PC Fan Failure Alarm


Need to monitor not only if a fan is running, but if it is running fast enough? Check out this PC fan failure alarm circuit. After several failed attempts using various circuits, they settled on a Schmitt trigger. They even show a couple variations including a manual reset and a relay instead of a buzzer.

[via HackedGadgets]

29 thoughts on “PC Fan Failure Alarm

  1. couldnt this have been done easier with a set of leds and reciever to see if the fan has stopped, normally the fan will be moving fast enough to allow normal circuit, but when it stops, it will cause an blade to break the circuit, as long as you put 2 seperate circuits spaced so that a blad cant cover both. would make the circuit much simpler. but cool design nevertheless.

    1. worthless is in the eye of the beholder. Just because your level of expertise is below the level where you would need this doesn’t make it worthless to everyone.

      I am building an external drive enclosure. I had a manufactured external enclosure but it suffered a fan failure which allowed it to damn near burn the house down. It got hot enough with 3 drives running that it fried its own PSU.

      it is hard to find a product in this category for under $1000 that has any sort of safe guard for this. the authors solution can be adapted easily and cheaply for my needs

  2. Well… even though it says PC fan failure alarm circuit, then there is numerous of other applications where this setup is usefull.

    mrgoogfan -> There is a big difference between software control and hardware control… its not always possible to control things properly with software… speedfan only works if the fans are connected to the motherboard, what if there is no fan-connections to the motherboard avaible?

    dennis -> no… not all bios’ can do this… calling this worthless is just stupid. Weird how people cannot appriciate.

    This circuit would be very usefull if operation with a cooling system for watercooling etc. where the cooling is not always close to the actual computer.

  3. I could have used this about 2 months ago. I only say that because just TODAY did I discover that a client’s PC, that has been BSOD’ing randomly, has likely been experiencing these problems because of a flaky CPU fan; it would start working with the slightest of vibrations, like opening the case. We’d gone through a new motherboard, several reinstalls of the OS/drivers and were ready to get a new hard drive until I just happened to see it throw an exception while it had booted on the install disk but did *not* have the hard drive hooked up at the time; only then did it dawn on me that the cpu fan might not be doing its job at that moment.

  4. “all BIOSes do this….
    worthless “hack””

    no they don’t, I have a machine at home that only monitors the cpu fan, that’s fine what about the case fan, the gpu fan the power supply fan etc etc

  5. PC fans are cheap, and for many projects, a convienient size, also typically more readily-available. This would be perfect for projects that utilize the commonly available pc fan, but don’t use a PC. I might consider this project if I ever getting around to my 900 mhz amplifier project. (Hard to get much distance out of mW)

  6. @dennis

    You’re one of those kids who started messing around with computers with a dual-core right?! oh… and windows :D

    Poor kid, doesn’t know 286, 386, 486Dx2 (ahh… good times)
    Now tell me… where in the hell could I monitor my good old 486 cooling fan JUST using softwarE?

  7. “flaky CPU fan” wow then why dont you just throw the fan out or the whole pc out it cost nothing to buy a new laptop…
    But obviously you don’t value your time.

    There are similar circuits in these internal hotswap racks like http://www.tigerdirect.com/applications/SearchTools/item-details.asp?EdpNo=3142352&CatId=285
    If anyone knows how to add another additional potmeter to it which controls the back 10cm fan let me know. It has a low and high rpm switch but even the low is too loud.

    The problem with the motherboards fan connectors that the fan control normally operated by the bios and it works in windows without installing any additional software but not in linux.
    In linux on most of the motherboards I used the fan was pushed on max rpm and it couldn’t be controlled cause there was no proper PWM driver.
    I2C is broken as well. The only part which works out of the box on it are the sensors in the cpu and gpu for temperature monitoring.

  8. Wow, this IS old, they had this circuit all over the place as project for people and kids since many decades, yes decades, it’s the kind of circuit that must be in many many books teaching electronics since way way back.

