QRP Direct Conversion SSB Transceiver For 80 meters
Many construction project designs have been published for "home brew" amateur radio equipment over the years and this is to be applauded. However these projects tend to suffer from some of the following difficulties,
|Functionally Incomplete - Transmit only and/or typically limited to CW Morse code rather than SSB voice.|
|Restricted frequency operation - Crystal frequency controlled, or limited "VCXO" frequency range (e.g. < ±500 ppm = ±1.85 kHz @ 3.7 MHz).|
|Overly ambitious - Excessive component count, mechanical complex, too many features and functions incorporated.|
|Lacking Scalability - Not modular, can't be easily added to, expanded or scaled down into simpler building blocks.|
The project (in progress) presents a complete, fully functional low power "QRP" transceiver based on commonly available components without undue complexity or design criticality. It is based on a modular approach and can be built "as is" or expanded to, or modified, or built in a cut down form. For example, the design can have an optional 20 W PEP RF Power Amplifier (RFPA) added, or have its receiver section replaced with an alternative design, or be built just as a receiver, or just as a transmitter etc.
This does involve some concessions of course. For example the design is not intended to be "state of the art" or the "world's best rig". Many companies already invest millions of dollars developing such products and it is unrealistic to expect that amateur radio "hobby" enthusiasts will have access to such resources and expertise. It is more appropriate to consider construction projects that provide educational opportunity in the field of radio communication theory without being loaded down with unnecessary construction effort and design complexity.
This project follows a different path. Its primary objectives are,
|It should illustrate general electronic principles associated with radio communication product design.|
|It should be fun to read about, build and use as an amateur radio rig.|
|It should operate as a standalone radio transceiver without the need for other radio equipment.|
|However it should also be possible to build a "cut down" version that may be just a receiver or a transmitter.|
|The design should be sufficiently "generic" so that electronic enthusiasts can introduce design modifications with a good chance of success.|
The design path adopted, based on these ambitions, is proposed to be,
|Single Popular Frequency Band Operation - Simplifies overall system design and reduces component count.|
|Low Frequency 80 Meter Amateur Band (3.5 MHz to 3.9 MHz) Operation - Allows a free running VCO instead of a Frequency Synthesizer.|
|Direct Conversion (DC) System Architecture - Allows transmit and receiver bandwidth control to be accomplished at Audio Frequencies.|
|Modular Design Approach - Main sub systems can be removed, replaced or modified without affecting other sub system areas.|
|Use Of Conventional, Readily Available Components - Can be leaded or SMD, but all must be available from Farnell, RS, DSE or companies that support low volume component sales (Minicircuits, Coilcraft, etc)|
|Simple Construction with Option Approaches - Can be built on Veroboard products, "dead ant" on bare single sided copper PCB, or on commercial 2 sided FR4 (e.g. Pad2Pad 3" by 4" SMD populated PCB modules)|
Further, a low transmit power, i.e. QRP approach is adopted to reduce cost, simplify thermal issues and reduce battery requirements for portable operation.
Having established a Direct Conversion approach, based on a minimum complexity and cost requirement, it remains to determine the trade-offs between Double Sideband (DSB) versus Single Sideband (SSB) voice modulation and demodulation capability.
Double Sideband (DSB) can be consider to be Amplitude Modulation (AM) with a suppressed carrier. Since the carrier energy doesn't convey information, this energy is effectively a waste of transmitter energy. In a low power QRP transmitter, with limited output power resource, it is best to preserve as much transmit energy as possible for conveying actual (voice) information and eliminate spectral energy contributions that detract from the overall transmit energy budget..
Amplitude Modulation "wastes" significant energy in it's carrier energy. For example, AM with an amplitude modulation index "m" as high as m = 100% (sine wave modulation is assumed here) has a carrier signal power that is four times higher (i.e. +6 dB) than either of its sideband components. As an example, consider a QRP transmitter with 1 Watt available output power. Its carrier energy will be (by definition) equal to 1 watt but each sideband will only have 250 mW energy. Further, a modulation index of m = 100% is not feasible, and m = 60% may be selected based on practical considerations (associated with simple, standard modulation architectures). This further reduces the energy in each sideband to PUSB = PLSB = 250mW × [ m / 100 ]2 = 90 mW.
