Why is S/N ratio dependent on impedance match? It would seem a ferrite loopstick antenna would deliver twice the voltage if connected to a high impedance rather than matching it to it's characteristic impedance. But I have heard this is not a good idea since the S/N ratio would degrade. Any truth to this idea? -Bill
S/N Ratio dependent on impedance match?
Started by ●January 27, 2013
Reply by ●January 27, 20132013-01-27
"Bill Bowden"> > Why is S/N ratio dependent on impedance match? > > It would seem a ferrite loopstick antenna would deliver twice the > voltage if connected to a high impedance rather than matching it to > it's characteristic impedance. But I have heard this is not a good > idea since the S/N ratio would degrade. Any truth to this idea?** No. FET ( source follower) pre amps for ferrite antennas usually have no gate resistor - cos it only reduces the Q and the signal level. .... Phil
Reply by ●January 27, 20132013-01-27
Bill Bowden wrote:> Why is S/N ratio dependent on impedance match? > > It would seem a ferrite loopstick antenna would deliver twice the > voltage if connected to a high impedance rather than matching it to > it's characteristic impedance. But I have heard this is not a good > idea since the S/N ratio would degrade. Any truth to this idea? > > -BillAbsolutely.. Jamie
Reply by ●January 27, 20132013-01-27
On Sat, 26 Jan 2013 21:13:59 -0800 (PST), Bill Bowden <bperryb@bowdenshobbycircuits.info> wrote:>Why is S/N ratio dependent on impedance match? > >It would seem a ferrite loopstick antenna would deliver twice the >voltage if connected to a high impedance rather than matching it to >it's characteristic impedance. But I have heard this is not a good >idea since the S/N ratio would degrade. Any truth to this idea? > >-BillIt would deliver four times the voltage if you run it through a 2:1 step-up transformer before you go into the hi-z amplifier, which would double the s/n ratio. Keep extending that idea, more and more step-up, until the input of the transformer stops looking like a high impedance. The best place to stop is when the transformer impedance matches the source impedance. But as Phil says, you don't want to kill the Q, so may not want to actually match impedances. And atmospheric noise usually dominates LF reception, so getting the best s/n in the electronics usually doesn't matter. -- John Larkin Highland Technology Inc www.highlandtechnology.com jlarkin at highlandtechnology dot com Precision electronic instrumentation Picosecond-resolution Digital Delay and Pulse generators Custom timing and laser controllers Photonics and fiberoptic TTL data links VME analog, thermocouple, LVDT, synchro, tachometer Multichannel arbitrary waveform generators
Reply by ●January 27, 20132013-01-27
On Jan 27, 5:42=A0pm, John Larkin <jjlar...@highNOTlandTHIStechnologyPART.com> wrote:> On Sat, 26 Jan 2013 21:13:59 -0800 (PST), Bill Bowden >> <bper...@bowdenshobbycircuits.info> wrote: > >Why is S/N ratio dependent on impedance match? > > >It would seem a ferrite loopstick antenna would deliver twice the > >voltage if connected to a high impedance rather than matching it to > >it's characteristic impedance. But I have heard this is not a good > >idea since the S/N ratio would degrade. Any truth to this idea? > > >-Bill >> It would deliver four times the voltage if you run it through a 2:1 step-=up> transformer before you go into the hi-z amplifier, which would double the=s/n> ratio. Keep extending that idea, more and more step-up, until the input o=f the> transformer stops looking like a high impedance. The best place to stop i=s when> the transformer impedance matches the source impedance. > > But as Phil says, you don't want to kill the Q, so may not want to actual=ly> match impedances. And atmospheric noise usually dominates LF reception, s=o> getting the best s/n in the electronics usually doesn't matter. >So, why does the s/n ratio double if you go through a transformer? Doesn't the transformer double everything going into it?> -- > > John Larkin =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0Highland Technology Incwww=.highlandtechnology.com=A0 jlarkin at highlandtechnology dot com> > Precision electronic instrumentation > Picosecond-resolution Digital Delay and Pulse generators > Custom timing and laser controllers > Photonics and fiberoptic TTL data links > VME =A0analog, thermocouple, LVDT, synchro, tachometer > Multichannel arbitrary waveform generators
