removal of Passive PFC inductor

[hkivan]

hi guys,

I have recently brought a ThermalTake pure power 480ADP with
200HKD. It should be a 2nd hand item and has only 7 days warrant.
But it looks beautiful and quite heavy. I attracted by its mirror like coating
,40A on 5v rail and the price. The label states the following, 550W peak, 5V = 40A, 12v=18A, 3.3V=30A.

When I touched the PSU case, it was quite hot, hotter than my Hedy PP450WHA & Delta Active Power 420. I removed the cover and found it crowded with many items. I also found a huge PFC inductor. I want to remove it to reduce heat and improve air flow, as I had do it before on my good old PP450WHA .

My old psu work normally. I done it by cut the two wires to the inductor
and short them afterward. But I would like to know more or can I modify a PSU to active PFC?

Secondly, can anyone tell me the way to increase the speed of the thermally
controlled 12cm fan inside the psu? It justs spin too slow.

[willawake]

Re: removal of Passive PFC inductor

i doubt if you could easily convert a passive pfc to active pfc, why not just buy an active pfc unit next time.

Secondly, can anyone tell me the way to increase the speed of the thermally controlled 12cm fan inside the psu? It justs spin too slow.

this is probably the source of the problem, thermally controlled fans in psus are often pushing far too little air because people want to buy silent psus. so you could instead give that fan 12v or 7v. the easiest way to do that is to take the 12v off a molex and feed wires back to the psu fan. you could also solder fan wires to the 12v on the bottom of the psu pcb.

if you think 12v is too much you could use a resistor or even better to connect the fan to a front panel fan controller then you will be able to play with the fan control how you like. Just be careful if you are soldering wires to important fans that when you work inside the case in the future that you check the wires do not get disturbed/broken because if the psu fan becomes disconnected and is running fanless you may have a big problem.

personally i could really use a fan controller that i could set up to have different fan speeds with a switch. the unit would allow tuning of all the fan speeds and then assigning those speeds to the switch so i would have the choice of silent/medium/gaming. it would be nice also to include a switch for lights so i could have the same choice no lights/few lights/all lights.

[tiresias]

Re: removal of Passive PFC inductor

Modifying a PSU to "Active PFC" would essentially require you to add a PCB with all the various components - which wouldn't turn out to be particularly cost-effective, particularly since the active PFC would not in any way increase the PSU's efficiency, as far as your electricity meter is concerned at least.

Certainly a passive PFC PSU can be modified in the way you describe , essentially downgrading it to "no PFC". There's a model of Fortron (ATX300-GTF) to which I do this procedure as a matter of routine, after one such PSU of mine killed itself when one of the wires leading to the PFC coil melted its insulation on a small, hot FET heatsink on the primary side, and shorted to said heatsink.

Then again, I find it hard to believe that this heatsink should have become so hot in the first place, so the FET in question may have been defective after all. **

As Will said, the easiest way to eliminate "temperature-controlled" fan problems is to simply wire a fan directly to the PSU's own 12vdc. With cheap PSUs, I just solder the wires from some old fan directly to the PCB; while in the case of "nicer" PSUs, I tend to like to use the EBM/Papst fans - they're effective and reliable, and most of their models come with a 4-pin "molex" attachment which you just plug into the PSU.

** Confession: I had "modded" that particular PSU to run a Compaq. It probably found this highly offensive.

[hkivan]

Re: removal of Passive PFC inductor

Thank you so much. Your replies are very informative.

[willawake]

Re: removal of Passive PFC inductor

looking at the fortron bluestorm there is the usual fan control board attached to the heatsink. i havent measured the input to that board but i am assuming it is 12v. it would be an easy job to solder the fan wires to the input there and bypass the fan control, if there is a similar accessible board on your psu



warning : dont mess around inside psus when they are on unless you know exactly what you are probing and what not to, the primary heatsink (near the big caps) is usually 160v to ground

[hkivan]

Re: removal of Passive PFC inductor

My fan is connected to a pcb but the pcb don't touch any heatsink!
Isn't it my fan is not a thermally controlled fan ?