  9. “All” bios DO NOT do this. Several modern bios do this, most monitor fan-speed but are incapable of alerting the user of a fan error. This is very useful for legacy computer that do not have this feature.

  10. @YaBa

    No, I’m 23, and most certainly was around in the days of 386,486 etc.

    But why would you want to monitor something that age?! For a start, most of them could run without fans if the heatsink was large enough, and also – they’re obsolete! Save up your pocket money and buy a modern computer.

  11. This is a simple circuit but by no means useless or obsolete, sure many modern motherboards come with fan monitoring sure you can use speedfan,but who said anything about a computer in the first place?
    I am thinking a high voltage h-bridge could be saved by a circuit such as this.Good job HAD.

    PS:Also no Arduino :P

  12. @dennis… because NOT EVERYTHING is about pc’s… there are other applications… home electronics, cooling systems, not to mention the numerous applications you could modify this circuit to control…

    This is not a computer-only site you know… and know that we are at it… loads of the older cpu’s are still being used in other applications, not only computers…

  13. Nice circuit; clear, simple, and just about a bare minimum of components for a reliable circuit.

    It basically comes down to this:
    AC couple the tacho signal, so it doesn’t matter if it gets stuck high or low, and use this signal to discharge a capacitor, which gets charged through a resistor. Finally a voltage comparator with a feedback circuit for hysteresis compares the voltage across the capacitor with the preset value.

    The only thing I don’t really like about it, is that it depends on the value of the decoupling capacitor and the transistor base current (which could use a resistor, btw) to get a slowly rising average voltage across the filter capacitor when the fan slows down. Since the duty cycle is fixed, the voltage across the decoupling capacitor needs to increase to Vcc-0.7 to get the transistor to close completely. This means it’s not entirely independent of the input frequency, above a certain frequency, the voltage across the filter capacitor won’t change any further.

    I don’t see an easy way to fix this without adding quite a few components.

    It’s useful for monitoring fans in things other than computers, and it’s a nice circuit for teaching things like hysteresis and low-pass filters.

  14. @dennis

    Don’t start a flamewar over a simple comment of mine.
    And BTW, i’m 28, and yes, like others say, NOT ONLY computers have fans we need to monitor.
    And yes, my 486 gives a good router (linux based) for a honeypot network ;)

  15. I think this counts as a good hack. That fan with a tach lead was OEM spec built for a PC more than any other end use. So scratch designing and building something that verifies pulse data intended to go to a PC is a hack.

    ANY method of feeding data about fan rotation into a monitoring device is better than none. The fans having an integral pulse out lead are “good enough” for some applications. Even though most of these brushless fans were destined by design for a PC, it’s been mentioned several different ways that not all fans are cooling a PC, so a stand-alone fan monitor can be a Very Good Thing.

    @MadScott: I’d like to see your designs for a scratch built Hot Wire sensor! They’re deceptive in how easy it first appears to do but nonlinear if built affordable. “Cheap, Rugged, Accurate- pick any 2” is my memory.

    It’s why we used to=still do use both fan pulses and “Sail Switch” airflow provers in safety affected situations. With a subset of safeties in HVAC that use VERY sensitive pressure switches.

  16. @dennis I still have a 486 machine which has been in constant use for well over 10 years now, its been ran as a dedicated firewall, or else as a mame machine depending on how I felt at the time, so you see there is plenty of life left in old pc’s…

    @YaBa stop copying off me, its my honeypot I tell you MINE MINE MINE :P I still remember the joy of overclocking a 386 :D oh the speed boost was amazing lmao :D

  17. I’ve tried contacting this guy as I am desperately in need of help, my system now has had the CPU fan stop to not restart again until the power had been reset, in the meantime the CPU cooler has gotten hot enough to fry an egg on, I need to have this circuit as well as one that monitors the TACH to try and figure out if bios is at fault telling the fan to stop for some stupid reason or if the fan is stopping of it’s own accord. So far one threadripper 2950 has fallen victim to this and it nearly happened to a 2nd one – I need an alarm.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.