I am planing to make a little more "stable" sircuit for square wave generations insted of NE555 using SG3525.
Using NE555 I have problems that is chaniging of frequency is related to duty cycle change: What I want is that duty cycle ALWAYS stays at the SAME PERCENTAGE no matter if I vary frequency..
What I want is resistor trimable square wave generator with freq and duty cycle change with resistors and switches for capacitor select (frequency range)
In the SG3525 datasheed it stays that MAX duty cycle is 50 %: So is that meant both A and B outpouts together 50 % or all together 50 %: ???
I mean if placing diode at the B output and conecting A and B across diode to the same otput cause I want duty cycle vary from 0 to 100 % (by trimmer change)
So than is that possbile to make vith SG3525 and is it good for that... Searching the net I didnt find any related schematics for such need
SG3525 is pretty cheap and can handle 300 khz frequency...
Using NE555 I have problems that is chaniging of frequency is related to duty cycle change: What I want is that duty cycle ALWAYS stays at the SAME PERCENTAGE no matter if I vary frequency..
What I want is resistor trimable square wave generator with freq and duty cycle change with resistors and switches for capacitor select (frequency range)
In the SG3525 datasheed it stays that MAX duty cycle is 50 %: So is that meant both A and B outpouts together 50 % or all together 50 %: ???
I mean if placing diode at the B output and conecting A and B across diode to the same otput cause I want duty cycle vary from 0 to 100 % (by trimmer change)
So than is that possbile to make vith SG3525 and is it good for that... Searching the net I didnt find any related schematics for such need
SG3525 is pretty cheap and can handle 300 khz frequency...
Medogrizli,
I have not used this chip since 1990, so my memory has faded a little. But, you could diode "OR" the outputs to get a full range duty cycle. The 50% max on each output is derived from F/F steering of the PWM signal to each output, sequentially. That is, each output is forced low every other osc cycle. So there is at least 50% off time for each output. The weakness would be that the positive edge will be xistor driven and the negative edge will be controlled by a pulldown and the load impedance unless you build an output stage buffer. You could also use a logic 2-in OR gate and parallel unused gates to increase output drive. Cmos would give better rail-to-rail swing at moderate output levels.
Controlling the duty cycle by controlling the error voltage input is possible, but you may find that it is not linearly proportional, nor entirely freq independent between low freq and high freq operation. Certainly at 100KHz and above I would expect delay times in the PWM logic to cause shifting of the duty cycle due to delays being a significant and growing portion of the osc period.
Another problem is that the outputs are not rail-to-rail and will shift quite a bit between no load and 100mA plus loading (the 555 has similar problems). Also, the chip power dissipation will be a factor for higher drive currents (see datasheet).
Depending on how high in freq you must go and how low an impedance you need to drive, a cmos 555 (say TI TLC555) can be made to be a good adjustable duty cycle generator up to 10KHz and not so good up to 100KHz if you run it at 15V. There is a standard hookup for a single pot duty cycle control for a 555 that goes back 25 to 30 years. It does have diodes which contribute to temperature sensitivity and to duty cycle shift at high freq operation.
The cmos 555 does exhibit good rail-to-rail output for loads up to 10mA and progressively poorer performance out to 100mA (worry about power dissipation at higher current) or so. The output stage does have asymmetry in the p and n channel mosfets so at high freq the rise time will be obviously slower than the fall time.
I hope this is helpful,
VSR
I have not used this chip since 1990, so my memory has faded a little. But, you could diode "OR" the outputs to get a full range duty cycle. The 50% max on each output is derived from F/F steering of the PWM signal to each output, sequentially. That is, each output is forced low every other osc cycle. So there is at least 50% off time for each output. The weakness would be that the positive edge will be xistor driven and the negative edge will be controlled by a pulldown and the load impedance unless you build an output stage buffer. You could also use a logic 2-in OR gate and parallel unused gates to increase output drive. Cmos would give better rail-to-rail swing at moderate output levels.