By the way, I attached additional heatsinks to the primary & secondary heatsinks
in my Hedy PP450WHA. They all contacted together but it runs without a hic.
If the primary heatsink are properly insulated from the FETs, will it get 160v to ground?
Anything else I should go with caution ?

[hkivan]

Re: removal of Passive PFC inductor

Again, when I connect more passive PFC coils to the existing one in series, will my PSU's power factor eventually grow up near 0.9 ?

Can I just connect a passive PFC coil to another electric device in order to make it
PFC capable ? Or, I just add some huge PFC coils to the mains, every electric device in my home appears to have PFC built within?

[Rainbow]

Re: removal of Passive PFC inductor

Don't connect the heatsinks together. You were lucky that it works - but it decreases safety.
There should be a thermistor somewhere that is used to monitor the temperature. It's sometimes near one of the heatsinks and bending it one way (so it touches the heatsink) or the other way (away from it) can be used to increase/decrease overall fan speed.

[tiresias]

Re: removal of Passive PFC inductor

They all contacted together but it runs without a hic.
If the primary heatsink are properly insulated from the FETs, will it get 160v to ground?

No - but it's still a good idea to try to maximise safety in that respect by keeping the primary and secondary sides isolated from each other by more than the (often very thin) layer of TIM under the FETs and diodes.

There was some talk about this issue on the "Watercooled powersupplies" thread : https://www.badcaps.net/forum/showthread.php?t=1723

Regarding the fan control board - it's largely up to you... if you wish to keep it for some reason you could always try to pull a 12v line from there; otherwise I would just remove it from your PSU entirely and wire the fan to the 12vdc output and ground.

[WeStSiDePLaYa]

Re: removal of Passive PFC inductor

if you want a thermally controlled fan to run full speed 24/7 just short the thermistor.

it will either be on the fan itself, or you will be able to see a diode somewhere.

ive done this with thermal controlled NMB fans, works great, remove the solder bridge and back to normal

[larrymoencurly]

Re: removal of Passive PFC inductor

I attached additional heatsinks to the primary & secondary heatsinks in my Hedy PP450WHA.

I'd get rid of any metal touching the high voltage heatsink (the one holding the high voltage transistors, usually located closest to the big capacitors) because a lot of times that heatsink is connected directly to high voltage (I've found this in Enermax, Antec, Delta, and Leadman/Powmax/Raidmax), but even if it isn't (at last some MaxPower/EverPower/Key Mouse, Fortron, Ultra/Wintech), only a thin layer of insulation between it and the high voltage transistors, too thin to trust your life to it. I've added metal to the high voltage heatsink, but I kept it at least 1/4" away from any other metal.

[hkivan]

Re: removal of Passive PFC inductor

Have you guys ever come across heatsinks that don't conduct electricity ?

I found that the primary & secondary heatsinks on my Hedy PP450WHA are
not electrically conductive. The are black in colour. I measured them by a DMM.
I don't know why ? Is there any coatings that make it non-conductive ?

The heatsinks I attached to them are conductive. But I used some non-conductive
thermal glue to attach them to the primary & secondary heatsinks.

[Galvanized]

Re: removal of Passive PFC inductor

Have you guys ever come across heatsinks that don't conduct electricity ?

Only to add to the information on this site, I will try to put a fine point on this post (not being grand ).

Many aluminum heat sinks are anodized. Why? For appearance sake because this will prevent unpleasent looking damage to the surface after initial production. Another reason would be to stop any corrosion but for computer use this should be a non-issue. For marine use=yes.

Many types of metal can be anodized. I will stay with aluminum. Anodizing can be left clear or dyed and sealed. Generally this is of the decorative type. Google will offer striking images of splash anodizing. Then there is hard anodizing. Depending on the trace elements alloyed in the aluminum the end color will be from dark green to black. The surface will grow in dimension when hard anodized, this is an issue for aluminum bushing type bearings and nut to bolt fit. Generally, the piece will be given extra clearance before the hard anodizing process.

