5015 SuperFan and UltraFan

Files in the Patreon shop (included with patronage): https://www.patreon.com/squirrelf3d/shop/squirrelfan-5015-fan-super-powered-by-180267

A single one of these fan is of course enough to provide plenty of air power, but running it at lower speed to reduce the noise and only a small tweak to the apollo to accommodate them is the short term goal.

Later will work on a shroud with one of those integrated into it.

I'm incrementally adding the SuperFan and UltraFan labeled model of the main body of the ApolloLander for a symmetrical fan output. ducts coming soon.

Superfan is the one that looks like a 5015, Ultrafan is the one with the wider opening, which gained 1g at full speed with the same blade design installed.

I also added a step file if you want to play with different blade design. (remember it's meant to be printed face down) for a flat contact with the motor

The files are labeled with a random name starting from A and going through the alphabet, just based on the design order. Next to the name is a weigh value, based on the experiment I did to measure each fan air pressure onto a scale placed 5mm above.


---------
This is a copy paste of the BOM and info txt file in the zip
so might have some redundant info from above.
---
These motors should theoretically work anywhere between 3.7v and 8.4v

Just make sure the ESC can support those voltage too, most ESC seem to run between 6v and 8.4v

One of these is of course plenty of air, but my hope is to run them at lower speed, and reduced noise, so would still be used on a dual fan setup.

New apollo ZIP will be updated to work with those fans, you will see an ApolloLander Mainbody labeled SuperFan, using the standard 5015 size model I made, and one called UltraFan with the wider fan opening. most hotends have the main body as of now.
Apollo backplate for symmetrical will be done by request.
You can otherwise use these with the Apollo as is, using the normal Apollo parts,

========== Tools  ========

you'll need a micro screwdriver, especially for Torx screws (I listed philips screws as well)
This is a cheap set I got that works great
https://www.aliexpress.us/item/3256806184008327.html

======= WORD OF WARNING ========

I will not be responsible for any injuries! Be warned we are talking about fans that can spin at 20K~30K RPM

Make sure you take safety precaution, in case the fan breaks apart, I don't expect them to do so, but we never know.

The blades I designed are attached on top and bottom by a ring for strength. But if you design a fan blade like standard 5015 that are only attached at the bottom, I would be afraid of them flying off from the high centrifugal force, velocity and air pressure, thus don't recommend that design approach. In short keep the bottom and top ring.

They will eat through a nail or finger like sandpaper if you touch them at high velocity. If trying to figure out the spinning direction have them spin at very low speed.

================================

====== few info about ESC and the motors =====

These motors are dumb, the are brushless 3 phase DC motor, there is no electronics in them, only magnets and electromagnets split in 3 groups (thus 3 wires), the ESC job is to power the group in sequence to get the motor spinning, but to do that it also reads the feedback generated from the idle set of electromagnets to determine the state the motor is in, and allows to figure out the perfect timing without actually using a normal tachometer setup. All that to say that each motor and ESC will become one entity, you cannot have one ESC for two motors.

I have not tested if a single PWM wire from a printer board can be connected to multiple ESC, there is a chance that's possible (not responsible if that's not the case and your burn up something)

These motors should theoretically work anywhere between about 3.0v and 8.4v

Just make sure the ESC can support those voltage too (rated 1S or 2S)

1S = 3v~4.2v (basically the range on a single cell battery between being full and almost empty)

2S = 6v~8.4v (the range of two cell battery in series)

Some ESC have low battery safety, so an ESC may not like 5.5v, as it will assume it's a two cell battery setup that is very empty.

All the ESC I tried can be calibrated by starting them at 100% speed, then down to 0% after a couple beeps

Some ESC can be programmed by starting it at 100% and waiting after a few beeps, some settings can be interesting, so take a look at them.

One of these fan provide of course plenty of air, but my hope is to run them at lower speed, and reduced noise, so would still be used on a dual fan setup. There is some PWM noise, so not as quiet as a PWM 4wire fan.