In Decibel terms, this implies that each information containing sideband is -10.5 dB lower than the transmitters output (carrier) power availability. DSB, in contrast, makes much better use of available transmitter power resources. Assuming complete carrier suppression, each sideband may be considered to contain 500 mW of energy - however this is not completely accurate. DSB transmit power is usually defined in terms of Peak Envelop Power (PEP). When single sinusoidal modulation is applied, each sideband will add in phase at certain times and the peak envelope will be four times the power contained in either sideband.
This phenomenon implies that each sideband will contain 250 mW PEP energy for 1 Watt PEP transmit power. This is superior to the previous case associated with AM. However it should be appreciated that a DSB transmitter draws variable energy from its power supply whilst an AM transmitter has a relatively high (an near constant) power supply energy drain. This attractive feature further motivates the avoidance of AM in QRP situations.
Note: AM is still used by some radio amateurs and is still a useful modulation format for radio communications. If energy efficiency is not a prime factor, other considerations may cause AM to a format of personal choice.
DSB represents a worthwhile energy improvement but still suffers from the transmission of an unnecessary "mirror image" sideband. Commercial SSB receivers will ignore one of these sidebands and therefore "loose" ½ of the available transmitted power from a DSB source (-3 dB lost energy). A DSB receiver would process both sidebands, but its RF bandwidth would be twice that of an equivalent SSB receiver. A DSB receiver is therefore more susceptable to interference and has an input noise floor that is 3 dB higher than a SSB receiver.
However, how much carrier and sideband suppression is required?
Commercial (high power) SSB transceivers target different technical requirements compared to a QRP approach. Since these products are intended to present the clearest possible signals across large distances they aim at achieving maximum legal output power and present receiver performance that extracts signal information despite high levels of background interference. Similarly, they must transmit clean spectrum so as not to add to overall spectral pollution (i.e. being a "good neighbor" policy)..
The ambitions for QRP are somewhat different. The primary objective is to make best use of finite power supply energy resource, such as that available from battery (or solar or wind) power. This suggests that SSB is the best choice for transmit. The actual amount of relative sideband suppression is less critical than compared with a high power commercial transmitter. For example, a commercial SSB transmitter may provide 100 Watts (+50 dBm) PEP wanted sideband power, and -50 dB suppression to the unwanted sideband. The unwanted sideband energy will therefore be 0 dBm PEP. In contrast, a 1 Watt PEP (+30 dBm) QRP transmitter would only require -30 dB unwanted sideband suppression to create the same level of unwanted (other) sideband interference of 0 dBm PEP.
From an energy utilization perspective, unwanted sideband suppression equal to -30 dB represents a minimal loss of wanted power (i.e. 0.1 %). Even -20 dB suppression corresponds to 99% energy transfer to the wanted sideband and only 1 % lost. A similar argument applies to unwanted carrier power - this is much less contributing to spectral wastage as it is very close to the main sideband signal.
This project will aim at suppression targets consistent with this reasoning so as not to "over design" a radio communication project that is intended to be fun to build, use and operate within a constrained financial (and time) budget.
On the other hand, how about the receiver approach? DSB and SSB architectures can have similar power supply energy requirements. DSB receivers are simpler but Direct Conversion (DC) approaches are still relatively simple for both formats.
It is constructive to note that DSB and SSB transceivers interoperate fairly well. A DSB receiver will demodulate a SSB signal and only suffer from some potential interference that may be present on the other sideband. A SSB receiver will demodulate a DSB transmission with a minor penalty of loosing ½ the available transmitted signal energy (i.e. information). A 1 Watt DSB transmission will appear to be equivalent to a 500 mW SSB transmission as an example.
However DSB transmission is very hard to demodulate with a DSB receiver (unless some use is made of the residual carrier energy to allow "synchronous" detection - which is an additional design complication).
From an energy perspective, transmitting SSB is a logical preference. This also allows compatibility with commercial SSB receivers and other QRP DSB or SSB receivers. Transmitting DSB, in contrast, is energy inefficient compared to SSB and prevents reception on other receivers that may only have have DSB demodulation capability.
