A frequency counter, along with a dip meter, is an essential item for DIY hams. Since you work based on the frequency displayed by the counter, it's natural that a highly accurate counter is desirable, but that doesn't mean it has to be absolutely accurate.

　In my opinion, counting errors up to the 10Hz range are not a problem at all. This discrepancy is only a problem when aligning the carrier point of an SSB transceiver, and as for the operating frequency, it is not a problem as we are not a professional broadcasting station (operating at a fixed frequency) except when operating at the band edge.

　As you know, the accuracy of a frequency counter is determined by the accuracy of the reference oscillator. If you absolutely need accuracy, you can solve this problem by spending money on the reference oscillator.

　This introduction has been quite long, but I'm afraid I might seem like a fossil if I were to introduce a homemade counter that combines an old-fashioned TTL decimal counter, so here I'd like to introduce a counter that I made relatively recently.

　The photo above shows a 4.5 frequency counter I made in May 1982. The main counter IC is an Oki Electric μPD851D. Those who have experience building their own frequency counters will know what digits are displayed on a 4.5 frequency counter.

　A 4.5/2 display is a counter where only the first digit shows either 1 or 0 (if it is 0, nothing is displayed due to the zero suppression function), and the remaining 4 digits count normally from 0 to 9. From the above, it is probably easiest to think of it as a counter with a 4-digit display.

　The maximum countable frequency is 1200 MHz. The input range can be switched between three levels: 0-100 MHz, 0-250 MHz, and 80-1200 MHz, as shown in the diagram below.

　This type was used during the transitional period between the days when TTL decimal counters were connected in multiple stages and assembled discretely, and the emergence of the current one-chip dedicated ICs, so it is anachronistic to assemble them using this method now.

　For this reason, in this introduction to homemade counters, we will only introduce the block diagram, and if you are going to make your own, we recommend an 8-digit counter that uses the following one-chip dedicated IC.

8-digit frequency counter

　The photo below shows an 8-digit frequency counter completed in November 1994. The main counter IC uses ICM7216B. ICM7216B is a dedicated counter IC that contains all the functions required to configure a counter.

　When I think back to the days when we connected multiple TTL decimal counters in stages and made a big fuss about reset pulses and strobe pulses, I feel nostalgic for the hardships we faced, and am amazed at the technological advances that have been made in this field.

　This 8-digit frequency counter consists of four components: the main counter circuit board, the preamplifier and prescaler circuit board, the reference oscillator circuit board, and the power supply circuit board. In the photo on the right below, the preamplifier and prescaler circuit board is located in the upper left, the reference oscillator circuit board in the upper right, the main counter circuit board in the lower left, and the power supply circuit board in the lower right.

How it feels when used

　It displays 8 digits simultaneously, so unlike the 4.5-digit counter mentioned above, there is no need to switch gate times to check all digits. Once you start using an 8-digit counter, you will no longer want to use a counter with decimal points.

　By the way, you may think that the reference oscillator circuit is rather poor, but I think it's a waste of money to spend money on this part. If you need accurate frequency measurements, I think there will be no problem for amateur use (with the understanding that there may be an error of around 10 Hz) if you calibrate the reference oscillator frequency with JJY (10 MHz standard radio wave) in advance.

10MHZ reference oscillator replacement

（Some content added February 17, 2023）

　Recently, I have been working on improving my own decades-old equipment. Last time, I made an SWR meter, but this time I replaced the frequency counter and the reference oscillator of a YAESU FT-817.

　This is the first update to the frequency counter in 26 years, since December 15, 1997. Recently, when I contacted a local station, they pointed out that their frequency seemed to be drifting. The cause of the frequency fluctuation is thought to be fluctuations in either the carrier frequency (15 MHz) or the local oscillator (6 MHz). Checking the fluctuations in each frequency with the frequency counter should help identify the cause.

　However, the homemade frequency counter introduced here has a simple circuit configuration using general parts for the 10MHz reference oscillator that determines the measurement frequency and stability. For this reason, the accuracy of the measurement results and the stability can be said to be low.

　I thought I would like to replace it if I could find one with a more accurate reference oscillator, so I previously purchased a 10MHz reference oscillator with a thermostatic oven at an online auction.However, the oscillator itself was quite large and could not fit into the case of the counter introduced here, so I had no choice but to install the reference oscillator unit externally.

　For these reasons, I wasn't very keen on installing the reference oscillator externally, which would make it difficult to use, so I left it alone, thinking I'd work on it at some point. Then I found something on Amazon that looked like it could be used as a 10MHz reference oscillator. The reference oscillator I found looked like the one below.