These will be adapted to fit the Apollo Lander shroud, for ender3 type printer ( http://patreon.com/squirrelf3d ) with a symmetrical dual fan setup.

I've had one faulty motor out of 4, faulty in the sense that it's extra noisy in an annoying way. So it happens.

Rev2 top/bottom part of the case are not compatible with the original case, the fan blades are unchanged.

Tolerance on rev 2 were increased to avoid fan blades scraping the bottom of the case

========== Motor - Micro 1104 4300KV ========

 - Motor (or any other sellers you find that sell these, Micro 1104 4300KV, as long as they are exactly the same the are this one, there are similar sized one with small difference, but this is not designed for those):

https://www.aliexpress.us/item/3256805971807081.html

https://www.aliexpress.us/item/3256803954722759.html

========== ESC (electronic speed controller) ========

One ESC per motor

This ESC worked from 4.2v to 5.1v, on a 25watt 5v PSU (voltage define the max speed, 4300rpm per volt):

 - https://www.aliexpress.us/item/3256805913773225.html

These Esc all worked at voltage between 5v and 8.4v

- 12amp model, https://www.aliexpress.us/item/3256805028458826.html

- 10amp model, https://www.aliexpress.us/item/3256803041698619.html

Only worked at 6v~8.4v

- 20 amp model, https://www.aliexpress.us/item/2251832793301329.html   <- this is my favorite noise wise, but has instant breaks, which isn't good for the buck converter, but hasn't burned out so might be ok as is, a firmware update will solve that, but that requires another board to flash it, will post more info soon once I go through the flashing.

- 20 amp model, https://www.aliexpress.us/item/3256806226017420.html   <--this is supposedly one of the better one of the lot here in term of reliability. But had a high 12Khz pitch when I tried, so I'm not using it.

========== Bottom Screws - four M1.4x3mm========

 - Bottom screws any "normal" M1.4x3mm will work, but the screw head will stick out the back.

These will be pretty flush against the back

M1.4x3mm countersunk - torx head

https://www.aliexpress.us/item/3256805610161003.html

M1.4x3mm countersunk - philips

https://www.aliexpress.us/item/3256805105462100.html

These will hold more thanks to the flat surface but stick out the back a bit, 0.3mm

M1.2x3mm flat heat - philips

https://www.aliexpress.us/item/3256805084387535.html

(2mm long screws will work too, but will only grab 0.8mm of threads instead of the full 1.5mm)

========== Top Screws - four M2x3mm counter sunk ========

M2x3mm countersunk - philips head

https://www.aliexpress.us/item/3256804078422852.html

These are Torx head, which I would recommend if you set of tools for it

M2x3mm countersunk - torx head

https://www.aliexpress.us/item/3256805610161003.html

========== buck converter ========

Buck converter (powered from the 24v PSU) worked great, I tested a 3amp buck I had on hand, but wouldn't recommend them as that's pushing the max amp the buck can support going from 0% to 100% speed.

Instead get a 5amp buck (5A XL4015 with one potentiometer on it, blue box, the other common version has two pot and complicates things for this usage)

https://www.aliexpress.us/item/2251832462782484.html

Use ONE BUCK PER MOTOR top play it safe and not overload the buck.

Set the output voltage to anywhere between 6.8v and 8.4v, the higher the faster the fan will spin. Some ESC will allow running down to 6amp

(One exception: one of the ESC listed further up will only run between 4.2v and 5.1v)

======= optional pi pico 5 USD ==========

Look for official retailer here: https://www.raspberrypi.com/products/raspberry-pi-pico/

Most other places inflated the price a lot

This would run klipper as if it was a secondary printer board, but have the fan connected to it. Haven't done that yet but I will soon, and document the process.