Further, a QRP receiver may benefit from having optimum sensitivity and interference immunity available from the use of SSB demodulation, but this would only apply for cases when a corresponding QRP transmitting station is being contacted. Contacts with amateur radio operators that use a commercial "rig" will be received with fairly reasonable signal strength due to their higher transmitted power. Given that QRP transceivers are far less common than commercial transceivers, the need for SSB reception is questionable.
The approach recommended is therefore to use SSB for transmit and DSB for receive. However it is still advantageous to consider SSB for reception providing some additional design complexity is tolerable. In a DC approach this requires a 90 degree phase quadrature LO (as used on Tx), additional receiver mixer and a 90 degree audio signal combiner (can be as simple as a Quad OpAmp, 4 capacitors and 12 resistors!).
The design presented will aim to achieve the following design specification targets. These are somewhat relaxed compared to commercial products but are not expected to interfere with normal radio communication capability or operational enjoyment. Improvements can of course be made, but at the expense of complexity or cost. Such modifications are encouraged of course, i.e. the presented project should be best seen as a demonstration product with fairly open-ended future potential depending on the electronic enthusiast's experience or interests.
|Specification Target||Value||Units||Explanatory Comments|
|Supply Voltage||12||Volts||10.6 V to 15.6 V, suitable for 12 Volt lead-acid gell cells or 10 series connected NICAD batteries|
|Supply Current||< 50
|mA||Receiver at low volume
Receiver at full audio output into an 8 Ohm speaker load (1 Watt rms audio @ 1 kHz sine wave)
Transmitter at full PEP output power using a single carrier
|Lowest Frequency||3.500||MHz||Varicap tuned LC based VCO, internal preset, subject to aging and temperature tolerances|
|Highest Frequency||3.900||MHz||Varicap tuned LC based VCO, internal preset, subject to aging and temperature tolerances|
|kHz||Main tuning potentiometer rotation of 0 to 270 degrees
Fine tuning potentiometer rotation of 0 to 270 degrees
|Transmit Power||1.0||Watts||Peak Envelope Power for SSB|
|Sideband Suppression||-20||dB||SSB operation is used (over DSB) to make best use of limited transmit power, not best spectral purity|
|Harmonic Suppression||-40||dB||Equivalent to -60 dB for a commercial amateur radio transmitter @ 100 Watt PEP output|
|Microphone Drive||2||mV rms||To suit typical magnetic microphones, but the design can be modified to use electrect microphones|
|Receiver Format||DSB||Double Sideband is appropriate for QRP transmitters as the remote station will have adequate power|
|Receiver Sensitivity||< 0.6||uV 10dB||Corresponds to SNR = 10 dB, Noise Figure NF < 15 dB and a DSB bandwidth of 6 kHz (i.e. ±3 kHz)|
|Receiver Selectivity||-35||dB@8 kHz|
|AGC dynamic Range||50||dB||For less than ±3 dB variation in audio output power using a 1 kHz carrier offset input frequency|
|Audio Output Power||0.8||Watts rms||Maximum available audio power into an 8 Ohm speaker at 1 kHz and 10% distortion (clipping).|
The system concept is kept purposefully simple in order to allow its operation to be clearly understood by prospective constructors and experimenters. Its main sub systems are
|A Receiver Path (SSB approach shown - can simplify to DSB as suggested previously by removing one mixer and the 90 degree audio combiner.|
|A Transmitter Path (SSB approach shown)|
|A Phase Quadrature (90 degree output) Free Running Varicap Tuned VCO|
|An Antenna Tx-Rx Switch|
The proposed system is based on a modular approach. For example the transmit path can be omitted if the project is only required to be a receiver. Further, the RF PA subsystem can be substituted for a higher power version if more output transmit power is required - on an additional power amplifier can be added between the RFPA output and the antenna Tx-Rx switch. Further the receiver can be simplified to DSB as mention beforehand. This also applies to the transmitter. If both are simplified to DSB, the the VCO may also then be replaced with a simpler "on f" design, avoiding the need for frequency division to produce 90 degree (phase quadrature) LO outputs.
Return to 80m QRP Transceiver Page 80 meter transceiver
Return to main page Ian Scotts Technology Pages
Return to main page
© Ian R Scott 2007 - 2008