Re: [Elecraft] DSP vs. non-DSP portable transceivers
"Good, Fast, Cheap ... choose two." 73, Fred ["Skip"] K6DGW Sparks NV DM09dn Washoe County On 5/23/2018 7:32 AM, Ignacy wrote: To put it mildly, with progress in computing, the implementation in DSP is easier, smaller, cheaper and mostly better. Programming is expensive. Same with tube and SS amplifiers. Most likely, hardware-wise SS is easier, cheaper and mostly better. Initial design and programming are expensive. Average age of ham radios is around 65 (QST). Old "truths" die slowly. Ignacy, NO9E -- Sent from: http://elecraft.365791.n2.nabble.com/ __ Elecraft mailing list Home: http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/mmfaq.htm Post: mailto:Elecraft@mailman.qth.net This list hosted by: http://www.qsl.net Please help support this email list: http://www.qsl.net/donate.html Message delivered to k6...@foothill.net __ Elecraft mailing list Home: http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/mmfaq.htm Post: mailto:Elecraft@mailman.qth.net This list hosted by: http://www.qsl.net Please help support this email list: http://www.qsl.net/donate.html Message delivered to arch...@mail-archive.com
Re: [Elecraft] DSP vs. non-DSP portable transceivers
To put it mildly, with progress in computing, the implementation in DSP is easier, smaller, cheaper and mostly better. Programming is expensive. Same with tube and SS amplifiers. Most likely, hardware-wise SS is easier, cheaper and mostly better. Initial design and programming are expensive. Average age of ham radios is around 65 (QST). Old "truths" die slowly. Ignacy, NO9E -- Sent from: http://elecraft.365791.n2.nabble.com/ __ Elecraft mailing list Home: http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/mmfaq.htm Post: mailto:Elecraft@mailman.qth.net This list hosted by: http://www.qsl.net Please help support this email list: http://www.qsl.net/donate.html Message delivered to arch...@mail-archive.com
Re: [Elecraft] DSP vs. non-DSP portable transceivers
On 5/22/2018 11:26 AM, Wayne Burdick wrote: The P3 panadapter, meanwhile, is actually a direct-sampling SDR, with its own DSP. It taps off the 1st IF, which is protected by the K3’s marrow RF band-pass filters. The P3’s noise floor is extremely low because it is also preceded by the K3’s preamp, LNA, and low-loss mixer. Yes. AND in the hands of a knowledgeable engineer/technician, a near lab-quality test instrument, with a display that can easily be calibrated to the signal level at the antenna input. I've used it to measure occupied bandwidth of transmitted signals to precision of a few Hz, and the SVGA board, with it's separate, higher res FFT, improves that to one Hz. I own two other dedicated spectrum analyzers -- an HP 8657A and a Rigol. They are not capable of these measurements, because they lack the frequency resolution (by a factor of about 500:1). Many other modern SDRs also have this capability when used with suitable software. 73, Jim K9YC __ Elecraft mailing list Home: http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/mmfaq.htm Post: mailto:Elecraft@mailman.qth.net This list hosted by: http://www.qsl.net Please help support this email list: http://www.qsl.net/donate.html Message delivered to arch...@mail-archive.com
Re: [Elecraft] DSP vs. non-DSP portable transceivers
If I was Eric and Wayne, I'd name the next base transceiver the K5, just because. On 5/22/2018 11:26 AM, Wayne Burdick wrote: Hi Buck, I’m not entirely sure how to parse your response to my posting. But I can cover several possible interpretations by pointing out that the K3S has all of the same DSP features as the KX2/KX3, and then some. The K3S also has the advantages of a superhet, with narrow crystal filters in the 1st IF protecting the ADCs. The P3 panadapter, meanwhile, is actually a direct-sampling SDR, with its own DSP. It taps off the 1st IF, which is protected by the K3’s marrow RF band-pass filters. The P3’s noise floor is extremely low because it is also preceded by the K3’s preamp, LNA, and low-loss mixer. 73, Wayne N6KR __ Elecraft mailing list Home: http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/mmfaq.htm Post: mailto:Elecraft@mailman.qth.net This list hosted by: http://www.qsl.net Please help support this email list: http://www.qsl.net/donate.html Message delivered to arch...@mail-archive.com
Re: [Elecraft] DSP vs. non-DSP portable transceivers
