Very interesting! I think the db difference is probably due variations in the preamp, maybe turn both off to see which is really more sensitive. But what is really weird is both had contacts that the other did not. I would think it would be 1 sided if there was a clear winner. That seems to point to something in the software, possibly a resource issue, or software glitch. I wonder how much it would differ if you ran 2 separate computers or pi's. For sure a head scratcher.... JMHO plus I didn't stay at Holiday Inn last night.....
Great concept and presentation. So a few observations: The use of a tee to isolate two receivers is not entirely valid. The input impedance of the two receivers is now in parallel with each other, and may be affecting one more than the other. Alternatively a power divider with matched ports combined with swapping the radios and from port 1 to port 2 and rerunning testing to eliminate a “best lane” drag race analogy. The 1db difference is a bit of a red herring since the measurement instruments are two independent, uncalibrated radios that could each be introducing their own errors. Frankly I’m pleasantly surprised they were within 1db, but actually don’t believe they are, or that they are representative. Without capturing the message traffic from the radios to the computer it’s a reach to say one radio “heard” something the other didn’t. A more fair statement would be that “the combination of radio, data transfer, and software stack” “heard” or more accurately “properly detected and decoded” different transmissions. The technique of bypassing CAT control also depends a lot on how the error handling of the application, operating system, and USB driver is coded. This could leave the radios “deaf” while the CAT control routines were “timing out” I wondered if the two instances running on these computer had been tested to show they had matching performance (like the “first one started” runs faster than the “second one launched”) I didn’t hear mention about the firmware versions of the radios in particular if you used the FT8 presets in the more recent 7300 firmware, and if the 991 has similar collections of settings to ensure you have incorporated “best practices” Sorry for the long list of pondering points, none are dismissive, all are for “consideration” 73 Rich
First let me say thank you for your considered, polite, and constructive discourse. It seems at times that the art of having a conversation is lost and then I see a comment like this one and my faith in humanity gets restored. I don't have a proper power divider so my testing is limited to the resources I can gather. If such a device is not terribly expensive I might be able to throw some $ at it and run the tests again. I had not considered parallelism on the input impedance but since your brought it up it makes sense that the feed lines would allow for one radio to affect the other. I wonder if putting in a diode in line on the center pin might solve the problem? There was no red herring intended with the 1db note. It is just what I observed in testing. You are correct in that the statement should be that the entire stack "heard". With that said the only difference in the setup is the radio. The cables are all the same, the software is all the same, the computer is the same. Therefor a level playing field in the testing environment. The software CAT control has nothing to do with the sound card input from the radio so the lack of it or the interrupt of it would have no effect on the software decodes. You can take the headphone output from the radio and connect it to your computer and run WSJT-X and tell it to use the sound card of your computer and it will decode digital modes. Firmwares - I should have made note of this. They are both running the latest firmware for each radio that I own (IC-7300 and FT-991A). The video with the X6100 does not have the latest firmware on the radio but since I was borrowing it I did not feel as though it would be my place to upgrade it to the latest version. The 7300 has a setting for FT8, the 991A does not. I matched as best as I could the settings on each otherwise to make it as close to a level playing field as I could. Again, thanks for the thoughtful feedback. 73, Carlos KD9OLN
Nice job Carlos! I guessed right and wish I could find an artical for you on the 991a recieve but was unable to find it for you. In the artical the author said that the first receive on the 991a is pulled back at the factory for some reason and I forget why. Amp 1 would be in line with the 7300 with the amp off on the 7300 and both about the same on amp2. If I can find the artical I will get it to you. It might prove the auther right if you did the test with the amp1 on with ft991a and off with the 7300. Not sure if the authur knew what he was talking about really.
Thanks very much. I have both rigs but have never tried the 991A on digital, use it for /P from vehicle, advantage to have VHF / UHF in one box for that. Very interesting presentation.
I was going to suggest testing against Xiegu :). (G90 or X6100). Certainly, the Japanese radios have a better receiver but it'd be interesting to see by how much. N5TAB
Let's make arrangements. My email address is my skydiver at my callsign dot com. I'd love to put the x6100 on the bench and compare it. Thank you very much for offering.
