|
Administrator
|
Hi all,
With regard to SM5BSZ's SDR wideband-noise performance table (http://www.sm5bsz.com/dynrange/dubus313.pdf): The reason the K3 has excellent wideband transmit noise suppression is that it is a down-conversion superhet with a very clean synthesizer and transmit chain. This was intentional. We had several customers in our K3 focus group--VIP contesters and DXers--who were within line-of-sight of each other's QTH's, up on a ridge. They were tired of hearing their respective phase noise when they all happened to be on the same band. They asked us to do something about this, and we accepted the challenge. To ensure as clean a signal from the transmit chain (DSP/DAC/TX mixer) as possible, we routed it through the same crystal filters used by the receiver, as well as through an additional 4-pole crystal filter that immediately follows the transmit mixer. The two filters in series do a great job of band-limiting transmit phase noise before it even gets to the mixer. While this basic architecture is used by other superhet transceivers, they may not have as clean a synthesizer as the K3, or they may lack the second filter in series with the transmit path, or they may have such complex conversion schemes that noise is allowed to leak in at various points. In some cases they use a DDS without a following PLL/VCO, or their VCO may be operating a high kHz/volt tuning ratio, leaving it susceptible to higher phase noise, especially on the higher bands. We went to extremes to keep the synthesizer's kHz/volt ratio as low as possible, with a combination of switchable hi-Q inductance (1 uH/0.25 uH) and lots of switchable fixed capacitors. The inductors and larger capacitors are switched with relays to eliminate any loss of Q in the VCO's tank circuit. The smaller capacitors are switched with PIN diodes, but because they're a small part of the total C, this has a negligible effect. The KX3 is no slouch, either, falling about in the middle of this particular evaluation table. But the KX3 doesn't have the benefit of routing its TX signal through crystal filters (there are no crystal filters in a "pure" SDR), so the noise performance is determined almost entirely by the transmit DAC and its following amplifiers. We made these as clean as possible within budgetary constraints. We could have used a much more expensive DAC and added more shielding and filtering, but that would have pushed the radio to a higher size, cost, and current drain than we wanted. It's nice to have further confirmation that, despite the relative simplicity of the KX3's transmit architecture, it still comes in ahead of many radios touted for their sophistication and high performance. 73, Wayne N6KR On Dec 27, 2013, at 5:00 PM, "Bruce Beford" <[hidden email]> wrote: > > You're not missing anything. It shows how good the receiver is. ('better' than the transmitter). It highlights the difference in performance between the receiver and the transmitter. But... does having a "way better" receiver than a transmitter necessarily make it the "best all around transceiver"? I don't believe so. > > > I still believe (personally) that my K3 is a better all around transceiver than my KX3. But- they each have their strengths. > > 73, > > Bruce N1RX ______________________________________________________________ Elecraft mailing list Home: http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/mmfaq.htm Post: mailto:[hidden email] This list hosted by: http://www.qsl.net Please help support this email list: http://www.qsl.net/donate.html |
|
I always get nice comments on my transmit signal using either K3 or
KX3, thanks Elecraft! The attention to low phase-noise in the LO was one of the primary reasons I chose the K3 to be the IF for my VHF/UHF/mw transverters. Since I am in the pursuit of "very" weak-signals this it highly important to receiver sensitivity, especially on eme. I can give an "A" report card on the K3. I have yet to use the KX3 very much on VHF but will once the 2M module in installed. The KX3 is my main 6m monitor. Nice the transmitter is equally as good. 73, Ed - KL7UW http://www.kl7uw.com "Kits made by KL7UW" Dubus Mag business: [hidden email] ______________________________________________________________ Elecraft mailing list Home: http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/mmfaq.htm Post: mailto:[hidden email] This list hosted by: http://www.qsl.net Please help support this email list: http://www.qsl.net/donate.html |
|
In reply to this post by wayne burdick
> > With regard to SM5BSZ's SDR wideband-noise performance table (http://www.sm5bsz.com/dynrange/dubus313.pdf): Something doesn't jive with the table in the above report. I spot checked by looking up the Flex5000A in July 2008 QST, fig 3. QST shows wideband noise never gets worse than -120dbc, although the report says it is -77 at 350 KHz. And (as I remember) the K3 report from QST showed phase noise way out at the extremes of the meausrement dropping off much more than this author reported, down into the range of -155 db. Am I reading something wrong? Apparently this report is not recent because the worst performer of late (from QST measurements) has been the FT3000. Too bad SM5BSZ didn't comment on it. Rick K2XT ______________________________________________________________ Elecraft mailing list Home: http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/mmfaq.htm Post: mailto:[hidden email] This list hosted by: http://www.qsl.net Please help support this email list: http://www.qsl.net/donate.html |
|
Yes I was confused by that too, it appears that Leif has adjusted the TX noise figures from the QST dBc/Hz to a bandwidth of 500 Hz which is what QST use for RX measurements.
