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

Re: Anyone built the Ensemble RXTX for 40-30- & 20 meters ?

Expand Messages
  • warrenallgyer
    David Your concerns are valid and there are enough variables here that I think we need more testing and confirmation before ratifying this change. I have put
    Message 1 of 98 , Jan 3, 2013
    • 0 Attachment
      David

      Your concerns are valid and there are enough variables here that I think we need more testing and confirmation before ratifying this change.

      I have put the new filter in place in the 40/30/20 Peaberry and did not receive the expected results, either on transmit or receive and I need further testing to understand why the bread-board performs differently than with the filter in place.

      With the filter in place 30 meters harmonics are not in spec on the spectrum analyzer and that needs more testing as well as I measured 12.5 dB of attenuation on the breadboard and normally that should be plenty.

      Finally you raise a very big point on the toroid winding strategy. I have always wound a toroid on about 300 degrees, leaving a big gap at the bottom. I need to find the reference for that and I am thinking it is either the ARRL handbook or the Micrometals site, but the rationale was that closing the gap increased the capacitance added by the toroid and it obviously reduces the total inductance as well. As you note, there is significant difference depending on how the toroid is wound.

      I recommend we not jump to a final conclusion on this yet. Sorry, we probably need to start a new topic as this will no doubt prompt a new flurry of comments.

      Warren Allgyer - W8TOD

      --- In softrock40@yahoogroups.com, David Turnbull wrote:
      >
      > I did some tests on a 40-30-20 build. Instead of an oscilloscope,
      > which adds probe capacitance (especially on 10X), I measured with a
      > VSWR bridge that uses OpAmps to drive a galvanometer. The wattmeter
      > was calibrated with a new Fluke 87V (factory calibrated to ±0.05%).
      > The full test setup is a Peaberry connected to an Oak Hills Research
      > WM-2 terminated on an Elecraft 50.0 ohm resistive load. I started by
      > setting HDSDR to output exactly 1.0 watt on 10 MHz.
      >
      > 7.1Mhz 1.0 W
      > 10.1Mhz 1.0 W
      > 14.1Mhz 0.9 W
      >
      > There was enough power available to get to 1.2W on 20m. I've built
      > about a dozen of these 40-30-20 models. I always test with a dummy
      > load at 14 MHz. Never a problem. I've looked at the filters from the
      > receive side as well using a calibrated signal generator and the
      > measurements always follow the filter model.
      >
      > I've seen customer-built units where the LPF inductors are poorly
      > spaced. This causes a big loss on 20m. It's important to space the
      > turns evenly around the entire 360 degrees of the toroid. If you have
      > a big gap at the bottom near the PCB then you will not be able to
      > reach 1W on 20m.
      >
      > I'm a little concerned that the new filter values are too tight for
      > 30m harmonics. Has anyone tested them with a spectrum analyzer yet?
      >
      > 73 David AE9RB
      > http://AE9RB.com/
      >
      > On Thu, Jan 3, 2013 at 10:10 AM, barry underwood
      > wrote:
      > > Paul,
      > >
      > > I just found time to have closer look at my Ensemble. The original power measurements I made were with a Watson power / SWR meter at the low end of it's range, so could be a little out. I double checked by measuring the voltage across a 50Ohm dummy load with an oscilloscope. The input was set to 2.4V P/P as per instructions, the input signal being generated by the M0KGK IQ software. The output was as follows:
      > >
      > > 7.1Mhz 1.5 Watts
      > > 10.1Mhz 1.7 watts
      > > 14.1Mhz 0.8 Watts
      >
    • R. R. (Robby) Robson
      Gentlemen: Please review the projects on the new WB5RVZ website. I have added a Revisions/Errata tab to the menu on each project. Not much populated there,
      Message 98 of 98 , Jan 7, 2013
      • 0 Attachment
        Gentlemen:

        Please review the projects on the new WB5RVZ website.
        I have added a "Revisions/Errata" tab to the menu on each project.  Not much populated there, but the RXTX info that was raised in this thread has been posted.

