A 'Starter Project' for 'Newbies/Interested Constructors' - A DC RX

A "Starter Project for after the RAE"

I was thinking about a 'starter project' for the 'newbies' after the RAE. It used to be something like a 'Crystal Set'. Because the AM transmitters are few and far between nowadays, it should be a 'simple DCRX.' And to make it more 'interesting' it should receive SSB, CW and PSK31 and RTTY.

Direct Conversion Receivers have been around for several decades.

Lets try putting an AI on the task...

Based on a search of technical literature and recent hobbyist projects, there are many articles and resources dedicated to direct conversion receivers (DCRs), ranging from historical overviews to modern, high-performance designs.

Academic and Technical Articles:

Numerous technical papers and articles exist on ScienceDirect, MDPI, and ResearchGate covering DCR design, specifically addressing challenges like DC offset, LO leakage, and I/Q imbalance.

Main Themes in Recent Articles:

Simplification:

Many designs focus on minimizing component counts, often using under four transistors.

Performance Improvement:

Articles discuss techniques to overcome traditional DCR limitations (hum, Microphonics) using I/Q demodulators and modern DSP.

Direct Conversion vs. Superhet:

Several articles compare the cost and complexity benefits of DCR, especially for integrated, zero-IF designs.

Overall, the number of articles is substantial (over 20+ prominent technical, academic, and hobbyist resources found in the search results).

So what do 'we' want to build?

Some design features:-

1) The ZS article used a 420mm loop antenna

2) Use an 'active antenna'

3) An IC "Balanced Mixer" to suppress the local oscillator leakage to the antenna.

*** Whoops! That might be a problem...

4) Which 'Band' do we want to use?

5) How sensitive should it be?

Example Picture showing simple construction method 1.

You can't get the wood, you know!

There 'used to be' many integrated circuits for the Direct Conversion Receiver.

SO42 - SL6440 - TIxxx - CA3086 NE602/NE612 and so on. Most are no longer in production or available for purchase. [Also fake ones are available!]

See the note 'Dilbert Cell.' [Not a typo]

A recent Youtube described how a 'Gilbert Cell' worked. Take a look. The simple version would do the job nicely - using 'readily available' - 'general pupose' transistors!

The Gilbert Cell (084d)

https://youtu.be/mQ36yy7mloA?si=2W3HRTWym6j-LcsR

Now we are scavengers

This is where I get on a 'soap box' and tell you about how many transistors I have recovered/scrounged/scavenged from old/wornout electronic circuits. The list includes ATX Power supplies, VHS recorder/players etc.

Just Found this:-

"There are many conflicting technical requirements for a good-quality front-end in an SW receiver. The noise figure and the intermodulation level should be low, the RF insulation between ports LO, RF and IF should be high, and some amplification is desirable. The Type SL6440 high level RF mixer from Plessey ensures a noise figure of around 10 dB, and offers sufficient suppression of the LO signal." - Oh dear! No longer available.

Dilbert Cell ???

In electronics, the Gilbert cell is a type of frequency mixer. It produces output signals proportional to the product of two input signals. Such circuits are widely used for frequency conversion in radio systems.[1] The advantage of this circuit is the output current is an accurate multiplication of the (differential) base currents of both inputs. As a mixer, its balanced operation cancels out many unwanted mixing products, resulting in a "cleaner" output. Gilbert cells can also be used as variable-gain amplifiers (VGA).[2]

It is a generalized case of an early circuit first used by Howard Jones in 1963,[3] invented independently and greatly augmented by Barrie Gilbert in 1967.[4] It is a specific example of "translinear" design, a current-mode approach to analog circuit design. The specific property of this cell is that the differential output current is a precise algebraic product of its two differential analog current inputs.

NE602/612 ICs

The NE602 and NE612 (including SA602/SA612 variants) are essentially identical, interchangeable Gilbert cell mixer/oscillator ICs often used in RF applications. Originally, the NE612 was introduced as a slightly updated, redesigned version of the NE602 for better high-frequency performance, but they are often identical in modern production and share the same specs, datasheet, and Pinout.

Both are designed by Signetics (now NXP/Philips). The distinction between "NE" (commercial) and "SA" (industrial/automotive) is more relevant than 602 vs. 612.

