A 30 Mtr Direct Conversion Receiver.

Title image of DCRX.


The receiver described here is based on a design built and used by Hans Summers (G0UPL) and Paolo (I1DFS) for the reception of QRSS signals in the 30 Mtr band. The receiver can be easily duplicated from the details which follow or simply used as a source of ideas for a design of your own. An example of the original receiver upon which this design is based appears on Hans (G0UPL) QRSS web pages here.

Hans (G0UPL) 30m QRSS Receiver.

It is a simple Direct Conversion design and while it does not yield state of the art performance it does provide a useful introductory receiver for qrss reception. In this version a number of enhancements have been incorporated to improve the basic performance. These enhancements include the following...

1)  A modified SA602 circuit to improve the short term frequency stability of the crystal oscillator.
2) A crystal "oven" to ensure excellent long term frequency stability of the crystal oscillator.
3) A separate signal frequency crystal bandpass filter to improve front-end selectivity and remove A.M. broadcast "breakthrough".
4) A separate audio frequency bandpass filter to further improve selectivity and reduce interference (QRM) from in-band signals.

The receiver is comprised of three modules, this was done in order to provide a quick method of replacing or upgrading sections of the receiver without having to fully rebuild the entire receiver each time the circuit was modified. The three modules in the system are...

The Direct Conversion Receiver Module.
The Signal Frequency Crystal Bandpass Filter/Amplifier module.
The A.F. Buffer Amplifier/A.F. Filter module.

Direct Conversion Receiver Module.

The first module to be described is the Receiver. This is a direct conversion design built around the popular SA602 device followed by an NE5534 low noise Op-Amp. The NE5534 is not one of the lowest noise devices available but it is very cost effective in this application. Because of the relatively high level of atmospheric and man made noise (QRM) in the 30 Mtr band a lower noise Op-Amp would not give much improvement over the 5534 used unless you happen to be using a very small or inefficient receiving antenna.

An image/link to the full schematic of the receiver module appears below.

DC-RX Schematic.

The circuit does require some explanation and astute readers will notice that the configuration of the crystal oscillator is somewhat unusual. An additional transistor has been added to the SA602 oscillator circuit, the BC238B transistor in conjunction with the transistor/resistor already present in the SA602 form a Darlington pair, this configuration offers a number of advantages. Firstly, the lower O/P impedance of the Darlington configuration makes it possible to increase the value of the capacitors in the feedback network such that they will "swamp" any stray or parasitic capacitance more effectively. Secondly the BC238B transistor forms a very effective buffer stage permitting the connection of a frequency counter (or QRSS TX etc) to the oscillator stage without "pulling" the frequency of the crystal oscillator. The idea for this modification arose after previous success with a crystal oscillator design given to me by Peter (DL6NL) which is currently used in my 30 Mtr QRSS TX. Looking at this oscillator circuit I realized it may be possible to incorporate elements of the circuit in this receiver design using some of the components which are internal to the SA602.

An image/link to the schematic of the crystal oscillator from DL6NL appears below for reference.

DL6NL Xtal Oscillator.

The circuit for the DC-RX has been constructed "ugly" style over a ground plane. Both the SA602 and the NE5534 are mounted in sockets just in case of "accidents" and to permit easy replacement in case of failure. An image of the DC-RX board appears below (left hand side image) and an internal view of the complete DC-RX module appears below on the right. The yellow foam (left hand image) is thermal insulation around the crystal to improve the operation of the crystal oven while the pink coloured foam shown in the right hand image is an additional layer of thermal insulation.

DC-RX Board.     DC-RX Boxed.

In the same compartment as the DC-RX is the crystal oven temperature control board, this part of the circuit does not require a ground plane therefore it has been constructed using strip board. The schematic for the crystal oven temperature controller is shown below (left hand side image) and an image of the oven controller board appears below on the right.

MK-2 Xtal Oven Schematic.     Mk-2 Oven control board.

A much more comprehensive description of the Mk-2 oven controller circuit, its operation and construction is described here.

The two boards (DC-RX and the Crystal Oven control board) are assembled in a small di-cast aluminum box measuring 30 x 40 x 120 mm. L.E.D.'s have also been fitted to indicate "Oven-on" and "Power-on". Connections for the Antenna I/P and A.F. O/P are via screened cables terminated with phono plugs. The current drawn by the DC-RX module is around 80 to 90 mA @ 12.5 Volts which includes the L.E.D's and the crystal oven. An image of the complete DC-RX module appears below.

Picture of DC-RX boxed with cover fitted.

Signal Frequency Crystal Bandpass Filter/Amplifier Module.

The direct conversion receiver module can be used on its own but I encountered frequent problems with A.M. broadcast stations "breaking through" from the lower frequency S.W. bands. Part of the problem is with the SA602 which is notorious for its poor large signal handling. Because of these "breakthrough" problems it was decided that a very narrow bandpass filter would be constructed using a quartz crystal as the bandpass element. This proved to be highly successful in removing the A.M. breakthrough. The design of this unit became a highly rewarding project in its own right and is the focus of a separate web page which can be found  here.

Crystal band pass filters for QRSS applications.

The schematic for the crystal bandpass filter/amplifier board appears below (left) with images of the assembled and boxed unit appearing below center and right.

Xtal BPF Schematic.  Xtal BPF in box (1 of 2)   Xtal BPF in box (2 of 2)

The signal frequency bandpass filter is constructed over a ground plane and housed in its own enclosure measuring 75 x 75 x 50 mm. The separate enclosure ensures freedom from unwanted signal "leakage" around the filter. The crystal bandpass filter module also includes a small R.F. amplifier of modest gain to compensate for the small loss in the crystal filter. The module is fitted with phono sockets for Antenna I/P and R.F. O/P. The R.F. O/P of the filter/amplifier module is connected to the antenna I/P of the QRSS DC receiver. The current drawn by this module is about 35 mA @ 12.5 Volts which includes the L.E.D.'s.

