I used a resistor once to make a graphics card fan quieter. It worked fine, until the friend I adjusted the computer for, let a pile of dust build up, then the fan stopped, and the gpu burned out. So you always have to remember, if you are slowing the fan down to the point where it just starts, that a bit of dust build up could make it stop entirely. Wherever possible now, I use larger fans, as they move more air with less rpm, and therefore are quieter.
Well, technically speaking the fan draws 0.25A only at 12V. With an in-series resistor this will be lower so a larger resistor is needed (30+ Ohm). It is better to thing the fan as a resistor of 12V / 0.25A = 48 Ohm. Then calculate from there the resistor needed for a 5V drop. This calculation can never be accurate. So either-way, yeah the converter is the best solution. The transistor with a variable resistor should be good too.
@@wishusknight3009 Or 7V from the difference between 5V and 12V pins on a Molex connector. Drawback: you can't connect the tacho pin then, you can't have the mainboard warn you that the fan has stopped.
@@SianaGearz that has potential to start a fire because some psus use return current for the protection circuit, and short-circuits could go undetected
@@devl547 I know, I’ve tried it back in the early 2000s. It “works.” But, it’s not a very polite hack. You’re dumping current from the 12V rail through the load and _into_ the 5V rail. It’s not supposed to be a current sink. The only reason it works is because other things on that rail are consuming that current and pulling the voltage back down to the regulated voltage. As long as the 5V rail is loaded adequately, it ... _should_ ... be fine. But it’s not particularly sound engineering. Also, I would only ever consider doing this with 2-pin fans. If you try and use the tach pin on a 3-pin fan, your motherboard is expecting it to be at Gnd potential, while your fan is referenced to 5V. The result could be bad tach readings, or damage to the motherboard speed sensor channel and/or fan. IMO: Just use a regulator. It’s what they’re for.
Thanks for the video, even if simple this was pretty enlightening - the step down modules are just what I needed for my next project! Very elegant solution.
if you have PWM fan, another option is build a 555 or opamp based PWM generator. the benefit of PWM control, you could run the fan much slower than DC control if needed
Very good explanation of some of the basics you would not believe how many new technicians that pass through the trades don't have a basic understanding of how a resistor actually works
If there are technicians who don't know that energy can't disapear and it will turn to heat, there is probably something wrong with your education system. 😀 But problem is that many pepople who work in some specialisation don't actually have education in that, but when you work with voltage, you can't do it without some paper which says that you are able to do that in my country.
there is a lot of PWM fans if you connect the PWM pin with GND for some reason the fan will go into low speed mode im not sure what logic behind it but it seems to be doing the job leaving the PWM floating means full speed
Yes, the calculation has to be different. At first the resistance of the fan: 12V : 0,25A = *48 Ohm* Then the new current for 7V at the fan is 7V : 48 V/A = *0,146 A* The pre resistor has the same current at 5 V R=U:I 5V : 0,146A = *34,2 Ohm* - I choose a 33 Ohm 2W resistor. Then the resistor will not become so hot.
Thank you very much too. To be honest, I can't help Ukraine. There is nothing I can do to help a whole country. But I try to help people, who need my help. I feel with Ukrainians just as much as I feel with Russians, who are also against this war and were forced to run away. Currently both Russian and Ukrainian refugees are living in my home side by side.
For older PCs with fans hanging on a molex connector, I do something similar connecting +5V to the ground pin of the fan and getting it to +7V, but in this case the fan has to be connecte to the mainboard, atherwise it would give an alarm. And wiring 5V to the fan in such case would be quite a messy operation.
В принципе можно поэкспериментировать. Последовательно с переменным резистором подключить терморезистор и его закрепить на основании подошвы кулера. Должно в зависимости от температуры изменять обороты вентилятора.
I used a buck converter on my 3D printer. The power supply is 12 volts, I took a leg down to 5 volts to power a raspberry pi. The pi runs octoprint, which adds brains to the printer, very useful.
Ah, those were good times in the 2000's😊Especially early 00's, almost every system had noisy fans, spinning way too fast like this! I remember a good article on a PC magazine back then, they built a silent Pentium 4 Northwood PC. It had all kinds of cool stuff in it, previously unheard of: Soundproofing material(bitumen+foam), big Zalman Copper CPU cooler, silent Enermax PSU etc. Really appealed to a youngster and encouraged to build stuff yourself. It would have been nice to control fans like this. I remember 5V and 7V connections for fans(switching the fan adapter pins with a small screwdriver at 4-pin molex connectors😇)Just being careful not short circuiting between 12V and 5V lines.
I've just bought totaly new coolers for socket 370 from some bankrupted shop, but those fans......jesus, it's 60x60mm size and it has like 5000 RPM or something, maybe even more, it's pretty loud even when you connect it to 5V. 🤣 BTW, all those myths about how terribly hot Pentium 4s were are just myths, compared to today CPUs, it's cold like an ice. Problem is that when socket 478 was released, people were used to coolers like from sc370 which was mostly just the simpliest tiny aluminium heatsink with some crazy fan, this was not enough to cool Pentium 4 CPUs, but this was fixed with newer coolers and temperatures were totaly ok then. When you put some more modern cooler on 3 GHz prescott, it will barely get's higher than 55°C in full load, which is crazy compared to temperatures which we have totday like 80+ even with giant coolers or AIO. 😀 "switching the fan adapter pins with a small screwdriver at 4-pin molex connectors" I am still doing it even in 2023, I have some adaptors modded to 5V only or 7V only for all 4 connectors. I hate that today computer shops already don't have any interesting stuff like rpm regulators and such things, but thankfully, most of today fans are 4pins and can be regulated in bios, at least something is easier today. But some motherboards will not allow to lower it enough, there can be limit like 30% or something, so sometimes you have to use "old hacks" even in 2023. 🙂
Thanks for the great video! You need a step-up converter for PSU transplants in workstations, where a 12V only PSU was used originally. Even the STANDBY voltage needs to be converted from 5V (regular PSU) to 12V (12V only PSU).
Dude, you should use that PWM pin to control the fan's speed, that's what it is for. A simple NE555 circuit would be great fit for that. Limiting the voltage when you have a fan supporting the PWM control is a lame solution at best.
Yes, true, but the answer is simple, I didn't have a 555 at hand and used, what I had instead. Furthermore, this can be still helpful for someone, who has a 3 or even 2 wires fan, or something else. Just like everything on my channel, it's never a competition for the best solution, but more like, "look guys, I had a problem and here is how I solved it" thing :)
So I like to add USB-C PD to old things like a Sony "Boombox" with a refurbished cassette deck in it. This way I can run old tech off that is common today, USB-C PD battery packs or wall adapters, not having to use alkaline or even rechargeable batteries on the go. I have a couple old computers and even a VHS Camcorder I've done this mod too. If the device can run directly off 9v or 12v or 15v I don't need to use a buck (or boost) converter, but in a few cases for something like a 6v device I will use a USB-C PD 9v trigger circuit, and buck it down to 6v with a little board like that. Never thought to use a buck converter on a fan. Great use case and bonus, adjustable. Great videos as always!
