Loading ...
Sorry, an error occurred while loading the content.

Re: Introduce myself and a short story about wideband loop matching (NIC based)

Expand Messages
  • claudio
    ... I ve found this interesting document about AL-1530S+ Wellbrook: http://www.nakatoyo.com/wellbrook/ALA1530S_PLUS/ALA1530S_c_4.pdf The words negative
    Message 1 of 12 , Sep 5, 2012
    • 0 Attachment
      --- In loopantennas@yahoogroups.com, "Chris Trask" <christrask@...> wrote:
      >
      > >
      > > > I've experimented with numerous approaches to an NIC amplifier for
      > > > active loop antenna applications, but have had no success to date. I
      > > > have a couple of ideas that I have yet to try.
      > > >
      > > > In the meantime, I did design a wideband amplifier for loop
      > antennas
      > > > that outperforms the Wellbrook and Pixel commercial products:
      > > >
      > > > http://www.home.earthlink.net/~christrask/Paper009.html
      > > >
      > >
      > >
      > >
      > > Hi Chris,
      > > good news, I'll be brief now but I'll give a more detailed report later
      > on.
      > > I added the augmenting circuit to my amp (you find it in the file section
      > as
      > > "IT9JCB's...etc") using an ideal VCVS as tranformer to simplify the job.
      > The
      > > ratio used is 5 ( best one).
      > > These are the interesting results:
      > >
      > > W/O AUG:
      > > emitter imput impedance ~1.2 Ohm, freq. dependent
      > > Gain 30 dB in 1-15MHz
      > > noise voltage rti 0.47 nV/sqrt(Hz)
      > >
      > > W AUG
      > > (I'd to increase the input transformer ratio from 3:1 to 5:1 to better
      > match
      > > the new low impedance)
      > >
      > > emitter imput impedance ~0.3 Ohm, freq. flat
      > > Gain 42 dB in 1-18MHz
      > > noise voltage rti 0.25 nV/sqrt(Hz)
      > >
      > > In few words we gained about 6 dB in SNR, furthermore it seems that the
      > > impedance seen by the loop is now much resistive and flat on frequency.
      > >
      >
      > Those are impressive results, and they attest to the improvement that
      > can be enjoyed by applying augmentation. Will you have more detailed
      > results later?
      >
      > >
      > > Is this the Andy's secret ?
      > >
      >
      > Don't think so. He can't use augmentation for commercial designs as I
      > hold the patents and nobody has asked for a license to use them.
      >
      > >
      > > Anyway, could you provide me more info about your results with NIC and
      > > some schematic, if possible, thanks in advance.
      > >
      >
      > Don't have anything to share in terms of NIC circuitry as all of mine
      > failed on the launch pad, which is typical of positive feedback circuitry.
      > I do have a couple of ideas still, but not the time to work on them.
      >
      > Chris
      >

      I've found this interesting document about AL-1530S+ Wellbrook:

      http://www.nakatoyo.com/wellbrook/ALA1530S_PLUS/ALA1530S_c_4.pdf

      The words "negative impedance" and "non-foster matching" are evident.
      Strangely, this doc is not present in the wellbrook site, that means that or have been removed or it's not still a valid doc.

      Claudio
    • lz1aq
      Hello Claudio, You can visit http://www.lz1aq.signacor.com/docs/wsml/wideband-active-sm-loop-antenna. htm
      Message 2 of 12 , Sep 10, 2012
      • 0 Attachment
        Hello Claudio,

         You can visit  http://www.lz1aq.signacor.com/docs/wsml/wideband-active-sm-loop-antenna.htm. The paper treats similar problems and might be of some help to you. I have some remarks about active wideband loop antenna:

         1. It is better to use Norton equivalent circuit for the small loop in the spice models. The current source becomes frequency independent and this corresponds to the physical nature - small loop loaded with resistance much smaller than its inductive impedance gives fixed current (short circuit current) irrespective of the frequency.

         2. The wideband loop amplifier is current to voltage converter. I will suggest giving its gain as a ratio of output voltage to input current. Gain= Uout/Iinp [ohms]. Dimension is ohms and I called this trans-resistive gain. The amplifier presented in my paper has a gain of approx. 650 ohms ( 1uA  @ input will give 650uV @output). It is now easy to compare different loop amplifiers gain irrespective of their input resistance.  It is straightforward to obtain the output noise voltage and trans-resistive gain in spice models.

         3. It is also convenient to compare their noise floor expressed in uV/m. Dividing the Uout-noise/TR-Gain we will easily obtain the input noise current and compare it to the loop current by applying the simple and quite accurate equation for the loop short circuit current from the  field intensity. So we can easily obtain (or compute) the noise floor of an amplifier with a specific loop expressed in uV/m. This is the most important parameter of the wideband loop and it is pity that it is usually not given in the loop specifications (except for some commercial professional cases).