Leif has not actually measured the K3 or KX3, just used the QST results (the graph of "composite noise" and converted them for his purpose of showing differences between receiver and transmitter noise. He has measured the particular brand that exhibited very poor noise sidebands on transmit. David Anderson GM4JJJ > On 30 Dec 2013, at 14:34, Rick Stealey <[hidden email]> wrote: > > >> >> With regard to SM5BSZ's SDR wideband-noise performance table (http://www.sm5bsz.com/dynrange/dubus313.pdf): > > > Something doesn't jive with the table in the above report. > I spot checked by looking up the Flex5000A in July 2008 QST, fig 3. > QST shows wideband noise never gets worse than -120dbc, although the report says it is -77 at 350 KHz. > And (as I remember) the K3 report from QST showed phase noise way out at the extremes of the meausrement dropping off much more than this author reported, down into the range of -155 db. > > Am I reading something wrong? Apparently this report is not recent because the worst performer of late (from QST measurements) has been the FT3000. Too bad SM5BSZ didn't comment on it. > > Rick K2XT > > ______________________________________________________________ > Elecraft mailing list > Home: http://mailman.qth.net/mailman/listinfo/elecraft > Help: http://mailman.qth.net/mmfaq.htm > Post: mailto:[hidden email] > > This list hosted by: http://www.qsl.net > Please help support this email list: http://www.qsl.net/donate.html Elecraft mailing list Home: http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/mmfaq.htm Post: mailto:[hidden email] This list hosted by: http://www.qsl.net Please help support this email list: http://www.qsl.net/donate.html |
|
I agree after rereading all of his post and charts twice they seem to be
flawed, again by someone without a real radio to test. He does seem somewhat confused about the measurements of the QST labs and that of Rob Sherwood both of which have had hands on testing. IMHO only, a Flex is a nice bench setup but any SDR radio needs to operate without a computer hooked to it all the time to function. I have not owned one but have considered a 1500 if I find one reasonably priced. 73, Fred/N0AZZ K3 Ser # 6730--KX3 # 5210--K2/100 # 6470-KAT100 P3/SVGA--KAT500--W2 Amps Elecraft KPA500 HF/6m--Alpha's 9500 HF--87A HF--Mirage B-5030-G 300+w--(2) B-5016-G's 165w 2m -----Original Message----- From: [hidden email] [mailto:[hidden email]] On Behalf Of David Anderson Sent: Monday, December 30, 2013 8:52 AM To: Rick Stealey Cc: [hidden email] Subject: Re: [Elecraft] [KX3] SDR transmitter issues? Yes I was confused by that too, it appears that Leif has adjusted the TX noise figures from the QST dBc/Hz to a bandwidth of 500 Hz which is what QST use for RX measurements. Leif has not actually measured the K3 or KX3, just used the QST results (the graph of "composite noise" and converted them for his purpose of showing differences between receiver and transmitter noise. He has measured the particular brand that exhibited very poor noise sidebands on transmit. David Anderson GM4JJJ > On 30 Dec 2013, at 14:34, Rick Stealey <[hidden email]> wrote: > > >> >> With regard to SM5BSZ's SDR wideband-noise performance table (http://www.sm5bsz.com/dynrange/dubus313.pdf): > > > Something doesn't jive with the table in the above report. > I spot checked by looking up the Flex5000A in July 2008 QST, fig 3. > QST shows wideband noise never gets worse than -120dbc, although the report says it is -77 at 350 KHz. > And (as I remember) the K3 report from QST showed phase noise way out at the extremes of the meausrement dropping off much more than this