Pacify Your PC - Part 6 : Got water cooling?

[choongsta]

i find i can maintain a fairly substantial overclock without the need for water cooling

[GoNz0]

I did talk about silencing PSUs in one of the earlier shows but it seems it didn't make the final cut. Anyway what I basically said was that although changing the fan in your existing PSU seems the easiest and cheapest option in some cases that could be the last thing you should do and here are some reasons why!

Firstly, the PSU will require a certain amount of airflow too keep it cool. Generally if you want to make something quieter you have to reduce the fan speed which in-turn may reduce the airflow (depending on the replacement fan). Therefore if you reduce the airflow it could cause the PSU to overheat and ultimately fail!

Second, opening up a PSU and unscrewing things can be very dangerous! If your screwdriver where to slip and touch the wrong components in the PSU you could get an almighty shock. Anything more than 20ma and you'll be pushing up daises!! One reason for this is because the capacitors (a components used in a PSU) holds a charge even when the PSU has been turned off. Although over a period of time the charge will dissipate. Take a look at this Google search if there are any doubts!

Finally, and if the second reason was not good enough you will invalidate your warranty, just as Pob255 said.

I hope this helps.

unless you hit the power on button with the psu unplugged from the mains, then its all discharged in a couple of seconds.

[Hicks12]

I thought capacitors did lose it over time, not instantly which is why i think its merely stored power which isnt enough to start it.

I didnt plan on opening the PSU as i want the warranty and so i was just hoping for advice on say rubber seats etc, does it work? SHould i shove a fan outside of the PSU next to the fan grill to help suck air out?.

[Pob255]

A capacitor is a basically like a battery that can hold a relatively small amount of charge but can release it very quickly so once it's charged up it will hold it's charge until something short circuits it (like a battery it also loses charge over time but as the amounts are small it loses it faster than a battery will)
So unless the psu automatically earths the capacitors when you turn it off (very unlikely) then they will still be charged and the large primary capacitors in a psu can hold quite a bit of charge.

A few years back my brother use a couple of very large industrial capacitors to make an arc solidering system.

http://en.wikipedia.org/wiki/Capacitor

As to slienceing the psu, basically it's buy a quieter psu, the addition of rubber gaskets and mounting blocks can also be used to reduce vibration from it going into the case, depends on what the sound issue is.

[TAKTAK]

just connect across both legs of a cap i.e. a screwdriver, et voila instantly (safely) discharged...

when you charge a cap the electrons basically go from one plate to the other, and then remain there, and then when you discharge the electrons balance out on both plates again

i.e.

at rest

  ¦ ¦ 
e ¦ ¦ e
e ¦ ¦ e
e ¦ ¦ e
  ¦ ¦ 

charging

electrons flow <

e &#166; &#166; 
e &#166; &#166; 
e &#166; &#166; 
e &#166; &#166; 
e &#166; &#166; 
e &#166; &#166; 

discharging

 electrons flow back >

  &#166; &#166; 
e &#166; &#166; e
e &#166; &#166; e
e &#166; &#166; e
  &#166; &#166;

if you discharge through a resistor then it takes longer for the cap to discharge (5 time periods)

so if you had a 50mA cap discharging through a resistor of 20K ohms then:
T=RC

T=50*20 (m and K cancel [x10^-3 and x10^3])

T=1000

the cap discharges by 63% each time period, so is effectively 0% after 5T (is exponential so it would take a while for it to be truly zero)

5*T=5000 seconds to discharge fully