Keyboard Shortcuts
ctrl + shift + ? :
Show all keyboard shortcuts
ctrl + g :
Navigate to a group
ctrl + shift + f :
Find
ctrl + / :
Quick actions
esc to dismiss
Likes
Search
Low noise high reverse isolation low distortion RF Amplifier
|
Hi Jeff, Can you clarify a bit more what part confuses you? Bruce designed this amplifier to buffer (and slightly amplify) low phase noise frequency references and the goals for it are low residual/additive phase noise, low distortion, and high reverse isolation. The high reverse isolation is needed to prevent oscillator injection locking when performing frequency stability measurements. The 40 dB or so of reverse isolation is much better than the original Norton amplifier but is still quite a bit less than other amplifiers with similar purposes. If you look at the isolation amplifiers NIST has designed for their atomic clocks the reverse isolation is north of 100 dB. It¡¯s not too difficult to find an opamp with high reverse isolation, but it¡¯s harder to find one that also has very low additive phase noise. Matt
|
|
Jeff/Matt,
how would you instrument the measurement for reverse isolation? ?Perhaps an sig gen feeding the ¡°output¡± and then look at the ¡°input¡± on a SA? ?I ask because I¡¯d like to measure my distribution amp. While I have not measured additive phase noise on it a similar line-up in a tripler resulted in an ultralight low flicker design. thanks Steve, KC4SW |
|
Hi Steve, Yep that¡¯s right. If you want a sweep across frequencies typically a VNA or spectrum analyzer would be used. But for a single frequency (which is often all you need for a frequency reference buffer or distribution amp) that would work just fine. Another measurement you may want to perform is to measure the isolation between outputs. Ie inject a signal into one output and measure the strength at another output. Not that you asked, but additive phase noise is trickier. I use an HP 3048A setup for this. Matt On Sat, May 13, 2023 at 4:29 PM Stephen Hanselman <kc4sw.io@...> wrote: Jeff/Matt, |
|
Oh and if you want a much better explanation of what this amplifier is all about than I can give you, you might ask on the time-nuts mailing list. The amplifier designer (Bruce Griffiths) along with a number of other individuals who are very knowledgeable and have built this or similar amplifiers are on that list. Matt On Sat, May 13, 2023 at 3:54 PM Matt Huszagh <huszaghmatt@...> wrote:
|
|
Matt, |
|
I think that the coax cables, and connectors, should be investigated in isolation, Non-linearities in cables and connectors are a well-known phenomenon.
toggle quoted message
Show quoted text
73 Jeff G8HUL -----Original Message-----
From: [email protected] <[email protected]> On Behalf Of Jeff Green Sent: 14 May 2023 01:19 To: [email protected] Subject: Re: [Test Equipment Design & Construction] Low noise high reverse isolation low distortion RF Amplifier The best I can do is to refer you to this web page PDF, Dr. Lankford's concept of best small antennas. ,%20LW,%20And%20SW%20rev%202.pdf Here is some of the text" "While studying active whip intercepts some time ago I discovered, much to my amazement, that long coax (50 feet) lead often degrades 2nd order intercepts of active whip antennas by 20dB or more and degrades 3rd order intercepts of active whip antennas by up to 10dB, depending on the type of active whip antenna. I have not studied the cases of longer coax lead in, or long coax lead in used with active dipoles, or long coax lead in used with (passive) noise reducing antennas. For active whips long (50 feet) twin lead lead in does not change 2nd or 3rd order intercepts. Also, I have not studied the cases for longer twin lead lead in with active whips, or for twin lead lead in used with active dipoles or (passive) noise reducing antennas. I rather expect that coax lead in will be a loser with respect to intercepts in all of those cases, while twin lead lead in will be a winner with respect to intercepts in those cases. Of course, if your antenna is not in a high RF environment, then it probably won't matter if you use coax lead in. On the other hand, more recently I have found that coax lead in can cause substantial man made noise in active whip antennas compared to twin lead lead in. It appears that the coax induced noise is via common mode, but unfortunately the noise is virtually impossible to eliminate completely throughout the MW band even with multiple common mode chokes. I am beginning to understand more clearly why active whip antennas have such bad reputations wrt man made noise. All except mine use coax lead in as well as DC power feed." Consider how an active antenna operates, hi-Z input stage, buffer for output. If coax can degrade 2nd order intercept by 20dB and degrade 3rd order intercept by 10dB, then clearly something is going on. I don't know near enough if Dr. Lankford's observations and conclusions are correct. I've read most of Dr. Lankford's papers and it appears he was meticulous and his results are accurate...if only in his location. |
|
I agree with Jeff. Properly functioning cables and connectors are not nonlinear and thus cannot produce the effects that Lankford reports (assuming that he has achieved a proper match to eliminate the obvious length-dependent effects). A 20dB degradation in intercept sounds like the result of a tarnished connector to me.