Ah, the age old problem of using a newer PWM fan on an older non-PWM fan header - super-speed blowy fans ahoy! (I had the same problem with having to replace smashed fans on my NZXT case caused by a mild housemove disaster, and found they now only do PWM fans - so I ended up using SpeedFan to manually bump it down until I eventually replaced the motherboard).
They don't do only PWM fans, you can still buy 3-pin fans, but to find some fans which have low RPM without regulating it is not that easy today. It's sad that today computer shops have literally nothing compared to 15 years ago and you have to order every shit from China. 😀
I have a whole bunch of such buck converters myself, too. Super cheap and really convenient. Whenever i do any project (arduino, esp32, stm32, etc) that i need multiple voltages for, i just provide one from bench supply and then use these little buck converters to get the other voltages.
as a crude alternative, you can take the 12v (yellow) "positive" wire from the PSU molex and the 5v (red) wire as "negative or ground", so 12 - 5 is 7v, which will reduce noise. I have exactly the same problem with a 1800MHz Athlon XP. Greetings.
To slow the fan of a Nvidia graphics card, I made a 4-to-3 converter, with 2 transistors, one small NPN to invert the PWM wire and one big PNP to actually drive the fan. I guess it can also work to use 3-wire fans on working 4-wire motherboard
IRLZ44N, TC4400, 555 timer, some caps, resistors and trimpot, if you wanna go fancy with temperature speeding up the fan then add a thermocouple to the mix.
I used 3pin fan to molex adapter , and on adapter switched contacts in molex to choose 5 or 7v . Later i used fan speed controler for case . Later i got better mobo with full pwm support/speed control . Only thing i remember it was an Asrock s478 , later i got Asus .
Наконец-то новое видео. Кстати, я тоже mini360 для вентилятора использовал, в модельном зарядном устройстве. Там китайцы посчитали, что подавать 16 вольт на вентилятор, рассчитанный на 12- отличная идея. Да и вообще, после КРЕНок такие модули- просто праздник. Не греются, ток в нагрузку выдают больший, ну разве что импульсными помехами шумят немного.
Nice! Looks like a very affordable solution :-) ... How about adding an external potentiometer and mount that to the back/front of the case for manual adjustment? ... Better yet, how about expanding this to, let's say, the CPU plus a couple of case fans for mainboards without fan headers?
I had used one of those booster ones from 5V to 7V for a Noctua NF-A14 fan that I put onto my Fritz!Box Cable 6490. The fan was powered by its USB port. Due to Gigabit internet, I now got a Fritz!Box Cable 6591 which runs much cooler without a fan.
Glad to see you are back i missed watching you video's i still have a bunch of old AGP cards including the ATI radieon 9800 Pro, I also have a Ti 4200 Pro. A bunch of different main boards. Anyways, Love the video's my friend be safe and God Bless.
you can use a 555 timer IC and build a PWM controller... tie grounds from fan power rail together with 555 circuit then run the fans PWM wire to the output of the 555. it will give you full range adjustability.
I might try this with my ML350 G6. The fans are really loud, it would be nice to get them down to a quiet level for a home environment. This might be the ticket.
I prefer using a Zener diode in antiseries (power + to diode cathode, diode anode to fan +). Usually I keep a good power margin, using a diode rated about 2*Vzener*Ifan (fan current at rated voltage, even though it will draw less), to compensate for higher ambient temperature and worse heat transfer due to heat shrink tubing. A 4.7 or 5.1V zener, rated 2 watts, usually works really well. Also I prefer this due to the reason that DC-DC converters have fairly large input/output caps, which unnecessarily stresses the circuitry on the board if it switches the fan on/off, and causes PWM to not work well if the board does it by modulating the power (for 3-pin fans). But of course, a temperature regulated circuit is the best solution :)
To make the transistor design even better use a linear potentiometer for the voltage divider on the FET design for variable voltage/speed control. This would be a great solution if you didn’t have a buck converter and would cost less than a dollar.
Yeah, I wanted to make the transistor solution first, but then found the buck-converters. The transistor solution would be probably even free, since that simple parts I'd be able to desolder from some scrap, which I have in my workshop. But it was just simpler to got with the ready to use converters, which I had anyway.
You could also use a zenerdiode in block-mode as I call it. In other words: put a zener on the red wire the wrong way! Voltage should be about 3.2v less.
You can make a PWM controller using a 555 timer, frequency needs to be about 25khz. You could use a temperature dependent resistor to vary the duty cycle depending on how hot the CPU heatsink is.
My motherboard has only one header for sysfan and since I live in a hot and humid jungle I need something good to cool my pc, I'm actually used to loud pcs, my pc in the early 2010's was a Phenom II x6 that was really loud cause it had to be 100% speed everytime, found some fans from dead PSU's and put them on the PC but I couldn't control the speed, when I was about to build something like that I realized I could just use all 3 fans in the same header and it's all good now
Wouldn't there be a way to make this temperature driven? Like instead of a resistor add something that changes resistance with heat? Maybe with a potentiometer to tweak or something...
I would make similar sized PCB with microcontroller for PVM signal, output voltage analog input and temperature probe input for radiator temperature, one transistor for PWM signal->FAN. For microcontroller voltage regulator would be needed. Write a program with if statement, if temperature goes UP, then increase fan speed (change PVM), with output voltage sense, minimum voltage can be set. Yes, its sound complicated, but to make such pcb in china, it costs just a few $, for example jlcpcb
Heres an idea. make a fan controler with one of those. I bought a used 212+ for cheap because it didnt come with the original fan, it came with a cheap 1200rpm chinese fan which wasnt fit for this heatsink, so i added an industrial 3000rpm fan which i had laying around. The problem is that while the fan that it came with had 3 pins, this industrial one didnt, it only had + and - so not only it didnt have a tachometer but it also wasnt possible to change the speed, so i wanted to be ablt to have a powerfull fan while also be able to control it and so i grabbed one of those tiny buck converters and hacked it a bit. First i grabbed the pwm signal from the motherboard and pass it though an RC filter, then that output goes to a 393 , then i have another variable resistor as the voltage reference and the output of the 393 opamp goes to a small PNP transistor that bridges the output of the buck converter to the middle of the voltage divider of the feedback circuit. It works great though i think a 393 is not the best option as even though its a 2A buck converter it cant handle 2A while on the fan controller. It can in fixed voltage mode but im sure there must be some noise in the circuit that is causing it to trigger earlier. I had the scope in a box when i made it so i didnt check it but my next prototipe will be with one of those 5A ones as one of the fans i got for the pc was too powerfull for the tiny fan controler (a big 130mm x 32mm 12v 1.9A server fan )
Can't you use the PWM from a fan to just drive a MOSFET or BJT directly? The fan will be powered in bursts but if you're lucky they're fine enough to where it isn't audible, this is approximately what the PWM pin does on these fans anyway. This mainboard doesn't have PWM outputs, that's the whole thing.