Reply by ●January 27, 20132013-01-27
On 1/27/2013 9:28 PM, Bill Bowden wrote:> On Jan 27, 5:42 pm, John Larkin > > <jjlar...@highNOTlandTHIStechnologyPART.com> wrote: >> On Sat, 26 Jan 2013 21:13:59 -0800 (PST), Bill Bowden >> > >> <bper...@bowdenshobbycircuits.info> wrote: >>> Why is S/N ratio dependent on impedance match? >> >>> It would seem a ferrite loopstick antenna would deliver twice the >>> voltage if connected to a high impedance rather than matching it to >>> it's characteristic impedance. But I have heard this is not a good >>> idea since the S/N ratio would degrade. Any truth to this idea? >> >>> -Bill >> > >> It would deliver four times the voltage if you run it through a 2:1 step-up >> transformer before you go into the hi-z amplifier, which would double the s/n >> ratio. Keep extending that idea, more and more step-up, until the input of the >> transformer stops looking like a high impedance. The best place to stop is when >> the transformer impedance matches the source impedance. >> >> But as Phil says, you don't want to kill the Q, so may not want to actually >> match impedances. And atmospheric noise usually dominates LF reception, so >> getting the best s/n in the electronics usually doesn't matter. >> > > So, why does the s/n ratio double if you go through a transformer? > Doesn't the transformer double everything going into it? >As the old saying goes, "One man's noise is another man's data." If you were looking at sferics, for instance, you'd want a low noise front end so as to make sure you were measuring the atmosphere and not the amplifier. At low signal levels, jacking up the input amplitude with a transformer is a win, until the transformer or the transformed amplifier input impedance starts to load down the signal. At MF, the input of a follower made from a BF862 looks like about 1/(j*omega*3.5 pF) to ground, with very little real part. So in principle you can go a fair way before that becomes a limitation. That approach is commonly used with FETs at lowish frequency, since their noise temperature is so very low, but their noise resistance is so very high. It's generally limited by the nonideal behaviour of the transformer. Cheers Phil Hobbs -- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 USA +1 845 480 2058 hobbs at electrooptical dot net http://electrooptical.net
Reply by ●January 28, 20132013-01-28
On Sun, 27 Jan 2013 18:28:34 -0800 (PST), Bill Bowden <bperryb@bowdenshobbycircuits.info> wrote:>On Jan 27, 5:42�pm, John Larkin > ><jjlar...@highNOTlandTHIStechnologyPART.com> wrote: >> On Sat, 26 Jan 2013 21:13:59 -0800 (PST), Bill Bowden >> > >> <bper...@bowdenshobbycircuits.info> wrote: >> >Why is S/N ratio dependent on impedance match? >> >> >It would seem a ferrite loopstick antenna would deliver twice the >> >voltage if connected to a high impedance rather than matching it to >> >it's characteristic impedance. But I have heard this is not a good >> >idea since the S/N ratio would degrade. Any truth to this idea? >> >> >-Bill >> > >> It would deliver four times the voltage if you run it through a 2:1 step-up >> transformer before you go into the hi-z amplifier, which would double the s/n >> ratio. Keep extending that idea, more and more step-up, until the input of the >> transformer stops looking like a high impedance. The best place to stop is when >> the transformer impedance matches the source impedance. >> >> But as Phil says, you don't want to kill the Q, so may not want to actually >> match impedances. And atmospheric noise usually dominates LF reception, so >> getting the best s/n in the electronics usually doesn't matter. >> > >So, why does the s/n ratio double if you go through a transformer? >Doesn't the transformer double everything going into it?If you're talking about noise that the antenna picks up from the world, the transformer multiplies that and the stuff that you consider to be "signal". In that situation, the transformer doesn't improve s/n. That's the usual case for LF radio. If the noise is inherent to the amplifier, as it tends to be many situations, then the s/n is optimized by proper impedance matching. That tends to be the case for microwave and exotic instrumentation and a lot of audio stuff. -- John Larkin Highland Technology Inc www.highlandtechnology.com jlarkin at highlandtechnology dot com Precision electronic instrumentation Picosecond-resolution Digital Delay and Pulse generators Custom timing and laser controllers Photonics and fiberoptic TTL data links VME analog, thermocouple, LVDT, synchro, tachometer Multichannel arbitrary waveform generators