Interesting test and setup! One thing that I saw during the intro were the volume levels - when you switched from the 991A to the 7300, the bar shows a lot lower volume @4:58. This might mean nothing or maybe a little. Not sure. I know the 7300 can set an output volume - this might be a way to neutralize this. Other things, but that's not from your video - the filters, are they both set at 3 KHz as was visible on the 991A? The 7300 does know a sharp and soft filter. Something might be just in the filter of one and outside of the filter of the other. Keep up the good work!
So I guessed correctly. What's interesting is more noise = more signal, but that doesn't mean ssb would yield same results. I think the 7300 has the advantage in HF because it's HF only...makes sense... I think one must compare 7300 to the ft710 or dx10? Thanks Carlos, great experiment (used to be a stats prof, you did well).
Thanks! I'll put the 710 and the DX10 against it if I can borrow them. I am working on that so be on the look out. There is an X6100 that got to me by mail that is next on the bench.
Not sure. My take is that Sherwood uses a laboratory environment and uses pristine stuff. This is a real world heads up test so we are at the mercy of the activity of the bands and the band conditions. I think that this is a fair test of the radio as it would be operated.
Great analysis. Although, I kept waiting to hear the Stone Cold Steve Austin music and then have a Kenwood crash the party at the end, LOL.
If a Kenwood shows up I'll make a point of doing that. :)
You know Xiegu is the real winner here
Only one way to tell for sure - We need a sacrificial G90!
Very interesting! I think the db difference is probably due variations in the preamp, maybe turn both off to see which is really more sensitive. But what is really weird is both had contacts that the other did not. I would think it would be 1 sided if there was a clear winner. That seems to point to something in the software, possibly a resource issue, or software glitch. I wonder how much it would differ if you ran 2 separate computers or pi's. For sure a head scratcher.... JMHO plus I didn't stay at Holiday Inn last night.....
So what you are saying is I need a followup video comparing them without a PreAmp running. Got it.
Enjoyed the video, thanks for posting 👍
Now I wish I had a spare FT-710 to loan you…😁
73, Don W6DSG
You and me both! You can always buy a spare and send it to me. :)
Great concept and presentation.
So a few observations:
The use of a tee to isolate two receivers is not entirely valid. The input impedance of the two receivers is now in parallel with each other, and may be affecting one more than the other. Alternatively a power divider with matched ports combined with swapping the radios and from port 1 to port 2 and rerunning testing to eliminate a “best lane” drag race analogy.
The 1db difference is a bit of a red herring since the measurement instruments are two independent, uncalibrated radios that could each be introducing their own errors. Frankly I’m pleasantly surprised they were within 1db, but actually don’t believe they are, or that they are representative.
Without capturing the message traffic from the radios to the computer it’s a reach to say one radio “heard” something the other didn’t. A more fair statement would be that “the combination of radio, data transfer, and software stack” “heard” or more accurately “properly detected and decoded” different transmissions.
The technique of bypassing CAT control also depends a lot on how the error handling of the application, operating system, and USB driver is coded. This could leave the radios “deaf” while the CAT control routines were “timing out”
I wondered if the two instances running on these computer had been tested to show they had matching performance (like the “first one started” runs faster than the “second one launched”)
I didn’t hear mention about the firmware versions of the radios in particular if you used the FT8 presets in the more recent 7300 firmware, and if the 991 has similar collections of settings to ensure you have incorporated “best practices”
Sorry for the long list of pondering points, none are dismissive, all are for “consideration”
73
Rich
First let me say thank you for your considered, polite, and constructive discourse. It seems at times that the art of having a conversation is lost and then I see a comment like this one and my faith in humanity gets restored.
I don't have a proper power divider so my testing is limited to the resources I can gather. If such a device is not terribly expensive I might be able to throw some $ at it and run the tests again. I had not considered parallelism on the input impedance but since your brought it up it makes sense that the feed lines would allow for one radio to affect the other. I wonder if putting in a diode in line on the center pin might solve the problem?
There was no red herring intended with the 1db note. It is just what I observed in testing.
You are correct in that the statement should be that the entire stack "heard". With that said the only difference in the setup is the radio. The cables are all the same, the software is all the same, the computer is the same. Therefor a level playing field in the testing environment.
The software CAT control has nothing to do with the sound card input from the radio so the lack of it or the interrupt of it would have no effect on the software decodes. You can take the headphone output from the radio and connect it to your computer and run WSJT-X and tell it to use the sound card of your computer and it will decode digital modes.