　　 10MHZ oven-equipped reference oscillator available for sale ※The lowest price at the moment is 2,011 yen

　 　

*Click on each photo to enlarge it.

10MHz reference oscillator unit with thermostatic chamber function	I bought it for 2,477 yen. The lowest price available at the moment is 2,113 yen.	Warm-up current: 0.75A for 5 minutes normal operation current: 0.3A

　I immediately started up the reference oscillator I purchased and measured its frequency with my homemade 8-digit counter. The measurement result was 10.000.030Hz, which is about 30Hz higher. I'm not sure if this 30Hz deviation is due to a frequency deviation on the reference oscillator side or an error on the counter side.

　However, since I couldn't think of any discrepancies in the reference oscillator I purchased this time, I readjusted the 10MHz reference oscillator of my homemade frequency counter so that it displayed 10.000.000HZ. I left it in this state for a few hours and checked the frequency fluctuations.

　The measurement results were more stable than expected, with only a 1-2Hz fluctuation observed after leaving it for about 5 hours. I don't know if this fluctuation is due to a deviation in the reference oscillator or the counter, but since the reference oscillator on the homemade counter is poor, I think it's probably due to drift in the reference oscillator on the counter side. Looking at these measurement results, considering the poor reference oscillator of the homemade counter, the fluctuation in the displayed frequency was fairly small, so I felt it wasn't completely useless.

　If you think about it, the counter's reference oscillation of 10 MHz is 1/10,000,000 of the original frequency, which is 1 Hz (1 second), and is used as the counter's gate time. Because of this, even if there is a slight drift in 10 MHz, the effect is 1/10,000,000. Therefore, it is thought that the effect of drift is significantly reduced.

　With the test complete, I replaced the counter's reference oscillator with the newly purchased temperature-compensated reference oscillator. Specifically, I removed the crystal from the counter's 10MHz reference oscillator circuit and connected the output of the newly purchased 10MHz reference oscillator with thermostatic oven function to the gate of the 2SK241 to which the crystal was connected via a 0.1μF capacitor. Once the circuit switch was complete, I immediately measured the 6MHz PLL output of the local oscillator of Unit 5.

　I set the center click volume, which adjusts below 1KHZ, to the center position, fine-tuned the correction volume, and set it to 6,000,000HZ before starting measurements. I left it in this state and checked the displayed frequency after 5 hours.

　The measurement results showed a very slow frequency fluctuation within the range of +2 to -3 Hz, but five hours after the measurement was stopped, it had dropped to -3 Hz, as shown in the photo below. The maximum value within the observation period was within 3 Hz of the fluctuation. The PLL circuit of No. 5 uses a varicap to adjust below 1 kHz, so I was worried that the frequency drift might be large, but the measurement results were quite good and the numbers were not particularly problematic.

　So, after confirming that it was working without any problems, I decided to put the newly purchased reference oscillator inside the case. Looking around the inside of the case, I couldn't find any empty space to install it as it was, so I installed it on top of the existing reference oscillator, like a bunk bed, as shown in the photo on the far right below.

　 　 　

*Click on each photo to enlarge it.

Remove the 10MHZ crystal andconnect it to the reference oscillator I purchased this time with a 0.1μF resistor.	While measuring the 6MHZ PLL output of unit No. 5, what is the drift after 5 hours?	The drift after 5 hours is -3Hz.The frequency fluctuation is less than expected.	Install the purchased reference oscillator board on top of the existing reference oscillator circuit board

　After replacing it, I felt like the measurements were very stable. My homemade counter, which had been a little unreliable in terms of measurements, is now something I can use with confidence. (Maybe that's a bit of an exaggeration?)

　If you have an older counter and the reference oscillation accuracy is not very good, or if you have built one using Akizuki's DIY counter kit and the reference oscillation is not very stable, you can upgrade the reference oscillation by investing just over 2,000 yen, so please give it a try.

"At the same time" I also replaced the reference oscillator of the FT-817

　I also replaced the 22.625MHz reference oscillator in my YAESU FT-817 with a highly stable TCXO. This was because I wanted to monitor and check the frequency fluctuations using the FT-817. However, if the frequency of the receiver being monitored is not stable, it will be impossible to compare.

　I had been thinking about replacing it for a while, but was hesitant because it cost about 9,000 yen at the time. However, when I checked Amazon last December, I found a compatible part for about 1,600 yen. The price difference with the genuine part was so large that I wondered if it would even be useful, but since it was only about 1,600 yen, I decided to buy it anyway.