        I am giving these suggestions since if we try to compare different amplifiers and antennas we should have common terms. For example I can not evaluate any active loop if  its noise floor (in uV/m) is not given. Evaluating and compareing different loops with on the air tests by saitsfied customers is not very accepatble for me.  

         4. The noise figures which you obtain in spice models are of the same order as in my models. I can say that in the flat frequency region the experimental results are very close to the spice models. My experiments with common base (CB) amplifiers show that usually the loop sensitivity is limited by the internal noise (not by the atmospheric) for frequencies above 10 MHz. In practice, 1 m diam. active wideband loop with L<=3 uH,  with simple CB amplifier,below 10MHz, at night times has an acceptable noise floor which is below the atmospheric noise level even in rural places. In order to obtain better reception during daytimes and higher frequencies we need some 3 - 10 dB lower active antenna noise floor. This sensitivity can be reached more easily with small active whip (electric) antenna ( see Fig.1.6 in  http://www.active-antenna.eu/tech-docs/1_ActiveAA_DandS_20.pdf   ). So the fight for every dB noise reduction in CB active loop amplifiers is very important.

         5. I have not tested the augmented amplifiers (suggested by Chris Trask) and your noise floor results with them are quite interesting.

         6. I am skeptic about the possibility to reduce substantially the active wideband  loop noise floor. The reason for that lies in the very small current of the loop. Probaly there are physical limits – I will give  a very approximate example:  In 1 m diam. loop with inductance of 2.4 uH,the induced short circuit current from 1uV/m external field in the flat frequency response region will be 1.1 nA. To simplify the case let's assumethat the loop is loaded with 3 ohms resistor which emulates the input resistance of the differential CB amplifier. The voltage drop across 3 ohmsload resistor will be 3.3 nV. From the other hand the thermal noise voltage at290 deg. K of 3 ohm resistor at BW of 1 KHz is 7 nV.  In this case we have 2.1 uV/m  equivalent noise floor  @ 1KHz BW of the loop which is terminated with 3 ohms load resistor - from my point of view this is the main factor that limits the noise floor of a wideband  loop terminated with low input impedance amplifier.

         7. The most effective way to reduce the loop inductance (and noise floor) is with Crossed Parallel loops - see my paper.With large crossed loops the noise floor can be reduced substantially and it is possible to have antenna limited by the atmospheric noise.

         8. I do not have experience with NIC - they are promising because they probably aviod the above mentioned limitations, but we should consider also the non-linear behavior of these circuits.

        9. The balanced scheme improves IP2 not IP3 (probably you have made a lapsus). 

         Wish you a success in your project and I will be glad if you publish your results.

        Chavdar,   LZ1AQ   lz1aq@... , www.lz1aq.signacor.com

         P.S.  I will  be on my holiday the next  7 days at a place where there are no telephones and Internet  :-) so I will answer any possible messages  when I am back.


      • claudio
        Hi Chavdar, liked very much your extensive comments, will give you back mines as soon later on (very busy now). I would like to reply now just about one point
        Message 3 of 12 , Sep 11, 2012
        • 0 Attachment
          Hi Chavdar,
          liked very much your extensive comments, will give you back mines as soon later on (very busy now). I would like to reply now just about one point that I think the most important.

