Afterthought: One motor option that was not mentioned is - drumbeat:) - a stepper motor! :) You can go as slow as you want w/o any gearboxes :) And - they have really tiny ones :

https://www.amazon.com/Acxico-2-Phase-U ... 09CTT8Z2G/


Well this option is a bit heavy on electronics and wiring - but if you're open to moving power supply and controller into the base you can have a commercial board + remote controller for speed for like $25 :) Driving multiple motors from the same controller can be done - but comes with some extra considerations...

FYI: If this helps - I've modeled in the computer the tiny step motor above together with rough 30mm prop model and scaled it up back to B-36 size with overlay of actual b-36 engine photo to show the volume it takes:



But soldering 4 wires to it and foam(?) mounts would take a microscope and lot of work and patience :) And probably good deal of remaining volume... :)

For fun I've posted a short animation of this motor with 30mm diameter prop here :) : https://www.veed.io/view/ed898e76-11f1- ... 633056050b

Perch is correct about the diode, but incorrect about the lower voltage vs pwm assessment. PWM will have a lower minimum running speed than varying the voltage.

The reason is that typically mechanical systems have a much slower impulse than electrical, and brush motors follow this well.

The PWM pulse is well over stall current, so it induces a movement in the near linear operating speed of the motor for a fraction of time. To the viable eye and the mechanical system's impulse response, the result is near constant rotation speed. The pulse off time will have the system still running but at a slower interpolated speed.

The voltage method has voltage that will have to overcome the stall current and start spinning. Because the back emf and running friction are much smaller the static friction/moment of the motor, this small-load & minimum speed is probably faster than is wanted in this application.

Here is some actual data for this.
https://cdn-learn.adafruit.com/download ... motors.pdf

The data also shows how there is the operating range of the pwm frequency, too slow than its slower than the mechanical response, and too short it won't be enough force under the curve to overcome the mechancal moment + resistance.

I think Perch may have come across a system where the pwm rates/cycle% was outside of the motor operating range, so he was observing a diminished voltage response. Only real bad thing about PWM is the possibility of audio resonance.

In Short
Voltage Controlled
+simpler circuit
+no possible audio resonance
-min speed from stall is higher, and running below it may burn out motors (say if battery is low)
-min speed will vary by motors (might be noting)

PWM
+min speed is lower
+variance in speed is lower
-more complex circuit
-possible audio resonance (may be nothing)

Thanks, guys. Bear in mind my last time working with this stuff was 1999, so I have a lot of refresher and catchup learning to do--frankly, with Perch finding an audio recording, I think my buddy would find a sound board with audio more impressive than all six having precise synchronization.

I'm not even sure they fully synched on the real thing - once airborne and up to cruise speed it was a standard practice to completely shutdown two and sometimes even three engines until needed for the "target-area dash" to save fuel. (And all ten firewalled weren't enough to outrun a 15-megaton "Runt"--the test-drop on the Mk 17, the drop plane had to be left isolated for a YEAR for the radiation to cool down for inspection and it was found "beyond repair." Granted, Convair and the crews TOLD the HMFIC's the bomb NEEDED a parachute retarder to slow its fall and the Brassholes vetoed it...)

Some more info: Props are 32mm (almost but not quite 16 straight-line from shaft centerline to one tip), each has a rotating mass "fresh off the sprue" of 0.117 grams. Will be a little less after I cut out the original shaft, hollow out the spinner inside for mount on the driveshaft and turn down the spinner outside from the kit "ogive" to more like the "cones" of the real thing.

How do I calculate mechanical load on a motor?

the load will mostly be from the air displacement it creates and will increase with RPM and is cubic in nature.
The other is from the rotational moment calculated by the distance of mass from the center of rotation.

In actuality, the weight is so small and the blades would probably shatter at speeds were air displacement matter.

I would just calculate for a no load condition, so you need at least the turn on voltage to overcome the static friction.

Another thing is, I'm considering brushing a thin layer of epoxy over the blades trying to reinforce them, and filling the hubs with same to compensate for the amount I'm going to have to take off reshaping them from bullets into cones.

[...]

Would something like this work?
DC 1.5-3.7V 42000RPM Coreless Micro DC Motor 4mmx8mm for Airplane Model RC Toys, Pack of 5 https://a.co/d/426zzRr

Physically fit but WAY too fast turning by a factor of 100. Remember, at full size they ran at only 900rpm (try to imagine those ten-foot paddles swinging through a full turn every FIFTEENTH of a second, if it were converted from rotational into straight-line movement those tips would be moving at the speed of a .45ACP bullet!) and for both structural and handler safety reasons we're looking for around 120-240rpm for the model. Remember, we're not trying to make this thing fly, just LOOK like it is! :)

I understand that, but I figure they can probably run a heck of alot slower as well. Maybe not 240 rpm, that's pretty slow!