PWM controlled by an LFO? new guy help!

[Zowno]

I'm currently designing a device that requires motor control and upon doing research it appears that a 555 timer PWM circuit is the best way to vary a motors speed. Upon setting the initial speed I need the device to speed up and slow down at a constant rate. Not so extreme as to stop and start again, just vary slightly up and down from the initial speed. I was wondering if one could wire a circuit such as this http://charlieslick.blogspot.com.au/2013/02/ultra-simple-8038-lfo-circuit-pcb.html to be able to vary the initial speed of the PWM circuit.

Many thanks

 

[(*steve*)]

my recommendation is that you use a triangular wave source at the PWM frequency you desire, and use a comparator to compare this with another low frequency waveform of your choice.

This will generate a PWM signal with mark/space ratio determined by the voltage of your low frequency signal.

ensure the low frequency waveform has peak voltages inside that for the high frequency triangular wave if you don't want dead regions above and below 100% and 0%.

 

[Zowno]

Thank you very much for the fantastic answer. For a new guy this has put my mind at ease.

A follow up query: All the triangle wave sources I've found have a variable frequency as well as amplitude. What amplitude would be suitable for this application?

 

[(*steve*)]

A follow up query: All the triangle wave sources I've found have a variable frequency as well as amplitude. What amplitude would be suitable for this application?

I might choose 1/3 and 2/3 Vcc. That way I could take the voltage from the top of the capacitor of a 555 astable which will correspond to an approximate triangular wave.

I could do the same for the lower frequency oscillator too.

Then feed both of these into a comparator.

 

[Zowno]

Thanks again. You're clearing away a lot of clouds for me.

If I apply a controlled voltage from the mains power source of this device to pin 5 of the 555 producing the triangle wave, as I understand this will be able to vary the frequency of the triangle wave and therefore the motor speed?

 

[Zowno]

Or would adding variable resistors give me the needed frequency sweep I require?

 

[(*steve*)]

No need to worry about shenanigans with pin 5.

Comparing a triangular wave with a changing voltage is all you need to produce a changing mark/space ratio.

And you can create that with another 555 operating at a lower frequency (you'll need to adjust the voltage swing to get the appropriate depth of PWM modulation though).

 

[Zowno]

Okay, I think I follow. I'm going to throw some arbitrary figures in here to get my point across. If I wanted to run a motor at say between 300 and 400 rpm and have it adjustable, but then have the lfo ramp up the rate by say 10 rpm then ramp down by 10 rpm from the initial speed but also have it vary to say 5 rpm adjustments or even 30, the circuit you've suggested would be sufficient? Presumably I'd have to be able to adjust the frequency of the original triangle wave and that would control the initial speed of the motor (the 300 - 400 range) then I'd use a comparator to compare that signal to a lower voltage triangle wave that would also be adjustable to adjust the rate of change (The 10/5/30 rpm ramp up and down). I'd then connect this to a PWM motor control instead of having a pot to adjust it? Am I understanding the theory correctly?

 

[(*steve*)]

If I wanted to run a motor at say between 300 and 400 rpm and have it adjustable, but then have the lfo ramp up the rate by say 10 rpm then ramp down by 10 rpm from the initial speed but also have it vary to say 5 rpm adjustments or even 30, the circuit you've suggested would be sufficient?

Maybe.

Without feedback, PWM might set the speed for a given load, but changing that load will change the speed.

Presumably I'd have to be able to adjust the frequency of the original triangle wave and that would control the initial speed of the motor (the 300 - 400 range)

No, the frequency does not control the speed (or the power).

The mark/space tatio controls the amount of power delivered to the motor and that amount of power (along with the load on the motor) determines the speed.

then I'd use a comparator to compare that signal to a lower voltage triangle wave that would also be adjustable to adjust the rate of change (The 10/5/30 rpm ramp up and down).

No, the voltage on the comparator causes the mark/space ratio to change. A varying voltage (such as the one I suggest would cause this to vary and hence the motor speed to change.

That varying voltage would need to be within carefully defined limits to cause a variation in speed that you require.

I'd then connect this to a PWM motor control instead of having a pot to adjust it?

No, this is a PWM controller. All you need to add is a switching element that can switch the current required by the motor. That might be a mosfet and a few assorted other components.

Am I understanding the theory correctly?

Not really :-(

Perhaps you'd be better off doing the whole thing with a microcontroller, perhaps a picaxe. Have you got any experience writing software?

 

[Zowno]

I really seem to have fallen flat on my face with this one.

Okay, let me try and simplify my thinking. The motor is a 4 pole induction motor. If I step down the mains power from whatever Australian power is to something more manageable, the 50hz power cycle will cause the motor to spin at 1200rpm. If I can create a circuit to vary the Hz and then route it into a comparator to compare with my LFO signal, would this be a suitable solution?

 

[Zowno]

I think I'm still not understanding. I'll give the whole thing a rethink and update this post when I think I have something workable.

Any areas you suggest I research?

 

[(*steve*)]

PWM for an induction motor is a far more advanced topic.

I suggest you research speed control of induction motors. From memory there was a circuit to do just this in a fairly recent Silicon Chip magazine.

Yeah, it was April 2012 http://archive.siliconchip.com.au/cms/A_112812/article.html

At the very least, see how they did it.

 

[Zowno]

Unfortunately because I'm not a subscriber I don't have access to that article.

I didn't think any of this would be so hard, quite a learning curve I have to say.

The motor in question is that of a reel to reel tape machine, they already have speed control and I thought it would be easy to adapt. I'm trying to create a machine similar to the machine Ken Townsend modified for his sessions with the Beatles to produce ADT. His machine used an oscillator to vary the motor speed but his had a big knob to manually vary the speed. I was hoping to replace that by an LFO that would automatically vary the speed.

I thought all of this would be possible, but maybe I'm wrong.

Thanks for all the help

 

[(*steve*)]

for variable speed control of an induction motor you will need variable frequency AC. That's not as trivial as you might think.

A relatively simple solution may be a variable frequency sine wave oscillator driving a power amplifier.

Could you get away with something other than a pure sine wave? Maybe.

maybe you could use pwm and an H bridge to generate the equivalent of a sine wave.

 

[Zowno]

Okay, that makes sense. Would I be able to subject the sine wave generator to the same comparator circuit as previously discussed? Or would the power amp output be subject to the comparator?

 

[Zowno]

I think I've misunderstood the role of the power amp. Would you be able to clarify?

 

[(*steve*)]

No, the amplifier directly powers the motor.