toggle quoted message
Show quoted text
-- Tom -- Prof. Thomas H. Lee Allen Ctr., Rm. 205 420 Via Palou Mall Stanford University Stanford, CA 94305-4070 On 5/14/2023 01:42, G8HUL wrote:
I think that the coax cables, and connectors, should be investigated in isolation, Non-linearities in cables and connectors are a well-known phenomenon. |
¿ªÔÆÌåÓýIf an amplifier isn¡¯t prone to oscillation (with high power) that might damage a vector network analyzer measuring S21 is a direct measurement of isolation.
|
¿ªÔÆÌåÓýI think with problematic interfaces on cables, the non linear effects will probably be on the order of 90 db or more down.
Whether this will even be measurable will certainly depend on your noise floor and the nature of the signals being measured.
In the cellular industry, running watts to tens of watts, and having antennae in close proximity made it start to become a concern, but not until other aspects of receivers were good enough.
I couldn¡¯t say whether there were any vendors with ?sales agenda to promote replacement of cable runs rather than just swapping equipment at both ends of a cable run.
|
|
I would disagree with Tom regarding " Properly functioning cables and connectors" not being non-linear. There is always some non-linearity, it just depends on the level. Even the slightest amount of contact resistance in the braid or at connector interface will cause unwanted products, perhaps at an extremely low level but they will be there.
toggle quoted message
Show quoted text
There is a lot of info on the subject on the web, but here is one example: < > Regards Jeff G8HUL -----Original Message-----
From: [email protected] <[email protected]> On Behalf Of Tom Lee Sent: 14 May 2023 12:28 To: [email protected] Subject: Re: [Test Equipment Design & Construction] Low noise high reverse isolation low distortion RF Amplifier I agree with Jeff. Properly functioning cables and connectors are not nonlinear and thus cannot produce the effects that Lankford reports (assuming that he has achieved a proper match to eliminate the obvious length-dependent effects). A 20dB degradation in intercept sounds like the result of a tarnished connector to me. -- Tom -- Prof. Thomas H. Lee Allen Ctr., Rm. 205 420 Via Palou Mall Stanford University Stanford, CA 94305-4070 On 5/14/2023 01:42, G8HUL wrote: I think that the coax cables, and connectors, should be investigated in isolation, Non-linearities in cables and connectors are a well-known phenomenon. |
|
You inject the test signal into the output of the RF Amp and connect the RF Amp Input port to the test set detector. On network analyzers not providing automatic switching of input and outputs, one simply reverses the connections on the analyzer. So the network analyzer signal source port? is connected to the RF Amp output and the network analyzer detector port is connected to the RF Amp Input.? This is fairly straight forward with small signal amplifiers as the RF Amp output level is usually below the damage threshhold of the analyzer's signal source port if the RF Amp should take off and oscillate. With RF Amps that can provide output levels above a nominal? 0 dBm (50 Ohm system) you need to insert attenuators at the output port of the amplifier or use a rf coupler which will provide enough attenuation if the amp should decide to oscillate. Unless verified, do not assume any amplifier is Unconditionally stable. Nothing will ruin your day faster than burping your network analyzer. Chuck WD4HXG
|
|
Well, everything is nonlinear to sone degree, but that¡¯s pedantry and not useful engineering. We¡¯re looking at a claimed 20dB degradation in intercept. That¡¯s well beyond anything that can be attributed to a proper cable/connector system.
toggle quoted message
Show quoted text
Cheers Tom Sent from my iThing; please forgive the typos and brevity On May 14, 2023, at 09:34, G8HUL <g8hul@...> wrote: |
|
I would disagree with Tom regarding " Properly functioning cables and connectors" not being non-linear. There is always some non-linearity, it just depends on the level. Even the slightest amount of contact resistance in the braid or at connector interface will cause unwanted products, perhaps at an extremely low level but they will be there.
