Project N0NAME

[dgrmkrp]

...after a long time at home.. with nothing to set me back (school hours, long night drinking sessions or such ), i've been doing 2 things mainly:
a nice new fan controller... which works in a very efficient way, i might add.. pic time:

..this is the 4th topology i tried.. this is getting very close to what i can do right now.. schematic is from wikipedia, indirect link, but i tweaked some values and changed a part here and there.. this is a 'pulse generator'

i control the circuit from one knob.. the ! and by varying this resistance (can tweak it to be a linear control or even a logarithmic control by adding some stuff), i vary the noise/speed/cfm of my fans and not just one! this baby should control up to a few dozens of watts (the v3 could keep even 120W under control, but i sacrificed some control), not just 10,15 or 20 watts..

this is how it works, for those that wanna know but don't feel like reading thru hours of materials all over the net and in.. the books.. scary place..

while normal fan controllers vary the amplitude of the voltage, giving the poor fan less than 12V.. this type of control gives the fan pulses, 12V in size, but varying in duration.. so, the fan receives it's nominal voltage needs, but is allowed to take a break too PWM (Pulse Width Modulation) in principle, can be used as a fan controller, a strobe light driver, an ignition coil driver or even an audio amp...

..this is the controller set on low.. this means the pot's resistance is close to zero, and the other components in the circuit determine the minimum duration of the pulse.. i don't wanna have a total off position because i plan on using some other options to tweak fan speed..

..turning the knob on the higher side of the resistance side .. we get a longer work time and a smaller (it can never be zero, but it gets really close) pause.. this means the fan will work a lot more and move more air.. generate more noise

i know i've taken extreme shots.. so. i also have a nice video of the circuit's response.. i turn the speed down in this video.. hope the effect is seeable

what i have to mention is that, as you can easily spot in the pics, the max voltage isn't 12V, but 11+.. this is because the components eat some voltage too.. that's why, if this thing passes the prototype stage, i'm gonna have a nice rotary switch to allow my fans running in off/5V/PWM/12V mode.. i still have some doubts regarding the functioning of this contraption, as fans have a variable resistance while working.. i know it's high, but not how high oh, another shortcoming (at leat in the fan drive theory) is the pulse frequency.. i tried building a fully customizable pulse generator.. frequency and duration.. and kinda failed.. or at least, i didn't get the response i wanted.. it did lower the apparent voltage to 6V.. but that wasn't my goal... i wanted 3% to 97% of 11V.. and failed on th elower end.

why would i want to control the frequency of pulses? because you can hear it indirectly.. the pulses translate into switching noises from the fans ICs, which are there even if u use DC, without PWM.. the problem could get more serious, and i've seen it with the other design(v3), which i built and tested.. by varying frequency, noise is also controlled.. if i where to use a frequency higher than what we could hear, i would have problems only with animal rights activist but could sleep at night my best design yet achieved only a frequency of 2KHz (need more than ten times this value to be in the safe zone.. >20KHz).. after i build this puppy (next week i think, when i get back together with my tools and have a nice long time to spare)..
oh, basically, i wanna get my fans really slow when needed and do that by reinventing the wheal to fit a larger one (mobos have some kind of PWM, think asus Q-fan or the newish board sporting 4-pin CPU coolers)

..the other thing i did during late nights was.. (warning, should be in the crazy idea thread, but i couldn't find one ) this.. heat-pipe extreme...

..start of the idea.. an all fanless design.. but with the ability to turn the things cooler.. the other side is a huge piece of aluminum with fins.. design not final..

..more heat pipes..

..err m.. looked at the specs of the oo so quiet ac3 fans and surprise.. the new ac12L are even more quiet.. so i used 2 of those.. and maybe added something to the video side of the story.. still working on this.. it looks good and it should do wonders (8mm heat pipes used all over, except on the ram).. but i'm thinking is this do-able? my work-flow in max is by nature mimicking the natural way.. but.. need some critique if anybody likes this, i can do blow-ups.. maybe someone with the money backing would try this i sure would like it, but no heat pipe maker close by (i'm gonna try and mail the company mentioned looong time ago.. again)

[Mitternacht]

But, is the motherboard gonna boot up if it doesn't sense a CPU fan? My computer won't even attempt to boot windows if there's no CPU fan detected.

[dgrmkrp]

depends on the mobo.. and there is a setting in the BIOS to allow for this (cpu fan alarm or something).. i actually did run almost fanless.. i had only the exhaust working on low.. and no other fan was running/connected.. and the case fan has no yellow wire.. so all i got was a red warning near the cpu fan rpm at POST time depends on the mobo, but if u think how many fanless heatsinks are out there.. and that there are water-coolers that also sport no fans at all... newer mobos should allow running nothing at all...

[Mitternacht]

depends on the mobo.. and there is a setting in the BIOS to allow for this (cpu fan alarm or something).. i actually did run almost fanless.. i had only the exhaust working on low.. and no other fan was running/connected.. and the case fan has no yellow wire.. so all i got was a red warning near the cpu fan rpm at POST time depends on the mobo, but if u think how many fanless heatsinks are out there.. and that there are water-coolers that also sport no fans at all... newer mobos should allow running nothing at all...

Only for the innovators.

[dgrmkrp]

=)) kinda came out funny =))
almost reminds me of a 486 board i had looong long time ago.. it had everything on-board.. and no fans at all.. even the ram was partly integrated i think intel and amd are going back to that design (integrate everything in the northbridge/cpu).. except with fans

[sirkillalot617]

hey did you draw those heatpipes if so waht did u draw them in

[dgrmkrp]

draw? if it were that simple.. i use 3d studio max... and the bends should be do-able in real life.. most of them.. the heat pipes are shapes, with an 8mm thickness, at first, tracing the 'path' of the heat pipe going thru the blocks and making contact with the radiator.. of course, to make it look like a heat pipe, i used a controlled reflective orange(copperish) color with a slight bump map that's how i drew them

[sirkillalot617]

ahh i get it now i think iv got that program somewere on my pc

[GT40_GearHead]

one problem.....

for a heatpipe to work the hot end has to under the level of the cold end, so the GPU cooler would not work, and the rest of the sistem for that matter, because you are looping the pipes downwoard
think about it, the way heatpipes work is this vapors rise from the bottom to the top were they condense and due to the cold walls, and fall back down in the hot bottom end

but i must say you did a great jog drawing them

[DaveW]

for a heatpipe to work the hot end has to under the level of the cold end, so the GPU cooler would not work, and the rest of the sistem for that matter, because you are looping the pipes downwoard
think about it, the way heatpipes work is this vapors rise from the bottom to the top were they condense and due to the cold walls, and fall back down in the hot bottom end

Actually, I'm pretty sure that the design of them means that hot vapor is forced through to the other end, so it doesn't matter which way up it is.

+Rep for another magnificent update.

-Dave