Need help with Frequency counter design

[matrix3297]

I am a student and I need a frequency counter design by thursday of this week. I have the displays wired with MAN74 7-segment displays being run by two 4511s and the decade counters are 4029s. I also have a 555 timer pulsing to show a count on the displays. I need to be able to read 5 different frequency ranges and don't really know how I'm going to do it. If anyone could help I would appreciate it.

 

[dipal_z]

As you know frequency is measured as number of cycles per second. Generally to measure the frequency the signal if passed to though a switch for pre-determined time and number of clock cycles are counted in that time slot. For example if you enable a switch for 1 second your counter will provide the exact frequency of the input waveform.

There are different ways to achieve this for example you can use that AND gate as switch, connect one input of the AND gate to input single and other input should be made high for 1 second. When one input of the AND gate is high its output depends on the state of the other input. Thus if you enable it for 1 second input signal will be available on its output for 1 second and you can feed it to the counters.

Some things you need to consider here:

1. When you want to measure high frequency you have to reduce the time for which gate is enabled otherwise your counters will overflow. For example enable it for 1 ms and your display will show frequency as KHz.
2. When you want to measure low frequency you have to increase the time for which gate is enabled. For example if you want to measure 0.5 Hz you will not see any clock transition in 1 second so at least you have to enable it for 2 seconds.
3. If your input signal is not square wave or not of sufficient amplitude you need to do the signal condition and amplification. This can be done with the help of Schmitt Triggers and Op-Amps.

I hope it will be helpful.

 

[matrix3297]

Need help with frequency counter

Thank you for your suggestion. I will try a schmitt trigger circuit, but do I need a different clock pulse for each frequency range? The ones I need are 9.9Hz-99Hz, 100Hz-9.9kHz, 1kHz-99kHz, 100kHz-990kHz, and 1MHz-9.9MHz. Or can I divide the frequency as it comes into the circuit with the AND gates? I have to read much higher frequencies than I am familiar with so this is kind of new to me. Also, is a 4029 an AND gate that I can use for the frequency division?

 

[Nigel Goodwin]

Re: Need help with frequency counter

Thank you for your suggestion. I will try a schmitt trigger circuit, but do I need a different clock pulse for each frequency range? The ones I need are 9.9Hz-99Hz, 100Hz-9.9kHz, 1kHz-99kHz, 100kHz-990kHz, and 1MHz-9.9MHz. Or can I divide the frequency as it comes into the circuit with the AND gates? I have to read much higher frequencies than I am familiar with so this is kind of new to me. Also, is a 4029 an AND gate that I can use for the frequency division?

You can either alter the gate time, or divide the input down, depending on the frequencies involved.

An AND gate does not do frequency division!, you need divide by 10 counters.

The resolution you get depends on your gate time and input frequency, with a 1 second gate time it will count in Hz, with a 10 second gate time it will count in 0.1Hz, and so. For high frequencies it's normal to use shorter gate times, so a 0.1S gate time will give only 10Hz resolution - but at RF frequencies that's usually more than you need (and probably better than the accuracy of your gate timing?).

 

[matrix3297]

Need help with frequency counter

I used a schmitt trigger circuit to condition the input signal and it worked to conver the sine wave intput to a square wave. Thanks dipal_z! Now I have to divide the input frequencies with a 4029 decade counter (not AND gate, sorry said the wrong name ) How can I use this chip to divide frequencies? Do I need to have the frequency dividers' outputs connected to the schmitt trigger input? If so, how do I do that? The accuracy is limited to a 2-digit display so it decreases as the input frequency increases.

 

[dipal_z]

Re: Need help with frequency counter

I used a schmitt trigger circuit to condition the input signal and it worked to conver the sine wave intput to a square wave. Thanks dipal_z!

I am glad that it was helpful.

Now I have to divide the input frequencies with a 4029 decade counter

Truly speaking 4029 is much more than what you need for frequency division. This is Presettable Binary/Decade Up/Down Counter while just a simple decade counter is enough for your purpose. Any way if you want to use 4029 you can wire it as given below.
1. Set preset enable pin “low” this will disable parallel inputs used to start from pre-determined count.
2. Set Binary/Decade pin to “low” this will configure it as binary counter.
3. It doesn’t matter whether it counts up or down as long as you want to use it as divider so its upto you to decide what to do with Up/Down Pin.
4. Once you configure it like this you will get following on outputs
a. Q1 = CLK/2
b. Q2 = CLK/4
c. Q3 = CLK/8
d. Q4 = CLK/10
e. Carry out = CLK/10
5. For dividing the frequency more than 10 times you need to cascade these counter please refer to Datasheets for possible configuration.

But as suggest by Nigel changing the gate time will be easier.

Do I need to have the frequency dividers' outputs connected to the schmitt trigger input?

It should be other way output of Schmitt trigger should go to CLK input of 4029.

The accuracy is limited to a 2-digit display so it decreases as the input frequency increases.

Two digits are not really sufficient for accurate measurement but you can change the gate time accordingly to get maximum possible accuracy on two digits, I hope you already got the logic behind it.

 

[aibelectronics]

reply

check this link up:**broken link removed** and:
**broken link removed** and:
http://hem.passagen.se/communication/freqcount.html

 

[panic mode]

the first link only shows counter part....
the whole article is here
**broken link removed**

 

[matrix3297]

Thanks everyone for all your help.

I ended up using a combination of change in time base using monostables that were driven by an astable multivibrator (555s) and frequency division using decade counters (4518s). I was able to count up to 2.5 Mhz with a 0.2 Vp-p input and up to 5 Mhz with a 2 Vp-p input. The whole thing is done now and looks terrible, but works pretty well. The main problem I had was the sequence of events using the pulser circuits (4011s). I was trying to reset the counter at the same time I was trying to latch the number with the 4029s and 4511s. I ended up just using the pulse from the latch to drive the pulse for the zero and it worked. Anyway thanks all.

Matrix3297

 

[mstechca]

There are different ways to achieve this for example you can use that AND gate as switch, connect one input of the AND gate to input single and other input should be made high for 1 second. When one input of the AND gate is high its output depends on the state of the other input. Thus if you enable it for 1 second input signal will be available on its output for 1 second and you can feed it to the counters.

an AND gate only works 1/2 the time. an XNOR gate is better.
An audio waveform is an AC signal, and can go positive and negative at any time. an AND gate only works if the waveform is positive and the internal clock is positive. Remember that both are also in sync if both are negative. an AND gate cannot do this. an XNOR gate can.