Hi Buck, I’m not entirely sure how to parse your response to my posting. But I can cover several possible interpretations by pointing out that the K3S has all of the same DSP features as the KX2/KX3, and then some. The K3S also has the advantages of a superhet, with narrow crystal filters in the 1st IF protecting the ADCs. The P3 panadapter, meanwhile, is actually a direct-sampling SDR, with its own DSP. It taps off the 1st IF, which is protected by the K3’s marrow RF band-pass filters. The P3’s noise floor is extremely low because it is also preceded by the K3’s preamp, LNA, and low-loss mixer. 73, Wayne N6KR http://www.elecraft.com > On May 22, 2018, at 11:09 AM, Buck wrote: > > And that my friends, tells you everything you need to know about the K4. > > Buck, k4ia > Honor Roll > 8BDXCC > EasyWayHamBooks.com > >> On 5/22/2018 11:55 AM, Wayne Burdick wrote: >> At the Dayton (yeah, OK, Xenia) Hamvention, just before giving a demo of the >> KX2, a bystander asked me, “Why would you need DSP in a portable >> transceiver?” He went on to mention a few currently available rigs he was >> considering that didn’t have DSP, including one very new one [an “upgraded” >> legacy model]. I’ll leave the entire list to the reader’s imagination. >> There are two basic reasons to use DSP: >> 1. As the core of the transceiver’s architecture, to reduce size and weight >> and increase flexibility >> 2. To provide a number of important features that have typically been seen >> only in desktop radios >> Explaining #1 is always a bit of a challenge since it dives into theory. >> Here’s the short version: The receiver in both the KX2 and KX3 uses a >> high-level quadrature mixer that down-converts the RF signal to baseband (0 >> kHz) or to a low I.F. (8 kHz). The resulting I and Q signals are fed to a >> pair of high-performance audio A-to-D converters, and all demodulation is >> then handled in DSP code. This configuration has immense versatility, >> reduces component count, and provides very consistent unit-to-unit >> performance in parameters such as selectivity. >> To explain #2, I prefer to just dive into the demo. There's a long list of >> functions that directly benefit those who operate portable, as well as >> anyone trying to maximize their communications capabilities. >> Here are the DSP functions that I try to include in demos (given enough >> time), along with their benefits for the operator: >> * Multiple Modes >> The KX2 and KX3 both support SSB, CW, AM, FM, and several data modes, >> without the need for supplementary hardware. All demodulation is handled by >> the DSP. And if we add a new mode in the future, it will be available to all >> KX-line radio owners free of charge with a simple firmware upgrade. >> * Bandwidth Control >> The DSP bandwidth can be adjusted from 50 Hz to 4000 Hz (5000 Hz in AM >> mode). The passband can also be shifted. Rigs without DSP are limited to the >> passband of their analog filters (including crystal filters). >> * Stereo Audio Effects and Dual Receive >> Both rigs demodulate two independent audio channels, allowing for simulated >> stereo, which reduces listening fatigue. >> Stereo is also used when DUAL RX is enabled; this allows the operator to >> listen to one signal in the left ear (VFO A) and another in the right (VFO >> B). This is essential for use with DX stations operating split, and greatly >> facilitates hunt ’n’ pounce contest operation where one receiver is parked >> on a station to be worked, while the other is used to locate the next one. >> * Adjustable AGC >> The KX2/KX3 include user-configurable AGC. Threshold, attack, decay, and >> hold time can all be adjusted. These are parameters that would require >> dozens of components in a non-DSP design. There’s also a short-duration >> pulse detector (AGC PLS). When this is turned on--the default--the DSP will >> seamlessly remove such pulses before they can pin your S-meter and cause a >> long RX recovery delay. (Rob Sherwood complemented us on this feature, which >> he found was missing on even high-end desktop transceivers. So, for in-house >> purposes, we call it “Sherwood mode” :) >> * Noise Blanking >> Many non-DSP radios lack noise blanking, or have simple blanking circuitry >> that can be easily overloaded. The KX-line DSP implements very effective >> blanking that can take out narrow pulses even at very low settings, and >> greatly attenuate irregular, high-duty cycle noise sources typical of light >> dimmers, electric fences, and wideband radar. Blanking can make the >> difference between a good QSO and not even hearing the other station. >> * Noise Reduction >> We often have to deal with stochastic (“white”) noise on the bands. Our >> DSP-based NR can be adjusted over a wide range to improve intelligibility, >> especially of speech signals. >> * Audio Peaking Filter (APF) >> CW operators often have to copy signals right at or even below the noise
Re: [Elecraft] DSP vs. non-DSP portable transceivers
