The "Pixie" 1 Watt CW Transceiver for the 30 to 80m Bands
These 40m CW Transceiver kits are widely available from China via e bay and other sources. A lot has been written about these on the net so I won't repeat it all here. The PCB measures only 2 inches square and is very well made. All the components are supplied including the BNC aerial socket, 2.5mm DC coaxial socket and two 3.5mm stereo jack sockets for the key and headphones, another earlier version is also available which uses small 2 pin sockets for the dc power and aerial output.
The kits are supplied for 40m band operation only and come supplied with a crystal for 7.023Mc/s.
I have made three additional units for the 30, 60 and 80m bands and this can be done with some simple component changes which are detailed below. The 30 and 60m units like the 40m one are fixed single crystal operation but the 80m board has been fitted with a crystal socket so that a variety of crystals (HC49U) can be selected making it more versatile. A picture of each unit is shown below and at first glance very little difference can be seen so they have each been clearly marked to avoid selecting the wrong transceiver.
I recommend that only a 12v Lithium battery be used as a mains power supply no matter how clean it's output is will always result in noise/hum problems on receive, here is the one I use which of course has a very long operational time as it is 6.8Ah.
These 40m CW Transceiver kits are widely available from China via e bay and other sources. A lot has been written about these on the net so I won't repeat it all here. The PCB measures only 2 inches square and is very well made. All the components are supplied including the BNC aerial socket, 2.5mm DC coaxial socket and two 3.5mm stereo jack sockets for the key and headphones, another earlier version is also available which uses small 2 pin sockets for the dc power and aerial output.
The kits are supplied for 40m band operation only and come supplied with a crystal for 7.023Mc/s.
I have made three additional units for the 30, 60 and 80m bands and this can be done with some simple component changes which are detailed below. The 30 and 60m units like the 40m one are fixed single crystal operation but the 80m board has been fitted with a crystal socket so that a variety of crystals (HC49U) can be selected making it more versatile. A picture of each unit is shown below and at first glance very little difference can be seen so they have each been clearly marked to avoid selecting the wrong transceiver.
I recommend that only a 12v Lithium battery be used as a mains power supply no matter how clean it's output is will always result in noise/hum problems on receive, here is the one I use which of course has a very long operational time as it is 6.8Ah.
The Pixie Kit as supplied from China showing all hardware and components
Note that only stereo headphones are connected to the output as if a mono plug is used this will effectively short circuit the audio output, a stereo or mono plug can be used for the key input however.
The rig produces one watt output on each of the bands and the most important change is in the RF output low pass filter, a change of capacitor is also needed in the oscillator circuit (C3). The component changes can be seen in the table shown below.
30m 60m 80m
C3 125 470 560 (Capacitors are in pico Farads)
C5 330 560 820
C6 330 560 820
L2 1.0 (orig) 2.2 2.2 (Inductors in micro Henries)
One of the problems with such a simple direct conversion receiver design is it's inherent lack of selectivity, this can only be improved in the audio stage so ideally a high "Q" Inductive Low Pass filter of about 800c/s makes a tremendous improvement and this is what I have done. The filter can be made so selective that "ringing" is present which I find very unpleasant to listen to and therefore this can be eliminated by fitting a damping pre set pot across the inductor to broaden it's response slightly to the point that the ringing effect is just eliminated. This approach however means obtaining a suitable pot cored inductor and the needed test equipment to set the filter up which is beyond the scope of many, I will however be giving the details of such a filter at the end of this article. A compromise can be made with the addition of a capacitor and a resistor to form a low pass audio filter which is easily added to the PCB. Looking at the circuit it can be seen that a degree of high pass filtering is obtained with the value of the audio coupling capacitor C10 but no attempt has been made to limit the audio HF response and so on relatively busy bands more than one CW station is usually heard making the QSO rather difficult. Fitting the new resistor and capacitor which I have called R8 and C12 make a low pass filter with an audio frequency roll off above about 1.5Kc/s limiting the receiver bandwidth and easing this problem. The resistor R8 is fitted in series with the original capacitor C10 and must be fitted exactly as shown below, the capacitor C12 can be fitted across pins 2 and 3 of the audio output IC socket marked U1.
R8 is 2.2K, 1/4 watt. C12 is 0.047 micro Farad miniature disc ceramic.
