouldn't be used. PERIOD. If they cut cost by the 2 PENNIES those bleeder resistors cost, they cut costs EVERYWHERE and it's a trash unit.

That said, I've been nailed by 240V on multiple occasions, and as long as you aren't grabbing it with both hands, even a cap bite is a non-issue. This level of caution is appropriate for CRT's, and I appreciate being cautious.

CORRECT
[Comparing scars]
In the early 80's I got hit by 400 vdc from a Radar unit.
We had matting down and I ground my arm so I only took it from my wrist to my forearm.
Once was more than enough, thank you very much!

SirNickity wrote:

Handling a PSU is never restricted to the secondary side. I don't think I've ever done anything in a PSU case without bumping up against a heatsink (which may or may not be connected to a pin on a transistor / diode / regulator package, or ground), or a wire connected to the power inlet or switch, or the through-hole bits on the bottom of the PCB.

IMO, it's either safe to handle, or it isn't, and you shouldn't be futzing with it if it isn't. 😀 It only takes a second to probe the reservoir cap pins, or at least cross them with a screwdriver. By the time you get the thing undressed, they should be at zero (or nearly zero) volts. If not, I would also be suspicious about what other corners they cut. Lingering HVDC is just bad design.

CORRECT

SirNickity wrote:

Oh get over yourself. It's cramped quarters in there, and sooner or later you're going to bump a knuckle into something metal. If your safety practice is "don't do that" without regard to whether the rest of the supply is safe, then you're going to get away with it right until you don't. It's a moot point anyway unless you're not working on a supply with the mains plugged in.

CORRECT

retardware wrote:
LIve PSU handling is not recommendable unless you have an insulation transformer. As I don't like getting shocked, I short the p […]
Show full quote

LIve PSU handling is not recommendable unless you have an insulation transformer.
As I don't like getting shocked, I short the primary caps using a wire with two alligator clamps.
This is better than accidentally short and damage some components.
If there is a noticeable zap, then be careful for the reasons the others already mentioned.

CORRECT, but you need more than just an insulation transformer.

Tiido wrote:

I have a 150ohm ceramic block power resistor with two pins on one side for safely discharging the big capacitors without damaging them or causing huge sparks.

That's a good plan.
.
.
The wet rag thing is more red-neck than I am, and that's pretty red-neck.
Yes, when I first heard of it I tried it a few times.
Not only does it not work well, if the caps are charged it's MORE likely to get you shocked than just leaving the caps alone.
The wet rag is a conductor and you would be holding a conductor with your fingers and pushing it against a live voltage source.
If fully charged the primary caps (the two in series) are at around 338 vdc.
This is why they are 200-250 v caps and why PSUs with single primaries use 400-450 v caps.
If poor choices have you using a simulated sine-wave BBU and a PFC PSU that 338 is closer to 380-390 vdc.
If you are going to do the wet rag you should be wearing electrical safety gloves rated for at least 500 volts.

Just a screw driver is not a great idea either but it's certainly safer (to the human) than a wet rag.

A proper shorting device is basically a screwdriver like thing with a lead that has a high wattage resistor in it and either a probe or a clip at the end of the wire. (The clip gets put on first when you use it.)
You can DIY one or amazon/ebay have scads of 'economy' (hobbyist?) variations for under $20.

The ones we used in power plants were about 3' long with a 8" to 10" long ¼"-3/8" copper alloy rod, the resistor was embedded in the glass handle and the clip lead was 8 gauge (IIRC) thin strand copper welding type wire.
Bear in mind the newer larger plants used an internal 4160v 3-phase grid while the older or smaller plants still used 440v 3-phase.

I have only once ever seen PSU primaries hold a charge and that one was in operation when lightning struck a power pole 50 feet from the guy's house.
If you check the voltage on the primaries and find a charge then PSU either has too much damage or it is too cheap to even bother with so just throw it out.

You should never see voltage on any cap anywhere in a modern PC because it's a design consideration.
When AT/ATX PSUs start-up they have timing sequences based on how fast the voltages come up.
Too slow a rise and the PSU thinks there is a short and shuts itself back off to protect downstream.
Too fast a rise and it thinks there is no load so it shuts itself off to protect itself.
Pre-charged caps when the PSU starts would muck the timing all up so PCs are designed to bleed the caps down when there is no power. Low end PSUs don't always do that right.

Where you can get into trouble working in a PSU is by powering it up with the cover off.
Those exposed heatsinks usually have power on them when power is applied and can it be at that 338-390 volts that is on the primary caps.
So, if you have to power it up to check something don't go randomly poking around in there.
.

GRUMPY OLD FART - On Hiatus, sort'a
Mann-Made Global Warming. - We should be more concerned about the Intellectual Climate.
You can teach a man to fish and feed him for life, but if he can't handle sushi you must also teach him to cook.