Controlling the duty cycle by controlling the error voltage input is possible, but you may find that it is not linearly proportional, nor entirely freq independent between low freq and high freq operation. Certainly at 100KHz and above I would expect delay times in the PWM logic to cause shifting of the duty cycle due to delays being a significant and growing portion of the osc period.
Another problem is that the outputs are not rail-to-rail and will shift quite a bit between no load and 100mA plus loading (the 555 has similar problems). Also, the chip power dissipation will be a factor for higher drive currents (see datasheet).
Depending on how high in freq you must go and how low an impedance you need to drive, a cmos 555 (say TI TLC555) can be made to be a good adjustable duty cycle generator up to 10KHz and not so good up to 100KHz if you run it at 15V. There is a standard hookup for a single pot duty cycle control for a 555 that goes back 25 to 30 years. It does have diodes which contribute to temperature sensitivity and to duty cycle shift at high freq operation.
The cmos 555 does exhibit good rail-to-rail output for loads up to 10mA and progressively poorer performance out to 100mA (worry about power dissipation at higher current) or so. The output stage does have asymmetry in the p and n channel mosfets so at high freq the rise time will be obviously slower than the fall time.
I hope this is helpful,
VSR
darkfenriz said:
I could use mosfet driver
sherelec
But I can use for example 12 V divided by resistors in series: So compared ti internal voltage source SG would change duty cycle tompared to that voltage?
than are there any other posibilites instead 555 or SG
Maybe it is somehow possible to make circuit with 555 so that DUTY cacle in percentage could be constant
Medogrizli,
I believe that at some increasing freq, a fixed voltage applied to the error amp (SG3525) will see shifts in the duty cycle. How much shift and at how high a freq, I do not know. You could experiment. You have not provided any specifications for your freq range of interest or how stable the duty cycle must be for changes in freq. Nor any specs for rise/fall time, drift with temp, etc.
If you search the internet for 555 application circuits or even some original Signetics 555 application/specsheets you should find a circuit for a one pot variable duty cycle 555 square wave generator. I know of a few techniques for making such a circuit more stable, but the 555 has limited drive and limited freq. As I indicated previously: a TLC555 might give pretty good results up to 10KHz if running at 15V, but less desirable performance as you go towards 100KHz.
There are PWM chips that are cmos, single output, 0-100% duty cycle, and good to several 100Khz. These are not as cheap as the SG3525, or as easy to find! Look at TI SMPS PWM controllers on the internet.
There have been several function generator chips designed. I do not have the numbers off the top of my head. I believe Intersil, Exar may be manufactures. An internet search for DIY function generators should give you lots of ideas. There are several kits available using these chips. Most of these are 0-100KHz type performance. Most optimistic 0-1MHz.
Finally, your idea is not bad, but without knowing your specifications; I suspect the SG3525 approach may have several features that you will find to be bugs.
Best Wishes,
VSR
I believe that at some increasing freq, a fixed voltage applied to the error amp (SG3525) will see shifts in the duty cycle. How much shift and at how high a freq, I do not know. You could experiment. You have not provided any specifications for your freq range of interest or how stable the duty cycle must be for changes in freq. Nor any specs for rise/fall time, drift with temp, etc.
If you search the internet for 555 application circuits or even some original Signetics 555 application/specsheets you should find a circuit for a one pot variable duty cycle 555 square wave generator. I know of a few techniques for making such a circuit more stable, but the 555 has limited drive and limited freq. As I indicated previously: a TLC555 might give pretty good results up to 10KHz if running at 15V, but less desirable performance as you go towards 100KHz.
There are PWM chips that are cmos, single output, 0-100% duty cycle, and good to several 100Khz. These are not as cheap as the SG3525, or as easy to find! Look at TI SMPS PWM controllers on the internet.
There have been several function generator chips designed. I do not have the numbers off the top of my head. I believe Intersil, Exar may be manufactures. An internet search for DIY function generators should give you lots of ideas. There are several kits available using these chips. Most of these are 0-100KHz type performance. Most optimistic 0-1MHz.
Finally, your idea is not bad, but without knowing your specifications; I suspect the SG3525 approach may have several features that you will find to be bugs.
Best Wishes,
VSR
Hi,
If all you want is square wave generation and nothing else, there are perhaps easier ways to do it.