The anodizing process can be read about here. Scroll down the page for a look at the surface offered. http://electrochem.cwru.edu/ed/encyc...-anodizing.htm (This site was posted here by another member, TY whoever you are)

For our purposes, the anodized surface is to be considered as glass. Glass insulates current flow as well as heat. So your black aluminum heat sink could be hard anodized or decorative anodized then dyed black. Blue or gold Zalman passive sinks and the outside of Lian Li cases will give an open reading on my DMM ohm meter. AMD, Intel and Zippy/Emacs sinks give a reading of 0 ohms because they are not anodized.

It would only be with some effort, that the probes of a DMM will break through this surface. I lap the anodizing off the base of all board heat sinks I install. The extruded aluminum is not very flat and the anodized surface insulates heat just a bit, not much but it does.

[hkivan]

Re: removal of Passive PFC inductor

Galvanized; Thank for your detailed info!
I learnt a lot in this forum.

By the way, I want to use my old PSU to supply only 12v to
graphic card. It only has a sense wire in 3.3v line. Will the
12v rail drop out of 5% limit when current rises?
Is there any problems with that? How can I use all 18A on this
rail ?

[Galvanized]

Re: removal of Passive PFC inductor

I was on firm ground with anodizing but detailed discussion of cross loading a SMPS is not my strong suit.
This is just my opinion and being a Yank, tend to be direct. Generally, the more bling on the outside of a PSU, the less sweetness is offered on the inside. Your TT offers 5V@40A & 12V@18A and for your use it would be better the other way around. Older systems needed high output on the 5V line but current systems are 12V dependent.

To get the full 18A@12V the 5V line is going to need a serious load on it. This is because only one transformer supplies all rails. If you read some PSU reviews at xbitlabs.com cross loading will be explained. They offer flash video of the cross loadings of tested PSUs. In accessories at PCP&C's home page a 1A resistor is shown for stabilizing the 5V line, your PSU may need to be loaded at 10A on the 5V to get the full 12V output. IDK for sure, as it would need to be done hands on. I only offer the PCP&C resistor as an example. Buy what you might need locally. If you need a ballast, mount it externally over the case exhaust fan, as it will get hot. Read this link as best you can, as it does offer more than I have. http://www.quepublishing.com/article...&seqNum=6&rl=1 This is a dated article but does offer a decent explaination.

I will sit corrected, by the well versed/informed, if what I offered was off the mark.

[Galvanized]

Re: removal of Passive PFC inductor

I will share some of my homework and bits&bobs

Print out the lapping tutorial here. http://www.easypckits.com/ He suggest 1/4 sheets. I use full sheets and never go beyond a well used sheet of 800grit. I used to rebuild diesel unit injectors & hydralic valve bodys, his tutorial is VG indeed.

Go to jonnyguru.com=good resource site. Look at the cross loading voltage issues he had with some of his tested PSUs. Note how low the 12 line drops when the 5&3.3V are lightly loaded. I would seriously consider at least loading your TT's 3.3&5v to minimum spec. It should be about 2~5A, you'll have to look it up.

Google,"adding a second PSU", here in The States directron.com comes up first hit. They offer many mini-tutorials.

I found some round ceramic ballasts yesterday that would be perfect for me. I have a server grade PSU that offers 5V@40A & 12V@32A. If I up-grade to a high wattage graphics card loading the 5V line abit might be inorder, just to stabilized the 12V line a tad. This *might*be an issue when overclocking, IDK. I would mount them using small aluminum clamp (like for retaining hoses) with a TIM interface on an aluminum bracket over the case exhaust fan, inside or outside the case IDK..Yet

The ballasts are 2 ohm @ 25W, costs $0.49 and look to be decent quality. One would load the 5V at 2.5A, two in parallel=5A. You know that. The nice thing is, the ballast would only be working at half it's rated capacity.

Nice little calculator http://www.sengpielaudio.com/calculator-ohmslaw.htm If I could read German, the rest of his site would get a look.