        Check it out.

        On Sun, Jan 6, 2013 at 9:43 AM, richard.lawn@... <rjlawn@...> wrote:
         

        I couldn't agree more about the need to make a clear errata page obvious and available on the website with dates. Tnx for making this suggestion.
        Rick
        W2JAZ



        --- In softrock40@yahoogroups.com, "hladikz" wrote:
        >
        > Hi Oms,
        >
        > I just ended RXTXE building and found those build changes for 40/30/20 version.
        >
        > I think that it will be very nice if you will add errata/changes subpage to your new documentation web. Date of changes will be also very usable. At this time if I am coreect there are two changes:
        >
        > 1. Crossing of ant TR leads
        > 2. changes of values for 40/30/20 LPF C's and L's (please add also previous number of winds for L, to allow avoid recounting of winds)
        >
        > Yet something missing?
        >
        > I also think that way of band variant selection is not very clear. Especially for beginners it is not simple to recognize that you must go to band subpage. It will be more friendly if there will be possibility to change bands variant on each page.
        >
        > On other way I think that Softrock DOC at all is excelent and very easy to understand.
        >
        > 73 Zdenek OK7DR
        >
        >
        >
        > --- In softrock40@yahoogroups.com, "R. R. (Robby) Robson" wrote:
        > >
        > > I have made the changes to the documentation. See
        > > http://wb5rvz.org/ensemble_rxtx/05_rf (be sure to select the 40/30/20m band
        > > option).
        > >
        > > On Thu, Jan 3, 2013 at 10:33 AM, pauldebono@
        > > wrote:
        > >
        > > > **
        > > >
        > > >
        > > > Tony,
        > > >
        > > > Thanks very much.
        > > >
        > > > I was the the one who raised the issue.
        > > > I had already started reducing the turns and was going to make the
        > > > capacitors variable to determine the optimum values.
        > > >
        > > > 73's
        > > >
        > > > Paul
        > > > 9H1FQ
        > > >
        > > >
        > > > --- In softrock40@yahoogroups.com, "Parks, Tony" wrote:
        > > > >
        > > > > Good Morning All,
        > > > >
        > > > > Sure looks like it would be appropriate to make the 40m/30m/20m LP filter
        > > > > component changes starting right away.
        > > > >
        > > > > I will get the schematic values changed today with the following changes:
        > > > >
        > > > > C24 and C26 each become 180pF
        > > > > C25 becomes 390pF
        > > > > L2 and L3 each become 0.6uH, 14 turns of #26 wire on T37-6 cores.
        > > > >
        > > > > A continue to appreciate the thoughtful input from the group and hope the
        > > > > kits are still providing a positive entry level SDR experience for many.
        > > > >
        > > > > Thanks and 73,
        > > > > Tony KB9YIG
        > > > >
        > > > > On Thu, Jan 3, 2013 at 8:15 AM, John Greusel wrote:
        > > > >
        > > > > > **
        > > >
        > > > > >
        > > > > >
        > > > > > My vote is that Tony's suggested change and Warren's empirical results
        > > > > > point strongly to this modest change as a good thing.
        > > > > > The prior variance in output for different builders makes me think it's
        > > > > > related to the bias differences (via Q6) that Warren uncovered
        > > > recently. In
        > > > > > other words, those of us that are experiencing near 1 watt output are
        > > > > > actually having much higher output with attenuation.
        > > > > >
        > > > > > John
        > > > > > KC9OJV
        > > > > >
        > > > > >
        > > > > > ------------------------------
        > > > > > *From:* warrenallgyer
        > > > > > *To:* softrock40@yahoogroups.com