The rig produces one watt output on each of the bands and the most important change is in the RF output low pass filter, a change of capacitor is also needed in the oscillator circuit (C3). The component changes can be seen in the table shown below.
30m 60m 80m
C3 125 470 560 (Capacitors are in pico Farads)
C5 330 560 820
C6 330 560 820
L2 1.0 (orig) 2.2 2.2 (Inductors in micro Henries)
One of the problems with such a simple direct conversion receiver design is it's inherent lack of selectivity, this can only be improved in the audio stage so ideally a high "Q" Inductive Low Pass filter of about 800c/s makes a tremendous improvement and this is what I have done. The filter can be made so selective that "ringing" is present which I find very unpleasant to listen to and therefore this can be eliminated by fitting a damping pre set pot across the inductor to broaden it's response slightly to the point that the ringing effect is just eliminated. This approach however means obtaining a suitable pot cored inductor and the needed test equipment to set the filter up which is beyond the scope of many, I will however be giving the details of such a filter at the end of this article. A compromise can be made with the addition of a capacitor and a resistor to form a low pass audio filter which is easily added to the PCB. Looking at the circuit it can be seen that a degree of high pass filtering is obtained with the value of the audio coupling capacitor C10 but no attempt has been made to limit the audio HF response and so on relatively busy bands more than one CW station is usually heard making the QSO rather difficult. Fitting the new resistor and capacitor which I have called R8 and C12 make a low pass filter with an audio frequency roll off above about 1.5Kc/s limiting the receiver bandwidth and easing this problem. The resistor R8 is fitted in series with the original capacitor C10 and must be fitted exactly as shown below, the capacitor C12 can be fitted across pins 2 and 3 of the audio output IC socket marked U1.
R8 is 2.2K, 1/4 watt. C12 is 0.047 micro Farad miniature disc ceramic.
The selection of a suitable frequency for each of the bands is entirely up to the operator but suitable QRP frequency crystals are available on e bay very cheaply making the total cost of each transceiver about £5 in Britain making the cost for all four transceivers around £20 so an extremely cheap way of becoming operational on CW on four very popular bands.
It should be noted that the RF output transistor Q2 is operating near to it's maximum parameters at one watt output and so the key should not be held down for more than about 20 seconds especially if the aerial system is presenting a poor SWR.
80m Transceiver with a crystal holder for multi channel operation
60m Transceiver
40m Transceiver
30m Transceiver
High "Q" Receiver Low Pass Filter for use with the Pixie.
This filter uses only one Inductor and three capacitors but is very effective as a high "Q" low pass filter for the Pixie receiver. If you look at the graph below you will see how sharp the filter is and when using an inductor as shown the "Peak" frequency can be adjusted slightly but 800 c/s is just right for most ops although with changes in the inductor and capacitor values this of course can be changed. The "Q" is so high that one problem is that undamped "Ringing" will manifest which is not pleasant to the ear and so the pre-set pot VR1 across the inductor is adjusted to just eliminate this issue while still maintaining a sharp response. The capacitor C1 effectively replaces the original capacitor C10 which is removed from the Pixie PCB offering some high pass filtering. The circuit below should be fitted as close as possible to the Pixie PCB not forgetting to fit the earth connection.
This filter completely transforms the operation of the Pixie making it truly useable.
Response Graph
Audio Low Pass Filter Circuit
The circuit below is fitted in place of the original capacitor C10 making sure that it is connected as shown with it's output connected to the LM386 audio IC, U1.
The circuit below is fitted in place of the original capacitor C10 making sure that it is connected as shown with it's output connected to the LM386 audio IC, U1.
Parts List
C1 0.047mfd Disc Ceramic
C2 0.22mfd Poly
C3 0.22mfd Poly
VR1 20K Pre-Set Pot
L1 250-300mH Tuneable Pot Cored Inductor
NB Other values of inductor and capacitors can be used as long as the same "Peak" frequency is arrived at but will not normally exhibit such a high "Q" and will be somewhat less effective.
C1 0.047mfd Disc Ceramic
C2 0.22mfd Poly
C3 0.22mfd Poly
VR1 20K Pre-Set Pot
L1 250-300mH Tuneable Pot Cored Inductor
NB Other values of inductor and capacitors can be used as long as the same "Peak" frequency is arrived at but will not normally exhibit such a high "Q" and will be somewhat less effective.