Shipping to Europe is $15USD. If you wish to see where these cheap but good ballasts are, PM me. The site has many interesting goodies, NIDEC 92mm fan w/tach wire at<<$5 and a NMB 120 w/hub mounted thermistor at <$5. If I post it they might evaporate

[jonnyGURU]

Re: removal of Passive PFC inductor

Only to add to the information on this site, I will try to put a fine point on this post (not being grand ).

Many aluminum heat sinks are anodized. Why? For appearance sake because this will prevent unpleasent looking damage to the surface after initial production. Another reason would be to stop any corrosion but for computer use this should be a non-issue. For marine use=yes.

Many types of metal can be anodized. I will stay with aluminum. Anodizing can be left clear or dyed and sealed. Generally this is of the decorative type. Google will offer striking images of splash anodizing. Then there is hard anodizing. Depending on the trace elements alloyed in the aluminum the end color will be from dark green to black. The surface will grow in dimension when hard anodized, this is an issue for aluminum bushing type bearings and nut to bolt fit. Generally, the piece will be given extra clearance before the hard anodizing process.

The anodizing process can be read about here. Scroll down the page for a look at the surface offered. http://electrochem.cwru.edu/ed/encyc...-anodizing.htm (This site was posted here by another member, TY whoever you are)

For our purposes, the anodized surface is to be considered as glass. Glass insulates current flow as well as heat. So your black aluminum heat sink could be hard anodized or decorative anodized then dyed black. Blue or gold Zalman passive sinks and the outside of Lian Li cases will give an open reading on my DMM ohm meter. AMD, Intel and Zippy/Emacs sinks give a reading of 0 ohms because they are not anodized.

It would only be with some effort, that the probes of a DMM will break through this surface. I lap the anodizing off the base of all board heat sinks I install. The extruded aluminum is not very flat and the anodized surface insulates heat just a bit, not much but it does.

Confirmed. I was just talking to a machinest that's going to help me with my PSU water block and I explained my concern of having one water block electrically conductive to two heat sinks. He told me that the annodizing is actually non-electrically conductive.

[jonnyGURU]

Re: removal of Passive PFC inductor

I've often just used the ballasts from an old car w/ points ignition.

Cost: Free!

[hkivan]

Re: removal of Passive PFC inductor

Finally, I did remove the PFC inductor & replaced
the original fan to a Tt 12cm 3 colour LEDs flashing fan.
I connected it to the m/b fan connector and it
runs at full speed of 2000 rpm. Now, the air coming out
from my Tt 480W is just a little bit warm and the case
is cool to touch. Thx guys for information.

[hkivan]

Re: removal of Passive PFC inductor

My Tt 480W comes with 2 teapo 820uF caps.
I want to solder additional caps parallel to them
as I have some caps sit around. I think this
way will help smooth out the I/P current a bit.
I learn of this from some forums & webs.

A few months ago, I done this to my huntkey 300W PSU.
I soldered wires on the back of PCB. The other ends
soldered to HEC caps. The HEC caps came from another
PSU. In this way,the origional teapo LXK 680uF and
the HEC LS 680uF worked in parallel and the
capacitance were doubled.

The PSU worked some days without problem.
Every rails were fine. Until some days later,
bad smell came out but it continued
in operation without halting up my PC. For
the sake of satety, I stopped it and got it out
from my PC.I tried to start it along. It never
start again. I found there were black spots around
the soldering positions. It seems that I had short
somethings but why it worked normally until
that day ? Can you guys explain this to me?
Are there any significant for 2 different brands
or kinds of caps in parallel? How do I know
my caps fail?

I don't want to buy new and larger caps.
I just want to use my exiting ones. In this way,
I won't waste my caps while improving the PSU.


Also, I replaced some I/P caps & transformer to my
5.1 PC speaker system. Unkown 25v 3300uf were
replaced by rubycon YK 25v 4700uf. The very hot 35VA 12v
EI transformer gave way to a 100VA 12v ring transformer.
I brought the ring transformer in shenzhen & it only costed
me around 50 HKD. Finally, a 8 cm fan attached to the
heatsink. Now my 5.1 speaker sound better and
run colder.