Hi,
When discussing antenna analysers there's always a discussion on the issue of measuring the sign of X. For some years now I've carted round an article out of the Power Amplifier Handbook titled High-Speed Microprocessor Controlled Antenna Matching Unit, in which the author implemented a phase angle sensor which output a logic one for inductive phase angle and logic 0 for capacitive. The circuit didn't look like it was too complex to implement, so I've always wondered why such a circuit isn't implemented in every antenna analyser. Mike VK1KCK _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
I'd be interested in the article if you have it in digital form, Mike.
I implement a sign detection algorithm in the Plot program for my LP-100 Digital Vector wattmeter. It works pretty well most of the time, but I added editing of curve points in case it makes an obvious mistake so that the curve can be "fixed" after scanning. I can't do it in the box, because it requires a number of frequency related datapoints to work. This would be no problem in an antenna analyzer, since the box controls the frequency, of course. I think the Autek VA1 uses a dither scheme with their oscillator to produce the variable frequency, even though it appears you are only testing at one frequency. It's narrow, but seems accurate with simple components like caps and coils, and with antennas which have a simple series or parallel resonance. The problem comes in when measuring an antenna at the end of a random length of feedline, where the phase slope can reverse without changing sign. A typical example of this curve would be a dipole which uses a "broadbanding" network with a "W" shaped curve at resonance. A better method for an analyzer is to use a signal source with quadrature outputs, so that I & Q samples can be generated. Some of the more expensive analyzers and VNAs do this. 73, Larry N8LP Mike Walkington wrote: > Hi, > > When discussing antenna analysers there's always a discussion on the issue > of measuring the sign of X. For some years now I've carted round an article > out of the Power Amplifier Handbook titled High-Speed Microprocessor > Controlled Antenna Matching Unit, in which the author implemented a phase > angle sensor which output a logic one for inductive phase angle and logic 0 > for capacitive. The circuit didn't look like it was too complex to > implement, so I've always wondered why such a circuit isn't implemented in > every antenna analyser. > > Mike > VK1KCK > > _______________________________________________ > Elecraft mailing list > Post to: [hidden email] > You must be a subscriber to post to the list. > Subscriber Info (Addr. Change, sub, unsub etc.): > http://mailman.qth.net/mailman/listinfo/elecraft > > Help: http://mailman.qth.net/subscribers.htm > Elecraft web page: http://www.elecraft.com > > > Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
In reply to this post by Mike Walkington-3
It is not too simple to measure phase with a additional circuitry. However,
it is simple to move the frequency of the antenna analyzer and see the Reactance change, IF it is increasing with frequency it is inductive reactance, (+); if decreasing with frequency increase, it is capacitive reactance, (-). Once you get used to that, and make a little reminder note to paste on the analyzer case, you can do the check easily and remember which type of reactance change you are seeing re frequency. Stuart K5KVH _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
That won't work, Stuart. I wrestled with this for awhile when developing
my LP-100 Plot program. You also need to know which side of resonance you're on, and whether it's a series or parallel resonance... at a minimum. Take a look at the plot at www.telepostinc.com/Files/OCF_Z_PH2.jpg. This is a section of an off center fed dipole as measured by my HP 87510A VNA. If you take the absolute value of the phase data (ie, have it bounce off zero as it would with a meter that can't determine phase sign), then you will see the problem. If you look at the phase in this plot, starting at the left edge, and remembering that all the negative values have to be reversed to look like a meter without sign detection, by your definition, the reactance makes the following changes from left to right edge... C, L, C, L, C, L, C, L In reality, as the VNA shows, there are fewer changes. The sequence is... C, L, C, L, C Without the visual aid, it can be confusing, to say the least. In my Plot software, I take into account a number of factors, based on the slopes of the magnitude and phase of impedance. If you don't do this, you could easily be exactly wrong ;-) 73, Larry N8LP Stuart Rohre wrote: > It is not too simple to measure phase with a additional circuitry. However, > it is simple to move the frequency of the antenna analyzer and see the > Reactance change, IF it is increasing with frequency it is inductive > reactance, (+); if decreasing with frequency increase, it is capacitive > reactance, (-). > > Once you get used to that, and make a little reminder note to paste on the > analyzer case, you can do the check easily and remember which type of > reactance change you are seeing re frequency. > > Stuart > K5KVH > > > _______________________________________________ > Elecraft mailing list > Post to: [hidden email] > You must be a subscriber to post to the list. > Subscriber Info (Addr. Change, sub, unsub etc.): > http://mailman.qth.net/mailman/listinfo/elecraft > > Help: http://mailman.qth.net/subscribers.htm > Elecraft web page: http://www.elecraft.com > > > Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
