On the quest to quiet down my Mini, I’ve done the Y motor damper - which works great. However, when idle, the Mini case fan is loud. After reading here about replacing the Taz fan with a Noctua, I ventured to try the same with the Mini case fan.
The Mini’s Rambo put’s out 24V to the stock fan. I also measured it and found that on power up, it hits 24V, then drops to about 15V. Although the Rambo1.3 likely uses PWM to drive the fan circuit voltage, after looking at the schematic, I am fairly certain the power output to the fan itself is not PWM (just the to the transistor).
I decided to take a gamble and pick-up a 12V Noctua NF-A8 FLX 80mm fan (rated as very quiet and efficient). Knowing it was 12V, I played around with a few options to drop the 24V. I first used a voltage divider (2x 100Ohm resistors) which worked well, but it’s very inefficient and the resistors got hot. However, this allowed for the variable voltage (somewhere between 7-12V) to allow the Rambo to adjust the speed of the fan as it does with the stock unit.
I then put a switching regulator on it that maintained the voltage at 12V. This works fine but will not allow for the Rambo to control the fan speed (the regulator keeps the voltage constant @ 12v).
Next step as to test the fan mounted in place of the old fan. This is where the game changed. The Mini case design for the fan inlet vent holes significantly amplifies the fan noise. I know this is somewhat of a duh moment, but I did not expect the Noctua to get THAT much more noisy once it was mounted. It was very quiet on it’s own, where the stock fan was notably more noisy.
At this point, I am back to SQ 1. The stock fan is still in there. I think the Rambo drives these fans at too high a RPM which could be reduced and still not cause a heat issue, so that is where I am going to look next - in the firmware to see if the speed can be dropped. Also, the Mini fan is on all the time, which is part of the issue. I am not sure if the firmware does any sort of temp monitoring, but I’m going to look at that too. I don’t like messing with FW because it’s just one more thing to remember when you upgrade, so my other option is to build a standalone circuit (I have lots of digisparks sitting around) that I can use to control the Noctua @ 12V based on a few temp sensors placed within the case (possibly on/near the Rambo as well as the power supply. With this approach, I think figure I would run the Noctua at about 7V which would be much more quiet - even when mounted.
That said, it would be interesting to know if LulzBot are looking at ways to better engineer the metal case and the fan inlet to reduce the noise the air makes as it is sucked in. I might even look at dremmeling out the fan slots and printing something to allow it to be more quiet. Some sort of duct…
Lesson’s Learned:
-
The Mini power supply does not have a fan. It uses convective cooling from the case fan. So, dont waste your time seeking to reduce noise in the power supply.
-
Don’t assume that replacing the fan with a better one will solve the problem.
-
If you do get the Noctua fan, you will need to widen the case mounting holes if you want to use the rubber inserts. (this will makse sense when you see the pkg contents)