　Replacing the parts was very easy and did not require any soldering or other work; the work was completed by simply pulling out the existing reference oscillator unit by hand and inserting the new unit by hand. When replacing the unit, the original unit had 3P and 4P through holes in the printed circuit board, with the part mounting surface facing up. The unit purchased this time did not have these holes, so it could only be mounted with the printed circuit board surface facing up, as shown in the photo below.

　I was worried about whether this was the right direction, but since I couldn't install it with the part mounting surface facing up, I installed it in the direction shown in the photo. Worried about whether it would work, I turned it on and tried it out, and it worked without any problems. After the replacement was complete, I immediately turned on the FT-817 and let it age for about an hour, then received a 10MHz JJY signal to check for any frequency discrepancies.

　Here I will explain the specific method I use to check this. First, receive a 10MHz JJY signal. In this case, it is easier to judge if the receiving mode is set to SSB mode. The mode can be either USB or LSB. If checking with the LSB setting, switch to USB after checking and check the deviation again. Also, the JJY signal to be received does not have to be 10MHz; 5MHz or 15MHz is also fine, and select the standard radio wave (JJY) with the strongest signal possible.

　In this case, if the received JJY signal is weak, you won't be able to hear a beat sound with a frequency of a few Hz. If the signal strength is around 9+, you should be able to hear a beat sound with a frequency deviation of a few Hz. If you turn up the volume while receiving a strong JJY signal, a beat sound will be heard in proportion to the frequency deviation. If the deviation is a few Hz, a popping sound will come out of the speaker. If you are adjusting the frequency by varying it, adjust the frequency so that the intervals between these popping sounds become wider. When the frequency is correct, the intervals between the popping sounds will become wider, and eventually you will hear a buzzing sound (zero beat sound), indicating that the frequency has been adjusted.

　As an aside, I also use this method when tuning the carrier frequency of an SSB transmitter. I place the receiver's antenna close to the transmitter's SSB generator to receive the SSB signal. At this time, I tune the receiver's frequency to the carrier frequency. In the case of my No. 5 transmitter, it is 15,000 MHz.

　In the case of a transmitter, there is a microphone, so you will hear a high-pitched howling sound. When it is perfectly aligned, you will hear a continuous beep. Of course, with this method, the accuracy of the receiver becomes an issue. If you use a general receiver, it goes without saying that the adjustment will be based on the frequency accuracy of that receiver.

　Using the above measurement method, the result was a deviation of about 0.5Hz. This degree of frequency deviation is not a problem for amateurs. It was a cheap 1,600 yen reference oscillator, but it was pretty accurate. I'm curious to see how much difference there is in accuracy between a genuine product that costs around 9,000 yen and a knock-off (sorry! compatible) that costs around 1,600 yen, but it may actually be almost the same.

　I found a blog that measured the frequency fluctuations of a standard TCXO and a compatible TCXO. The standard TCXO is not a TCXO, so it seems to have quite a bit of frequency fluctuation. On the other hand, the compatible TCXO also showed almost no frequency fluctuation, and there were no stability issues considering the price, so it seems there are benefits to replacing it.

Comparison of frequency fluctuations between standard mounted products and the compatible TCXO that was replaced.　*Measurement results of frequency fluctuations between standard mounted products and the compatible TCXO that was replaced are reported.

　　 Comparison of frequency fluctuations between standard mounted and replaced compatible TCXO　*Measurement results of frequency fluctuations of a compatible TCXO replaced with a standard mounted product have been reported.

　　 Reference oscillator sold on Amazon *Compatible product The lowest price at the moment is 1,474 yen

　 　 　

*Click on each photo to enlarge it.

The reference oscillator is in the lower left corner. The existing part has been pulled out and removed.	Enlarged view of the reference oscillator mounting area. 4P on the left and 3P on the right are connected via pins.	The 4P socket is visible on the right side of the new reference oscillator to be replaced.	3pと4pのピンに向きを合わせて 基準発振器を差込んで交換完了

　If you own a YAESU transceiver FT-817/857/897 and haven't upgraded your reference oscillator yet, why not consider upgrading it now, as it's only an investment of about 1,600 yen? The price includes tax, and shipping and convenience store transfer fees are free.

　So, to address the frequency drift issue with my homemade transceiver, I measured and readjusted the frequency using an FT-817 with a replaced reference oscillator and a frequency counter with a replaced reference oscillator. I also checked the frequency drift for a little longer, but as far as I could see, there was no significant frequency drift.

　I plan to have the local station that pointed out the frequency drift listen to it and get a report on whether there is any frequency discrepancy or drift. I'm checking it with measuring equipment that I believe has improved in accuracy (the receiver is also a good measuring instrument), so I don't think there's a problem.

Return to Home Page