author reported, down into the range of -155 db. > > Am I reading something wrong? Apparently this report is not recent because the worst performer of late (from QST measurements) has been the FT3000. Too bad SM5BSZ didn't comment on it. > > Rick K2XT > > ______________________________________________________________ > Elecraft mailing list > Home: http://mailman.qth.net/mailman/listinfo/elecraft > Help: http://mailman.qth.net/mmfaq.htm > Post: mailto:[hidden email] > > This list hosted by: http://www.qsl.net Please help support this email > list: http://www.qsl.net/donate.html Elecraft mailing list Home: http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/mmfaq.htm Post: mailto:[hidden email] This list hosted by: http://www.qsl.net Please help support this email list: http://www.qsl.net/donate.html ______________________________________________________________ Elecraft mailing list Home: http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/mmfaq.htm Post: mailto:[hidden email] This list hosted by: http://www.qsl.net Please help support this email list: http://www.qsl.net/donate.html |
|
In reply to this post by Rick Stealey
Although it is not shown on the graph nor in the text, the ARRL report shows the noise in dBc/Hz, i.e. a bandwidth of 1 Hz. This must be adjusted to the appropriate bandwidth. Add 27 dB to get to 500 Hz and about 33 dB to get to a typical SSB bandwidth. And the worst noise is about -110 dBc/Hz out about 350 kHz from the carrier. Add 33 to that and you get -77 dBc as shown by SM5BSZ. And the one poster the other day who claimed his Flex 1500 did not have have this issue is incorrect. According to the ARRL lab reports *ALL* Flex models are afflicted. 73 de dave ab9ca/4 On 12/30/13 8:34 AM, Rick Stealey wrote: > >> >> With regard to SM5BSZ's SDR wideband-noise performance table (http://www.sm5bsz.com/dynrange/dubus313.pdf): > > > Something doesn't jive with the table in the above report. > I spot checked by looking up the Flex5000A in July 2008 QST, fig 3. > QST shows wideband noise never gets worse than -120dbc, although the report says it is -77 at 350 KHz. > And (as I remember) the K3 report from QST showed phase noise way out at the extremes of the meausrement dropping off much more than this author reported, down into the range of -155 db. > > Am I reading something wrong? Apparently this report is not recent because the worst performer of late (from QST measurements) has been the FT3000. Too bad SM5BSZ didn't comment on it. > > Rick K2XT > > ______________________________________________________________ > Elecraft mailing list > Home: http://mailman.qth.net/mailman/listinfo/elecraft > Help: http://mailman.qth.net/mmfaq.htm > Post: mailto:[hidden email] > > This list hosted by: http://www.qsl.net > Please help support this email list: http://www.qsl.net/donate.html > . > Elecraft mailing list Home: http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/mmfaq.htm Post: mailto:[hidden email] This list hosted by: http://www.qsl.net Please help support this email list: http://www.qsl.net/donate.html |
|
In reply to this post by Fred Smith-2
On Mon, Dec 30, 2013 at 10:01 AM, Fred Smith <[hidden email]>wrote:
> ...have considered a 1500.. ========== If you operate CW, you won't like it. Get a KX3. Tony KT0NY ______________________________________________________________ Elecraft mailing list Home: http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/mmfaq.htm Post: mailto:[hidden email] This list hosted by: http://www.qsl.net Please help support this email list: http://www.qsl.net/donate.html |
| Free forum by Nabble | Edit this page |