toggle quoted message
Show quoted text
There is a lot of info on the subject on the web, but here is one example: < > Regards Jeff G8HUL -----Original Message-----
From: [email protected] <[email protected]> On Behalf Of Tom Lee Sent: 14 May 2023 12:28 To: [email protected] Subject: Re: [Test Equipment Design & Construction] Low noise high reverse isolation low distortion RF Amplifier I agree with Jeff. Properly functioning cables and connectors are not nonlinear and thus cannot produce the effects that Lankford reports (assuming that he has achieved a proper match to eliminate the obvious length-dependent effects). A 20dB degradation in intercept sounds like the result of a tarnished connector to me. -- Tom -- Prof. Thomas H. Lee Allen Ctr., Rm. 205 420 Via Palou Mall Stanford University Stanford, CA 94305-4070 On 5/14/2023 01:42, G8HUL wrote: I think that the coax cables, and connectors, should be investigated in isolation, Non-linearities in cables and connectors are a well-known phenomenon. |
¿ªÔÆÌåÓýHi Jeff:The paper on Gore's web page is talking about high-power, multichannel systems. The Langford paper is talking about the second order intercept of an active antenna and the feed line that comes after it, i.e. very low power levels.? But details are missing, like what length of line and how the measurement was made. -- Have Fun, Brooke Clarke, N6GCE axioms: 1. The extent to which you can fix or improve something will be limited by how well you understand how it works. 2. Everybody, with no exceptions, holds false beliefs. -------- Original Message --------
I would disagree with Tom regarding " Properly functioning cables and connectors" not being non-linear. There is always some non-linearity, it just depends on the level. Even the slightest amount of contact resistance in the braid or at connector interface will cause unwanted products, perhaps at an extremely low level but they will be there. There is a lot of info on the subject on the web, but here is one example: Regards Jeff G8HUL -----Original Message----- From: [email protected] <[email protected]> On Behalf Of Tom Lee Sent: 14 May 2023 12:28 To: [email protected] Subject: Re: [Test Equipment Design & Construction] Low noise high reverse isolation low distortion RF Amplifier I agree with Jeff. Properly functioning cables and connectors are not nonlinear and thus cannot produce the effects that Lankford reports (assuming that he has achieved a proper match to eliminate the obvious length-dependent effects). A 20dB degradation in intercept sounds like the result of a tarnished connector to me. -- Tom -- Prof. Thomas H. Lee Allen Ctr., Rm. 205 420 Via Palou Mall Stanford University Stanford, CA 94305-4070 On 5/14/2023 01:42, G8HUL wrote:I think that the coax cables, and connectors, should be investigated in isolation, Non-linearities in cables and connectors are a well-known phenomenon. 73 Jeff G8HUL -----Original Message----- From: [email protected] <[email protected]> On Behalf Of Jeff Green Sent: 14 May 2023 01:19 To: [email protected] Subject: Re: [Test Equipment Design & Construction] Low noise high reverse isolation low distortion RF Amplifier The best I can do is to refer you to this web page PDF, Dr. Lankford's concept of best small antennas. ,%20LW,%20And%20SW%20rev%202.pdf Here is some of the text" "While studying active whip intercepts some time ago I discovered, much to my amazement, that long coax (50 feet) lead often degrades 2nd order intercepts of active whip antennas by 20dB |
|
In order to measure the IMD levels, unless the performance of the amplifier is terrible, it will have to be operated near it upper limit otherwise the IMD products will be so small as not to be visible on a spectrum analyser; so the signals concerned will not be that small, certainly in the range where very poor connector and cable non-linearities might play a part.
toggle quoted message
Show quoted text
Regards Jeff -----Original Message-----
From: [email protected] <[email protected]> On Behalf Of Brooke Clarke via groups.io Sent: 15 May 2023 20:07 To: [email protected] Subject: Re: [Test Equipment Design & Construction] Low noise high reverse isolation low distortion RF Amplifier Hi Jeff: The paper on Gore's web page is talking about high-power, multichannel systems. The Langford paper is talking about the second order intercept of an active antenna and the feed line that comes after it, i.e. very low power levels. But details are missing, like what length of line and how the measurement was made. -- Have Fun, Brooke Clarke, N6GCE axioms: 1. The extent to which you can fix or improve something will be limited by how well you understand how it works. 2. Everybody, with no exceptions, holds false beliefs. -------- Original Message -------- I would disagree with Tom regarding " Properly functioning cables and connectors" not being non-linear. There is always some non-linearity, it just depends on the level. Even the slightest amount of contact resistance in the braid or at connector interface will cause unwanted products, perhaps at an extremely low level but they will be there. There is a lot of info on the subject on the web, but here is one example: < > <> Regards Jeff G8HUL -----Original Message----- From: [email protected] <mailto:[email protected]> <[email protected]> <mailto:[email protected]> On Behalf Of Tom Lee Sent: 14 May 2023 12:28 To: [email protected] <mailto:[email protected]> Subject: Re: [Test Equipment Design & Construction] Low noise high reverse isolation low distortion RF Amplifier I agree with Jeff. Properly functioning cables and connectors are not nonlinear and thus cannot produce the effects that Lankford reports (assuming that he has achieved a proper match to eliminate the obvious length-dependent effects). A 20dB degradation in intercept sounds like the result of a tarnished connector to me. -- Tom -- Prof. Thomas H. Lee Allen Ctr., Rm. 205 420 Via Palou Mall Stanford University Stanford, CA 94305-4070 On 5/14/2023 01:42, G8HUL wrote: I think that the coax cables, and connectors, should be investigated in isolation, Non-linearities in cables and connectors are a well-known phenomenon. 73 Jeff G8HUL -----Original Message----- From: [email protected] <mailto:[email protected]> <[email protected]> <mailto:[email protected]> On Behalf Of Jeff Green Sent: 14 May 2023 01:19 To: [email protected] <mailto:[email protected]> Subject: Re: [Test Equipment Design & Construction] Low noise high reverse isolation low distortion RF Amplifier The best I can do is to refer you to this web page PDF, Dr. Lankford's concept of best small antennas. ,%20LW,%20And%20SW%20rev%202.pdf Here is some of the text" "While studying active whip intercepts some time ago I discovered, much to my amazement, that long coax (50 feet) lead often degrades 2nd order intercepts of active whip antennas by 20dB |
|
Once you determine the signal levels, from the amplifier sample port, as presented to the SpecAn, simulate that input with two isolated signal generators (Combined via 3dB hybrids and isolating pads) to make sure the measurement result is not dominated by Specan internal IM generation.