        > > > > > *Sent:* Thursday, January 3, 2013 5:31 AM
        > > > > >
        > > > > > *Subject:* [softrock40] Re: Anyone built the Ensemble RXTX for 40-30- &
        > > >
        > > > > > 20 meters ?
        > > > > >
        > > > > >
        > > > > > I have bread-boarded and tested Tony's suggested filter with caps of
        > > > > > 180,390,180 pF and T37 toroids with 14 turns, again terminated in 47
        > > > ohms
        > > > > > and fed directly by the output of an MFJ antenna analyzer.
        > > > > >
        > > > > > 7 MHz: Reference at 0 dB
        > > > > > 10 MHz: -0.1 dB
        > > > > > 14.2 MHz: -.7 dB
        > > > > > 16.3 MHz: -3.0 dB
        > > > > > 20 MHz: -12.3 dB
        > > > > > 21 MHz: -15.2 dB
        > > > > >
        > > > > > Warren Allgyer - W8TOD
        > > > > >
        > > > > > --- In softrock40@yahoogroups.com, "warrenallgyer" wrote:
        > > > > > >
        > > > > > > I just calculated different results with the external filter. The
        > > > > > > results reported previously were a comparison of simultaneous
        > > > readings
        > > > > > > on the input and the terminated output of the filter.
        > > > > > > The new readings are looking at the terminated output only and using
        > > > the
        > > > > > > value at 7MHz as the reference. Doing it this way I get
        > > > > > > 7 MHz: 0 dB10 MHz: -0.1 dB13.5 MHz: -3.3 dB14.0 MHz: -4.8 dB14.2 MHz:
        > > > > > > -5.4 dB 14.5 MHz: -6.6 dB
        > > > > > > The output of the analyzer, when terminated and measured separately
        > > > > > > across the range 7 - 14.5 MHz is flat within 0.2 dB.
        > > > > > > The discrepancy between the early results with simultaneous input and
        > > > > > > output measurements is the input impedance of the filter changes as
        > > > the
        > > > > > > filter begins to engage and that causes the voltage to rise. These
        > > > > > > latter results are more appropriate and still show a significant
        > > > problem
        > > > > > > with the specified turns counts and cap values
        > > > > > > I will build Tony's suggested filter and model it the same way.
        > > > > > >
        > > > > > >
        > > > > > > Warren Allgyer - W8TOD
        > > > > > >
        > > > > > > --- In softrock40@yahoogroups.com, "Parks, Tony" wrote:
        > > > > > > >
        > > > > > > > 16 turns on a T37-6 core is supposed to result in an 0.8uH inductor
        > > > > > > > according to the calculator on Diz's website.
        > > > > > > http://toroids.info/T37-6.php
        > > > > > > >
        > > > > > > > Checking the 40m/30m/20m LP filter design with AADE Filter Design
        > > > v4.5
        > > > > > > I
        > > > > > > > see about 1dB of additional attenuation at 14.3 MHz compared to
        > > > lower
        > > > > > > > frequencies.
        > > > > > > >
        > > > > > > > My recent experience with a freshly built 40m/30m/20m RXTX board
        > > > was
        > > > > > > the
        > > > > > > > 20m output was down about 1.5 dB.
        > > > > > > >
        > > > > > > > The capacitors supplied in the RXTX kit are 5%. Increasing
        > > > capacitor
        > > > > > > and
        > > > > > > > inductor values by 5% results in 2.5 dB 20m attenuation. Increasing
        > > > > > > > capacitor values by 5% and inductor values by 10% results in about
        > > > > > > 3.2 dB
        > > > > > > > of additional attenuation.
        > > > > > > >
        > > > > > > > The design seems to be reasonable but some kit builder's experience
        > > > > > > > suggests that the filter should be changed. Changing the 220pF
        > > > > > > capacitors
        > > > > > > > to 180pF, the 470pF capacitor to 390pF and the inductors to 0.6uH,
        > > > (14
        > > > > > > > turns on a T37-6 core), results in a LP filter with about the same