@@SianaGearz while the intel spec says the pwm signal is around 25khs the problem is that most fans have capasitors in the power input, so any pulse gets filtered. Using pwm with one of those would give you a very logarismic looking speed cuve. (already tried that) the pwm on the fans controls the switching transistor directly so theres no capasivite load on them. But to an externat transistor the fans (at least the big ones and more powerful) at quite capasitive
@@laharl2k I have seen that, for that reason i have jammed in an inductor and a capacitor before the fan on one of the things i built where i use PWM fan control. Inductor... i forget, a shitty tiny little one, somewhere between 50 and 330 µH, i don't remember, with maybe around 2 Ohm series resistance; actually i think i used two, one on each wire; and a capacitor of i think 470µF which was WAAAAY too generous, please choose a smaller one :D The thought that this is total overkill did cross my mind, but the cap just landed in my hand when i grabbed it so i decided to try, and was too lazy to replace it with anything later.
Warum hast du nicht einfach ein Slotblech mit einem einzelnen Lüftersteuerungspotentiometer dazwischen gesteckt, dass wäre einfacher gewesen und du hättest es im Nachhinein bei geschlossenem Gehäuse regeln können!?
Nice explanation of how everything worked ! No point over complicating things.. i did think you were going to go down the Arduino route or something with a thermistor :)
oh nice thank you so much for this video. Fan on my Tualatin 1.4GHz runing at max speed like hell and in bios no settings for speed change. I will buy som modules in Czech Republic cost one 0.77 EUR so is much cheaper than Noctua fan i planned 😊
Welcome back!!!!! Hey Necroware. I have a very rare dauphin orasis tablet, it is a pentium 266 tilamook that wont POST. I have the schematics but drawing a blank trying to fix it. Would you be interested in taking a look?
Or you can use 5 volt from motherboard)) Or 3.3V on ATX... or 5V minus 1V fall thru diod :))) Good variant is use NMOS transistor with capacitor and two resistors, to make fan run full speed at power on and slowly go down...)
5V is too little, usually the fan Stopps at about 4,5-5V completely, so something between 6V and 9V is ok. But basically, yes, there are plenty of ways, that's why it makes fun to tinker with.
That was a PWM fan. Although this would work on ****ANY**** computer fan, but now do a video utilizing the PWM signal!!!? I'm thinking a 555 circuit, but whatevah!
Wellp. Not sure if your calceulations correct. You see, adding resistor in series is equal to creating a voltage devider of resistor and a fan, so first thing what you must - is to calculate actual resistance of a fan. lets do that. 12/0.25=48 Ohm Ok. Now we can see if we add 48 Ohm resistor we will half the voltage to 6V. So dancing around 40-70 Ohm resistor you can adjust desired fan speed. And the last but not least - lets calculate currrent for a resistor of 48 Ohm - 12/(48+48)=0.125A As you can see adding resistor in circuit always limits current too. In my practice usually 60 Ohm resistor is enougth to make quiet any laud fan and 0.5W resistor is more than enougth.
Yes, you are absolutely right :) I had to take voltage division into the calculation, instead of just taking the current. In my case that wouldn't help though, since all the resistors I had were 250mW. That wouldn't be enough in any case and I'd had to order some.
@@necro_ware Actually fan motor isn`t so linear, and operating on lower voltage may gradually lover current too. You should take measurements of a current drown by the fan. About the wattage of a resistor - i`ve never bothered about that, its just works. BTW you can take two resistors in parallel (120+120=60 Ohm) to double wattage.
@@Evhen_Velikiy Yes, that's also a good idea. Twice the resistance of each would half the current in parallel and result in lower power consumtion. Next time :D
Those adaptors with resistors which they sell are mostly useless, most of fans need less than 6V to make it silent, 7-9V is still too much and when I need to lower voltage that much with resistor, it can get pretty hot and also you just waste that energy, I prefer just to connect it to 5V from molex connector, but in some cases, I use resistors too, for example on old GPUs where I don't want to have some wires going from GPU to motherboard or PSU, but those fans are that small with such a low current that resistor can slower it without getting hot. I count nothing, I just try few resistors and keep what works the best, from my experiences, it's mostly between 100-200 ohm when you need to slower down super fast fan, but for some little fans, even 50 ohm can be too much. But when you need 200 ohm resistor, it's probably really better just to connect it to 5V from PSU than heat your room with some resistors. 😀 But I noticed that word "silent" means something totaly different for different people, for me, it means that I can't hear fans at all, when I clearly hear it, it's not silent, lowering it from 12V to 9 or 7V mostly just turns terrible noise to just noise. 😀 Most of old fans have possiblity to acces soldered contacts for wires, so you can also connect 2 fans to series which is pretty good solution for some purposes, but annoying is that you can't just disconnect that fan and use it somewhere else then. Much easier would be if producers of those fans were not idiots and RPM was normal from the factory, when it's 5000+, you just have to do something with that, you can't listen such noise.
Nice! Did you use some kine of buck converter as well, or just an LM7905? I also made some unusual -5V supply for the mainboards with ISA slots with Phil from PhilsComputerLab two years ago or so. You can find a video on my and his channel about it. Just search for the Voltage Blaster.
Just oil it! Seriously! My machine wouldn't be kicking right now without some good old Wahl clipper oil! Remove the rubber stopper under the fan's sticker, place 2 drops, give it a spin with your fingers to work it in, replace rubber stopper and sticker, and there you go. Good as new!
Or for providing 7 Volts without any resistor being used you just disconnect your fan, pull the black and red wires off the fan connector, and then connect red one to the yellow wire of the Molex connector (+12V) and black one to the red wire of the Molex connector (+5V). Since these voltages have the same ground (black wire) voltages will be substracted one from another and we will get the needed 7 Volts (12-5=7)
Used to be able to go down to Radioshack and get a linear voltage regulator, a pack of resistors, and a variable resistor. Now you can order a $1 buck converter from China. Ahh, such progress.
In thailand you can still do that, not radio shack ofc but an electronics and stuff store. Discrete regulator components or the aliexpress buck converters boards from a walk in store.
You still can do this in Germany too. Well kind of, currently due to the electronics crisis some parts are hard to find and everything is super expensive, but possible.
