## Multiple Feedback Band-Pass filter

Band-pass filters are filters that are designed to let only a certain set of frequencies pass through. Overtime a myriad of different topologies have been developed by some very creative designers. Passive vs active, RC-filters vs LC filters, combinations of low pass and high pass filters, with gain and without gain, etc. One of the topologies that are often used for CW filters in receivers is the multiple feedback band-pass filter :

Continue reading Multiple Feedback Band-Pass filter

## 40m DC-receiver – What’s next?

Now that the basic receiver is working, what do I want to add to consider the receiver ‘complete’ ? This is the current set-up :

The following things I would like to add:

• S-meter : shows the strength of the received signal. Based on AF signal
• CW filter : Multiple Feedback Band-Pass filter, 1 or 2 stages, centered around 750Hz
• Audio Amp – current the bench utility amp is used with a gain of 20 dB.

So the receiver is alive and kicking and I have received amateurs from The Netherlands, UK and Italy on a simple 10m wire antenna. However, the VFO is comprised of a Siglent SDG1032X function generator from work. Hardly a self-contained solution!
After an article by Bill Merea in the GQRP Sprat magazine, I’ve decided to go for a digital VFO. The clock generator is based on a Si5351 module. The I2C interface is controlled via an Arduino Nano board together with a Groove I2C LCD.

## Attenuator

An attenuation network is often used in devices like a (digital) voltmeter, in oscilloscopes, etc. It offers the possibility to expand the input range of the measurement tool and thus makes it more versatile. Usually the attenuation network is build with a couple of resistor as shown in fig. 1. If the input resistance of the rest of devices is considered indefinite (as for example with a J-FET) the voltage is Vo=Vi*R1/(R1+R2). This works very well with DC-voltages.

However, with AC-voltages, things are not quit that simple. There are parasitic capacitances that ruins this simple setup. Figure 2 shows what is happening. For a DC-voltage this setup still works. However, with an AC-voltage it now becomes different. The total resistance is now the parallel impedance of the resistor and the capacitance and is thus lower. The meter will thus show a lower reading and the accuracy of the instrument so carefully constructed is down the drain….

However, this can easily solved by adding some more capacitors. Yes, you read it correct even more capacitors!

$R_1C_1 = \tau_1$ $R_2C_2 = \tau_2$ $Z_1 = R_1//C_1$ $Z_1 = R_1//Z_{C1} = { {R_1 {{1}\over {j \omega C_1} }} \over {R_1 + {{1}\over {j \omega C_1} }} } = {{R_1}\over {1+j \omega R_1C_1} }$ $Z_2 = R_2//C_2$ Likewise : $Z_2 = R_2//Z_{C2} = {{R_2}\over {1+j \omega R_2C_2} }$ ${{V_o}\over{V_i}} = {{Z_2}\over{Z_1+Z_2}} = {{ {{R_2}\over {1+j \omega R_2C_2} }}\over {{{R_1}\over {1+j \omega R_1C_1} }+{{R_2}\over {1+j \omega R_2C_2} } }}$ $\tau_1 = \tau_2 = \tau$ ${{V_o}\over{V_i}} = {{ {{R_2}\over {1+j \omega \tau} }}\over {{{R_1}\over {1+j \omega \tau} }+{{R_2}\over {1+j \omega \tau} } }} {{1+j \omega \tau}\over {1+j \omega \tau}}$ ${{V_o}\over{V_i}} = {{R_2} \over {R_1 + R_2}} qed$

## Milliohm meter

This is a simple but accurate milliohm meter with a range of 0 – 2000 milliohm (= 2 Ohm). It has a typical accuracy of 2%-3%. Unlike other designs this circuits doesn’t use large currents to measure these small resistances. So there is no risk to damage components. A small alternating current is used to excite the resistance under test. This AC voltage is then amplified with a common OpAmp AC gain block. Therefor the OpAmp DC offset voltage doesn’t come into play. It can be used to measure contact resistance of coax connectors, relays, switches, etc. It is powered from 2 9V / 6LR61 batteries.