In reply to this post by Stuart Rohre
> IF it is increasing with frequency it is inductive > reactance, (+); if decreasing with frequency increase, it is capacitive > reactance, (-). This is not quite true. The reactance change over freq is something like a sawtooth waveform. Assume we start at reactance=zero (resonance) and freq is increased. Reactance then increases and is inductive up to some max, then decreases (while still inductive) until it crosses zero and becomes capacitive. Note that it was decreasing and inductive. If we continue to increase the freq. the reactance continues capacitive and increases (while still capacitive and freq is increasing) until a max is reached where it begins decreasing. It then returns to zero and we begin the sequence over again. The reversal from inductive max to capacitive max happens quickly, but the reactance can be decreasing while inductive and increasing while capacitive. We have to know where we are in the cycle. 73 de dave ab9ca Stuart Rohre wrote: > It is not too simple to measure phase with a additional circuitry. However, > it is simple to move the frequency of the antenna analyzer and see the > Reactance change, IF it is increasing with frequency it is inductive > reactance, (+); if decreasing with frequency increase, it is capacitive > reactance, (-). > > Once you get used to that, and make a little reminder note to paste on the > analyzer case, you can do the check easily and remember which type of > reactance change you are seeing re frequency. > > Stuart > K5KVH > > > _______________________________________________ > Elecraft mailing list > Post to: [hidden email] > You must be a subscriber to post to the list. > Subscriber Info (Addr. Change, sub, unsub etc.): > http://mailman.qth.net/mailman/listinfo/elecraft > > Help: http://mailman.qth.net/subscribers.htm > Elecraft web page: http://www.elecraft.com > > > _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
I was referring to a narrowband case of examining a simple antenna to see if
it is long or short. One with well defined sharp peaks and valleys of impedance, and one principal resonance which you can estimate from the dipole physical length. Seems to work using my Autek or MFJ 269. I was not starting from a totally black box, or unknown case. Stuart K5KVH _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
In reply to this post by N8LP
Larry,
I see no inconsistency with Stuart's statement in your plot. Without looking at the specific phase angle - if the absolute magnitude of the reactance increases with increasing frequency, then the load is inductive, it it decreases, then the load is capacitive. The poles of the black curve on your plot do line up with the zero crossings for the phase. So if I look at any point on your plot, I see that the absolute value of the magnitude of the reactance will increase with increasing frequency anytime the load has positive reactance (inductance), and the absolute magnitude of the reactance will decrease with increasing frequency any time the load is capacitive. Yes, a simple (or complex) phase angle detector working without regard for the sign will not provide the information required, but 'dithering' the frequency and looking only at the relative magnitude of the reactance between 'dither' points can provide the sign of the reactance. 73, Don W3FPR > -----Original Message----- > > That won't work, Stuart. I wrestled with this for awhile when developing > my LP-100 Plot program. You also need to know which side of resonance > you're on, and whether it's a series or parallel resonance... at > a minimum. > > Take a look at the plot at www.telepostinc.com/Files/OCF_Z_PH2.jpg. This > is a section of an off center fed dipole as measured by my HP 87510A > VNA. If you take the absolute value of the phase data (ie, have it > bounce off zero as it would with a meter that can't determine phase > sign), then you will see the problem. If you look at the phase in this > plot, starting at the left edge, and remembering that all the negative > values have to be reversed to look like a meter without sign detection, > by your definition, the reactance makes the following changes from left > to right edge... > > C, L, C, L, C, L, C, L > > In reality, as the VNA shows, there are fewer changes. The