Here is another idea: speedfan. On these computers the BIOS didn't regulate the fan, but it was still possible to regulate it, you just had to use software.
I must be dreaming, another video! This isn't the start of a streak, is it...? A small correction: The calculation at the start involving a resistor is flawed, it doesn't work out that way. If you introduce another load in the circuit, the overall load and therefore current changes and so you can not use the fan's rated current anymore. What you can do is use the rated current at 12V to calculate the fan's equivalent DC resistance first (Rfan=12V/0.25A=48Ohm) and then use a voltage divider formula to calculate the needed resistor to drop the voltage across the fan to 7V. (5V/7V=R/48Ohm -> R=48Ohm*5/7=~34.3Ohm) Hope this makes sense (and is correct) :D Edit: Oh and the resistor should be rated for: P=U*I=U*U/R=5V*5V/34.3Ohm=~0.73W
This should get closer, but it still isn’t really accurate. DC fans are not resistive loads, so it may react very differently. Particularly since the fan’s controller IC may not have a predictable curve of current vs input voltage.
@@nickwallette6201 I tried with a fan I had at hand and it came close enough. But does a non-PWM fan really have any other electronics like a controller inside?
@@necro_ware oh ok then..i didnt get to see that video..it was on my youtube suggestions this one...i guess you cant see the io chip on older hardware as they dont report much back to the user
@@theodordan680Yeah, kind of. To be precise, this board has the rotation sensor, which is reported both in BIOS and SpeedFan, but there's no way to control the rotation speed. I guess, on this model there's no way to do it in software.
This would have doubled the value of the machine and no fun 😄 I have used Noctua in one of my AM2 "Phenom" machines. Just to try, looks cool, works cool but too easy - just swap. The original fan was beyond repair but the metal block is pretty good.
@@necro_ware Wenn ich mit so einem Lüfter fertig war, dann *hatte* der einen Molex-Anschluss und lief dauerhaft mit 7 Volt. 😀 Auf einer LAN streikte mal der Chipsatz-Lüfter unseres Jungstars. Damit er weiterzocken konnte (die Läden hatten alle schon lange geschlossen), schmierten wir Butter in das Lager. Ich bin ein Freund pragmatischer und dauerhaft-provisorischer Lösungen. 👍 🙂
Sicher :) In diesem Fall müsste ich aber dann mit dem Daueralarm des PC-Speakers leben, weil der Lüfter nicht angeschlossen ist. Im großen und ganzen bin ich auch für pragmatische, aber doch möglichst ordentliche Lösungen. Was das Bier angeht, ein Freund von mir hatte Kühlschrank in seinem Zimmer neben dem PC stehen und hatte stets eine kühle Bierdose auf der CPU stehen, die er dann regelmäßig ausgetauscht hat. Der PC war definitiv lautlos :D
@@necro_ware Hey, das ist ne Retro-Machine! Die macht man ja auch hübsch! Ich kann Dir mal ein Bild von meinem C64 schicken, der innen aussieht, als sei eine Kabel-Fabrik explodiert. Damals war das normal, heute nicht mehr vorstellbar. Speaking of vorstellbar: warmes Bier? 🙂 Warum hat es den Rechner nicht in den Kühlschrank gestellt?
@@JendaLinda I did that with an Arduino and it completely escalated into a full on fan controller with hardware fan curves and sensor value mixing (matrix) for my water cooling setup. :D
I used a resistor once to make a graphics card fan quieter. It worked fine, until the friend I adjusted the computer for, let a pile of dust build up, then the fan stopped, and the gpu burned out. So you always have to remember, if you are slowing the fan down to the point where it just starts, that a bit of dust build up could make it stop entirely. Wherever possible now, I use larger fans, as they move more air with less rpm, and therefore are quieter.
Long time no video.... missed you
Well, technically speaking the fan draws 0.25A only at 12V.
With an in-series resistor this will be lower so a larger resistor is needed (30+ Ohm).
It is better to thing the fan as a resistor of 12V / 0.25A = 48 Ohm.
Then calculate from there the resistor needed for a 5V drop.
This calculation can never be accurate.
So either-way, yeah the converter is the best solution.
The transistor with a variable resistor should be good too.
or just wire 5v to the fan -
@@wishusknight3009 Or 7V from the difference between 5V and 12V pins on a Molex connector. Drawback: you can't connect the tacho pin then, you can't have the mainboard warn you that the fan has stopped.
@@SianaGearz or that....
@@SianaGearz that has potential to start a fire because some psus use return current for the protection circuit, and short-circuits could go undetected
The calculation can be accurate if one has a amperage vs voltage curve for the fan.
Glad to see you are back. I hope things are going well considering the circumstances.
A neat trick to get the fan running on 7V without using resistors is to connect it between +5V (red) and +12V (yellow) on the PSU molex connector.
Yeah, i nearly commented but i see that you already wrote it
Also, the the 5v will become ground and 12v become pozitive
I’m not sure that the 5V rail will appreciate being a current sink ...
@@nickwallette6201 actually it works ok, that's a really old trick
@@devl547 I know, I’ve tried it back in the early 2000s. It “works.” But, it’s not a very polite hack.
You’re dumping current from the 12V rail through the load and _into_ the 5V rail. It’s not supposed to be a current sink. The only reason it works is because other things on that rail are consuming that current and pulling the voltage back down to the regulated voltage.
As long as the 5V rail is loaded adequately, it ... _should_ ... be fine. But it’s not particularly sound engineering.
Also, I would only ever consider doing this with 2-pin fans. If you try and use the tach pin on a 3-pin fan, your motherboard is expecting it to be at Gnd potential, while your fan is referenced to 5V. The result could be bad tach readings, or damage to the motherboard speed sensor channel and/or fan.
IMO: Just use a regulator. It’s what they’re for.
So glad that you're back. would love to see and hear more about usage of the linear voltage regulator.
Thanks for the video, even if simple this was pretty enlightening - the step down modules are just what I needed for my next project! Very elegant solution.
if you have PWM fan, another option is build a 555 or opamp based PWM generator. the benefit of PWM control, you could run the fan much slower than DC control if needed
Exactly what I do. Works great. Efficient
@4:53 You don't need powerful resistors, but you'll need a powerful transistor :)
Very good explanation of some of the basics you would not believe how many new technicians that pass through the trades don't have a basic understanding of how a resistor actually works
If there are technicians who don't know that energy can't disapear and it will turn to heat, there is probably something wrong with your education system. 😀 But problem is that many pepople who work in some specialisation don't actually have education in that, but when you work with voltage, you can't do it without some paper which says that you are able to do that in my country.
there is a lot of PWM fans if you connect the PWM pin with GND for some reason the fan will go into low speed mode
im not sure what logic behind it but it seems to be doing the job leaving the PWM floating means full speed
Yes, the calculation has to be different. At first the resistance of the fan: 12V : 0,25A = *48 Ohm* Then the new current for 7V at the fan is 7V : 48 V/A = *0,146 A* The pre resistor has the same current at 5 V R=U:I 5V : 0,146A = *34,2 Ohm* - I choose a 33 Ohm 2W resistor. Then the resistor will not become so hot.
