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

Re: Fw: Re: Fw: Fw: Re: Fw: Re: Fw: Re: Fw: Re: [softrock40] FST switches

Expand Messages
  • Bill Dumke
    Kees, If you are tlking about the minus input terminal, that is supposed to be virtual ground. I don t think what you are looking at is anything valid.
    Message 1 of 2 , Mar 28, 2006
      Kees,

      If you are tlking about the minus input terminal, that is supposed to be
      virtual ground. I don't think what you are looking at is anything
      valid. Either look at the output or what feeds the input series resistor.

      Bill WB5TCO

      windy10605@... wrote:

      >Measured at the Op Amp input with a resistor to ground representing the
      >Op Amp input impedance and various capacitors to ground. The smaller the
      >impedance the lower the averaged DC bias level, 70ns period (14Mhz)
      >square wave (xxx7474 output) controlling the switches, 10 ohm resistor
      >from bias to the switch to dampen the charge time.
      >
      >73 Kees K5BCQ
      >
      >--------- Forwarded message ----------
      >From: Bill Dumke <billd@...>
      >To: softrock40@yahoogroups.com
      >Date: Mon, 27 Mar 2006 22:50:34 -0600
      >Subject: Re: Fw: Fw: Re: Fw: Re: Fw: Re: Fw: Re: [softrock40] FST
      >switches
      >Message-ID: <4428C09A.2010408@...>
      >References: <20060327.212302.2400.37.windy10605@...>
      >
      >Measured at the op amp output or right on the capacitors? A long time
      >ago with the SoftRock 40 I looked at each switched capacitor and what I
      >saw was a nice clean symmetric ramp up ramp down on each cap. I didn't
      >look at the op amp output, though.
      >
      >Bill WB5TCO
      >
      >windy10605@... wrote:
      >
      >
      >
      >>You know, the other thing is that if you have a low impedance input to
      >>the OpAmps (like the SR series) you are constantly charging/discharging
      >>the capacitors with the bias level ....looks real nasty on the scope
      >>(ringing, overshoot, etc). If you have a very high impedance OpAmp you
      >>don't do that and only see the small signal fluctuations (and a slight
      >>switching effect).
      >>
      >>73 Kees K5BCQ
      >>
      >>--------- Forwarded message ----------
      >>From: windy10605@...
      >>To: softrock40@yahoogroups.com
      >>Date: Mon, 27 Mar 2006 20:46:08 -0600
      >>Subject: Fw: Re: Fw: Re: Fw: Re: Fw: Re: [softrock40] FST switches
      >>Message-ID: <20060327.204609.2400.36.windy10605@...>
      >>
      >>Bill,
      >>
      >>Yes, I built a Digital DDS using the AD9851 for use with the SR-4
      >>(pictures on the web site). I use the --two-- comparator outputs and
      >>drive two (offset) divide by 2 FFs. That is the one I mentioned which
      >>falls off at around 18Mhz based on signal generator measurements.
      >>
      >>A good broadband transformer is the Mini Ckts T1-1T ...seems to work
      >>
      >>
      >very
      >
      >
      >>well across the whole HF spectrum. Brad Thompson has these
      >>for $0.50. Also has the HP 12dB MMICs.
      >>
      >>I'm doing some testing of the switches, starting with the FST3126.
      >>Someone may have done this before and has data ?? I'm using Vcc into the
      >>switch, switching it with a 14Mhz square wave, and measuring the outputs
      >>into various load terminations; C=0.047uF, C=.001uF, C=47pf etc and
      >>various load resistances which discharge the capacitor. So far all the
      >>capacitor values show right at 2ns "on" switch delay and 4ns "off"
      >>
      >>
      >switch
      >
      >
      >>delay. You can see the knee where the switch is turned off and the
      >>capacitor discharge is controlled by the load resistance. Very difficult
      >>to tell how much the switch is still "on" on at that time. The most
      >>readable data is with the 47pf load capacitor and a 300-600 ohm load
      >>resistor (very similiar to their test circuit). Then this is reading at
      >>the RF frequency not the Audio frequency. Quite a bit of ringing and
      >>distortion due to all the tack wires and leads. Trying to figure out a
      >>better test setup.
      >>
      >>73 Kees K5BCQ
      >>
      >>--------- Forwarded message ----------
      >>From: Bill Dumke <billd@...>
      >>To: softrock40@yahoogroups.com
      >>Date: Mon, 27 Mar 2006 19:35:37 -0600
      >>Subject: Re: Fw: Re: Fw: Re: Fw: Re: [softrock40] FST switches
      >>Message-ID: <442892E9.10600@...>
      >>References: <20060326.191201.2344.7.windy10605@...>
      >> <d3c938ee0603271024t13335a99i718ab365cb32b022@...>
      >> <7.0.1.0.0.20060327180151.01c5ae90@...>
