rising PWM frequency

[Matteo Ferrari]

Hi.
I need to design a circuit to rise a 100 Hz PWM input signal to a 25
kHz PWM keeping the same duty cicle. I thought about filtering the 100
Hz input to obtain a linear control to regenerate a brand new 25 kHz
PWM. Isn't it possible to simply step up frequency, with a PLL
syntetizer for example? In this case, will the duty cicle at 25 kHz
mantain the same ratio?
Thanks.

 

[Homer Simpson]

Hi.
I need to design a circuit to rise a 100 Hz PWM input signal to a 25
kHz PWM keeping the same duty cicle. I thought about filtering the 100
Hz input to obtain a linear control to regenerate a brand new 25 kHz
PWM. Isn't it possible to simply step up frequency, with a PLL
syntetizer for example? In this case, will the duty cicle at 25 kHz
mantain the same ratio?
Thanks.

Pll won't track reference duty cycle. You have to construct brand new
25KHz signal and modulate it with filtered original signal.

 

[Luhan Monat]

Hi.
I need to design a circuit to rise a 100 Hz PWM input signal to a 25
kHz PWM keeping the same duty cicle. I thought about filtering the 100
Hz input to obtain a linear control to regenerate a brand new 25 kHz
PWM. Isn't it possible to simply step up frequency, with a PLL
syntetizer for example? In this case, will the duty cicle at 25 kHz
mantain the same ratio?
Thanks.

Hi,

Why are you doing this?

Your answer may help others provide alternative solutions.

 

[mdp]

You could design two PLLs, one locked to the leading edge, one locked to the
falling edge of the incoming signal. Using simple logic gates and a divide
by N PLL design, you can produce a multiple of the incoming frequency with
the same duty cycle. It gets more complicated if you want the edges locked
to each other in some fashion. Probably harder than it's worth.

 

[Genome]

You could design two PLLs, one locked to the leading edge, one locked to the
falling edge of the incoming signal. Using simple logic gates and a divide
by N PLL design, you can produce a multiple of the incoming frequency with
the same duty cycle. It gets more complicated if you want the edges locked
to each other in some fashion. Probably harder than it's worth.

That is smart......

DNA

 

[Robert Monsen]

Hi.
I need to design a circuit to rise a 100 Hz PWM input signal to a 25
kHz PWM keeping the same duty cicle. I thought about filtering the 100
Hz input to obtain a linear control to regenerate a brand new 25 kHz
PWM. Isn't it possible to simply step up frequency, with a PLL
syntetizer for example? In this case, will the duty cicle at 25 kHz
mantain the same ratio?
Thanks.

Using a PIC at running at 20MHz, you'll have accuracy to 1 in 200 on the
25kHz duty cycle. There are faster uC available if this isn't good enough.

If you don't need to vary the duty cycle while the system is running, you
can simply modify the clock being fed to the PIC. The duty cycle will remain
constant.

Regards,
Bob Monsen

 

[Matteo Ferrari]

Hi,

Why are you doing this?

Your answer may help others provide alternative solutions.

I need to drive a 12V DC electric motor with a 25 kHz PWM signal. The
100 Hz PWM signal come from another chip and has variable duty cycle.
Wider is the duty cycle faster the engine runs. Normally I use a
microcontroller to do the raising job, but this time I would a cheaper
and simpler solution.
The output signal will feed a BJT totem pole generating the needed
current to switch a big driving MOSFET.

Thanks a lot, Matteo.



__________
|frequency |25 kHz PWM _ _
100Hz PWM ---> | riser |------------------->|_(M)_|
30% duty cycle |__________|30% duty cycle

 

[Ban]

Matteo Ferrari wrote:
|| Hi.
|| I need to design a circuit to rise a 100 Hz PWM input signal to a 25
|| kHz PWM keeping the same duty cicle. I thought about filtering the
|| 100 Hz input to obtain a linear control to regenerate a brand new 25
|| kHz PWM. Isn't it possible to simply step up frequency, with a PLL
|| syntetizer for example? In this case, will the duty cicle at 25 kHz
|| mantain the same ratio?
|| Thanks.

Matteo,
all what you want is only possible if your 100Hz signal has no feedback from
the motor, there is no loop. Otherwise the additional dead time could make
the loop unstable and might even blow your power devices or the motor.

If this is just to control the speed for a model train or something, you
could do as you wanted, filter the PWM with a looong time constant, and feed
the DC to a comparator which has a 25kHz triangle voltage as a reference.
You could do that all with a dual comparator. If you want I could draw up a
small schematic.
Well, you can just add a pot to this modulator and don't need the old 100Hz
circuit at all.

 

[Matteo Ferrari]

Matteo Ferrari wrote:
|| Hi.
|| I need to design a circuit to rise a 100 Hz PWM input signal to a 25
|| kHz PWM keeping the same duty cicle. I thought about filtering the
|| 100 Hz input to obtain a linear control to regenerate a brand new 25
|| kHz PWM. Isn't it possible to simply step up frequency, with a PLL
|| syntetizer for example? In this case, will the duty cicle at 25 kHz
|| mantain the same ratio?
|| Thanks.

Matteo,
all what you want is only possible if your 100Hz signal has no feedback from
the motor, there is no loop. Otherwise the additional dead time could make
the loop unstable and might even blow your power devices or the motor.