Very simple and effective solution! Will have to give this a try one day. Thanks for sharing 👍
Didn’t watch the video yet, but so glad to see you’re back! Hope you’re feeling better too!
Thank you for supporting Ukraine! Glad to see that you are posting videos again.
Thank you very much too. To be honest, I can't help Ukraine. There is nothing I can do to help a whole country. But I try to help people, who need my help. I feel with Ukrainians just as much as I feel with Russians, who are also against this war and were forced to run away. Currently both Russian and Ukrainian refugees are living in my home side by side.
Thanks to you I'm using this solution on my every retro build for the power supply fan.
I use to go the complex route like you did, then I just started wiring up the fans to the +5v rail which worked perfectly.
For older PCs with fans hanging on a molex connector, I do something similar connecting +5V to the ground pin of the fan and getting it to +7V, but in this case the fan has to be connecte to the mainboard, atherwise it would give an alarm. And wiring 5V to the fan in such case would be quite a messy operation.
The PSU also has a 5V rail. I usually use a 3 way (ON-OFF-ON) switch to switch between 5V, 12V or off, and mount it on the case somehow.
Very elegant and professional looking result! Thanks for sharing
Very good and clear explanation.
Technically you could use 555 timer to make PWM signal that you could apply to pin 4 of the fan.
Yes, I thought about that too, but as I said, I dug through what I had at hand and came out with this solution.
@@necro_ware should have used the transistor dropper with a thermistor and then the fan speed would increase should temp rise
В принципе можно поэкспериментировать. Последовательно с переменным резистором подключить терморезистор и его закрепить на основании подошвы кулера. Должно в зависимости от температуры изменять обороты вентилятора.
I used a buck converter on my 3D printer. The power supply is 12 volts, I took a leg down to 5 volts to power a raspberry pi. The pi runs octoprint, which adds brains to the printer, very useful.
Ah, those were good times in the 2000's😊Especially early 00's, almost every system had noisy fans, spinning way too fast like this! I remember a good article on a PC magazine back then, they built a silent Pentium 4 Northwood PC. It had all kinds of cool stuff in it, previously unheard of: Soundproofing material(bitumen+foam), big Zalman Copper CPU cooler, silent Enermax PSU etc. Really appealed to a youngster and encouraged to build stuff yourself. It would have been nice to control fans like this. I remember 5V and 7V connections for fans(switching the fan adapter pins with a small screwdriver at 4-pin molex connectors😇)Just being careful not short circuiting between 12V and 5V lines.
I've just bought totaly new coolers for socket 370 from some bankrupted shop, but those fans......jesus, it's 60x60mm size and it has like 5000 RPM or something, maybe even more, it's pretty loud even when you connect it to 5V. 🤣 BTW, all those myths about how terribly hot Pentium 4s were are just myths, compared to today CPUs, it's cold like an ice. Problem is that when socket 478 was released, people were used to coolers like from sc370 which was mostly just the simpliest tiny aluminium heatsink with some crazy fan, this was not enough to cool Pentium 4 CPUs, but this was fixed with newer coolers and temperatures were totaly ok then. When you put some more modern cooler on 3 GHz prescott, it will barely get's higher than 55°C in full load, which is crazy compared to temperatures which we have totday like 80+ even with giant coolers or AIO. 😀
"switching the fan adapter pins with a small screwdriver at 4-pin molex connectors" I am still doing it even in 2023, I have some adaptors modded to 5V only or 7V only for all 4 connectors. I hate that today computer shops already don't have any interesting stuff like rpm regulators and such things, but thankfully, most of today fans are 4pins and can be regulated in bios, at least something is easier today. But some motherboards will not allow to lower it enough, there can be limit like 30% or something, so sometimes you have to use "old hacks" even in 2023. 🙂
Thanks for the great video! You need a step-up converter for PSU transplants in workstations, where a 12V only PSU was used originally. Even the STANDBY voltage needs to be converted from 5V (regular PSU) to 12V (12V only PSU).
I used a few buck converters to convert a battery powered bubble machine to a 12V bubble spewing powerhouse.
Dude, you should use that PWM pin to control the fan's speed, that's what it is for. A simple NE555 circuit would be great fit for that.
Limiting the voltage when you have a fan supporting the PWM control is a lame solution at best.
Yes, true, but the answer is simple, I didn't have a 555 at hand and used, what I had instead. Furthermore, this can be still helpful for someone, who has a 3 or even 2 wires fan, or something else. Just like everything on my channel, it's never a competition for the best solution, but more like, "look guys, I had a problem and here is how I solved it" thing :)
So I like to add USB-C PD to old things like a Sony "Boombox" with a refurbished cassette deck in it. This way I can run old tech off that is common today, USB-C PD battery packs or wall adapters, not having to use alkaline or even rechargeable batteries on the go. I have a couple old computers and even a VHS Camcorder I've done this mod too. If the device can run directly off 9v or 12v or 15v I don't need to use a buck (or boost) converter, but in a few cases for something like a 6v device I will use a USB-C PD 9v trigger circuit, and buck it down to 6v with a little board like that. Never thought to use a buck converter on a fan. Great use case and bonus, adjustable. Great videos as always!
Ah, the age old problem of using a newer PWM fan on an older non-PWM fan header - super-speed blowy fans ahoy!
(I had the same problem with having to replace smashed fans on my NZXT case caused by a mild housemove disaster, and found they now only do PWM fans - so I ended up using SpeedFan to manually bump it down until I eventually replaced the motherboard).
They don't do only PWM fans, you can still buy 3-pin fans, but to find some fans which have low RPM without regulating it is not that easy today. It's sad that today computer shops have literally nothing compared to 15 years ago and you have to order every shit from China. 😀
I have a whole bunch of such buck converters myself, too. Super cheap and really convenient.
Whenever i do any project (arduino, esp32, stm32, etc) that i need multiple voltages for, i just provide one from bench supply and then use these little buck converters to get the other voltages.
Me either! Very useful small guys ;)
Those buck converters are a charm, i used a chain of them and boosters to get positive and negative 5v and 15v out of a regular 3.6v Li-Ion...
as a crude alternative, you can take the 12v (yellow) "positive" wire from the PSU molex and the 5v (red) wire as "negative or ground", so 12 - 5 is 7v, which will reduce noise. I have exactly the same problem with a 1800MHz Athlon XP. Greetings.