Reply by ●January 28, 20132013-01-28
On Jan 27, 7:37=A0pm, Phil Hobbs <pcdhSpamMeSensel...@electrooptical.net> wrote:> On 1/27/2013 9:28 PM, Bill Bowden wrote: > > > > > > > On Jan 27, 5:42 pm, John Larkin > > > <jjlar...@highNOTlandTHIStechnologyPART.com> wrote: > >> On Sat, 26 Jan 2013 21:13:59 -0800 (PST), Bill Bowden > > >> <bper...@bowdenshobbycircuits.info> wrote: > >>> Why is S/N ratio dependent on impedance match? > > >>> It would seem a ferrite loopstick antenna would deliver twice the > >>> voltage if connected to a high impedance rather than matching it to > >>> it's characteristic impedance. But I have heard this is not a good > >>> idea since the S/N ratio would degrade. Any truth to this idea? > > >>> -Bill > > >> It would deliver four times the voltage if you run it through a 2:1 st=ep-up> >> transformer before you go into the hi-z amplifier, which would double =the s/n> >> ratio. Keep extending that idea, more and more step-up, until the inpu=t of the> >> transformer stops looking like a high impedance. The best place to sto=p is when> >> the transformer impedance matches the source impedance. > > >> But as Phil says, you don't want to kill the Q, so may not want to act=ually> >> match impedances. And atmospheric noise usually dominates LF reception=, so> >> getting the best s/n in the electronics usually doesn't matter. > > > So, why does the s/n ratio double if you go through a transformer? > > Doesn't the transformer double everything going into it? > > As the old saying goes, "One man's noise is another man's data." =A0If yo=u> were looking at sferics, for instance, you'd want a low noise front end > so as to make sure you were measuring the atmosphere and not the amplifie=r.> > At low signal levels, jacking up the input amplitude with a transformer > is a win, until the transformer or the transformed amplifier input > impedance starts to load down the signal. =A0At MF, the input of a > follower made from a BF862 looks like about 1/(j*omega*3.5 pF) to > ground, with very little real part. =A0So in principle you can go a fair > way before that becomes a limitation. > > That approach is commonly used with FETs at lowish frequency, since > their noise temperature is so very low, but their noise resistance is so > very high. =A0It's generally limited by the nonideal behaviour of the > transformer. > > Cheers > > Phil Hobbs >Yes, thanks Phil. I think you are saying that jacking up the signal with a transformer is a good idea until it degrades the Q due to non- ideal conditions of the transformer? I guess any reduction in S/N ratio would result from a lowered Q and wider bandwidth? So, if you can maintain the Q, there will be no change in S/N ratio? Is that a correct statement? -Bill
Reply by ●January 28, 20132013-01-28
On Sat, 26 Jan 2013 21:13:59 -0800, Bill Bowden wrote:> Why is S/N ratio dependent on impedance match? > > It would seem a ferrite loopstick antenna would deliver twice the > voltage if connected to a high impedance rather than matching it to it's > characteristic impedance. But I have heard this is not a good idea since > the S/N ratio would degrade. Any truth to this idea?So: kinda what Phil said, unless you're working with some bizarre non- voice radio or no-tune thing that would get screwed up by a too-high Q. And kinda what John said; if it's a loopstick antenna then you're working at MF, and at MF the atmospheric dominates all but the worst radios. But while it was mentioned that the best S/N ratio isn't to be had at a perfect impedance match, no one said where it _can_ be found. Amplifiers -- most specifically RF amplifiers, but audio ones, too -- have an optimal impedance for the best noise performance. It varies by the amplifier, but it's basically the the noise voltage of the amplifier reflected to the input, divided by the noise current reflected to the input. If you can present the amplifier with that impedance without losing any power in the coupling stages, then you're doing pretty good. But, you get back to the fact that you're working at MF, and your signal is dominated by atmospheric noise, so it doesn't matter so much. -- Tim Wescott Control system and signal processing consulting www.wescottdesign.com