sequence is... > > C, L, C, L, C > > Without the visual aid, it can be confusing, to say the least. > > In my Plot software, I take into account a number of factors, based on > the slopes of the magnitude and phase of impedance. If you don't do > this, you could easily be exactly wrong ;-) > > 73, > Larry N8LP > > > > Stuart Rohre wrote: > > It is not too simple to measure phase with a additional > circuitry. However, > > it is simple to move the frequency of the antenna analyzer and see the > > Reactance change, IF it is increasing with frequency it is inductive > > reactance, (+); if decreasing with frequency increase, it is capacitive > > reactance, (-). > > > > Once you get used to that, and make a little reminder note to > paste on the > > analyzer case, you can do the check easily and remember which type of > > reactance change you are seeing re frequency. > > > > Stuart > > K5KVH > > > No virus found in this outgoing message. Checked by AVG Free Edition. Version: 7.5.446 / Virus Database: 268.18.11/723 - Release Date: 3/15/2007 11:27 AM _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
Thanks Don, my work around has worked for simple antennas like dipoles,
loops, verticals, etc. His OCF antenna has an interesting sharp resonance which must be the principle one, but also a non harmonic resonance which is interesting. I had not looked at a horizontal OCF (off center feed) one before with my 269 nor Autek. But, I use the trick to evaluate new wires I hang with good success. Stuart K5KVH _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
In reply to this post by Don Wilhelm-3
Folks,
There is an error in the reply I sent last night (below) that is very important - it makes a big difference. When increasing or decreasing the frequency to determine the sign of the reactance: One must observe the increasing or decreasing of the IMPEDANCE (Z) and not just the reactance. I had erroneously stated the reactance, and Larry helped get my mind straightened out - make that impedance instead of reactance and it will be correct. In other words, when using an MFJ259B - look at the "Z" value when increasing or decreasing the frequency, and not the "X" value. The notes in the MFJ259B manual recommending how to determine the sign of the reactance are placed under the instructions for measuring Capacitance and Inductance, and they do state that one can look at the reactance value. While that works fine for a pure reactance, it does not work for complex impedances, where one must look at the magnitude of the IMPEDANCE (Z). In other words, switch the MFJ259B to IMPEDANCE MODE and do the 'frequency increase/decrease thing' before attempting to affix a sign to the value of the reactance parameter. 73, Don W3FPR > -----Original Message----- > > > Larry, > > I see no inconsistency with Stuart's statement in your plot. Without > looking at the specific phase angle - if the absolute magnitude of the > reactance increases with increasing frequency, then the load is inductive, > it it decreases, then the load is capacitive. > > The poles of the black curve on your plot do line up with the > zero crossings > for the phase. So if I look at any point on your plot, I see that the > absolute value of the magnitude of the reactance will increase with > increasing frequency anytime the load has positive reactance (inductance), > and the absolute magnitude of the reactance will decrease with increasing > frequency any time the load is capacitive. > > Yes, a simple (or complex) phase angle detector working without regard for > the sign will not provide the information required, but 'dithering' the > frequency and looking only at the relative magnitude of the reactance > between 'dither' points can provide the sign of the reactance. > > 73, > Don W3FPR > > > -----Original Message----- > > > > That won't work, Stuart. I wrestled with this for awhile when developing > > my LP-100 Plot program. You also need to know which side of resonance > > you're on, and whether it's a series or parallel resonance... at > > a minimum. > > > > Take a look at the plot at www.telepostinc.com/Files/OCF_Z_PH2.jpg. This > > is a section of an off center fed dipole as measured by my HP 87510A > > VNA. If you take the absolute value of the phase data (ie, have it > > bounce off zero as it would with a meter that can't determine phase > > sign), then you will see the problem. If you look at the phase in this > > plot, starting at the left edge, and remembering that all the negative > > values have to be reversed to look like a meter without sign detection, > > by your definition, the reactance makes the following changes from left > > to right edge... > > > > C, L, C, L, C, L, C, L > > > > In reality, as the VNA shows, there are fewer changes. The > sequence is... > > > > C, L, C, L, C > > > > Without the visual aid, it can be confusing, to say the least. > > > > In my Plot software, I take into account a number of factors, based