Electronics > Beginners
Portable Low Frequency square wave generator circuit - Solved - page 3
Zero999:
--- Quote from: spec on January 31, 2019, 09:59:29 am ---The frequency control is by a single potentiometer to sweep from below 1Hz to above 1kHz.
Note that the 555 timer is a CMOS type- an ordinary 555 will not do.
http://www.ti.com/lit/ds/symlink/lmc555.pdf
http://www.ti.com/lit/ds/symlink/ucc21520.pdf
https://www.onsemi.com/pub/Collateral/FDD3672-D.pdf
https://www.onsemi.com/pub/Collateral/FDD3682-D.pdf
--- End quote ---
The original poster wanted 12V peak to peak, not 24V, but that's an easy mistake to make, since it wasn't clear and it's trivial to reduce the supply voltage to +/-6V.
R1 is a little on the low side. At 1k, the asymmetric resistances of the output transistors will contribute to the timing, taking the duty cycle away from 50%. It would be better to opt for a higher resistance, such as 10k and have a switch to select between two capacitors for two ranges, say 1Hz to 30Hz and 30Hz to 1kHz. This would also make sense from an ease of use perspective, as 1Hz to 1kHz with a single potentiometer would be quite difficult to adjust accurately, even with a ten turn pot.
--- Quote from: soldar on January 31, 2019, 10:26:09 am ---
--- Quote from: Zero999 on January 27, 2019, 11:50:03 pm --- The problem with the 555 timer is it's difficult to get a 50% duty cycle output.
--- End quote ---
There are several circuit designs with 555 timer that allow anywhere from 0% to 100% duty (with logical limits at both ends of the range) and there is no problem getting 50% duty cycle. OTOH it is easy to start with double the frequency and then divide by 2 and this results in 50% duty cycle exactly always and every time.
--- End quote ---
It depends on how stringent your 50% duty cycle requirements are. The CMOS variant isn't perfect, as the on resistance of the high and low side output transistors isn't symmetrical. Going from the graphs on the datasheet, when VDD = 5V and a junction temperature of 25oC, the P-MOSFET has an on resistance of 300R and N-MOSFET 60R. It's better if the impedance seen by the output stage is as high as possible.
https://www.nxp.com/docs/en/data-sheet/ICM7555.pdf
In this case, a CMOS 555 timer is probably more than good enough, so long as the output isn't loaded too much. As mentioned above, it has an output impedance of well over 50 Ohm, so will need buffering: a 7555 + 50 Ohm resistor will not do! A MOSFET driver IC can be used to buffer the output.
I'd probably go for a counter + divider IC such as the 4060. It can use a smaller, closer tolerance, higher stability capacitor and the ranges can be switched by changing between different divider outputs. The frequency could be varied between 500Hz to 16kHz, giving 31.25Hz to 1kHz on the Q3 output and 0.97Hz and 31.25Hz on the Q8 output. To take into account for component tolerances, set the oscillator minimum and maximum frequencies to be a little lower than 500Hz and higher than 16kHz: aim for 475Hz to 16.8kHz. Again it will need to be buffered, as the 4060 has far too higher output impedance.
Lots of information about calculating the resistor and capacitor values can be found using a search engine.
https://assets.nexperia.com/documents/data-sheet/HEF4060B.pdf
https://www.brighthubengineering.com/diy-electronics-devices/123626-making-flasher-timer-oscillator-circuits-using-a-single-chip/
https://dmohankumar.wordpress.com/2012/05/13/design-your-circuit-part-ii-cd-4060-timer/
soldar:
--- Quote from: Zero999 on January 31, 2019, 11:43:49 am --- It depends on how stringent your 50% duty cycle requirements are.
--- End quote ---
Yup. I am assuming "not too stringent". Again, a divider solves that issue if needed.
--- Quote from: Zero999 on January 31, 2019, 11:43:49 am --- In this case, a CMOS 555 timer is probably more than good enough, so long as the output isn't loaded too much. As mentioned above, it has an output impedance of well over 50 Ohm, so will need buffering: a 7555 + 50 Ohm resistor will not do! A MOSFET driver IC can be used to buffer the output.
--- End quote ---
The traditional bipolar version will source or sink up to 200 mA with a little voltage drop at each end which may be admissible or can be compensated with a bit higher supply voltage.
In other words, you can make something better and more complex but if you want something quick and dirty which will work I think a single bipolar 555 will do it. A split supply of 7.5 + 7.5 volts and very few more components.
I would make sure where it is being connected to because ... people.
I use bipolar 555 timers because I have a stock of them enough to last me several lifetimes. :)
Zero999:
The split power supply should be asymmetrical to match the saturation voltages of the bipolar 555.
How would you get the duty cycle to 50%, with an adjustable frequency from 1Hz to 1kHz, using the plain old bipolar 555?
spec:
--- Quote from: Zero999 on January 31, 2019, 11:43:49 am ---The original poster wanted 12V peak to peak, not 24V, but that's an easy mistake to make, since it wasn't clear and it's trivial to reduce the supply voltage to +/-6V.
--- End quote ---
No mistake. The output impedance is 50R forward terminated and the output voltage is +-6V8, as the OP informed me that he wants a little over 6V. The output voltage can be adjusted to any voltage from 2V to 12V by changing the two zener diodes at the output.
soldar:
--- Quote from: Zero999 on January 31, 2019, 01:28:45 pm --- The split power supply should be asymmetrical to match the saturation voltages of the bipolar 555.
--- End quote ---
Back to "not too stringent" because if we are talking "stringent" then, yes, there are better, more sophisticated, designs. I am guessing the input is for some counting circuit which is "not too stringent" and will accept a certain range. If we have strict requirements then we need a stabilized power supply because batteries will not do.
--- Quote from: Zero999 on January 31, 2019, 01:28:45 pm --- How would you get the duty cycle to 50%, with an adjustable frequency from 1Hz to 1kHz, using the plain old bipolar 555?
--- End quote ---
Well, there are several circuit designs which produce a 50% duty cycle and where the on time and down time depend on the same resistor, like 50% Duty Cycle Astable Oscillator halfway down the page. Vary R2 and the period changes but the duty cycle shouldn't.
Again, this is a quick and simple solution, not the best or most sophisticated. I think it would be enough for this type of application though.
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version