If this is just to control the speed for a model train or something, you
could do as you wanted, filter the PWM with a looong time constant, and feed
the DC to a comparator which has a 25kHz triangle voltage as a reference.
You could do that all with a dual comparator. If you want I could draw up a
small schematic.
Well, you can just add a pot to this modulator and don't need the old 100Hz
circuit at all.

There is no feedback at all. I need only to control a DC motor
rotating speed.
Yesterday I built a low pass filter using a 100 uF capacitor and a 10k
resistance. It feeded a comparator with a 25 kHz to obtain the new PWM
signal. Now I've got to make it as linear as possible refferred to
duty cycle amplitude. It should work.

 

[Ban]

Matteo Ferrari wrote:
||
|| There is no feedback at all. I need only to control a DC motor
|| rotating speed.
|| Yesterday I built a low pass filter using a 100 uF capacitor and a
|| 10k resistance. It feeded a comparator with a 25 kHz to obtain the
|| new PWM signal. Now I've got to make it as linear as possible
|| refferred to duty cycle amplitude. It should work.

This is a super-linear triangle generator. The 22n polypropylene capacitor
is charged /discharged by 1.1mA current source/sink which is switched by a
diode ring. The comparator is a dual part with totem pole O/P like LMC6762
Input should come from +5V/0V buffered 100Hz PWM. With the 2k7 resistor you
can adjust the modulation frequency.

+---------------+-----------+--------------+--------o +5V
| | | |
.-. .-. .-. .-.
| |330 330| | 100k| | | |
| | | | | | | |100k
'-' '-' '-' '-'
| | | |

| 2N2904 |< | |

|-+---------| | Input 100Hz |
/| | |\ | ___ | |\ Output 25kHz
| | | | o-|___|-+--|+\
+---+ | | 43k | | >--o
| +--|<-+->|-+------)--------------)--|-/
.-. | 1N4148 | | | |/
| | | | | |
| |2k7 +----------)------)----------+ |
'-' | | | | |
| | 1N4148 | | |\ | |
| +->|-+--|<-+------)----|-\ | |
+---+ | | | | >--+ |
| | | | +----|+/ | |
\| | |/ | | |/ | |
|-+--------| | | ___ | |
<| 2N2902 |> | +---|___|--+ +----+
| | | | 43k | |
.-. .-. | .-. .-. | +
| | | | ---100k| | 100k| | ###
| |330 330| | --- | | | | ---
'-' '-' |22n '-' '-' |10u
| | | | | |
+--------------+-----+------+--------------+----+----o GND
created by Andy´s ASCII-Circuit v1.24.140803 Beta www.tech-chat.de

 

[Ban]

Ban wrote:
|| Matteo Ferrari wrote:
||||
|||| There is no feedback at all. I need only to control a DC motor
|||| rotating speed.
|||| Yesterday I built a low pass filter using a 100 uF capacitor and a
|||| 10k resistance. It feeded a comparator with a 25 kHz to obtain the
|||| new PWM signal. Now I've got to make it as linear as possible
|||| refferred to duty cycle amplitude. It should work.
||
|| This is a super-linear triangle generator. The 22n polypropylene
|| capacitor is charged /discharged by 1.1mA current source/sink which
|| is switched by a diode ring. The comparator is a dual part with
|| totem pole O/P like LMC6762 Input should come from +5V/0V buffered
|| 100Hz PWM. With the 2k7 resistor you can adjust the modulation
|| frequency.
||
|| +---------------+-----------+--------------+--------o +5V
|| | | | |
|| .-. .-. .-. .-.
|| | |330 330| | 100k| | | |
|| | | | | | | | |100k
|| '-' '-' '-' '-'
|| | | | |
|| >| 2N2904 |< | |
|| |-+---------| | Input 100Hz |
|| /| | |\ | ___ | |\ Output 25kHz
|| | | | | o-|___|-+--|+\
|| +---+ | | 43k | | >--o
|| | +--|<-+->|-+------)--------------)--|-/
|| .-. | 1N4148 | | | |/
|| | | | | | |
|| | |2k7 +----------)------)----------+ |
|| '-' | | | | |
|| | | 1N4148 | | |\ | |
|| | +->|-+--|<-+------)----|-\ | |
|| +---+ | | | | >--+ |
|| | | | | +----|+/ | |
|| \| | |/ | | |/ | |
|| |-+--------| | | ___ | |
|| <| 2N2902 |> | +---|___|--+ +----+
|| | | | | 43k | |
|| .-. .-. | .-. .-. | +
|| | | | | ---100k| | 100k| | ###
|| | |330 330| | --- | | | | ---
|| '-' '-' |22n '-' '-' |10u
|| | | | | | |
|| +--------------+-----+------+--------------+----+----o GND
|| created by Andy´s ASCII-Circuit v1.24.140803 Beta www.tech-chat.de
||

+5V
o
|
.-.
| |7k5
| |
'-'
|
.-. |\
| |<------+------|+\ Output 25kHz
| |10k | | >---o
'-' | +-|-/
| | | |/
.-. | + |
| |7k5 ### |
| | --- |
'-' 10u| |
| | |
=== === |
GND GND |
o
triangle 1.5.....3.5V
If you want a linear control, you can put a pot here and don't need the
other 100Hz circuit any more.