I've read about folks using all sorts of materials to mod their PC's including wood, cardboard, etc. Also cutting holes in power supply enclosures. I am involved with electrical product safety in my daily work and feel it prudent to advise against using non UL flame-rated materials inside a PC case. Also, the PC power supply is designed with the metal enclosure to provide both shock protection and to act as a fire barrier. Many of us leave our PC's on 24/7. Think about the possible results of an overheated silent mod power supply catching fire, then dripping molten material on a cardboard duct or onto flammable case damping material!

As for power supply modification, use common sense. Drastically reducing airflow and disabling or modifying thermal fan control may result in a much quieter power supply but is it worth it at the cost of safety? By the way, I DID do a fan swap on my Sonata TP380S, but instead of using a low speed fan I stuck with a Panaflo M1BX, wired to the stock thermal control and tach lead (output to mobo). The noise improvement is substantial and I do not feel I've compromised safety.

Please use common sense and before modding consider the impact of the mod wrt electrical shock and increased fire hazard.

You raise a very good point. That's why I've said before that I recommend AGAINST the idea of modding a PSU with a fan that has considerably lower airflow than the stock fan. If you need a quieter PSU, I believe the best way to achieve this is by purchasing one that is quiet in the first place, and that has been properly engineered for the airflow they specify. There's no substitute for proper engineering, in my opinion.

And I'd hate to say this, but you do realize that fire insurance won't mean a thing if they determine that the cause of the fire was YOUR unapproved modifications to your UL-listed power supply, right?

yeah, common sense really, slower the fan -> less cooling -> more heat buildup -> higher risk of damage and shorter lifespan

Personally, I know PSU modding is far out of my league. Fortunately, the PSU is the least challenging aspect of quieting down a typical computer. There are a number of economical PSUs which give great stock results out-of-box.

Compared to a good stock PSU choice, DIY PSU modding doesn't save money, isn't easy, and doesn't have much noise reduction potential. Thus, I think most SPCR's are satisfied with never attempting to mod a PSU.

But sometimes it is also the last component standing in the way of a very quiet system, especially a system that doesn't need to have the latest, most power-hungry components.

Having said that, I'll agree perhaps modding a PSU is not the place to start quieting most systems. It was one of the last tweaks that I made, but swapping out the <25dBA stock Adda fan in my Seasonic PSU for an L1A made quiet...err, quite a difference (I also decoupled the fan with EAR isolators and cut out the restrictive grill at the same time, so perhaps an isolated Adda fan would have been enough ).


Modding the PSU, IMO, really depends on (a) the threshold of your noise tolerance, and (b) your concern about your homeowner's insurance (and perhaps (c) your confidence in your modding skills...oh and (d) concern about a warranty). But since when are "saving money" and being "easy" criteria for quieting a computer?

Anyone doing a PSU fan swap really should try to match CFM between the stock and replacement fan.

On a side note: It is interesting to me how this homeowners insurance thing comes up from time to time. For those of you who are bored, try searching on insurance and you'll find a few threads....including one from about this time last year.

I am sure this is true for most, however, I just finished the final phase of my silencing project (OK, I lie there is never a final phase).

Phase one was buying a 3700AMB, the cheapest high quality case with good ariflow.
Phase two was cutting the grillles and hardwiring the 120mm Antec fan to 5v.
Phase three was among other things, a PSU mod.

I simply removed the rear 80mm, took the grill off the bottom 92mm, soft mounted it (ear plugs), ran the fan lead from the 92mm out the small gap created by the soft mounting, and attached it to a fanmate running somewhere around 7-8v.

Cost: PSU I already had, fanmate $7, ear plugs less than $1.
Difference: Subjectively less than half the noise easily.
Difficulty: The only tools were a screwdriver.

I chose to keep the 92 because I figured it would move more air with less noise. The air coming out the back feels just as cool as before.

A quiet PSU that would satisfy me would be at least $85-100 canadian plus tax, almost as much as my case.

I always thought insurance did cover stupidity...

Didn't intend to rehash old stuff, just thought it a good service to raise awareness. Silent PC modding can be done safely, just need to use good common sense.

Funny that aphonos' 1st thought is homeowners insurance. Conversely my primary concern is that my family members are not shocked, injured or killed when the PC fails or catches fire in the middle of the night.

*gasp* You make me out to be some kind of monster

My comments were in response to zuperdee's remarks. He first raised the insurance issue. Not me. My first thought was actually about noise level.

And as I reread the thread in order to defend myself , I see:

Hmmmmm. I suppose that depends on what you mean by "proper." A properly engineered PSU by SPCR standards would be efficient, have an unimpeded airflow, and a isolated, quiet fan with load/temperature controls that adjusted smoothly and accurately. Not much proper engineering out there, IMO, which is why we end up modding.