on > > the slopes of the magnitude and phase of impedance. If you don't do > > this, you could easily be exactly wrong ;-) > > > > 73, > > Larry N8LP > > > > > > > > Stuart Rohre wrote: > > > It is not too simple to measure phase with a additional > > circuitry. However, > > > it is simple to move the frequency of the antenna analyzer and see the > > > Reactance change, IF it is increasing with frequency it is inductive > > > reactance, (+); if decreasing with frequency increase, it is > capacitive > > > reactance, (-). > > > > > > Once you get used to that, and make a little reminder note to > > paste on the > > > analyzer case, you can do the check easily and remember which type of > > > reactance change you are seeing re frequency. > > > > > > Stuart > > > K5KVH > > > > > > -- > No virus found in this outgoing message. > Checked by AVG Free Edition. > Version: 7.5.446 / Virus Database: 268.18.11/723 - Release Date: 3/15/2007 > 11:27 AM > > _______________________________________________ > Elecraft mailing list > Post to: [hidden email] > You must be a subscriber to post to the list. > Subscriber Info (Addr. Change, sub, unsub etc.): > http://mailman.qth.net/mailman/listinfo/elecraft > > Help: http://mailman.qth.net/subscribers.htm > Elecraft web page: http://www.elecraft.com > -- > No virus found in this incoming message. > Checked by AVG Free Edition. > Version: 7.5.446 / Virus Database: 268.18.11/723 - Release Date: > 3/15/2007 11:27 AM > No virus found in this outgoing message. Checked by AVG Free Edition. Version: 7.5.446 / Virus Database: 268.18.11/723 - Release Date: 3/15/2007 11:27 AM _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
Thanks for the clarification, Don. As I mentioned in my private email,
to eliminate all ambiguity, you really need to look at the slopes of both impedance magnitude and phase (or alternatively resistance and reactance). This will cover you for both sides of either series or parallel resonance. This is what I do with my Plot software. You could certainly do it in your head, but it's little much to remember. If you have an analyzer, like the MFJ, which displays both R and X, it might make sense to print a little "what if..." sequence on a laminated card and tape it to the back of the unit to help with going through the scenarios when calculating sign. Based on my algorithm, I could print up a sample for anyone interested. Or you could use the two plots I referenced in this discussion to make your own. Remember though, that my plots show correct sign, and you have to reverse all the negative values to arrive at the correct input data for an analyzer that doesn't show sign. 73, Larry N8LP Don Wilhelm wrote: > Folks, > > There is an error in the reply I sent last night (below) that is very > important - it makes a big difference. > > When increasing or decreasing the frequency to determine the sign of the > reactance: > One must observe the increasing or decreasing of the IMPEDANCE (Z) and not > just the reactance. > > I had erroneously stated the reactance, and Larry helped get my mind > straightened out - make that impedance instead of reactance and it will be > correct. In other words, when using an MFJ259B - look at the "Z" value when > increasing or decreasing the frequency, and not the "X" value. > > The notes in the MFJ259B manual recommending how to determine the sign of > the reactance are placed under the instructions for measuring Capacitance > and Inductance, and they do state that one can look at the reactance value. > While that works fine for a pure reactance, it does not work for complex > impedances, where one must look at the magnitude of the IMPEDANCE (Z). > > In other words, switch the MFJ259B to IMPEDANCE MODE and do the 'frequency > increase/decrease thing' before attempting to affix a sign to the value of > the reactance parameter. > > 73, > Don W3FPR > > >> -----Original Message----- >> >> >> Larry, >> >> I see no inconsistency with Stuart's statement in your plot. Without >> looking at the specific phase angle - if the absolute magnitude of the >> reactance increases with increasing frequency, then the load is inductive, >> it it decreases, then the load is capacitive. >> >> The poles of the black curve on your plot do line up with the >> zero crossings >> for the phase. So if I look at any point on your plot, I see that the >> absolute value of the magnitude of the reactance will increase with >> increasing frequency anytime the load has positive reactance (inductance), >> and the absolute magnitude of the reactance will decrease with increasing >> frequency any time the load is capacitive. >> >> Yes, a simple (or complex) phase angle detector working without regard for >> the sign will not provide the information required, but 'dithering' the >> frequency and looking only at the relative magnitude of the reactance >> between 'dither' points can provide the sign of the reactance. >> >> 73, >> Don W3FPR >> >> >>> -----Original