Earlier versions of the uBITX transceiver kits were characterized by poor intermodulation performance during transmission. This means that the transmitted signal would occupy a much wider bandwidth on the band than is permissible and could cause interference with neighboring frequencies. Intermodulation products are measured using a two-tone test. This article describes a modification that will significantly improve the quality of the transmitted signal.
Dočítate sa:
Two-tone test of an SSB transmitter
Warren WA8TOD performed very detailed measurements using a two-tone SSB test. The two-tone test is based on the principle that two low-frequency modulation signals, such as 700 Hz and 1900 Hz, are fed into the SSB transmitter. New signals, known as intermodulation products, are generated by the transmitter’s nonlinearities. In practice, 3rd- and 5th-order IMD products are most commonly measured. Based on their curve, it is then possible to estimate the generation of higher-order IMD products as well.
What is the source of IMD products in the uBITX?
WA8TOD found that the main source of IMD products is the bidirectional amplifiers. To simplify the design, the original circuit has been modified, which results in greater frequency dependence and, more importantly, worse IMD characteristics.
The second cause is the characteristics of the mixers and active components. By their very nature, these generate additional IMD products and thereby affect individual parts of the transmitter.
Methods for Improving IMD Products in the uBITX
This has prompted several designers to implement a series of modifications. Newer versions of the uBITX transceiver already incorporate certain improvements. In particular, there have been changes to the frequencies in the uBITX, the transistors in the amplifiers, adjustments to the amplifier’s operating point, and the addition of an LC filter after the second bidirectional amplifier.
Warren tested replacing the bidirectional amplifiers with simple MMIC amplifiers. He achieved a significant improvement in IMD products. However, within the broader generated spectrum, even this modification does not fully resolve the purity of the transmitted signal within the shortwave spectrum. Therefore, it is advisable to use either an additional 45 MHz filter (version 3) or an LC filter.
IMD Modification and Measurement of IMD Products on the uBITX
First, I performed an improvised measurement of the uBITX’s IMD products. I used an online dual-tone signal generator from the website http://www.onlinetonegenerator.com/binauralbeats.html. The first signal was a sine wave at a frequency of 700 Hz. The second signal had the same waveform at a frequency of 1900 Hz. The frequency difference between these signals was therefore 1200 Hz, or 1.2 kHz.
This signal was fed into a transceiver set to SSB transmission. A 1200/50-ohm divider connected via a 1 nF capacitor was inserted after the bidirectional amplifier at the 45 MHz intermediate frequency. A low-cost SDR receiver with an RTL2832 chip was connected to this point. SDR Console was used as the software. Although this is a makeshift and preliminary measurement, it can be very helpful.
IMD Measurements of uBITX Products
First, the IMD products of the unmodified amplifier were measured, with the output power set to 2 W using a wattmeter:
IMD products of the original uBITX circuit
Both modulation signals are visible at frequencies of 44,991.8 kHz and 44,993 kHz. The fact that the frequency is a few kHz lower is likely due to SDR frequency drift. On both sides, several peaks are visible; we are most interested in those at 44,994.2 kHz (i.e., 1.2 kHz above the top frequency). This is 3rd-order IMD, and if we trust the reading, it is only -12 dB below the level of the modulation signals. Another 1.2 kHz higher (44,995.4 kHz) is the 5th-order IMD. It is approximately -22 dB below the level of the modulation signals.
IMD products with the MSA-1105 MMIC amplifier and a 45 MHz filter at 3 W output power
Subsequently, a modified amplifier with a filter was added. This amplifier has a slightly higher gain, so the output power approached 3 W. A slightly different bandwidth is set for the measurement here, but it is immediately apparent that the resulting signal is cleaner. 3rd-order IMD is approximately -23 dB below the level of the modulation signals, and 5th-order IMD is approximately -33 dB below the level of the modulation signals.
The last measurement was performed with an amplifier and filter, but with modulation adjusted so that the output power was again 2 W:
IMD products with the MSA-1105 MMIC amplifier and a 45 MHz filter at 2 W output power
3rd-order IMD is approximately -34 dB below the level of the modulation signals, and 5th-order IMD is approximately -41 dB below the level of the modulation signals. This is already a very clean signal for a simple transceiver and fully suitable for operation on amateur radio bands.
Modification to the uBITX to improve IMD performance
The original bidirectional amplifier with transistors Q20, Q21, and Q22 was removed. Capacitors C20 and C22 were desoldered from the circuit board. The amplifier was built on a piece of universal circuit board. Both stages are populated with MSA-1105 MMICs (from my personal stock). The gain of these stages could be approximately 2 × 12 dB.
MSA-1105 MMIC amplifier and 45 MHz filter
PCB in the uBITX
This is followed by a transformer, whose role is to match the output impedance of the amplifiers (50 ohms) to the impedance of the 45 MHz crystal filter (eBay). This is estimated to be 600 ohms. The transformer has 2 turns on the primary winding and 7 turns on the secondary winding. The output transformer is identical, but is wired in reverse. This again results in an impedance of 50 ohms. A BN 43-2402 dual-slot core is used (not required). The output is connected directly to the mixer (R27 is not used either).
Circuit Diagram
Circuit diagram of the MMIC amplifier with a filter for uBITX
The amplifier is mounted above the main uBITX printed circuit board. This ensures relatively short connection traces.
Conclusion
The modification described here can significantly improve the IMD of the uBITX transceiver. Ideally, the first bidirectional amplifier at 12 MHz should also be replaced, which would further improve the quality of the output signal.
At the same time, a method for measuring the IMD of SSB transmitters is presented, which is accessible to a wide range of ham radio operators since it does not require any hard-to-find equipment.
References
http://www.onlinetonegenerator.com/binauralbeats.html
https://ubitx.net/2018/09/15/details-of-the-anti-spurs-mod-added-45mhz-filter/
https://ubitx.net/2018/10/21/the-last-challenge-reducing-imd/
https://ubitx.net/2018/09/15/on-the-matter-of-ubitx-imd/
https://ubitx.net/2018/09/15/imd-using-an-alternative-mmic-amplifier/
https://www.sdr-radio.com/download
Update May 30, 2020 – Change of the first amplifier to an SNA586 MMIC
This significant improvement in the quality of the transmitted signal definitely justifies using an MMIC instead of the Q40–Q42 amplifiers. Due to the low frequency, a different MMIC, the SNA586, would be used. Measurement results:
IMD products with the MSA-1105 and SNA586 MMIC amplifiers and a 45 MHz filter at 2W power
At 2W power, 3rd-order IMD products are suppressed by a fantastic -45 dB!
IMD products with the MSA-1105 and SNA586 MMIC amplifiers and a 45 MHz filter at 5 W
IMD products with the MSA-1105 and SNA586 MMIC amplifiers and a 45 MHz filter at 10 W power
If we increase the drive level, the IMD products will increase. Here, one might consider adjusting the transmitter’s drive level using the RV1 trimmer. Since IMD products are generated in the first amplifiers, increasing the drive level of an underpowered output stage could improve the quality of the output signal.
However, it can be said that if we limit the output power to 3W or less, it is possible to drive the output stage with such a transmitter without fear of poor signal quality.