Reply by ●January 28, 20132013-01-28
On 01/27/2013 11:28 PM, Bill Bowden wrote:> On Jan 27, 7:37 pm, Phil Hobbs > <pcdhSpamMeSensel...@electrooptical.net> wrote: >> On 1/27/2013 9:28 PM, Bill Bowden wrote: >> >> >> >> >> >>> On Jan 27, 5:42 pm, John Larkin >> >>> <jjlar...@highNOTlandTHIStechnologyPART.com> wrote: >>>> On Sat, 26 Jan 2013 21:13:59 -0800 (PST), Bill Bowden >> >>>> <bper...@bowdenshobbycircuits.info> wrote: >>>>> Why is S/N ratio dependent on impedance match? >> >>>>> It would seem a ferrite loopstick antenna would deliver twice the >>>>> voltage if connected to a high impedance rather than matching it to >>>>> it's characteristic impedance. But I have heard this is not a good >>>>> idea since the S/N ratio would degrade. Any truth to this idea? >> >>>>> -Bill >> >>>> It would deliver four times the voltage if you run it through a 2:1 step-up >>>> transformer before you go into the hi-z amplifier, which would double the s/n >>>> ratio. Keep extending that idea, more and more step-up, until the input of the >>>> transformer stops looking like a high impedance. The best place to stop is when >>>> the transformer impedance matches the source impedance. >> >>>> But as Phil says, you don't want to kill the Q, so may not want to actually >>>> match impedances. And atmospheric noise usually dominates LF reception, so >>>> getting the best s/n in the electronics usually doesn't matter. >> >>> So, why does the s/n ratio double if you go through a transformer? >>> Doesn't the transformer double everything going into it? >> >> As the old saying goes, "One man's noise is another man's data." If you >> were looking at sferics, for instance, you'd want a low noise front end >> so as to make sure you were measuring the atmosphere and not the amplifier. >> >> At low signal levels, jacking up the input amplitude with a transformer >> is a win, until the transformer or the transformed amplifier input >> impedance starts to load down the signal. At MF, the input of a >> follower made from a BF862 looks like about 1/(j*omega*3.5 pF) to >> ground, with very little real part. So in principle you can go a fair >> way before that becomes a limitation. >> >> That approach is commonly used with FETs at lowish frequency, since >> their noise temperature is so very low, but their noise resistance is so >> very high. It's generally limited by the nonideal behaviour of the >> transformer. >> >> Cheers >> >> Phil Hobbs >> > > Yes, thanks Phil. I think you are saying that jacking up the signal > with a transformer is a good idea until it degrades the Q due to non- > ideal conditions of the transformer? I guess any reduction in S/N > ratio would result from a lowered Q and wider bandwidth? So, if you > can maintain the Q, there will be no change in S/N ratio? Is that a > correct statement? > > -BillIt depends on the relative contributions of the background noise at the antenna and the RF amp. With a quiet input or a relatively noisy RF amp, a transformer is a win because the signal goes up and the noise basically doesn't. Once the antenna noise dominates the RF amp's noise, increasing the transformer ratio stops helping. Along the way, the transformer contributes reactance as well as both loss and noise. (The two are related by the fluctuation-dissipation theorem, which shows that any process that can dissipate power will also introduce noise. You can derive this from statistical mechanics in a page or two, or from the second law of thermodynamics in about three lines of algebra.) The loss will show up as a reduction of Q, just as you say, but also the reactance will detune the antenna some. Tapping the antenna down on a tank circuit is one possible method, but the last time I used a loopstick was when I was about 14, so I don't have a good feel for their characteristics. (It was a JW Miller part that cost me a whole week's allowance, $5.) Inductors used to be the biggest ripoff in electronics, bar high-end audio and scientology. $1 for a radial-lead Delevan choke, forsooth. Cheers Phil Hobbs -- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 hobbs at electrooptical dot net http://electrooptical.net