And that my friends, tells you everything you need to know about the K4. Buck, k4ia Honor Roll 8BDXCC EasyWayHamBooks.com On 5/22/2018 11:55 AM, Wayne Burdick wrote: At the Dayton (yeah, OK, Xenia) Hamvention, just before giving a demo of the KX2, a bystander asked me, “Why would you need DSP in a portable transceiver?” He went on to mention a few currently available rigs he was considering that didn’t have DSP, including one very new one [an “upgraded” legacy model]. I’ll leave the entire list to the reader’s imagination. There are two basic reasons to use DSP: 1. As the core of the transceiver’s architecture, to reduce size and weight and increase flexibility 2. To provide a number of important features that have typically been seen only in desktop radios Explaining #1 is always a bit of a challenge since it dives into theory. Here’s the short version: The receiver in both the KX2 and KX3 uses a high-level quadrature mixer that down-converts the RF signal to baseband (0 kHz) or to a low I.F. (8 kHz). The resulting I and Q signals are fed to a pair of high-performance audio A-to-D converters, and all demodulation is then handled in DSP code. This configuration has immense versatility, reduces component count, and provides very consistent unit-to-unit performance in parameters such as selectivity. To explain #2, I prefer to just dive into the demo. There's a long list of functions that directly benefit those who operate portable, as well as anyone trying to maximize their communications capabilities. Here are the DSP functions that I try to include in demos (given enough time), along with their benefits for the operator: * Multiple Modes The KX2 and KX3 both support SSB, CW, AM, FM, and several data modes, without the need for supplementary hardware. All demodulation is handled by the DSP. And if we add a new mode in the future, it will be available to all KX-line radio owners free of charge with a simple firmware upgrade. * Bandwidth Control The DSP bandwidth can be adjusted from 50 Hz to 4000 Hz (5000 Hz in AM mode). The passband can also be shifted. Rigs without DSP are limited to the passband of their analog filters (including crystal filters). * Stereo Audio Effects and Dual Receive Both rigs demodulate two independent audio channels, allowing for simulated stereo, which reduces listening fatigue. Stereo is also used when DUAL RX is enabled; this allows the operator to listen to one signal in the left ear (VFO A) and another in the right (VFO B). This is essential for use with DX stations operating split, and greatly facilitates hunt ’n’ pounce contest operation where one receiver is parked on a station to be worked, while the other is used to locate the next one. * Adjustable AGC The KX2/KX3 include user-configurable AGC. Threshold, attack, decay, and hold time can all be adjusted. These are parameters that would require dozens of components in a non-DSP design. There’s also a short-duration pulse detector (AGC PLS). When this is turned on--the default--the DSP will seamlessly remove such pulses before they can pin your S-meter and cause a long RX recovery delay. (Rob Sherwood complemented us on this feature, which he found was missing on even high-end desktop transceivers. So, for in-house purposes, we call it “Sherwood mode” :) * Noise Blanking Many non-DSP radios lack noise blanking, or have simple blanking circuitry that can be easily overloaded. The KX-line DSP implements very effective blanking that can take out narrow pulses even at very low settings, and greatly attenuate irregular, high-duty cycle noise sources typical of light dimmers, electric fences, and wideband radar. Blanking can make the difference between a good QSO and not even hearing the other station. * Noise Reduction We often have to deal with stochastic (“white”) noise on the bands. Our DSP-based NR can be adjusted over a wide range to improve intelligibility, especially of speech signals. * Audio Peaking Filter (APF) CW operators often have to copy signals right at or even below the noise floor. At times like this, our APF can dramatically improve copy without adding significant ringing. APF uses a special filter shape for this purpose, with a 30-Hz-wide peak and broad, low-delay skirts. * Text Decode/Encode DSP allows the KX2/KX3 to demodulate and display CW, RTTY, and PSK31/63 signals directly on the LCD (or in the case of the KX3, as several lines of text on the PX3 panadapter). These modes are also directly supported for transmit purposes, using a keyer paddle (a keyboard can also be used, in conjunction with KX2/3 Utility and a notebook computer). * Receive and Transmit EQ 8-band EQ is provided to allow the operator to tailor both receive and transmit audio to meet special hearing requirements, optimize for use with particular microphones, etc. * CWT (CW/data tuning [auto-spot] and mini-spectral display) The DSP sends the MCU information abou
[Elecraft] DSP vs. non-DSP portable transceivers