Easy way to start a fire since you could trick the short-circuit protection on psus that monitor the ground current
To slow the fan of a Nvidia graphics card, I made a 4-to-3 converter, with 2 transistors, one small NPN to invert the PWM wire and one big PNP to actually drive the fan. I guess it can also work to use 3-wire fans on working 4-wire motherboard
IRLZ44N, TC4400, 555 timer, some caps, resistors and trimpot, if you wanna go fancy with temperature speeding up the fan then add a thermocouple to the mix.
I used 3pin fan to molex adapter , and on adapter switched contacts in molex to choose 5 or 7v . Later i used fan speed controler for case . Later i got better mobo with full pwm support/speed control . Only thing i remember it was an Asrock s478 , later i got Asus .
Наконец-то новое видео.
Кстати, я тоже mini360 для вентилятора использовал, в модельном зарядном устройстве. Там китайцы посчитали, что подавать 16 вольт на вентилятор, рассчитанный на 12- отличная идея.
Да и вообще, после КРЕНок такие модули- просто праздник. Не греются, ток в нагрузку выдают больший, ну разве что импульсными помехами шумят немного.
Nice! Looks like a very affordable solution :-) ... How about adding an external potentiometer and mount that to the back/front of the case for manual adjustment? ... Better yet, how about expanding this to, let's say, the CPU plus a couple of case fans for mainboards without fan headers?
Back in the days I usually preferred to use a zener Diode instead of a resistor. No hassle with calculating resistor values.
Zener in backwards? Hm, also an interesting idea.
We used to cheat and plug the fan ground wire into the +5 v rail (effectively providing 7v to fan).
I had used one of those booster ones from 5V to 7V for a Noctua NF-A14 fan that I put onto my Fritz!Box Cable 6490. The fan was powered by its USB port. Due to Gigabit internet, I now got a Fritz!Box Cable 6591 which runs much cooler without a fan.
Yeah, that is a step-up or booster converter, also very useful. I used a step-down converter.
@@necro_ware Yep those modules come in all sorts of flavors, even extremely high powered ones as well!
Glad to see you are back i missed watching you video's i still have a bunch of old AGP cards including the ATI radieon 9800 Pro, I also have a Ti 4200 Pro. A bunch of different main boards. Anyways, Love the video's my friend be safe and God Bless.
little boxes with a pot on it became popular in the early 2000's since fan control on your motherboard and 4-pin fans didn't exist.
you can use a 555 timer IC and build a PWM controller... tie grounds from fan power rail together with 555 circuit then run the fans PWM wire to the output of the 555. it will give you full range adjustability.
i didnt realize this was an older video... others have already stated a PWM controller could be used. i didnt read before i commented. :)
I'm using LM2596, since years, and had never any problems!
I might try this with my ML350 G6. The fans are really loud, it would be nice to get them down to a quiet level for a home environment. This might be the ticket.
Last time I used a VR was on a 6600gt to give it more core voltage. I miss the old days of overclocking.
I prefer using a Zener diode in antiseries (power + to diode cathode, diode anode to fan +). Usually I keep a good power margin, using a diode rated about 2*Vzener*Ifan (fan current at rated voltage, even though it will draw less), to compensate for higher ambient temperature and worse heat transfer due to heat shrink tubing. A 4.7 or 5.1V zener, rated 2 watts, usually works really well.
Also I prefer this due to the reason that DC-DC converters have fairly large input/output caps, which unnecessarily stresses the circuitry on the board if it switches the fan on/off, and causes PWM to not work well if the board does it by modulating the power (for 3-pin fans).
But of course, a temperature regulated circuit is the best solution :)
I love using LM317.
They are not PWM, so not as efficients, but work perfectly in this case.
А еще с DC360 mini можно использовать терморезистор вместо подстроечного, тогда напряжение и обороты кулера будут меняться автоматически....
Отличная идея!
I have way to many 18v power supplies and using a buck converter to get the desired voltage is a good way to recycle.
To make the transistor design even better use a linear potentiometer for the voltage divider on the FET design for variable voltage/speed control. This would be a great solution if you didn’t have a buck converter and would cost less than a dollar.
Yeah, I wanted to make the transistor solution first, but then found the buck-converters. The transistor solution would be probably even free, since that simple parts I'd be able to desolder from some scrap, which I have in my workshop. But it was just simpler to got with the ready to use converters, which I had anyway.
You could also use a zenerdiode in block-mode as I call it. In other words: put a zener on the red wire the wrong way! Voltage should be about 3.2v less.
Best solution: little inexpensive mcu (pi pico), add some simple code to control the pwm pin and adjust it via usb serial or via a thermistor.
You can make a PWM controller using a 555 timer, frequency needs to be about 25khz. You could use a temperature dependent resistor to vary the duty cycle depending on how hot the CPU heatsink is.
I didnt see your comment till just now... i said the same thing!
My motherboard has only one header for sysfan and since I live in a hot and humid jungle I need something good to cool my pc, I'm actually used to loud pcs, my pc in the early 2010's was a Phenom II x6 that was really loud cause it had to be 100% speed everytime, found some fans from dead PSU's and put them on the PC but I couldn't control the speed, when I was about to build something like that I realized I could just use all 3 fans in the same header and it's all good now
Wouldn't there be a way to make this temperature driven? Like instead of a resistor add something that changes resistance with heat? Maybe with a potentiometer to tweak or something...
now this is what i call "edutainment"
Is there a simple way to manually control the PWM line instead of the supply voltage itself?
Yes, you can use a 555 timer in an adjustable circuit.
I would make similar sized PCB with microcontroller for PVM signal, output voltage analog input and temperature probe input for radiator temperature, one transistor for PWM signal->FAN. For microcontroller voltage regulator would be needed. Write a program with if statement, if temperature goes UP, then increase fan speed (change PVM), with output voltage sense, minimum voltage can be set. Yes, its sound complicated, but to make such pcb in china, it costs just a few $, for example jlcpcb
Use a NTC with a MOSFET to create a temp controlled fan. Easy.
I had a Pentium 4 with a noisy fan, i mixed some zippo fuel and 2T oil, the bearing were absolutely silent
I would be curious if your solution would work on some of my PA amps that use 24v fans. you mentioned they were capable up 23 volts though. ?
You can also buy or build a PWM controller.
I would have used a thermocouple as a variable resistor in the voltage divider
Heres an idea. make a fan controler with one of those.
I bought a used 212+ for cheap because it didnt come with the original fan, it came with a cheap 1200rpm chinese fan which wasnt fit for this heatsink, so i added an industrial 3000rpm fan which i had laying around. The problem is that while the fan that it came with had 3 pins, this industrial one didnt, it only had + and - so not only it didnt have a tachometer but it also wasnt possible to change the speed, so i wanted to be ablt to have a powerfull fan while also be able to control it and so i grabbed one of those tiny buck converters and hacked it a bit.
