Exploding fan issues and PWM noise

I just obtained these two massive 120mm fans for my Huxley. I cant find the exact datasheet, but they are 12v, manufactured by Etri, and have the part number 225DS1LP 13510. They are quite literally my biggest fans. (Pun intended )

Anyhow, while I was setting one up in Slic3r and Pronterface, I noticed the very loud whistling that was likely the pwm signal on the Huxley for regulating the speed. This was tested at power setting S200 and below. I attempted to see if the whistling would stop if I send it the Huxley's full fan speed, which I thought was 12 volts, not 19v.
Immediately after I send M106 S255, the fan rockets into warp speed and breaks free of the mounts I had it in. It landed intake side first on the Y axis motor and motor mount and sheered off all the blades, rocketing them to every corner of my room

Long story short, I was wondering if there is a way I can eliminate the whistling/buzzing caused by the pwm signal.

Edited 6 time(s). Last edit at 11/19/2012 08:07PM by Physics_Dude.

I found the specs for model 225DS-1LP11 (not LP13). It should be roughly the same.
Voltage range = 7.0 - 13.8
Current = 710mA (Can the Huxley handle that?)
Input = 8.52w
Speed = 2800rpm
Airflow = 120cfm

Quote

I was wondering if there is a way I can eliminate the whistling/buzzing caused by the pwm signal

You can raise the PWM frequency into an inaudible range, above 20 kHz. For starters, see the FAST_PWM option in Teacup firmware: [github.com]

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

I would prefer to not to alter the printer's PWM frequency for now. It doesn’t seem too healthy for the thing.

I do have a couple L78S12 7812CT voltage regulators, I'll see what those can do tonight.

Update: The regulator seems to have made the fan a bit more happy and quiet, but there is still a very noticeable whistle/buzz. Adding a 25v 10uF cap on the fan's input reduces the sound, but causes the fan to sputter. Increasing the capacitance results in less sound, slightly more sputtering, and a limited fan control. (S100 and up = same speed)

I'm still looking for answers.

Edited 1 time(s). Last edit at 11/19/2012 08:33PM by Physics_Dude.

Correcting myself on something...
The sputtering was due to the regulator (7812CT). Apparently the extra load from the caps caused to the regulator to get a little too hot and drop current. This was tested at around S100 in the G-code.
Sending a steady 16 volts, or 100% duty cycle, to the regulator under load without filtering proved this point.

On a related note, in that last test, I clocked the regulator's case temperature at 115C before cutting the power.

You are sending a square wave PWM feed to a Voltage regulator? I realy don't think they are made to take that. They are made for a smoothed DC supply.
Do you need 120mm fans? Are you trying to cool it or move it across the room?

Brian.

A very simple method to reduce a known voltage to a lower, also known voltage is to run the current through diodes. The 1N4004 reduces ("forward voltage") by 0.75 to 0.85 V, so you'd need 6 to 7 diodes in series. This won't change the buzzing, but it would allow you to use the fan with on-off control.

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

At that point I would seriously start look into getting a 4-wire PC fan. Designed to accept 25KHz PWM signals at 5 V max, steady 12 V input, and a tach output that you're free to ignore (or not). Unless you've got a whole mess of two wire fans for free, they don't really cost all that much more considering how easy they would be to integrate for our purposes.

You can only go from 20% to 100% with this, but that should work swimmingly.

Alright everyone, my problem is solved 100%!

To start, I sourced some new parts. I am using two NJM7812FA regulators in parallel, each sporting a small clip-on heat sink. They are in parallel to distribute the heat evenly across the two heat sinks which would otherwise be too small for a single regulator's needs. The datasheets on this regulator also highlight "Excellent Ripple Rejection" which slightly comforts me a bit in thinking that the PWM shouldn’t disturb them much.

With that, a small 35v 100uF cap on the fan's input eliminates all remaining buzzing. So much so that even with the fan stalled to a stop at 50% duty cycle, virtually no sound is produced. It also retains the ability to give me the full range of control for the fan's speed.

Lastly, all the circuitry is discretely hidden on the unseen bottom edge of the fan.