Hello to all,
I suppose that since most never hear themselves on-the-air, and darn few ever hear what their own transmitter does on freqs +/- a few khz (or worse +/- 10 to 20khz), the actual transmit IMD and spectral purity of our signals tend to get over-looked...
and that is a shame!But, the simple fact is that the limiting factors in most of our HF
receivers these days are:
a) The
transmit products (IMD for SSB and digital modes and wide-band noise/transmitter phase-noise, primarily effecting CW) of all the other stations on-the-air....
b) Local noise levels / RFI (caused by all the misc RF radiating products around us these days)
The limiting factors are NOT how good the 3rd-order IMD spec is on your receiver...although in some very rare instances (
probably < 0.1% of hams), this can be a factor, but even then not the only one...
I hope this info here helps some of my fellow hams understand what our modern HF rigs are doing to pollute the airwaves, even if they're operated with "good amateur practice", in accordance with the factory operations manual, etc...(now, if you crank-up the mic gain on most rigs, things will get worse....and many times, they'll get really bad....but, if you start with a bad rig, things get really bad very quickly...and if you start with a good rig, things might be "okay")
Almost 4 years since I started this thread, and after three years of inactivity, I'm sure there will be one or two hams here that will not appreciate bringing this thread back to life....but, there are a few reasons that I'm doing this.
1) The basic reasons are:
a) Tornado spawned by Hurricane Irma took down two of my 3 large trees (and a few smaller ones) and all my wire antennas supported there, as well as destroyed my rotor and damaged antennas on my tower (but I do have a new tower on the ground)...so I'm considering a station rebuild and was thinking about buying a new rig as well....
Updating my research....and then during some recent discussions with friends about transmit IMD, I found them repeating the oft-heard mantra of "high-voltage SSPA's have clean transmit signals", but here (like many things in life) this myth just persists.

b) New radios on the market, many of them costing 1000's of $$$$, and only the Apache Labs (ANAN) software pre-distortion equipped SDR transceivers have significantly improved transmit IMD. (sad, but true!)
c) A clarification of sorts showing that 2-tone testing is actually very representative...and that white noise testing can make explaining results easier to laypersons, as the effects are easier to show on the analyzer screen, using white noise....(although, dynamic testing can show deficiencies of power supplies, etc. that static two-tone testing / white noise testing do not show, 2-tone testing DOES show what is happening! and 3-tone tests from VHF/UHF TV-broadcast amplifiers that I've looked at years ago, also correlated well to 2-tone testing...)
From one of Rob Sherwood's papers, here is a 2-tone test overlaid with a white noise test (of his Icom IC-781)...
Have a look:

{BTW, during my discussions with friends, one asked why I didn't just post all the IMD specs of all the rigs??? Well, that's a LOT of data to sort thru, and I don't have the time for that...but then thought, I could do so (like I did for HF Amps) for at least a dozen or so rigs....Which ones?? I took suggestions, and also had some in mind that I might want to buy.}
For IMD comparisons of HF Amps, have a look here:
https://www.eham.net/ehamforum/smf/index.php/topic,100600.0.html2) Those that just want the brief nitty-gritty...here 'ya go:
a) Even now in 2018, there is still no direct correlation that "12 volt PA's splatter, and hi-voltage PA's are clean"....yes, some hi-voltage PA rigs produce better IMD figures than some 12 volt PA rigs, but it is not a direct correlation...
b) With the exception of the "pre-distortion" enabled Apache Labs ANAN rigs, the Icom M-802 HF Marine/Ham Transceiver still produces better IMD figures (by a significant amount) than ALL current HF ham rigs (or any made in the past 40+ years), and the Icom is a 12vdc rig, using a simple 12vdc PA!
http://www.eham.net/reviews/detail/6205But, some minor improvements from Kenwood with their TS-590SG (at ~ $ 1400 USD) and for those with hefty wallets their TS-990S (at ~ $7000 USD), and those who want to take out a second mortgage to buy a rig might look at the IC7851 (at ~ $14,000 USD!!!)...these might allow some like me looking for "a real radio with knobs" to at least not pollute the airwaves too much....BUT..
I still wonder why Icom (and others) just won't make a decent transmitter and PA...heck the Icom M-802 sells for ~ $1800....but it appears that most hams are really getting shafted with crappy transmitters!
c) This discussion is about transmit IMD and spectral purity, not about receivers....but many seem to be picking rigs based on their position on a list (assuming they can afford them)...I'm still surprised by the focus on "Sherwood's List"....now Rob Sherwood is a GREAT guy!! A GREAT ham! And, has done wonders for my favorite brand of HF radio, R.L. Drake!! I love the guy!