toggle quoted message
Show quoted text
Lester B Veenstra K1YCM M?YCM W8YCM 6Y6Y W8YCM/6Y 6Y8LV (Reformed USNSG CTM1) lester@... 452 Stable Ln Keyser WV 26726 USA GPS: 39.336826 N 78.982287 W (Google) GPS: 39.33682 N 78.9823741 W (GPSDO) Telephones: Home: +1-304-289-6057 US cell +1-304-790-9192 Jamaica cell: +1-876-456-8898 -----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of G8HUL Sent: Tuesday, May 16, 2023 3:57 AM To: [email protected] Subject: Re: [Test Equipment Design & Construction] Low noise high reverse isolation low distortion RF Amplifier In order to measure the IMD levels, unless the performance of the amplifier is terrible, it will have to be operated near it upper limit otherwise the IMD products will be so small as not to be visible on a spectrum analyser; so the signals concerned will not be that small, certainly in the range where very poor connector and cable non-linearities might play a part. Regards Jeff -----Original Message----- From: [email protected] <[email protected]> On Behalf Of Brooke Clarke via groups.io Sent: 15 May 2023 20:07 To: [email protected] Subject: Re: [Test Equipment Design & Construction] Low noise high reverse isolation low distortion RF Amplifier Hi Jeff: The paper on Gore's web page is talking about high-power, multichannel systems. The Langford paper is talking about the second order intercept of an active antenna and the feed line that comes after it, i.e. very low power levels. But details are missing, like what length of line and how the measurement was made. -- Have Fun, Brooke Clarke, N6GCE axioms: 1. The extent to which you can fix or improve something will be limited by how well you understand how it works. 2. Everybody, with no exceptions, holds false beliefs. -------- Original Message -------- I would disagree with Tom regarding " Properly functioning cables and connectors" not being non-linear. There is always some non-linearity, it just depends on the level. Even the slightest amount of contact resistance in the braid or at connector interface will cause unwanted products, perhaps at an extremely low level but they will be there. There is a lot of info on the subject on the web, but here is one example: < > <> Regards Jeff G8HUL -----Original Message----- From: [email protected] <mailto:[email protected]> <[email protected]> <mailto:[email protected]> On Behalf Of Tom Lee Sent: 14 May 2023 12:28 To: [email protected] <mailto:[email protected]> Subject: Re: [Test Equipment Design & Construction] Low noise high reverse isolation low distortion RF Amplifier I agree with Jeff. Properly functioning cables and connectors are not nonlinear and thus cannot produce the effects that Lankford reports (assuming that he has achieved a proper match to eliminate the obvious length-dependent effects). A 20dB degradation in intercept sounds like the result of a tarnished connector to me. -- Tom -- Prof. Thomas H. Lee Allen Ctr., Rm. 205 420 Via Palou Mall Stanford University Stanford, CA 94305-4070 On 5/14/2023 01:42, G8HUL wrote: I think that the coax cables, and connectors, should be investigated in isolation, Non-linearities in cables and connectors are a well-known phenomenon. 73 Jeff G8HUL -----Original Message----- From: [email protected] <mailto:[email protected]> <[email protected]> <mailto:[email protected]> On Behalf Of Jeff Green Sent: 14 May 2023 01:19 To: [email protected] <mailto:[email protected]> Subject: Re: [Test Equipment Design & Construction] Low noise high reverse isolation low distortion RF Amplifier The best I can do is to refer you to this web page PDF, Dr. Lankford's concept of best small antennas. ,%20LW,%20And%20SW%20rev%202.pdf Here is some of the text" "While studying active whip intercepts some time ago I discovered, much to my amazement, that long coax (50 feet) lead often degrades 2nd order intercepts of active whip antennas by 20dB |
to navigate to use esc to dismiss