========== Final steps ========

The ESC will need calibration, so set the speed to 100% ( maximum_pulse_width: 0.002 ), power them on, they will beep a 2~3 times, and within a couple seconds, set the speed to 0% (minimum_pulse_width: 0.001), you will hear more beeps and with a few second will be usable and calibrated (I believe the calibration is stored on the ESC so should survive power off)

Klipper:

--------

#you can edit the name of course, just remember to also change it inside the gcode_macro M106 below

[servo squirrelfan]

pin: PB0

# PWM output pin controlling the servo. This parameter must be

# provided.

maximum_servo_angle: 255

# The maximum angle (in degrees) that this servo can be set to. The

# default is 180 degrees.

minimum_pulse_width: 0.001

# this is a standard values used by ESC, keep as is

# The minimum pulse width time (in seconds). This should correspond

# with an angle of 0 degrees. The default is 0.001 seconds.

maximum_pulse_width: 0.002

# this is a standard values used by ESC, keep as is

# The maximum pulse width time (in seconds). This should correspond

# with an angle of maximum_servo_angle. The default is 0.002

# seconds.

initial_angle:0

# Initial angle (in degrees) to set the servo to. The default is to

# not send any signal at startup.

#intercept and override the M106 gcode command (the standard part cooling fan speed gcode)

#Slicer will emit M106 command without the P parameter

#M106 S0 #will power off all fans, [fan] defined fans and the squirrelfan/servo fan

#M106 S100 #will set both [fan] defined fans and the squirrelfan/servo fan speed to max speed, or comment out the line "M106.1 S{params.S|int}" in the {% else %} statement to only power the superfan/servo fan

#You can use the P parameter to control only one type of fan, P1 for [fan] type fan, and P2 for the squirrelfan/servo fan, like so "M106 P2 S100"

#or simply use with values between 0 and max angle set in the servo section.

#SET_SERVO SERVO=superfan ANGLE=255

#below you edit the ANGLE value, I left a 1.0, but you can change that to 2.0 or nay other factional value to reduce the top speed. if you divide by 2, so M106 S255 would mean 50% of the potential max speed of the squirrelfan/servo fan, but you can still issue a "M106 S510" for the purpose of calibration without klipper error.

[gcode_macro M106]

rename_existing: M106.1

gcode:

    {% set s = params.S|float %}

    {% if params.S is defined %}

      {% if params.P is defined %}

          {% if params.P|int == 1 %}

            M106.1 S{params.S|int}

          {% endif %}

          {% if params.P|int == 2 %}

            SET_SERVO SERVO= squirrelfan ANGLE={s/1.0}

          {% endif %}

      {% else %}

          M106.1 S{params.S|int}

          SET_SERVO SERVO= squirrelfan ANGLE={s/1.0}

      {% endif %}

    {% endif %}

============================================================
=============== Blade design notes (copy paste for the zip) ==========

The gram values in the filenames are useless outside of this experiment, there are just a way to compare fan blade performance between each other.

For reference a GDStime dual ball bearing 24v 6K rpm, would have a result around 2.5~3g, so even the worse fin design is better than the GDStime. Unsurprisingly as this motor is rated 4700k/v, so at 5v would spin at about 23K RPM, at 4v would be around 19K RPM

I used a small fan mount I made wrapped around a precision scale to get the active air pressure blowing onto the scale in a repeatable fashion.

With fan output 5mm above the scale, aiming the airflow down to its center.

Values would vary about 0.25g, I would select a value closer to the low end. Repeatability isn't totally consistent, but would be within 0.25g at least, so consider that the variability of my test.

So two fans that are recored as pushing 14.75g and another as 15g, may actually be equal.

----------

FAN blade Design 

In the step file you will find a fan without blades, has a motor cap, with a skirt as your base (there is an alternative model with a smoother slop, I included it just in case, but so far resulted in being worse by about 1.0g to 0.75g on similar fin designs)

You will also find an upper ring, and a lower one, which defines the outer side of the fan blades, everything should be contained within it.

The top ring opening should remain the same to remain compatible with 5015 fan openings.

The fan is meant to be printed upside down, so keep that in mind regarding support-less printing.

-----------

For symmetrical fan setup remember to mirror them in the slicer in either X or Y direction