How to modify a PWM signal in real time?

[math667]

Hi,

Is anybody would know how to modify in real time a PWM signal. I would need a hardware solution, I don't want to use a computer to make the job.

I would like to decrease the duty cycle of the signal by a specific percentage. For example, if the signal is on for 30% of the time, I would by adding a kind of magical box, to get a signal that is on for only 25% of the time. And this should be, of course, good for any duty cycle...

Thanks in advance

 

[fenugrec]

Re: PWM signal treatment

Using a microcontroller it would be fairly easy, and it'll need much less hardware than if you want to do it with gates or whatever...

Take a look at Microchip PICs, a 10F202 should be enough, and it's as small as a SOT23-6 package!

 

[on1aag]

Re: PWM signal treatment

Hi math667,

That's an easy one, use a non-retriggerable monostable multivibrator
like CD4047 and adjust the pulse duration to your own liking.

Regards,

on1aag.

 

[fenugrec]

Re: PWM signal treatment

Hi math667,

That's an easy one, use a non-retriggerable monostable multivibrator
like CD4047 and adjust the pulse duration to your own liking.

Regards,

on1aag.

For a of a fixed frequency & duty cycle, yes. But it won't work if you
want, say 50% *of* a variable duty cycle, as the OP said he wanted.

 

[math667]

Re: PWM signal treatment

The thing is because I don't know too much about electronic hardware.

The application of this is. I got a PWM signal from a controler to a motor and I would like to decrease the duty cycle of the signal from the controler to make the motor believe he must go slower. So I only want to add a kind of magical box between the controler and the motor

Thanks

 

[fenugrec]

Re: PWM signal treatment

Can't you act on the software instead? It'll be much easier to change the speed directly from the PC than to intercept the signal already going to the motor.

Especially if the only PWM you have access to is the one going directly to the motor terminals, in which case you need to add the circuitry to drive the load.

 

[math667]

Re: PWM signal treatment

No, sadly, I can't get access to the computer driving the system, the only thing I can do is intercept the signal between the two components... but it is not really a motor, it was just for example.... I need to intercept a 0-8V signal...

Thanks

 

[fenugrec]

Re: PWM signal treatment

You'll need a uC for that. It's the best solution... check out Microchip's low-pin count devices https://www.microchip.com/ParamChartSearch/chart.aspx?branchID=1033&mid=10&lang=en&pageId=74
if you need your 'magic box' to be small.
Or if you don't want to bother with small packages, get yourself
a 16F628A (or whatever you want, just do your best
to avoid dinosaurs, like 16C* and 16F84s...)

 

[on1aag]

Re: PWM signal treatment

Hi math667,

Try this !

Regards,

on1aag.

 

[fenugrec]

Re: PWM signal treatment

Hang on... Wasn't your goal to take an X dutycycle signal
and reduce it to a 30%-of-X (for example) dutycycle?
Because if that's what you want then the above circuit
won't do...

If I read correctly the datasheet of CD4047 :
( https://focus.ti.com/lit/ds/symlink/cd4047b.pdf )
"Monostable Multivibrator features :
-Output pulse *independent* of trigger pulse duration
"

EDIT
I read the schematic a bit too quick (and haven't simulated
or tried the circuit, either) -- it should limit the dutycycle
to a fixed value (not generate a fixed-value dutycycle as I
first thought)

Give it a try, see if it works for you...

 

[on1aag]

Re: PWM signal treatment

Hi Fenugrec,

You're absolutely right, but the man wanted a simple circuit.
It is obvious that he hasn't the knowledge to use/program a pic.
You read his reaction to your proposal, his face probably turned
green. This circuit is easy to build and cheap as well. Now he has
something to try out and he can observe how his machine will
react to the shortened pulse. There might be a feedback loop ?
Maybe he only has problems when the speed is too high, maybe
he accepts the shortcomings of the circuit and adjust it by hand ?
At least he's got something. But I think he's gonna go for a mechanical
solution, that is something he can grasp.

Regards,

on1aag.