Edit: Hope my insurance agent doesn't read SPCR.

If he did, he'd stay mum -- he's probably done ALL the mods!

I agree it's good to be prudent. Yet another reason to promote (one again) a fresh air intake vent to any lower airflow PSU -- keeps it cooler.

Well, i dare to say, that i never saw PSU on fire
And i will be even more provocative to say, that when the PSU get over their limits, you notice some smoke and after mostly audiable crack the main voltage switcher gone to components silicon heaven, and that it is
boma, please don't scare peoples too much, these things are unlikely to burn at all Electronic just sedently dies, and that's all.

However i do agree that common sense in modding especially PSU is a requirment. Using flamable materials around or even into PSU is a big no-no.

Anyway, my PSU mod created a fanless PSU for my fanless machine:
http://trodas.morgus.cz/galerie.php?p=p ... 7&d=1&v=v2
...so you see that is possible do a PSU mod witch works well (now for almost a year) and with acceptable temperatures - soon i was able to provide precise measurments

Hehe--boy, this thread really has lit up a storm, eh?

Actually, Radeonman was the first one to raise the insruance matter:

I just felt compelled to make the point I did because I used to work for a company that managed homeowner associations, and in that job, I had to know at least a little bit about how insurance works. I can assure you that insurance companies generally like to know that the risk of fire is minimized as much as possible.

In fact, I do believe the UL was originally created BY INSURANCE COMPANIES expressly for this purpose.

You never saw it -> it can't happen.

Lm - you are right, the fact i did not saw / experienced it firsthand don't allow me to say that this is impossible, however i allow me to say (based on my electronic knowledge, experience and degree), that it's very unlikely

On the THG site is the PSU on fire as fake example, don't worry

...and yup, there is a fool-proof method, how to set your PSU on fire - grab a PSU, move it to safe place - away at least 5 feets from anything that can fire - open it using fireaxe or screwdriver - whatever you prefer, then spray it a with gasoline and carefully (using long matches for fireplaces and some face protection in case of suddent "blam" before you expected it is wise ) set it on fire
This way you see your PSU on fire, but don't be much surprised that once the capacitors leak and some PVC isolation end up in smoke, that the fire dies, because there is really not much flamable material on the PSU - well, once you run out of gasoline

Clearly you don't read Tom's Hardware


Personally I think that the fire hazard from even a poorly modded PC is so tiny as to be insignificant, there's simply too many failsafes that would have simultaineously fail for enough heat to build up in there, over a long enough period of time, for a fire to start. Combine that with the lack of fuel inside a case, and the chances of a widespread fire starting inside the PC is incredibly low. You are actually much more likely to have a fire start st the wall outlet.


But take my opinion with a grain of salt....I am the guy who built a wooden computer case after all.

The closest thing I saw to a PC fire was not due to any modding on my part AFAIK. (yeah right, you say...) Or even a PSU. Despite the upmteen PSUs I've tested, messed with, modded and tried to fry, the most burning I got from a PSU was on me -- I let myself get zapped with the back of my hand just to see if the HS in a PSU were live. In that case, they were.

But the worst was a strange Seagate B-IV hard drive that stopped working one day. I opened up the PC, rebooted and was listening for the drive -- nothing. And even as I was watching, I saw the paper label on the drive start to develop a yellow spot in the center, which then started to turn brown... and I thought, wow, is that burning? and touched the drive -- and instantly got a burn on my finger. Yanked the power cord. The drive was burning hot, and the center of the label had a centimeter diameter circle that was singed brown. And I had a little burn blister on the tip of my finger -- to slow with the ice.

What would have happened if I was not there, who knows. I have never been able to figure out what happened, and the Seagate RMA folks didn't even reply. But they did send me a new drive.

I am a great believer in safe practices.

I am an equally great believer in the rigorous determination of safe practices.

At this point, I am unaware of objective evidence related to fire hazards in computer usage. It would be a great help if some group could carry out studies to give us a better idea of magnitude of risk related to various computer modifications.

The SPCR group wisdom seems to be that these modifications, carried out by reasonably experienced workers in a cautious and thoughtful way, have not resulted in calamities.

What I would like to know is the denominator. Is the hazard 1/10,000, 1/1,000,000, 1/1,000,000,000? At some risk point, most of us would just get in the car, buckle up and drive...

I don't know about hard numbers, but I do believe I would agree with MikeC and others who say that the risk of an actual FIRE resulting from failure of an electronic component is *VERY* unlikely. Indeed, I seem to recall that part of the UL certification requires that no single failure on its own can lead to a fire hazard.