Reply 14 of 35, by retardware

User metadata
Rank Oldbie
Rank
Oldbie
PCBONEZ wrote:
gdjacobs wrote:
mdog69 wrote:

I would consider 50v or less as safe.

The 20 to 30V range is more widely considered safe, assuming dry skin conditions, no probe stabbing, etc. 50V isn't particularly dangerous, but in certain circumstances it could pose some cardiac threat. However, any PSU worth it's salt will have bleeders in shunt with the bulk caps that will deplete any residual charge very quickly.

CORRECT
.

Just want to add that I have read that under the right conditions - when "connecting" the "appropriate" nerve ends accidentally ("probe stabbing", getting pinched by charged wires etc) even voltages less than 10V can be fatal.

PCBONEZ wrote:
[Comparing scars] In the early 80's I got hit by 400 vdc from a Radar unit. We had matting down and I ground my arm so I only to […]
Show full quote

[Comparing scars]
In the early 80's I got hit by 400 vdc from a Radar unit.
We had matting down and I ground my arm so I only took it from my wrist to my forearm.
Once was more than enough, thank you very much!
.

That time I was at school and collected defective TVs from the streets, carried them home (usually around 30kg apiece) and repaired and sold them if they were still good enough.
I once slipped with the scope probe when adjusting something in the convergence unit, touched a part and got well zapped. Back then there was no mains insulation in TVs. I am not sure which component I touched, according to the schematics the area in question it probably was in the range of 500 to slightly over 800V.

The jolt felt like as if I had hit that exposed nerve in the elbow super hard.
That increased my respect for high voltages a lot.
It was even worse than that damaged anode cable that once zapped me (luckily with the TV turned off).

Reply 15 of 35, by gdjacobs

User metadata
Rank l33t++
Rank
l33t++
derSammler wrote:
gdjacobs wrote:

The 20 to 30V range is more widely considered safe, assuming dry skin conditions, no probe stabbing, etc. 50V isn't particularly dangerous, but in certain circumstances it could pose some cardiac threat.

There's no need to "consider" what is safe. For AC, 50V is the limit. For DC (which is what we deal here with, since caps don't output AC), the limit is 120V. Unless you are a child or have heart issues, that is.

Normally, a PSU that has not been plugged in for hours is safe to work on, especially if you only replace a fan on the secondary side anyway.

I hope you're not in charge of anyone's safety procedures! The most widely quoted number for impedance between hand and foot is 500 ohms. DC impedance will be more due to skin capacitance. 100 mA resulting from 50V applied is at the low end of lethality by cardiac fibrillation. IIRC, the trip threshold for non US RCDs is 30 mA to provide some margin for safety.
https://m.littelfuse.com/~/media/protection-r … _whitepaper.pdf

All hail the Great Capacitor Brand Finder

Reply 16 of 35, by appiah4

User metadata
Rank l33t++
Rank
l33t++

Wow very informative guys thanks. I will just check the primary caps for voltage using a multimeter and dump it if I find any 😀

Retronautics: A digital gallery of my retro computers, hardware and projects.

Reply 17 of 35, by PCBONEZ

User metadata
Rank Oldbie
Rank
Oldbie
appiah4 wrote:

Wow very informative guys thanks. I will just check the primary caps for voltage using a multimeter and dump it if I find any 😀

Make sure to set the range to something over 400vdc just in case.
.

GRUMPY OLD FART - On Hiatus, sort'a
Mann-Made Global Warming. - We should be more concerned about the Intellectual Climate.
You can teach a man to fish and feed him for life, but if he can't handle sushi you must also teach him to cook.

Reply 18 of 35, by appiah4

User metadata
Rank l33t++
Rank
l33t++
PCBONEZ wrote:
appiah4 wrote:

Wow very informative guys thanks. I will just check the primary caps for voltage using a multimeter and dump it if I find any 😀

Make sure to set the range to something over 400vdc just in case.
.

Right.. I wouldn't want to destroy my $5 chinese multimeter 🤣 I will be cautious though, I promise. (Send my family your condolences if I don't post for a week or so 😎 )

Retronautics: A digital gallery of my retro computers, hardware and projects.

Reply 19 of 35, by PCBONEZ

User metadata
Rank Oldbie
Rank
Oldbie
appiah4 wrote:
PCBONEZ wrote:
appiah4 wrote:

Wow very informative guys thanks. I will just check the primary caps for voltage using a multimeter and dump it if I find any 😀

Make sure to set the range to something over 400vdc just in case.
.

Right.. I wouldn't want to destroy my $5 chinese multimeter.........

Oh come-on....
20 years from now it will be a collectors item and some nut on ebay will give you $150 for it.
.

GRUMPY OLD FART - On Hiatus, sort'a
Mann-Made Global Warming. - We should be more concerned about the Intellectual Climate.
You can teach a man to fish and feed him for life, but if he can't handle sushi you must also teach him to cook.