      >>
      >>Kees,
      >>
      >>You once built a Softrock VFO using I believe the AD9850 DDS. Did you
      >>ever check the conversion gain in that unit versus frequency? I thought
      >>
      >>
      >
      >
      >
      >>you said the SoftRock 40 gave out at about 18 MHz. Was this based on
      >>your own data?
      >>
      >>Also I was wondering if you had come up with a broadband input
      >>transformer to use with the SoftRock in place of the front end filter
      >>for use with a signal generator? It would be interesting making some
      >>measurements versus frequency.
      >>
      >>Bill WB5TCO
      >>
      >>KY1K wrote:
      >>
      >>
      >>
      >>
      >>
      >>>I'd sure like to hear what Dan Tayloe has to say about this issue. I
      >>>think it's incredibly interesting that sdr-1000 rigs don't lose
      >>>conversion gain as the frequency goes up...but that Spice models show
      >>>the conversion gain should be dropping.
      >>>
      >>>I can certainly understand why the noise floor should degrade as the
      >>>frequency increases-as explained by Dan on this list. But, I'm having
      >>>a problem understanding why the sdr-1000 doesn't show conversion loss
      >>>at higher frequencies.
      >>>
      >>>It's an interesting problem for sure!
      >>>
      >>>GL to all.
      >>>
      >>>Art
      >>>
      >>>
      >>>
      >>>
      >>>
      >>>
      >>>
      >>>>On 3/27/06, windy10605@... <mailto:windy10605@...>
      >>>><windy10605@... <mailto:windy10605@...>> wrote:
      >>>>
      >>>> The model runs OK at much lower values than .01uF (.001uf and
      >>>> even .0005uf at the higher 30Mhz frequencies) but you have to
      >>>> increase the series resistance (more noise) to maintain the same
      >>>> LPF roll-off and that interacts to increase the capacitance back
      >>>> up again. A completely different LPF technique may have to be
      >>>>
      >>>>
      >>>>
      >>>>
      >>used.
      >>
      >>
      >>
      >>
      >>>>In this case an instrumentation opamp (which have a high input
      >>>>reistance) with balanced inputs might be suitable. Configuring it
      >>>>with a high gain, so that the gain resistor (R in between the two
      >>>>input opamps) is low, and thus the noise contribution of the
      >>>>amplifier is low. My experimental spice model shows that lowering
      >>>>the integrator caps and shortening switch-on time, leads to less
      >>>>conversion loss.
      >>>>
      >>>>To simulate the possible "bucket-effect" as mentioned by Dan, I put a
      >>>>C in front of the switch. For f>30Mhz you can see an effect in the
      >>>>form of conversion loss in the mixer. However, this is not the effect
      >>>>that noise will increase with higher f as reported by sdr1k owners,
      >>>>but it is unclear if this noise is caused by the switch. In the spice
      >>>>model of softrock there is a conversion loss for higher frequencies,
      >>>>but the model also show that this effect can be mitigated by applying
      >>>>your idea.
      >>>>
      >>>>
      >>>>
      >>>>
      >>>
      >>>
      >>>
      >>>
      >>------------------------------------------------------------------------
      >>
      >>
      >>
      >>
      >>>YAHOO! GROUPS LINKS
      >>>
      >>> * Visit your group "softrock40
      >>> <http://groups.yahoo.com/group/softrock40>" on the web.
      >>>
      >>> * To unsubscribe from this group, send an email to:
      >>> softrock40-unsubscribe@yahoogroups.com
      >>>
      >>>
      >>>
      >>>
      >>>
      >><mailto:softrock40-unsubscribe@yahoogroups.com?subject=Unsubscribe>
      >>
      >>
      >>
      >>
      >>>
      >>> * Your use of Yahoo! Groups is subject to the Yahoo! Terms of
      >>> Service <http://docs.yahoo.com/info/terms/>.
      >>>
      >>>
      >>>
      >>>
      >>>
      >>>
      >>>
      >>------------------------------------------------------------------------
      >>
      >>
      >>
      >>
      >>
      >>Yahoo! Groups Links
      >>
      >>
      >>
      >>
      >>
      >>
      >>Yahoo! Groups Links
      >>
      >>
      >>
      >>
      >>
      >>
      >>Yahoo! Groups Links
      >>
      >>
      >>
      >>
      >>
      >>
      >>
      >>
      >>
      >>
      >>
      >>
      >>
      >>
      >
      >
      >
      >Yahoo! Groups Links
      >
      >
      >
      >
      >
      >
      >Yahoo! Groups Links
      >
      >
      >
      >
      >
      >
      >
      >
      >
      >
      >
    Your message has been successfully submitted and would be delivered to recipients shortly.