Astable NE555 Frequency Calculations

[charlie.n255]

I am working on my first design for a Soil Moisture probe and am having problems seeing my calculated and measured frequency values agreeing with each other.

I have followed the design stated on the datasheet and am using RA = 470, RB = 1k6 and cap at 470pF. My calculated values come out at about 840khZ but I am measuring only around 500kHz.

Could this be down to just tolerances in the capacitors and resistors or is this common for the NE555?

[xrunner]

Could this be down to just tolerances in the capacitors and resistors or is this common for the NE555?

Playing with a 555 calculator and putting into play some tolerances of your components, I think it's not out of the question for your observations. Measure your components and try the actual values in a 555 calculator.

[bob91343]

The frequency is very approximate.  It depends on the voltage thresholds, which can vary over a wide range.  Measure your components and then try a different chip.  Also, check the frequency vs. the supply voltage.

This is not an accurate circuit and the fact that you are only off by about 30% is par for the course.

[charlie.n255]

Thank you all for your help  , even though I do not think this circuit needs to be accurate in generating frequencies what is some alternative topologies that I could try for a more accurate output?

[mansaxel]

Thank you all for your help  , even though I do not think this circuit needs to be accurate in generating frequencies what is some alternative topologies that I could try for a more accurate output?

The time-nut reply is GPSDO and divider. (because one has the reference and the 10MHz distro already. ) 

[StillTrying]

I have followed the design stated on the datasheet and am using RA = 470, RB = 1k6 and cap at 470pF. My calculated values come out at about 840khZ but I am measuring only around 500kHz.

The calculations are accurate when everything is near the mid range, 840 kHz is definitely near the limits, sometimes NE555s are advertised as only good for up to 500 kHz.

An LT simulation gives 665 kHz.

RA is low at 470R, wastes a lot of current when DIS is on. IF the output is lightly loaded so always a good squarewave you can use it to drive the timing resistor and save quite a bit of supply current, 11 mA down to 3.5 mA in the simulation.

If the output current is enough, CMOS 555 versions are faster and more accurate at 800 kHz to 2 MHz speeds.

[xrunner]

Thank you all for your help  , even though I do not think this circuit needs to be accurate in generating frequencies what is some alternative topologies that I could try for a more accurate output?

The time-nut reply is GPSDO and divider. (because one has the reference and the 10MHz distro already. ) 

Well yea if you want to go over-board you can get one of these rigs - Si5351A Clock Generator Breakout Board - 8KHz to 160MHz

https://www.adafruit.com/product/2045

Get a nano and use it to control the frequency very simply. We can over-engineer anything around here. 

[Etesla]

https://ohmslawcalculator.com/555-astable-calculator

LTSpice also has been fairly accurate if you're using a less standard topology. Long term stability also relies heavily on the type of capacitors that you choose.

[bob91343]

Find a cheap crystal near the desired frequency and make an oscillator with one chip.

[Zero999]

I have followed the design stated on the datasheet and am using RA = 470, RB = 1k6 and cap at 470pF. My calculated values come out at about 840khZ but I am measuring only around 500kHz.

The calculations are accurate when everything is near the mid range, 840 kHz is definitely near the limits, sometimes NE555s are advertised as only good for up to 500 kHz.

An LT simulation gives 665 kHz.

RA is low at 470R, wastes a lot of current when DIS is on. IF the output is lightly loaded so always a good squarewave you can use it to drive the timing resistor and save quite a bit of supply current, 11 mA down to 3.5 mA in the simulation.

If the output current is enough, CMOS 555 versions are faster and more accurate at 800 kHz to 2 MHz speeds.

I agree about using the CMOS version.

Note that the standard LTSpice model doesn't simulate the TTL 555 output stage accurately, so the simulated frequency and duty cycle for the circuit on the right will be slightly wrong. It's better for simulating the CMOS versions, but only if the resistor values aren't too low.

[StillTrying]

Note that the standard LTSpice model doesn't simulate the TTL 555 output stage accurately, so the simulated frequency and duty cycle for the circuit on the right will be slightly wrong.

That was the 601kHz, ~520kHz in practice bit, I simulated 1.1V off the 8V output high with 2 diodes to get a rough idea of the lower frequency.