Having said this, I do believe the original point of this thread was that certain modifications CAN pose a fire hazard, for example: using cardboard for ducting. THAT is the kind of thing you need to be careful about.

It is my opinion that there's nothing inherently wrong with modifying a PSU in itself. However, *IF* you are going to do so, then you should follow some common sense guidelines, such as:

1) Make sure you at least have SOME idea what you are doing, and make sure you take prudent precautions. For example: don't use combustable materials.

2) Even if there is no fire hazard, it is still a good idea to take additional measures to make sure you aren't running the risk of killing your valuable electronic parts. For example: don't use fans that have significantly less airflow than the original fan, unless you have made some other modifications that you know will make the new airflow rate okay. (E.g., using bigger heatsinks, bigger intake vents, etc.)

3) ALWAYS keep in mind that if you modify your PSU, you will most likely risk voiding the warranty. This is a totally reasonable policy, since the manufacturer typically cannot ensure that YOUR modifications were well engineered. They can only reasonably be expected to know what modifications of THEIRS are okay. After all, THEY are the qualified, certified, and paid engineers, and as hobbyists, we are NOT.

aphonos, please accept my apologies, I certainly have no intent to offend.

In regards to risk of fire associated with PC's, I agree it is low. However, the reason it is low is due to the work of safety agencies such as UL. You will find that printed wiring boards (pwb) and all plastic parts in PC's and other electronic equipment are UL flame rated. Pwb?s are almost always rated 94V0, which means if a component mounted on the board fails and burns, the pwb will self extinguish once the fuel load of the failed component is exhausted. Metal Oxide Varistors, the most common component used to protect electronics from lightning surge are known to explode if hit by a severe enough surge. In well designed equipment this "fire" is contained and self extinguishes.

UL will perform a "locked rotor" test on a PC power supply to ensure that it does not become a fire hazard in the event of a fan failure. The power supply may be destroyed, but it will not pass UL Recognition if it fails in a manner that presents risks to the user. The primary risks associated with electronics equipment are fire and shock. I suspect that commercially available damping material intended for installation into a PC will carry UL Recognition and an appropriate flame rating. Obviously this will generally not hold true for household materials fitted into a PC and most notably for known flammable materials installed into the PC.

Common sense goes a long way here. If you?re fabricating that PD duct, cardboard may be fine for prototyping, but reach for the sheet metal, polycarbonate or lexan for the permanent solution.

As some of you may recall, I made some of the comments recently in another thread that bomba is now making.

The thing that got me thinking was that nightclub fire in Rhode Island a few years ago when sparks (not direct flame) set the foam on fire that was on the back wall of the stage. Over 100 people died in that fire. The thing that was so striking to me was that the foam on the wall was specifically designed for sound absorption.

Certainly many kinds of sound absorption foam is perfectly safe inside a PC, but maybe not all foam.

m0002a, you raise a VERY good point there--one I never thought of before.

I wonder--does anyone here know if things like AcoustiPack have a UL rating for fire safety?

m0002a - yes, i recall this unfortunate accident, however is the club owner fault. I used to work with guy, who run a cinema and from old ages, there is in the roof lot's and lot's of foam for acustic absorbing
Dunno who projected and build the cinema, it must be well into the deep cumunism times, however when safety checking person found it few years ago, it almost (from what i hear) fainted
Then he after drinking some water slowly explained, that if the electrical installation into the roof get shortcut and set on fire, the foam will on fire act like a napalm (!) - the dips of flaming foam will falling down to peoples being still on fire - and they stick (!) burning to peoples clotches, hairs and skin... Just like napalm...

Well, the cinema got very soon roof changes and it's safe now. No foam anymore. But because no-one tried pyrotechnic experiments into the cinema for it's lifetime, and the electricity is well-made - it worked like that for very very long time, it might be even over 20 years

So foam is pretty dangerous thing and cannot be installed on roofs for safety reasons

Draw conclusions about using foam into computer for yourself

...well, i believe myself and don't sleeping under my computer, so i using it, but as you can read above - it's dangerous and when set on fire, it acting very much like napalm...

Hahahahaha! "Symbolic representation only - Power supplies might not catch fire, but you can certainly burn them out."

If nothing else, THG is good for laughs.

laughs, but not much else

And when that article was first published, that caption was not there. They added it at a later time.

From AcoustiProduct's site:

My mistake. Sorry.


No offence taken. You're right. Family member's security certainly should come before concern for insurance policies.

/still trying to figure out a way to do that fireworks mod inside my silent PC