# Frequency Counter 2010-01-05

I built this frequency counter many years ago, maybe about 1976. I designed the entire thing myself as a student and built it somewhat shoddily but it is still working today. I only had some shabby hand-drawn schematics and I decided to take the trouble to draw them anew. Doing the re-drawing brought back some pleasant memories.

Many values of components are not shown because they were not in the original drawings. Also maybe some are not reliable because it was all an experiment and may have changed later.

At that time crystals for oscillators were *very* expensive and I happened to have a crystal for 5230 KHz so I built the whole time base circuit around it which makes the circuit much more complex than it would be if I had used a 1 MHz crystal.

In the upper right you can see the oscillator which outputs 5230 KHz. That signal is fed to a chain of 7493 binary dividers. A big nand gate detects when 5230 (binary: 1010001101110) and sets a the reset bistable latch formed by two nand gates. This resets the counters. When the nor gates below detect that all outputs are back to zero the reset latch is reset and the count starts again. This means all this circuitry is dividing by 5230 and we have a 1 KHz signal out. With a 1 MHz oscillator all we would have to do is divide by 1000 which is much easier.

The 1 KHz signal is further divided three times by 10 so that we have available 1000, 100, 10 and 1 Hz signals available at the range switch. Let us suppose the switch is using the 1 Hz signal. The 7473 divides it by two so that we have a one second half cycle.

The gates and circuit at the bottom right work as follows. For one second (half cycle) the pulses to be counted are allowed through nor gate 456. When the output of the 7473 changes the pulses are blocked and the counting needs to be transferred to the latches so a transfer signal is generated. Then we need a slightly delayed signal to reset the counters and start the count all over again.

So, basically, the output of this section is three signals: pulses (frequency) to be counted, transfer of count to latches and reset of counters for next cycle.

The range selection switch can switch the gate frequency so that it is divided by ten in each step and this means the frequency readout is multiplied by ten. The same switch selects the dot in the corresponding display so that the dot always shows KHz.

Let us now look at the counter - latch - display section in the lower left. It has five digits and each digit has a 7490 counter, 7475 latch, 7447 BCD to 7 segment driver and FND507 common anode 7 segment display so there is a total of 20 ICs (including the displays). All this was wired by hand on a very basic PCB.

Above that we have two input amplifiers. I could not get one amplifier to work well on all the range I needed so I made two amplifiers, one for LF and one for HF and I would just switch the input to whichever one worked best. They have considerable overlap. There was no need to switch the outputs because a nor gate takes care of that. I have drawn the same nor gate twice because it helps with the understanding and because I like it and it is *my* schematic and I can do whatever I want. So there.

Finally, in the upper left we have the +5 V power supply which uses a 7805 regulator.

You can see that it is all divided into separate, logical sections. I mounted them separately on separate boards of about 7.5 cm square. The bottom board has the counters, latches, BCD to 7 segment decoders and then there is a vertical board attached to it with the displays so it is all one unit.

A similar, vertical, board on the back has the 5230 KHz oscillator. And there are two more horizontal boards above the counters so there are in total two vertical boards and three horizontal boards and they are all soldered together so that it would be unthinkable to attempt any repair now. The day it fails is the day it goes in the trash.

I took some time and got some pleasure from re-drawing the schematic. It is done in such way that it can be divided vertically in two and each half printed on a sheet of paper. As you can see there are only four signals which cross that boundary so each block is quite independent of the rest.

The effort I put into designing, developing, testing and finishing this project was enormous but at that time buying a commercial frequency counter was totally out of reach for a poor student like me. The best thing was the learning and experience I got out of the project.

Today you could buy a single IC which could do what I needed 20 to do. The entire counter, latch, BCD to 7 segment driver and display could be bought cheaply as a single unit. But designing and building this taught me a lot about digital electronics.

Although I now have a HP-Agilent 5315 counter, I still have in my workshop this unit I built and, in fact, I use it quite often for quick frequency checks as it is simpler to use.

I have spent some time redrawing this because I want to get rid of paper and have as much information as I can in my computer. It took some time to draw but I enjoyed it.

Today it would not make sense to build this (heck, I don't think it made sense *then*) but if you are studying digital electronics studying and understanding how this works will teach you a lot.
Piort
Author
HS3
Views
319
First release
Last update
Rating
0 ratings