Firmwares - I should have made note of this. They are both running the latest firmware for each radio that I own (IC-7300 and FT-991A). The video with the X6100 does not have the latest firmware on the radio but since I was borrowing it I did not feel as though it would be my place to upgrade it to the latest version. The 7300 has a setting for FT8, the 991A does not. I matched as best as I could the settings on each otherwise to make it as close to a level playing field as I could.
Again, thanks for the thoughtful feedback.
73,
Carlos
KD9OLN
Nice job Carlos! I guessed right and wish I could find an artical for you on the 991a recieve but was unable to find it for you. In the artical the author said that the first receive on the 991a is pulled back at the factory for some reason and I forget why. Amp 1 would be in line with the 7300 with the amp off on the 7300 and both about the same on amp2. If I can find the artical I will get it to you. It might prove the auther right if you did the test with the amp1 on with ft991a and off with the 7300. Not sure if the authur knew what he was talking about really.
It is worthy of a test to be sure. I'll see if I can squeeze it into a video.
I Googled how to make a boring presentation interesting... got different results :D.
Interesting video Carlos, well done!
Hindsight being 20/20 I should have used a different adjective in my search. Thanks Hayden, glad you liked it.
Thanks very much. I have both rigs but have never tried the 991A on digital, use it for /P from vehicle, advantage to have VHF / UHF in one box for that. Very interesting presentation.
You are spot on. The ability to have HF/VHF/UHF in one box that goes full power is without question a massive advantage.
@@LifeAtTerminalVelocity I have an 857 as well which is a bit more compact but have been to the Dark side of nice displays 🙂
I have an 857 that I plan to put on the bench too. Only a two zebra lines on it so I'll count my lucky stars.
I love my 991-a but this hit me right in the feels.. lol thanks for the comparison!
I'm in the same boat. I really expected this to be much closer than it turned out to be. I still like both radios very much.
I was going to suggest testing against Xiegu :). (G90 or X6100). Certainly, the Japanese radios have a better receiver but it'd be interesting to see by how much. N5TAB
If you have one I can test let me know.
@@LifeAtTerminalVelocity I'll ship you my X6100 if you can get it back to me in a couple weeks.
Let's make arrangements. My email address is my skydiver at my callsign dot com. I'd love to put the x6100 on the bench and compare it. Thank you very much for offering.
Very interesting comparison between the two radios. Looking forward to more content from you. Subscribed.
Thanks for the sub!
Carlos great video and I really liked the way you explained your method and test parameters. I think the data is sound.
Thanks Don!
Carlos you did great on the video! I like this "amateur" approach to testing amateur radio 👍
Thanks Joe👍
Interesting test and setup!
One thing that I saw during the intro were the volume levels - when you switched from the 991A to the 7300, the bar shows a lot lower volume @4:58. This might mean nothing or maybe a little. Not sure. I know the 7300 can set an output volume - this might be a way to neutralize this.
Other things, but that's not from your video - the filters, are they both set at 3 KHz as was visible on the 991A?
The 7300 does know a sharp and soft filter. Something might be just in the filter of one and outside of the filter of the other.
Keep up the good work!
I got them as close to each other as possible. The bandwidth on the 7300 is set to 3kHz by using Filter 1 and selecting the "FT8" preset on the radio.
So I guessed correctly. What's interesting is more noise = more signal, but that doesn't mean ssb would yield same results. I think the 7300 has the advantage in HF because it's HF only...makes sense... I think one must compare 7300 to the ft710 or dx10? Thanks Carlos, great experiment (used to be a stats prof, you did well).
Thanks! I'll put the 710 and the DX10 against it if I can borrow them. I am working on that so be on the look out. There is an X6100 that got to me by mail that is next on the bench.
Check the comparative receiver on a QDX! :-)
I'm finishing the build of the QDX this week (I hope) so that is absolutely possible.
The real winner is Kyle!😆👍
Isn't he always though?
I think the Icom is operating dirty. Because Sherwood reported differently?
Not sure. My take is that Sherwood uses a laboratory environment and uses pristine stuff. This is a real world heads up test so we are at the mercy of the activity of the bands and the band conditions. I think that this is a fair test of the radio as it would be operated.