          --- In loopantennas@yahoogroups.com, "lz1aq" <lz1aq@...> wrote:
          >
          > Hello Claudio,
          > You can visit
          > http://www.lz1aq.signacor.com/docs/wsml/wideband-active-sm-loop-antenna.\
          > htm
          > <http://www.lz1aq.signacor.com/docs/wsml/wideband-active-sm-loop-antenna\
          > .htm> . The paper treats similar problems and might be of some help to
          > you. I have some remarks about active wideband loop antenna:
          >
          > 1. It is better to use Norton equivalent circuit for the small loop in
          > the spice models. The current source becomes frequency independent and
          > this corresponds to the physical nature - small loop loaded with
          > resistance much smaller than its inductive impedance gives fixed current
          > (short circuit current) irrespective of the frequency.
          >
          > 2. The wideband loop amplifier is current to voltage converter. I will
          > suggest giving its gain as a ratio of output voltage to input current.
          > Gain= Uout/Iinp [ohms]. Dimension is ohms and I called this
          > trans-resistive gain. The amplifier presented in my paper has a gain of
          > approx. 650 ohms ( 1uA @ input will give 650uV @output). It is now easy
          > to compare different loop amplifiers gain irrespective of their input
          > resistance. It is straightforward to obtain the output noise voltage
          > and trans-resistive gain in spice models.
          >
          > 3. It is also convenient to compare their noise floor expressed in
          > uV/m. Dividing the Uout-noise/TR-Gain we will easily obtain the input
          > noise current and compare it to the loop current by applying the simple
          > and quite accurate equation for the loop short circuit current from the
          > field intensity. So we can easily obtain (or compute) the noise floor of
          > an amplifier with a specific loop expressed in uV/m. This is the most
          > important parameter of the wideband loop and it is pity that it is
          > usually not given in the loop specifications (except for some commercial
          > professional cases).
          >
          > I am giving these suggestions since if we try to compare different
          > amplifiers and antennas we should have common terms. For example I can
          > not evaluate any active loop if its noise floor (in uV/m) is not given.
          > Evaluating and compareing different loops with on the air tests by
          > saitsfied customers is not very accepatble for me.
          >
          > 4. The noise figures which you obtain in spice models are of the same
          > order as in my models. I can say that in the flat frequency region the
          > experimental results are very close to the spice models. My experiments
          > with common base (CB) amplifiers show that usually the loop sensitivity
          > is limited by the internal noise (not by the atmospheric) for
          > frequencies above 10 MHz. In practice, 1 m diam. active wideband loop
          > with L<=3 uH, with simple CB amplifier,below 10MHz, at night times has
          > an acceptable noise floor which is below the atmospheric noise level
          > even in rural places. In order to obtain better reception during
          > daytimes and higher frequencies we need some 3 - 10 dB lower active
          > antenna noise floor. This sensitivity can be reached more easily with
          > small active whip (electric) antenna ( see Fig.1.6 in
          > http://www.active-antenna.eu/tech-docs/1_ActiveAA_DandS_20.pdf ). So
          > the fight for every dB noise reduction in CB active loop amplifiers is
          > very important.
          >
          > 5. I have not tested the augmented amplifiers (suggested by Chris
          > Trask) and your noise floor results with them are quite interesting.
          >
          > 6. I am skeptic about the possibility to reduce substantially the
          > active wideband loop noise floor. The reason for that lies in the very
          > small current of the loop. Probaly there are physical limits – I
          > will give a very approximate example: In 1 m diam. loop with
          > inductance of 2.4 uH,the induced short circuit current from 1uV/m
          > external field in the flat frequency response region will be 1.1 nA. To
          > simplify the case let's assumethat the loop is loaded with 3 ohms
          > resistor which emulates the input resistance of the differential CB
          > amplifier. The voltage drop across 3 ohmsload resistor will be 3.3 nV.
          > From the other hand the thermal noise voltage at290 deg. K of 3 ohm
          > resistor at BW of 1 KHz is 7 nV. In this case we have 2.1 uV/m
          > equivalent noise floor @ 1KHz BW of the loop which is terminated with 3
          > ohms load resistor - from my point of view this is the main factor that
          > limits the noise floor of a wideband loop terminated with low input
          > impedance amplifier.
          >


          here it is:

          assume that in some way we are able to reduce the CB amp impedance to just...say 0.3 Ohms, the thermal noise voltage will reduce about 20 dB, but we must include in these simple analysis the noise current of the amp itself.
          Anyway reducing the input impedance to sub-ohms values gives the advantage, now we have to understand how to do this without increase the noise (not simply increasing the TR Ic !!), one way is the Chris Trask's patented technique, but I found that it works as adviced only using ideal VCVS, a transformer gave little improvement, just yesterday I tried with an active solution and the initial results are encouraging.
          The other two advantages of this augmented circuits are:

          1. being Rin so low, it acts as short to ground, so even umbalanced loops can be used with lees sensitivity to electric fields.
          2. the augmengting tecnique has some effect of linearization on the input impedance, being Rs linear the ratio RL/Rs is more linear too.
          I observed about 7-10 dB of improvement in 3th order products.


          > 7. The most effective way to reduce the loop inductance (and noise
          > floor) is with Crossed Parallel loops - see my paper.With large crossed
          > loops the noise floor can be reduced substantially and it is possible to
          > have antenna limited by the atmospheric noise.
          >
          > 8. I do not have experience with NIC - they are promising because they
          > probably aviod the above mentioned limitations, but we should consider
          > also the non-linear behavior of these circuits.
          >
          > 9. The balanced scheme improves IP2 not IP3 (probably you have made a
          > lapsus).
          >
          > Wish you a success in your project and I will be glad if you publish
          > your results.
          >
          > Chavdar, LZ1AQ lz1aq@... <mailto:lz1aq@...> ,
          > www.lz1aq.signacor.com <http://www.lz1aq.signacor.com/>
          >
          > P.S. I will be on my holiday the next 7 days at a place where there
          > are no telephones and Internet :-) so I will answer any possible
          > messages when I am back.
          >

          Continuing to work.

          73s Claudio
          IT9JCB
        Your message has been successfully submitted and would be delivered to recipients shortly.