Fan connections... variable speed fan not working

[dcastor]

I have a red light on my RAMBo board... it's LED5...can ANYONE tell me what that indicates?

Thanks.

Dave C

[Jules]

These fans are either on (100%) or they are off. (These are not actually variable. S3D works with a lot of different printers though, so they show that variable speed option.)

There is normally a little red light showing inside the electronics case when the printer is on, at the SD card bay.

[insta]

These fans are either on (100%) or they are off. (These are not actually variable. S3D works with a lot of different printers though, so they show that variable speed option.)

There is normally a little red light showing inside the electronics case when the printer is on, at the SD card bay.

Which fans ... ? The bed-cooling fan is, to an extent, variable speed.

[Jules]

These fans are either on (100%) or they are off. (These are not actually variable. S3D works with a lot of different printers though, so they show that variable speed option.)

There is normally a little red light showing inside the electronics case when the printer is on, at the SD card bay.

Which fans ... ? The bed-cooling fan is, to an extent, variable speed.

Bed fan. Never been able to get variable to work.....it's always been either off or 100% on. Setting the S3D cooling amount to 60% for instance just turns the fan on for that particular layer. (Never noticed it going slower.)

[jsc]

I did some investigation into bed fan control in the early days of this forum. Marlin controls the fan speed using PWM, turning the current to them on and off at a very fast rate. Turns out fans aren't just simple DC motors these days, and some don't take very well to this kind of control. Slowing down the PWM rate from its default very high setting to something much slower can often get those fans to be controllable. Requires a bit of poking around in the firmware, and that requires that you get the proper firmware for your exact machine. See: viewtopic.php?f=7&t=1496

Also something that should work, but I've never tried, is putting together a little two component low pass RC filter to smooth the current. I leave that as an exercise for any of you EE guys with oscilloscopes to test.

[pyronaught]

PWM is a pretty common way to control speed on DC motors and should be very responsive to different pulse widths on any DC motor. If it is not working then I would suspect the software isn't implementing it correctly.

[ednisley]

PWM is a pretty common way to control speed on DC motors

These days, fans aren't good old dumb DC motors with carbon brushes. They're now brushless DC (BLDC) motors with an embedded microcontroller that chops the DC into multi-phase AC that drives the rotor.

https://en.wikipedia.org/wiki/Brushless ... tric_motor

When you apply PWM to a BLDC fan's power supply, you're basically killing the microcontroller between each pulse. The only reason the speed goes down is that the microcontroller takes a while to reboot and resynchronize the rotor's position. That's why only low-frequency PWM works at all: there's enough time for a reboot during every pulse. As the duty cycle drops, the pulses become so short that the controller can't start up, so the fan stops.

Also, "smoothing the current" doesn't work, because the filter will drop the fan's terminal voltage below the BLDC controller's brownout point, whereupon it shuts down. It will, however, deliver more-or-less constant RPM as the voltage drops, because it's desperately trying to maintain the preset RPM.

There is no way to control a BLDC motor with the PWM found in RepRap-style firmware. Use 100% all the time and be done with it.

[pyronaught]

Dang, I wouldn't have thought these el-cheapo fans would have all that going on inside. I'm going to hack the next one that fails apart and see if it has all that tech.

I have a couple bed fans that just won't run for whatever reason. I'm printing the same gcode file on 9 machines and the fan won't run on two of them. Haven't tried switching fans around yet to see if it is a fan issue vs a wiring/rambo/software issue.

[ednisley]

all that going on inside

Even on the low end, a complex silicon chip costs less than all the fiddly mechanical parts it replaces, probably because it also offsets the assembly labor: fans have one moving part with a single bearing!

[ksevcik]

There is no way to control a BLDC motor with the PWM found in RepRap-style firmware. Use 100% all the time and be done with it.

I've been thinking about hacking together a proper 12V supply + pins to make it possible to use actual PWM controlled fans. 2A DC-DC converter for $15, teeny perf board for $8, plus miscellaneous wiring and 0.1 headers. Wiring the fan's PWM pins straight to the PWM pins on the RAMBO should work fine, as long as you avoid ones using a timer that's already claimed for the heatbed/etc. There's gotta be an application for more than 0%, less than 100% fan speed.