The circuit is build up with 5 separate blocks that can be build and tested sequentially. A square wave oscillator with 50% duty cycle and two complementary outputs. The non-inverted output switches the 10 mA current-source on and off. This current is passed through the unknown resistance. The voltage drop is then amplified 200 times. The output of OpAmp is then rectified by the synchronous rectifier (more on this later). The DC voltage still contains some switching artifacts and is then smoothed out by a 3-pole low pass filter. The resulting smoothed DC voltage can then be offered to a 3 1/2 DMM or an analog meter.

## 40m DC receiver – “It’s alive!”

With a three day Pentecost weekend and my wife away on a camping trip with my son, I had some time to look into the receiver again. With the previous series of tests, I’ve found that product detector was working fine and featured good conversion gain. The LO oscillator injection was flawed because the 74LS132 was behaving incorrectly. I could use my work-horse the PM5134 20MHz function generator. Because the frequency setting is a bit course, I risked the chance of missing a signal in band and thus still not knowing the outcome. Therefor I borrowed the Siglent SDG1032X DDS function generator from my work.

I decided to focus on the hart of the receiver, the product detector, first. I hooked up my random length wire antenna to the gate of the J-FET, the function generator to the control inputs of the switches and a utility bench amplifier to the OpAmp output. (Max gain of 20dB) . I started at 7 Mhz and slowly increased the frequency up to 7.3Mhz. Although I definitely received some signals, it all sounded heavily distorted, except for CW. Connecting a scope to the output of the OpAmp showed some serious clipping. I decided to lower the OpAmp gain by increasing the resistor back again to 2k2. The gain is now Av = 46x = 33dB. This definitely improved the situation, weaker signals were now understandable but strong signals still featured clipping. Therefor I decided to add a RF gain pot to the input of the JFet. After playing around a bit, I’ve found that setting this 1/4 is sufficient in most cases.

With these changes I was able to receive signals all over the 40m band. Within 25 minutes I picked up signals from hams from the UK, Italy, Netherlands and Germany and possible a couple from the USA. By this time it was already 2345h, time to shuffle off.

Next, I’ve added the audio low pass filter to the output of the OpAmp. This definitely improved the audio quality! Much of the annoying high frequency noise was gone. CW signals are now much better to copy for only a smaller part of the band is received. There is still some 50Hz/100Hz hum picked up. Which is almost unavoidable with a receiver lying open on the bench. I realized that there are only low pass filter functions in the gain chain. The corner frequency for the OpAmp is set by Rx and Cx and is set at 0.7Hz… I’ve decided to decrease Cx to 0.22uF, the -3dB point is now set at 330Hz, much better!

Recalculating all RC filter points again, I realized that the input filter to the difference amplifier, also needed to be changed. The current filter point is set by R6+R7 and (C4 + C5//C6) and is only 2.6kHz. Which is fine for CW listening but cuts of a tad too much for voice signals. Therefor C4 was lowered to 4n7, giving a -3dB frequency of 5.1 kHz.

Next step is to build a local oscillator: one VFO to rule them all!

## Double-balanced cross-coupled product detector

The double-balanced cross-coupled product detector had a brief stint of popularity in the 70s and 80s. It’s popularity quickly faded once integrated product detectors like the Plessey SL640, Motorola MC1496/1596 and the CA3028A came on the market. Offering ease of use and further integration. My attention was first drawn to the cross-coupled product, when casually browsing some Technical Topics columns from Pat Hawker G3VA from the 80s. In his June/July 1980 article briefly mentions the product detector. Out of curiosity, I decided to build it and do some tests with it.
The nice thing about this whole circuit it can be build without any transformers! The only thing needed is a 1mH choke that be bought of the shelf for a couple of cents.

Continue reading Double-balanced cross-coupled product detector

## JFET testing

For an upcoming RF amplifier project I needed a couple of JFETs. The spread in device characteristics is quit broad. For the BF256B Vp ranges from -0.5 to -8.0V ( a factor 16! ) and IDSS ranges from 6 mA to 13 mA. Since, I bought a small stock of them, I decided to satisfy my curiosity and measure to static device characteristics.