        > > > Z
        > > > > > > but
        > > > > > > > the 1dB of additional attenuation is at 18 MHz.
        > > > > > > >
        > > > > > > > Thoughts?
        > > > > > > >
        > > > > > > > Tony KB9YIG
        > > > > > > >
        > > > > > > > On Thu, Jan 3, 2013 at 4:14 AM, Orin wrote:
        > > > > > > >
        > > > > > > > > **
        > > > > > > > >
        > > > > > > > >
        > > > > > > > > --- In softrock40@yahoogroups.com, "warrenallgyer" wrote:
        > > > > > > > > >
        > > > > > > > > > Further to the above post:
        > > > > > > > > >
        > > > > > > > > > I have breadboarded the 43/30/20 filter externally, feeding it
        > > > > > > from an
        > > > > > > > > MFJ
        > > > > > > > > > antenna analyzer and terminating it on the breadboard with a 47
        > > > > > > ohm
        > > > > > > > > resister. I
        > > > > > > > > > have dual scope probes on for simultaneous measurement on the
        > > > > > > input and
        > > > > > > > > output
        > > > > > > > > > of the filter.
        > > > > > > > >
        > > > > > > > > Will you antenna analyzer measure inductance @ 14MHz? If so, what
        > > > > > > does it
        > > > > > > > > measure your inductors at. My 259B will give a result +/- 10nH
        > > > or so
        > > > > > > which
        > > > > > > > > is close enough for these filters.
        > > > > > > > >
        > > > > > > > >
        > > > > > > > > >
        > > > > > > > > > The 6 dB point of my breadboard filter is 12.5 MHz. At 14.2 MHz
        > > > > > > the
        > > > > > > > > output is 13
        > > > > > > > > > dB down.
        > > > > > > > > >
        > > > > > > > > > Toroids are T37 Yellow each 16 turns. Caps are 220 pf input and
        > > > > > > output
        > > > > > > > > shunts
        > > > > > > > > > with 470 pf. in the center shunt.
        > > > > > > > >
        > > > > > > > > I put those values into Elsie (800nH for the inductors) and got
        > > > > > > -1.3dB at
        > > > > > > > > 14.2MHz assuming 50 ohms input and your 47 ohms ouput. I tried
        > > > > > > adding 10%
        > > > > > > > > to the inductors (880nH) and got -2.6dB at 14.2MHz. 16 turns on a
        > > > > > > T37-6
        > > > > > > > > should give 768nH though and Elsie give -1dB at 14MHz in that
        > > > case.
        > > > > > > > >
        > > > > > > > > Another thing to try would be 14 turns on an T37-2 if you have
        > > > any
        > > > > > > around
        > > > > > > > > (I do from unused options on my Ensemble). The -2s have been more
        > > > > > > reliable
        > > > > > > > > for me in getting the 'advertised' inductance.
        > > > > > > > >
        > > > > > > > > Anyone wondering what "Elsie" is, it's an LC filter
        > > > design/network
        > > > > > > > > analysis program. I'm using the free student/hobbyist version.
        > > > > > > Details are
        > > > > > > > > here: http://tonnesoftware.com/elsie.html . I wish I could
        > > > justify
        > > > > > > > > spending the money for the full version.
        > > > > > > > >
        > > > > > > > > Orin.
        > > > > > > > >
        > > > > > > > >
        > > > > > > > >
        > > > > > > >
        > > > > > >
        > > > > >
        > > > > >
        > > > > >
        > > > > >
        > > > > >
        > > > >
        > > >
        > > >
        > > >
        > >
        > >
        > >
        > > --
        > > Cheers,
        > > Robby
        > >
        > > Richard R. (Robby) Robson
        > > LTC USA (RET)
        > > www.wb5rvz.com
        > > www.wb5rvz.org
        > > www.rafino.org
        > >
        >




        --
        Cheers,
        Robby
         
        Richard R. (Robby) Robson
        LTC USA (RET)
      Your message has been successfully submitted and would be delivered to recipients shortly.