Message----- >>> >>> That won't work, Stuart. I wrestled with this for awhile when developing >>> my LP-100 Plot program. You also need to know which side of resonance >>> you're on, and whether it's a series or parallel resonance... at >>> a minimum. >>> >>> Take a look at the plot at www.telepostinc.com/Files/OCF_Z_PH2.jpg. This >>> is a section of an off center fed dipole as measured by my HP 87510A >>> VNA. If you take the absolute value of the phase data (ie, have it >>> bounce off zero as it would with a meter that can't determine phase >>> sign), then you will see the problem. If you look at the phase in this >>> plot, starting at the left edge, and remembering that all the negative >>> values have to be reversed to look like a meter without sign detection, >>> by your definition, the reactance makes the following changes from left >>> to right edge... >>> >>> C, L, C, L, C, L, C, L >>> >>> In reality, as the VNA shows, there are fewer changes. The >>> >> sequence is... >> >>> C, L, C, L, C >>> >>> Without the visual aid, it can be confusing, to say the least. >>> >>> In my Plot software, I take into account a number of factors, based on >>> the slopes of the magnitude and phase of impedance. If you don't do >>> this, you could easily be exactly wrong ;-) >>> >>> 73, >>> Larry N8LP >>> >>> >>> >>> Stuart Rohre wrote: >>> >>>> It is not too simple to measure phase with a additional >>>> >>> circuitry. However, >>> >>>> it is simple to move the frequency of the antenna analyzer and see the >>>> Reactance change, IF it is increasing with frequency it is inductive >>>> reactance, (+); if decreasing with frequency increase, it is >>>> >> capacitive >> >>>> reactance, (-). >>>> >>>> Once you get used to that, and make a little reminder note to >>>> >>> paste on the >>> >>>> analyzer case, you can do the check easily and remember which type of >>>> reactance change you are seeing re frequency. >>>> >>>> Stuart >>>> K5KVH >>>> >>>> >> -- >> No virus found in this outgoing message. >> Checked by AVG Free Edition. >> Version: 7.5.446 / Virus Database: 268.18.11/723 - Release Date: 3/15/2007 >> 11:27 AM >> >> _______________________________________________ >> Elecraft mailing list >> Post to: [hidden email] >> You must be a subscriber to post to the list. >> Subscriber Info (Addr. Change, sub, unsub etc.): >> http://mailman.qth.net/mailman/listinfo/elecraft >> >> Help: http://mailman.qth.net/subscribers.htm >> Elecraft web page: http://www.elecraft.com >> -- >> No virus found in this incoming message. >> Checked by AVG Free Edition. >> Version: 7.5.446 / Virus Database: 268.18.11/723 - Release Date: >> 3/15/2007 11:27 AM >> >> > -- > No virus found in this outgoing message. > Checked by AVG Free Edition. > Version: 7.5.446 / Virus Database: 268.18.11/723 - Release Date: 3/15/2007 > 11:27 AM > > _______________________________________________ > Elecraft mailing list > Post to: [hidden email] > You must be a subscriber to post to the list. > Subscriber Info (Addr. Change, sub, unsub etc.): > http://mailman.qth.net/mailman/listinfo/elecraft > > Help: http://mailman.qth.net/subscribers.htm > Elecraft web page: http://www.elecraft.com > > > Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
In reply to this post by Don Wilhelm-3
Don Wilhelm wrote: > Folks, > > There is an error in the reply I sent last night (below) that is very > important - it makes a big difference. > > When increasing or decreasing the frequency to determine the sign of the > reactance: > One must observe the increasing or decreasing of the IMPEDANCE (Z) and not > just the reactance. > > > that phase = 0 at each inflection point. i.e., if d|z|/df = 0 where f is frequency and the d symbols refer to differentiation there must be a phase zero, which can be seen to be true, since only an impedance with a net zero reactive component will have no change in impedance with frequency. This is one of the three definitions of parallel resonance. The interested reader can find the other two definitions in Terman's Radio Engineer's Handbook, amongst other places. Jack www.cliftonlaboratories.com _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
In reply to this post by Don Wilhelm-3
I did not mention that obviously, one has to watch both the impedance
variation and the reactance term to use my approximation to finding the sign of the reactance. Larry's antenna is a good exception to the performance of simpler antennas. Most ham antennas have well defined harmonic related resonances, and dips to lower impedance at a resonance. Another case where you might have a non harmonic resonance is if you are measuring paralleled and non harmonic band antennas such as a 15m dipole in parallel on one feeder with a 17m dipole. If you are dealing with a one band or fundamental antenna, the rule of thumb is not too hard to use. Stuart K5KVH _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
Free forum by Nabble | Edit this page |