First i grabbed the pwm signal from the motherboard and pass it though an RC filter, then that output goes to a 393 , then i have another variable resistor as the voltage reference and the output of the 393 opamp goes to a small PNP transistor that bridges the output of the buck converter to the middle of the voltage divider of the feedback circuit.
It works great though i think a 393 is not the best option as even though its a 2A buck converter it cant handle 2A while on the fan controller. It can in fixed voltage mode but im sure there must be some noise in the circuit that is causing it to trigger earlier. I had the scope in a box when i made it so i didnt check it but my next prototipe will be with one of those 5A ones as one of the fans i got for the pc was too powerfull for the tiny fan controler (a big 130mm x 32mm 12v 1.9A server fan )
Can't you use the PWM from a fan to just drive a MOSFET or BJT directly? The fan will be powered in bursts but if you're lucky they're fine enough to where it isn't audible, this is approximately what the PWM pin does on these fans anyway.
This mainboard doesn't have PWM outputs, that's the whole thing.
@@SianaGearz
while the intel spec says the pwm signal is around 25khs the problem is that most fans have capasitors in the power input, so any pulse gets filtered. Using pwm with one of those would give you a very logarismic looking speed cuve. (already tried that)
the pwm on the fans controls the switching transistor directly so theres no capasivite load on them. But to an externat transistor the fans (at least the big ones and more powerful) at quite capasitive
@@laharl2k I have seen that, for that reason i have jammed in an inductor and a capacitor before the fan on one of the things i built where i use PWM fan control. Inductor... i forget, a shitty tiny little one, somewhere between 50 and 330 µH, i don't remember, with maybe around 2 Ohm series resistance; actually i think i used two, one on each wire; and a capacitor of i think 470µF which was WAAAAY too generous, please choose a smaller one :D The thought that this is total overkill did cross my mind, but the cap just landed in my hand when i grabbed it so i decided to try, and was too lazy to replace it with anything later.
Warum hast du nicht einfach ein Slotblech mit einem einzelnen
Lüftersteuerungspotentiometer dazwischen gesteckt, dass wäre
einfacher gewesen und du hättest es im Nachhinein bei
geschlossenem Gehäuse regeln können!?
Auch möglich, alles kann nix muss ;)
@@necro_ware Ok wenn man es so sieht!😁
Ive always just wired the fan between the 5 and 12 on the PSU.
I just used a pot hacked into a slot shield back in those days ^^
Nice explanation of how everything worked ! No point over complicating things.. i did think you were going to go down the Arduino route or something with a thermistor :)
No need to overengineer things ;)
......or you install SpeedFan in Windows....Works in many Athlon and P4 boards. (:
See 8:38 ;) Unfortunately that seems not to work on this board.
oh nice thank you so much for this video. Fan on my Tualatin 1.4GHz runing at max speed like hell and in bios no settings for speed change. I will buy som modules in Czech Republic cost one 0.77 EUR so is much cheaper than Noctua fan i planned 😊
Welcome back!!!!! Hey Necroware. I have a very rare dauphin orasis tablet, it is a pentium 266 tilamook that wont POST. I have the schematics but drawing a blank trying to fix it. Would you be interested in taking a look?
Dam, maybe need get few of those, just in case.
Or you can use 5 volt from motherboard)) Or 3.3V on ATX... or 5V minus 1V fall thru diod :)))
Good variant is use NMOS transistor with capacitor and two resistors, to make fan run full speed at power on and slowly go down...)
5V is too little, usually the fan Stopps at about 4,5-5V completely, so something between 6V and 9V is ok. But basically, yes, there are plenty of ways, that's why it makes fun to tinker with.
That was a PWM fan. Although this would work on ****ANY**** computer fan, but now do a video utilizing the PWM signal!!!? I'm thinking a 555 circuit, but whatevah!
Yeah! I'll put it onto my wish list.
You may want to look at your microphone setup, there's a lot of hiss noise in the background.
Wellp. Not sure if your calceulations correct. You see, adding resistor in series is equal to creating a voltage devider of resistor and a fan, so first thing what you must - is to calculate actual resistance of a fan. lets do that. 12/0.25=48 Ohm Ok. Now we can see if we add 48 Ohm resistor we will half the voltage to 6V. So dancing around 40-70 Ohm resistor you can adjust desired fan speed. And the last but not least - lets calculate currrent for a resistor of 48 Ohm - 12/(48+48)=0.125A As you can see adding resistor in circuit always limits current too.
In my practice usually 60 Ohm resistor is enougth to make quiet any laud fan and 0.5W resistor is more than enougth.
Yes, you are absolutely right :) I had to take voltage division into the calculation, instead of just taking the current. In my case that wouldn't help though, since all the resistors I had were 250mW. That wouldn't be enough in any case and I'd had to order some.
@@necro_ware Actually fan motor isn`t so linear, and operating on lower voltage may gradually lover current too. You should take measurements of a current drown by the fan. About the wattage of a resistor - i`ve never bothered about that, its just works. BTW you can take two resistors in parallel (120+120=60 Ohm) to double wattage.
@@Evhen_Velikiy Yes, that's also a good idea. Twice the resistance of each would half the current in parallel and result in lower power consumtion. Next time :D
Those adaptors with resistors which they sell are mostly useless, most of fans need less than 6V to make it silent, 7-9V is still too much and when I need to lower voltage that much with resistor, it can get pretty hot and also you just waste that energy, I prefer just to connect it to 5V from molex connector, but in some cases, I use resistors too, for example on old GPUs where I don't want to have some wires going from GPU to motherboard or PSU, but those fans are that small with such a low current that resistor can slower it without getting hot.
I count nothing, I just try few resistors and keep what works the best, from my experiences, it's mostly between 100-200 ohm when you need to slower down super fast fan, but for some little fans, even 50 ohm can be too much. But when you need 200 ohm resistor, it's probably really better just to connect it to 5V from PSU than heat your room with some resistors. 😀
But I noticed that word "silent" means something totaly different for different people, for me, it means that I can't hear fans at all, when I clearly hear it, it's not silent, lowering it from 12V to 9 or 7V mostly just turns terrible noise to just noise. 😀
Most of old fans have possiblity to acces soldered contacts for wires, so you can also connect 2 fans to series which is pretty good solution for some purposes, but annoying is that you can't just disconnect that fan and use it somewhere else then.
Much easier would be if producers of those fans were not idiots and RPM was normal from the factory, when it's 5000+, you just have to do something with that, you can't listen such noise.