If you need a function generator that does sine, square and triangle wave generation, I justed posted an article that should show up any second about an inexpensive function generator based on the Atmel AVR microcontroller. It generates rock-solid functions up to almost 300khz. Excellent for testing and really easy to build and control. Here it is...
http://www.myplace.nu/avr/minidds/
eris
If all you want is square wave generation and nothing else, there are perhaps easier ways to do it.
If you need a function generator that does sine, square and triangle wave generation, I justed posted an article that should show up any second about an inexpensive function generator based on the Atmel AVR microcontroller. It generates rock-solid functions up to almost 300khz. Excellent for testing and really easy to build and control. Here it is...
http://www.myplace.nu/avr/minidds/
eris
i just builded a squarewave generator with a sg3525 ,it depends a little on what you want to use it for .
you can variate duty cycle very easy with a resistor devider bridge from vref .
but if you dont have any specific reason for choosing a pwm modulator then you can do it a lot simpler with a 555 .
the sg3525 is in itself not hard to handle with a total componentcount of 15 or 20 parts you can make a pwm generator with inverted outputs adjustable deadtime & dutycycle .
it just depends on what do you want to do with it .
you can variate duty cycle very easy with a resistor devider bridge from vref .
but if you dont have any specific reason for choosing a pwm modulator then you can do it a lot simpler with a 555 .
the sg3525 is in itself not hard to handle with a total componentcount of 15 or 20 parts you can make a pwm generator with inverted outputs adjustable deadtime & dutycycle .
it just depends on what do you want to do with it .
i can measure that for you tomorrow .
the requirements for my project were ofcourse difrent then yours .
(i was developing a ac chopper)
but for evaluation purposes i made a squarewave generator with variable dutycycle ,deadtime & frequency based on a sg3525 .
only difrence is that i needed 2 inverted outputs but as you said a diode solves all in that aspect .
tomorrow i will give my dutycycle potentiometer a swing and measure frequency & dutycycle .
in a way i think the sg 3525 is a good choice for a square wave generator .
i do have a schematic of a pwm generator with 555 but its in polish .
(if you like i can picture it and send you)
but with a 555 you will miss a few options that are available with the 3525 (deadtime for instance)
goodluck and i'll let you know tomorrow
the requirements for my project were ofcourse difrent then yours .
(i was developing a ac chopper)
but for evaluation purposes i made a squarewave generator with variable dutycycle ,deadtime & frequency based on a sg3525 .
only difrence is that i needed 2 inverted outputs but as you said a diode solves all in that aspect .
tomorrow i will give my dutycycle potentiometer a swing and measure frequency & dutycycle .
in a way i think the sg 3525 is a good choice for a square wave generator .
i do have a schematic of a pwm generator with 555 but its in polish .
(if you like i can picture it and send you)
but with a 555 you will miss a few options that are available with the 3525 (deadtime for instance)
goodluck and i'll let you know tomorrowi measured the frequency change while variating my dutycycle and in my case it was 0,1 Hz.
In my opinion thats neglectable .
what did change my frequency is toying with the deadtime .
my design has a potentiometer for setting the dead time and that makes quite a change in frequency.
i would say fix the deadtime or just get rid of it and i dont see any problem .
goodluck
In my opinion thats neglectable .
what did change my frequency is toying with the deadtime .
my design has a potentiometer for setting the dead time and that makes quite a change in frequency.
i would say fix the deadtime or just get rid of it and i dont see any problem .
goodluck
the sg3525 in practice gives max duty cycle on each output of 45% to prevent cross conduction.
your best bet is the 4060 cmos ic (st data sheet here- http://www.st.com/stonline/products/literature/ds/1967.pdf
this is ideal as it can be rc controlled or crystal controlled, and has the dividers built in to guarantee 50% duty cycle.
regards
your best bet is the 4060 cmos ic (st data sheet here- http://www.st.com/stonline/products/literature/ds/1967.pdf
this is ideal as it can be rc controlled or crystal controlled, and has the dividers built in to guarantee 50% duty cycle.
regards
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.