40m Direct Conversion Receiver

https://community.element14.com/technologies/open-source-hardware/b/blog/posts/simple-dcr-assembling-a-7-mhz-40m-direct-conversion-radio-receiver-part-1

https://qrp-labs.com/images/news/dayton2019/FDIM2019ConfProceedings.pdf

A 40m Direct Conversion Receiver project to upgrade from ZR to ZS

Hannes Coetzee, ZS6BZP, B.Eng Elektronic (Pretoria)

Also in :-

https://zs6wr.co.za/anode/AnodeJuly2010.pdf - Hannes Coetzee ZS6BZP

A 'Work in Progress' JB ZS6WL 2026-05-12

RESISTANCE is futile - 'the Borg'

And REACTANCE is imaginary. 

But resistance does 'work'. A resistor 'dissipates' heat when powered. Examples are a 'Bar Heater' - 1 to 2kWatts. A fan heater or a hair dryer. Some of which can be used as a 'heat gun'!

Right now the 'issue' is how to measure power. My long time friend and fellow radio amateur wanted to 'calibrate' a transceiver to S9. What is S9 you ask? Well the convention amongst amateurs is S9 is equivalent to 50 micro volts. Where? At the receiver input. Usually assumed to be 50 Ohms. But it doesn't have to be so low...

Like a lot of signal generators the output level is around zero dBm. Oh sorry! That is 1 milli Watt into 50 Ohms.  (even worse!) What is that in Volts? Time to get the calculator out. 

Power = ( V (r.m.s.)^2) / R  [That is V squared - and r.m.s. because this is an a.c. wave not d.c.]

which when transposed gives  Power * R = ( V r.m.s.  ^2)

Or  V r.m.s. =  SQR( Power * R)   [ SQR mean the square root - see a spreadsheet program/app]

When I worked it out I was gratified to get the same figure as I remember from a long time ago! The SQR of (50 * .001) is 0.2236 volts r.m.s. So the ratio of S9 (50uV) to 0.2236 is 4472...

So how do you get to 50uV from a signal at .2236 Volts? You use an attenuator. 

But his signal generator only had four 20 dB switched attenuators on the output. At this point we MUST discuss the use of dBV. This is because of the use of decibels and a constant impedance/resistance circuit. It is a 'convention' that radio frequency measurements in general use a input and output impedance of 50 Ohms. It didn't used to be... in the early days (last century) measurements at audio frequencies were done at 600 Ohms (in/out).

A voltage ratio of 4472 in dBV is 20 * log10( 4472) which is :- 73.0 dBV Which of course is impossible to achieve with his switched attenuator. He can get to 60 or 80 dB below the output...

So he needs a smaller step switched attenuator... Maybe a 1, 2, 4, 8 dBV attenuator?

Having got to this point I remembered that I had made a switched version many years ago using small toggle switches and resistors. Finding it was no problem. Drawing the circuit - no problem. Calculating the resistor values - time to hit the books. No not the Internet. 

Miniature Toggle Switch Version [Circa 1980]

Well the original was calculated using the Handbook by Scroggie and a new scientific calculator. Using Pi type attenuators at 50 Ohms. I tested it on a Spectrum Analyser and it was certainly 'good enough' for h.f. and above. Might not be accurate at 400 MHz or higher. But if made today with surface mount components might just be accurate. 

The calculated values             Nearest standard value

Loss    Shunt R        Series R 

1 dB    869.549        5.769        820        5.6

2 dB    436.211        11.615      470        12

4 dB    220.971        23.848      220        22

8 dB    116.143        52.844      120        51

Yes I know not very accurate - but close enough for Radio Amateur use. If you really - really want to - you could test a hundred of the standard values to find a close value. The sort of task given to students and apprentices last century. Oh and the 10 Ohm resistor? Well it is actually two in parallel. As I couldn't find a 5.6 Ohm resistor.

 

What about some power?

So someone gave you 20 1K 2 Watt (carbon film) resistors. You put them all in parallel inside a tin with the input connected to a good connector. Yes at h.f. an SO239 is fine. Now you have a QRP 'dummy load'. As an added 'feature' you might want to lift one of the resistors off the ground connection and connect it to another SO239. Put a 56 Ohm resistor across it to make it a known output resistance. This gives you the means to measure the output power into the 'dummy load' reduced by a simple calculated amount.  

What about the carbon film resistors? Yes they are slightly inductive. But as there are 20 of them in parallel, the inductance is 1/20th. So of no significance. 

Power rating? Well it won't be 40 Watts. But 25 should be alright for a short time. Do not pour in oil. Or like a colleague did many years ago only to make the match go from not bad to diabolical! Rather find a 8cm fan and a 12 Volt battery. Add vents and blow the resistors cool.