I made an ATX -5V adaptor (-5V Volt is missing in the atx 2.0 spec). I even made a video about it ;)
ruclips.net/video/W1I-X_wj6Eo/видео.html
Nice! Did you use some kine of buck converter as well, or just an LM7905? I also made some unusual -5V supply for the mainboards with ISA slots with Phil from PhilsComputerLab two years ago or so. You can find a video on my and his channel about it. Just search for the Voltage Blaster.
@@necro_ware I used the B0505S it is an isolated 5V to 5V converter. It is a mini DC/DC powersupply Does it use BUCK? ;)
Just oil it! Seriously! My machine wouldn't be kicking right now without some good old Wahl clipper oil! Remove the rubber stopper under the fan's sticker, place 2 drops, give it a spin with your fingers to work it in, replace rubber stopper and sticker, and there you go. Good as new!
You must be new to my channel ;) Welcome!
Or for providing 7 Volts without any resistor being used you just disconnect your fan, pull the black and red wires off the fan connector, and then connect red one to the yellow wire of the Molex connector (+12V) and black one to the red wire of the Molex connector (+5V).
Since these voltages have the same ground (black wire) voltages will be substracted one from another and we will get the needed 7 Volts (12-5=7)
I use this method for the older machines, where fans have molex connectors, but in this case it would be too messy for my taste.
@@necro_ware and start a fire because you could trick the short-circuit protection on psus that monitor ground current
@@duke_of_oz That's a good point, never thought about that. I'll have to check out, what will happen if I short 5 and 12V on a good PSU :)
Used to be able to go down to Radioshack and get a linear voltage regulator, a pack of resistors, and a variable resistor. Now you can order a $1 buck converter from China. Ahh, such progress.
In thailand you can still do that, not radio shack ofc but an electronics and stuff store. Discrete regulator components or the aliexpress buck converters boards from a walk in store.
You still can do this in Germany too. Well kind of, currently due to the electronics crisis some parts are hard to find and everything is super expensive, but possible.
Here is another idea: speedfan.
On these computers the BIOS didn't regulate the fan, but it was still possible to regulate it, you just had to use software.
I showed it in the previous video if I remember right. On this board speed fan didn't work for the CPU, but I use it for the GPU fan.
Nice, simple and works well :)
It’s лАуд, not лоуд. Лоуд means a bit different thing :) but you cool 😎 as usual! Looking forward for new videos!
Yeah, sorry for the pronunciation, turned out to be more load, than loud :D
Low noise adapter? Good luck. If your PC shuts down in the middle of playing Crysis ...
I must be dreaming, another video! This isn't the start of a streak, is it...?
A small correction: The calculation at the start involving a resistor is flawed, it doesn't work out that way. If you introduce another load in the circuit, the overall load and therefore current changes and so you can not use the fan's rated current anymore. What you can do is use the rated current at 12V to calculate the fan's equivalent DC resistance first (Rfan=12V/0.25A=48Ohm) and then use a voltage divider formula to calculate the needed resistor to drop the voltage across the fan to 7V. (5V/7V=R/48Ohm -> R=48Ohm*5/7=~34.3Ohm)
Hope this makes sense (and is correct) :D
Edit: Oh and the resistor should be rated for: P=U*I=U*U/R=5V*5V/34.3Ohm=~0.73W
This should get closer, but it still isn’t really accurate. DC fans are not resistive loads, so it may react very differently. Particularly since the fan’s controller IC may not have a predictable curve of current vs input voltage.
@@nickwallette6201 I tried with a fan I had at hand and it came close enough. But does a non-PWM fan really have any other electronics like a controller inside?
you can always use speedfan to do exactly the same thing on all fans...without the extra work
No, unfortunately you can't. I mentioned it in the previous video.
@@necro_ware oh ok then..i didnt get to see that video..it was on my youtube suggestions this one...i guess you cant see the io chip on older hardware as they dont report much back to the user
@@theodordan680Yeah, kind of. To be precise, this board has the rotation sensor, which is reported both in BIOS and SpeedFan, but there's no way to control the rotation speed. I guess, on this model there's no way to do it in software.
@@necro_ware did you try other sw?
@@theodordan680 yes, I tried couple of things, but I forgot to try it in Linux to be honest
Or you could connect to the -5v and +5v for 10v. Or just use the +5v rail?
My solution - buy Noctuas with low noise adapters. lol
That would be too boring for me ;)
This would have doubled the value of the machine and no fun 😄 I have used Noctua in one of my AM2 "Phenom" machines. Just to try, looks cool, works cool but too easy - just swap. The original fan was beyond repair but the metal block is pretty good.
At the time, we solved it much more simply: plus to +12V and minus to +5V = 7 volts. Problem solved. 😀
That was only possible with the fans with molex connectors. I do the same for the older PCs too to get +7V.
@@necro_ware Wenn ich mit so einem Lüfter fertig war, dann *hatte* der einen Molex-Anschluss und lief dauerhaft mit 7 Volt. 😀 Auf einer LAN streikte mal der Chipsatz-Lüfter unseres Jungstars. Damit er weiterzocken konnte (die Läden hatten alle schon lange geschlossen), schmierten wir Butter in das Lager. Ich bin ein Freund pragmatischer und dauerhaft-provisorischer Lösungen. 👍 🙂
Sicher :) In diesem Fall müsste ich aber dann mit dem Daueralarm des PC-Speakers leben, weil der Lüfter nicht angeschlossen ist. Im großen und ganzen bin ich auch für pragmatische, aber doch möglichst ordentliche Lösungen. Was das Bier angeht, ein Freund von mir hatte Kühlschrank in seinem Zimmer neben dem PC stehen und hatte stets eine kühle Bierdose auf der CPU stehen, die er dann regelmäßig ausgetauscht hat. Der PC war definitiv lautlos :D
@@necro_ware Hey, das ist ne Retro-Machine! Die macht man ja auch hübsch! Ich kann Dir mal ein Bild von meinem C64 schicken, der innen aussieht, als sei eine Kabel-Fabrik explodiert. Damals war das normal, heute nicht mehr vorstellbar. Speaking of vorstellbar: warmes Bier? 🙂 Warum hat es den Rechner nicht in den Kühlschrank gestellt?
@@Hessi Auch ne gute Idee :)
coool ill use these to shut up a power mac g5 who refuses to calibrate but runs just fine lol ill use the buck converters to slow down the fans on it
i control my fans by tl317
Or get a motherboard with a better temperature regulation.
Or use a Ryzen PC and close this channel ;)
Do you have such a regulator left? I could need one ;)
Great idea and very cheap improvement!
I thought you will be building a temperature controlled PWM regulator.
No need to overengineer things ;)
@@necro_ware It's not so complex, can be built using a 555 timer.
@@JendaLinda I did that with an Arduino and it completely escalated into a full on fan controller with hardware fan curves and sensor value mixing (matrix) for my water cooling setup. :D