But, geeze guys...."the list" isn't the "be all / end all" of how to judge HF rigs....and if you don't believe me, just read what Rob himself writes!! Please remember that these "85 to 90db radios" are specified for serious CW operations/pileups/contesting...and for SSB operations, even in pileups and SSB contesting, receivers with 15 to 20db worse close-spaced IMD specs are adequate, and these other factors (mentioned in the quotes below) are important!

Quotes from Rob [I added some
bold type for emphasis]:
{What level of performance do we need close-in on CW for a radio to perform well most of the time? I think 85 dB will suffice most of the time. Certainly one may want a 100 dB dynamic range radio, but other factors of a transceiver’s performance are very important, too. Ten-Tec receive audio is better (cleaner) than Elecraft K3 audio, for instance.
On SSB transmitted intermodulation products from QRM 3 to 5 kHz away is usually far above the LO phase noise (RMDR) of today’s top receivers.
Thus the “holy grail” of wanting a 100 dB radio is only a CW pile-up issue.
If every other feature or specification of a radio was top notch, it would seem logical to pick a 100 dB radio over an 85 or 90 dB radio, however this is rarely the case. With good firmware, Ten-Tec and Elecraft have made their DSP radios much less susceptible to having the AGC “load up” or “over react” to impulse noise (clicks, tick and pops). No Japan, Inc. radio at the moment has figured this out.........
How a ham picks one transceiver over another is likely all over the map. For me if the ergonomics are poor, or if the receive audio is fatiguing to listen to, then that radio falls off my selection list. At the end of the day, hopefully whatever we buy we enjoy using.
I sold an expensive radio about 10 years ago that worked OK, but I just didn’t like it compared to my 15 year old radio of the same brand.
73, Rob, NC0B
(15-JUL-2014) }
{ DO NOT OVER-RATE DYNAMIC RANGE
One thing that I need to stress is [that] the amateur community has become obsessed with which radio has a close-in dynamic range a few dB higher than another.
Decades ago Tom Rauch W8JI and I were saying that a close-in dynamic range (DR3) of 80 dB would perform well most of the time. Back then with all the up-conversion radios on the market, most had a DR3 around 70, with some in the 60s.
The Orion I was the first commercial radio to go back to what we now call "down conversion" and it had a DR3 over 90 dB. The amateur community now has a good selection of radios that have a DR3 value in excess of 85 dB.
It is rare that an 85 dB radio will not be adequate in a CW pileup. [and from other papers/talks: a ~ "70db radio" works well, in SSB service/pileups/contesting.]
All the T-T products except the Omni-VII are 90 dB or better, along with products from most of the other OEMs.
Once you have decided you want a 90 or better radio, for example, then there are lots of other important parameters to consider, such as:
• clean receive audio for low fatigue in a contest,
• a good AGC (which T-T has recently improved in respect to handling impulse noise),
• the ease of use (user interface),
• reliability,
• quality of service and long-term parts support,
• long term firmware support,
• lack of ALC overshoot, (a problem with the TS-590S and IC-7410 for example) when driving a linear.
• The list goes on.
You don't buy a car with one specification, like horse power. The 427 cubic inch Corvette from years ago had lots of horse power but didn't handle very well!
73, Rob, NC0B
(16-DEC-2013) }
{ THE BIG PICTURE:
With 10 or more radios with 85 to 105 dB DR3s at 2 kHz, it is time to look at the big picture:
Clean receive audio,
clean transmitter IMD,
a good AGC that doesn’t go nuts over an impulse click, tick or pop,
a reliable radio that doesn’t beak all the time,
an ALC that doesn’t overshoot and fault your linear amplifier, or even worse blow your amp.
And when service is needed, good and reasonably fast factory service.
How about long term support of radios out of production, such as on-going firmware updates and parts availability. (Unfortunately there is only so much anOEM can do about supplying out-of-production chips, PA transistors, LCD screens, etc.)
At the end of the day, do you enjoy using your radio on the air?
I hate to tell you how many radios I have had the good fortune to borrow and use on the air in contests that I really would not want to have to use on a daily basis. I sold one very expensive radio that just wasn’t a good fit for me. I took the money and put up two more towers and Yagi antennas.
Hopefully when you turn off your rig after a contest, a DX contact or just a rag chew, you feel that “that was fun”. This is a hobby, and using your radio should be enjoyable.
73, Rob, NC0B
(21-DEC-2013) }
3) Okay... Now, let's look at some modern rigs....and some of my favorites...
These are 3rd, 5th, 7th and 9th order transmit IMD figures, in db below PEP, with the rigs operated at the spec'd max output (typically 100 watts) unless otherwise noted.
{FYI, unless otherwise specifically notated, all of these Transmit 2-tone IMD test figures are from published ARRL Product Review Tests...in some circumstances they found an "outlier" (worst band) and also listed "typical" as well as this "worst case", and in those cases, this data is also included...(also note that some rigs may only have old spectral scans and/or data listed for 3rd and 5th order IMD, but every effort is made to provide as much real data as I can.) }
[Please note the asterisks, etc....as these are ones I have personally owned (***), have used extensively(**), used briefly and liked(*).....and/or interested in(+)....also please note
I use italics to highlight comment/opinion, everything else is factual data.]
+ Apache Labs ANAN-8000D [no ARRL test yet, so these figures are from Adam Farson, VA7OJ/AB4OJ, from Sept 2017] (50vdc LDMOS PA)
33 / 40 / 54 / 60 (with NO pre-distortion) at 200 watts
67 / 70 / 70 / 70 (with pre-distortion) at 200 watts
Sorry, I cannot copy the analyzer spectrum scan of the ANAN-8000D, as this is proprietary data of Adam's....but you can have a look for yourself on page #33 of this page
http://www.ab4oj.com/sdr/apache/anan8000dle_notes.pdf+ Apache Labs ANAN-100D (12vdc PA)
29 / 35 / 39 / 44 (worst case, NO pre-distortion)
38 / 38 / 44 / 52 (typical, NO pre-distortion)
49 / 56 / 59 / 60 (worst case, with pre-distortion)
52 / 54 / 60+ / 60+ (typical, with pre-distortion)
*** Icom M-802 (12vdc PA) {nice HF marine and ham rig...few "user" adjustments, but a real nice rig w/ GREAT transmit IMD!}
47 / 50 / 58 / 60 (at 150 watts) {3 to 6db better at 100 watts}
Have a look:

*** Drake TR-7 (12vdc PA) {although not the cleanest IMD, one of my favorite rigs of all time, sweet SSB audio xmt and rx, and no modern ALC issues....one of the best SSB receivers (and darn good CW as well) in a ham rig, ever!}
38 / 44 / 51 / 58 (at 100 watts)
33 / 38 / 45 / 50 (at factory spec'd 150 watts)
** Kenwood TS-830s (vacuum tube PA, two 6146's) {sweet rig all-around, my 2nd favorite}
32 / 52 / 70 / 70+
Collins 32S-3 (vacuum tube PA, two 6146's)
42 / 53 / 65 / 76
Have a look:

** Icom IC-781 (28vdc PA) 150-watt out {not the cleanest IMD, but a great SSB rig, both xmt and rx}
37 / 39 / 47 / 50 (old ARRL test) at 150 watts out
35 / 43 / 49 / 54 (from Rob Sherwood's paper)
Have a look:
 {keep ALC to zero, and this is fine Class A xmtr}<br /><br />27 / 48 / 52 / 50 in Class B<br /><br />45 / 64 / 72 / 80 in Class A (ARRL test)<br /><br />48 / 74 / 82 / 86 in Class A (from Rob Sherwood's paper)<br /><br />Have a look:<br />[img]http://i63.tinypic.com/20rmfbp.jpg)
* Yaesu FT-102 (vacuum tube PA / three 6146b's)
40 / 40 / 50 / 60 (at 150 watts out)
** Icom IC-765 (12vdc PA) {a great rig in it's day, and darn good even by 2018's standards...used the 765 and 761 for many hours each day for months on end, 6 - 7 days/wk, at MARS station in 1990/91...}
40 / 45 / 48 / 50
** Icom IC-761 (12vdc PA) {a great rig in it's day, and darn good even by 2018's standards...used the 765 and 761 for many hours each day for months on end, 6 - 7 days/wk, at MARS station in 1990/91..}
31 / 45 / 48 / 50
* Icom IC-756ProII (12vdc PA)
30 / 40 / 48 / 50
+ Kenwood TS-590SG (12vdc PA)
31 / 38 / 48 / 55 (worst case)
42 / 38 / 48 / 58 (typical)
+ Kenwood TS-990 (50vdc PA)
31 / 46 / 52 / 57 (worst case)
39 / 46 / 54 / 56 (typical)
The rigs above are ones that I've used / been interested in...
Note the asterisks above, as these are ones I have personally owned (***), have used extensively(**), used briefly and liked(*).....and/or interested in(+)
The rest below were all suggested by some friends...and/or prominently discussed by many hams.
Elecraft K3 (12vdc PA)
33 / 40 / 48 / 50
Have a look:

Elecraft K3s (12vdc PA)
30 / 38 / 41 / 48 (typical)
35 / 36 / 48 / 62 (specially tuned transmitter and PA, by Elecraft engineering dept, after the above disappointing results were reported, prior to publication, see Nov 2016, QST, page 50, for details)
Icom IC-7300 (12vdc PA)
30 / 37 / 44 / 58
Icom IC-7851 (50vdc PA)
29 / 40 / 48 / 70 (worst case)
36 / 52 / 49 / 61 (typical)
Icom IC-7800 (50vdc PA)
32 / 49 / 52 / 52
Icom IC-7600 (12vdc PA)
31 / 35 / 41 / 48
Hilberling PT-8000A (50vdc PA)
29 / 41 / 50 / 52 (worse case)
35 / 48 / 54 / 59 (typical)
Icom IC-706 (12vdc PA)
33 / 31 / 38 / 44
Icom IC-706mkIIG (12vdc PA)
30 / 33 / 37 / 43
Yaesu FT-857 (12vdc PA) {from Aug '03 QST}
25 / 40 / 50 / 52
Yaesu FT-857D (12vdc PA) {from Nov '04 QST}
21 / 32 (3rd and 5th order IMD was all that was reported)
Yaesu FT-897 (12vdc PA)
23 / 37 / 47 / 50
Icom IC-735 (12vdc PA)
33 / 39 / 43 / 47
Yaesu FTdx-5000 (50-vdc PA)
30 / 48 / 46 / 47 (class B)
43 / 64 / 68 / 72 (class A) [but, you must NOT have any ALC, just like the older FT-1000MP MkV, otherwise you're almost defeating the advantage of Class A operation]
Have a look at how the ALC screws things up, zero ALC vs. half-scale ALC:

Yaesu FT-dx-3000 (12vdc PA)
27 / 40 / 42 / 52 (worse case)
31 / 40 / 45 / 52 (typical)
Yaesu FT-991 (12vdc PA)
22 / 32 / 39 / 45 (worse case)
26 / 37 / 41 / 46 (typical)
Yaesu FTdx-1200 (12vdc PA)
32 / 35 / 42 / 50 (worse case)
37 / 38 / 44 / 50 (typical)
Kenwood TS-590S (12vdc PA) (un-modified units suffer from serious ALC-overshoot and other ALC issues)
29 / 32 / 42 / 52
Flex-5000 (12vdc PA)
34 / 40 / 48 / 54
Flex-6500 (12vdc PA)
32 / 43 / 48 / 51 (worse case)
39 / 42 / 49 / 55 (typical)
Flex-6700 (12vdc PA)
32 / 40 / 44 / 55
Ten Tec Orion (12vdc PA)
32 / 42 / 52 / 58
Ten Tec Orion II (12vdc PA)
28 / 42 / 44 / 52
Ten Tec Omin VI+ (12vdc PA)
26 / 43 / 45 / 54
And, a few 10-watt QRP rigs:
Elecraft K2 (12vdc PA) 10-watt QRP
29 / 40 / 50 / 62
Elecraft KX3 (12vdc PA) 10-watt QRP
30 / 40 / 51 / 55
Ten Tec Argonaut (12vdc PA) 10-watt QRP
30 / 36 / 47 / 51 (worse case)
37 / 42 / 58 / 59 (typical)
Flex-1500 (12vdc PA) 10-watt QRP
22 / 38 / 48 / 48
I hope this helps some understand what our modern HF rigs are doing to pollute the airwaves, even if they're operated with "good amateur practice", in accordance with the factory operations manual, etc...(now, if you crank-up the mic gain on most rigs, things will get worse....and many times, they'll get really bad....but, if you start with a bad rig, it's bad no matter what and things get really bad very quickly...and if you start with a good rig, things are better from the start and might be "okay")
The 40-odd radios above represent most of the popular HF rigs on the ham bands today and sold in the past 30 - 40 years, as well as a few that are very new (and some so obscenely expensive that few hams will ever operate one)....oh, and 25 of the above 40 rigs, also find themselves in the top 30 of the rigs listed by "narrow-spaced (2khz) 3rd-order rec IMD" on Rob Sherwood's list, maybe now some will see the red herring that buying a rig from a "list" can be?

73,
John, KA4WJA