The direct conversion receiver and filter/amplifier modules together provide a very effective QRSS receiver, a further improvement can be made with the addition of the A.F. Buffer Amplifier/A.F. Filter described below.

A.F. Buffer Amplifier/A.F. Filter Module.

Note: (Added 25/10/06) The unit about to be described was added to the DC-RX in an attempt to further improve the performance though I have to admit the improvement was hardly noticeable except under conditions of heavy QRM when signals close to the QRSS sub band would confuse or desensitize Argo. Since this unit was built I have had some time to evaluate it more fully and would suggest that using op-amps with a lower noise figure would give improved performance. My feeling is that the A.F. filter is a worth while optional extra for those times when QRM is a problem.

The A.F. Buffer Amplifier/A.F. Filter module was added to the receiver to ensure that the software A.G.C. in Argo was not "confused" by unwanted signals which appear in the audio passband of the DC-RX. These signals arise from other users of the 30 Mtr band (CW, RTTY  etc) which are within the bandpass of both the front end (RF) filter and the audio stages of the receiver. These unwanted signals can cause "desensitization" of Argo such that wanted signals appear weaker than they actually are. Initially a low pass audio filter was considered but it was soon realized that with little or no additional complexity an audio bandpass filter could be used which would offer better performance. The low frequency roll-off of the bandpass A.F. filter helps to reduce any possible power line noise from reaching the P.C. sound-card and also offers attenuation of unwanted signals below the 100 Hz window used for QRSS.

It was also felt that the original DC RX design was perhaps lacking in AF gain. To much gain risks overloading the sound-card on strong signals but my feeling was that a little more audio gain could be tolerated. I came to this conclusion for two reasons, the atmospheric noise was barely audible which made me wonder if I was achieving optimum sensitivity and secondly the additional gain from another low noise amplifier stage would lift the wanted signal level far above the noise level which may result from the Op-Amp's used in the active bandpass filter. A full schematic for the A.F. Buffer Amplifier/A.F. Filter is shown below.

A.F. Filter Schematic.

The center frequency of the A.F. Filter was chosen to match the center of the QRSS sub-band as it appears in Argo. This frequency may not be the same on your DC RX and may require adjustment to suit the frequency of the crystal you use. For use with Argo frequencies between 1 and 2 kHz should be suitable. Fine tuning of the A.F. filter can be performed by adjustment of the two resistors (Rx2 and Rx3 in the circuit diagram), adjusting these resistors individually can also permit "stagger" tuning of the two cascaded filter stages. This can be useful if a wider B.W. is required. In my version of the filter I found a "Q" of 8 to be more than enough with higher "Q" values causing excessive "ringing" of the filter.

Note: With the component values shown in the schematic the filter may appear to "ring" a little with CW or RTTY signals but this causes no problems with QRSS signals due to the slow nature of the QRSS modes.

If you wish to build this filter and want to "trim" to a different center frequency or wish to experiment with different values of "Q" or filter gain then the formulas below may be of interest.

Filter calculation image.   Filter calculation components image.

The A.F. Buffer Amplifier/A.F. Filter module is constructed on strip board and mounted in a small metal enclosure measuring 130 x 70 x 35 mm which provides screening. The I/P and O/P connections are via phono sockets. This module draws about 22 mA @ 12.5 Volts which includes the "Power-on" L.E.D. Two images of the completed module appear below.

AF filter board view.   AF filter boxed.

With all three modules interconnected the current drawn is around 140 to 150 mA @ 12.5 Volts. Below is an image of the complete interconnected QRSS receiver system comprising all three of the modules described above.

All modules together.

Test Results.

Theory states that a simple direct conversion receiver will suffer a -3dB penalty due to the unwanted audio "image" but in practice this penalty is far from obvious. Indeed the receiver currently outperforms any other receiver here at M0AYF for QRSS reception. The receivers frequency stability (a very important factor in the reception of QRSS signals) has exceeded expectations. I plotted the change of frequency with respect to time starting with a cold crystal oven at switch-on. As a reference I used a signal phase locked to the 60 kHz MSF frequency standard located here in the UK. The graph appears below.

Frequency-Time graph.

The graph is not particularly accurate since very few points have been plotted but it shows that the frequency drifts only a few Hz in the first few minutes after switch-on from a cold start. After around 40 minutes the crystal oven temperature stabilizes and the frequency remains substantially constant thereafter. The maximum "drift" in frequency from switch-on is around 8 Hz in a positive direction, after about 40 minutes the "drift" is below measurable limits here at M0AYF.

While I have no way to measure sensitivity or dynamic range of the receiver I can confirm that in use I can hear atmospheric noise which suggests the receiver is sensitive enough. I can also confirm that so far no signal overload problems have been encountered and all traces of A.M. breakthrough have been removed thanks to the crystal bandpass filter which hides many of the problems caused due to limitations in the SA602. Below are a two screen captures from Argo of several QRSS signals received using the DC-RX system described above.


DL6JAN Hell.

If you find this receiver design interesting then you may also be interested in the recently completed advanced 30 Mtr QRSS RX designed by Andy (G4OEP) which has many interesting features including the use of a crystal lattice filter as a signal frequency bandpass filter and a "H-Mode" mixer. A link to this excellent design appears below.

Advanced 30 Mtr QRSS receiver (lower third of web-page) designed by Andy (G4OEP)

Well, that’s about it, thank you for reading this and please send any questions, comments or "heckles" etc to the e-mail address linked below.

e-mail QSL


Des (M0AYF)