At the Dayton (yeah, OK, Xenia) Hamvention, just before giving a demo of the KX2, a bystander asked me, “Why would you need DSP in a portable transceiver?” He went on to mention a few currently available rigs he was considering that didn’t have DSP, including one very new one [an “upgraded” legacy model]. I’ll leave the entire list to the reader’s imagination. There are two basic reasons to use DSP: 1. As the core of the transceiver’s architecture, to reduce size and weight and increase flexibility 2. To provide a number of important features that have typically been seen only in desktop radios Explaining #1 is always a bit of a challenge since it dives into theory. Here’s the short version: The receiver in both the KX2 and KX3 uses a high-level quadrature mixer that down-converts the RF signal to baseband (0 kHz) or to a low I.F. (8 kHz). The resulting I and Q signals are fed to a pair of high-performance audio A-to-D converters, and all demodulation is then handled in DSP code. This configuration has immense versatility, reduces component count, and provides very consistent unit-to-unit performance in parameters such as selectivity. To explain #2, I prefer to just dive into the demo. There's a long list of functions that directly benefit those who operate portable, as well as anyone trying to maximize their communications capabilities. Here are the DSP functions that I try to include in demos (given enough time), along with their benefits for the operator: * Multiple Modes The KX2 and KX3 both support SSB, CW, AM, FM, and several data modes, without the need for supplementary hardware. All demodulation is handled by the DSP. And if we add a new mode in the future, it will be available to all KX-line radio owners free of charge with a simple firmware upgrade. * Bandwidth Control The DSP bandwidth can be adjusted from 50 Hz to 4000 Hz (5000 Hz in AM mode). The passband can also be shifted. Rigs without DSP are limited to the passband of their analog filters (including crystal filters). * Stereo Audio Effects and Dual Receive Both rigs demodulate two independent audio channels, allowing for simulated stereo, which reduces listening fatigue. Stereo is also used when DUAL RX is enabled; this allows the operator to listen to one signal in the left ear (VFO A) and another in the right (VFO B). This is essential for use with DX stations operating split, and greatly facilitates hunt ’n’ pounce contest operation where one receiver is parked on a station to be worked, while the other is used to locate the next one. * Adjustable AGC The KX2/KX3 include user-configurable AGC. Threshold, attack, decay, and hold time can all be adjusted. These are parameters that would require dozens of components in a non-DSP design. There’s also a short-duration pulse detector (AGC PLS). When this is turned on--the default--the DSP will seamlessly remove such pulses before they can pin your S-meter and cause a long RX recovery delay. (Rob Sherwood complemented us on this feature, which he found was missing on even high-end desktop transceivers. So, for in-house purposes, we call it “Sherwood mode” :) * Noise Blanking Many non-DSP radios lack noise blanking, or have simple blanking circuitry that can be easily overloaded. The KX-line DSP implements very effective blanking that can take out narrow pulses even at very low settings, and greatly attenuate irregular, high-duty cycle noise sources typical of light dimmers, electric fences, and wideband radar. Blanking can make the difference between a good QSO and not even hearing the other station. * Noise Reduction We often have to deal with stochastic (“white”) noise on the bands. Our DSP-based NR can be adjusted over a wide range to improve intelligibility, especially of speech signals. * Audio Peaking Filter (APF) CW operators often have to copy signals right at or even below the noise floor. At times like this, our APF can dramatically improve copy without adding significant ringing. APF uses a special filter shape for this purpose, with a 30-Hz-wide peak and broad, low-delay skirts. * Text Decode/Encode DSP allows the KX2/KX3 to demodulate and display CW, RTTY, and PSK31/63 signals directly on the LCD (or in the case of the KX3, as several lines of text on the PX3 panadapter). These modes are also directly supported for transmit purposes, using a keyer paddle (a keyboard can also be used, in conjunction with KX2/3 Utility and a notebook computer). * Receive and Transmit EQ 8-band EQ is provided to allow the operator to tailor both receive and transmit audio to meet special hearing requirements, optimize for use with particular microphones, etc. * CWT (CW/data tuning [auto-spot] and mini-spectral display) The DSP sends the MCU information about how far off the nearest signal is from the user’s selected sidetone or mark tone frequency. We use this to enable quick, one-tap auto-spotting of signals, as well
