Brushless DC motor drive.

[boylesg]

Surely you could setup 3 x 555 based multivibrators and introduce delays between the by passing 2 of the square waves through additional components. But I am having great difficulty in finding out how to do it.

Can anyone point me in the direction of a suitable circuit?

 

[BradtheRad]

Flip flops can be cascaded to form a shift register. Three of them rotate pulses to 3 outputs.

By varying the incoming duty cycle, you can change the 'on' times of the outputs.



Output waveforms depend on which logic gates you use, and which output you tap at the flip flops.

 

[boylesg]

Bugger this. By the time I frig around with assorted ICs and different methods it would be more cost effective to just buy one of these and have done with it.

**broken link removed**

The fact that info on the web seems quite sketchy and various examples on youtube seem far from convincing, it probably means that it is just really not that easy to achieve without a professional package.

I saw a hard drive engine working well with one of the above type devices though so I will give it a try.

 

[d123]

I saw somewhere in the Tony Kuphaldt lessons in those ibiblio e-books - to summarise the concept: a description of using a CD4017 to create a three phase cyclical signal interfaced with a transformer or something else to power a light aircraft AC propeller motor. I'll try to find the example, but don't hold your breath, I read it a few months ago...

Found it! It's page 410, Ring/Johnson Counters, (and just before that the 4017 is mentioned, same thing), maybe you can adapt the 4017/counter to your idea.

https://www.ibiblio.org/kuphaldt/electricCircuits/Digital/DIGI.pdf

 

[boylesg]

I saw somewhere in the Tony Kuphaldt lessons in those ibiblio e-books - to summarise the concept: a description of using a CD4017 to create a three phase cyclical signal interfaced with a transformer or something else to power a light aircraft AC propeller motor. I'll try to find the example, but don't hold your breath, I read it a few months ago...

Found it! It's page 410, Ring/Johnson Counters, (and just before that the 4017 is mentioned, same thing), maybe you can adapt the 4017/counter to your idea.

https://www.ibiblio.org/kuphaldt/electricCircuits/Digital/DIGI.pdf

I feel I understand the concept well enough. But achieving it, and getting the timing just right so that they spin like a regular dc motor, is another matter.

 

[BradtheRad]

I believe the waveforms need to be PWM, particularly at slower frequencies.

If not done, then you're applying ever-longer pulses to each coil as you slow down the rpm, thus drawing overmuch Amperes.

This means my post #2 schematic will not work, except when you wish to run at maximum speed.

 

[boylesg]

I believe the waveforms need to be PWM, particularly at slower frequencies.

If not done, then you're applying ever-longer pulses to each coil as you slow down the rpm, thus drawing overmuch Amperes.

This means my post #2 schematic will not work, except when you wish to run at maximum speed.

It should be easy with one of those ESC things. All I would need is an astable 555 at 500Hz with a pot as R1 and R2 and a power supply.

 

[d123]

Hi, I actually agree with you - re: that link to the electronic speed controller, good price too; the time spent messing around with devices and getting timing almost but never perfectly right is only worth if for the sake of the learning experience and fun involved.

I suspect that even using metal film resistors and appropriate dielectric capacitors, both with the same temperature coefficient as the timer, and for example the LMC555 would still be not quite perfect timing, and probably cost more than the ESC you linked to.

(Off and on I'm making a "toy" with three little DC motors, 555s and 4017s, and testing and tinkering show me that to co-ordinate their timing to the millisecond as I would have liked just isn't going to happen, it's a clunky labour of love more than a precision device...).

Your way seems the most sensible time and cost-wise.