Message-ID: <4F982D4A.50405@freenet.de>
Date: Wed, 25 Apr 2012 18:58:50 +0200
From: wolf_dl4yhf <dl4yhf@freenet.de>
User-Agent: Mozilla/5.0 (Windows NT 5.1; rv:11.0) Gecko/20120327 Thunderbird/11.0.1
MIME-Version: 1.0
To: rsgb_lf_group@blacksheep.org
References: <FE434AB14D2442DA9517B6F7304CE6E3@AGB> <FB72BFCD7B564064B0C93A0FD2DA9448@Extensa> <CAHAQVWNqFmyrEA5Jjq0EieKQO4G=7L8jxmdPu_JP_EsxuqdBvA@mail.gmail.com> <4F981FFA.20201@freenet.de> <4F98254C.9050300@iup.uni-heidelberg.de>
In-Reply-To: <4F98254C.9050300@iup.uni-heidelberg.de>
Subject: Re: LF: Generating 8970 Hz carrier with Spectrum Lab ?
Content-Type: multipart/alternative;
 boundary="------------030306060007080502050104"
Sender: owner-rsgb_lf_group@blacksheep.org
Precedence: bulk
Reply-To: rsgb_lf_group@blacksheep.org
X-AOL-SCOLL-URL_COUNT: 0  
x-aol-sid: 3039ac1d61884f982e24223d
X-AOL-IP: 195.171.43.25
X-AOL-SPF: domain : blacksheep.org SPF : none


This is a multi-part message in MIME format.
--------------030306060007080502050104
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 8bit

Am 25.04.2012 18:24, schrieb Stefan Schäfer:
> Hi Wolf,
>
> Am 25.04.2012 18:02, schrieb wolf_dl4yhf:
>>
>> Much easier and more reliable to build a dedicated hardware, clocked 
>> by a 10 MHz reference, using a PLL, DDS, or similar.
> Really easier?
Definitely easier than building a control loop (Regelkreis) which 
involves the soundcard's input, output, and a bunch of system-dependent 
latencies. Ok, the loop doesn't need to be fast, but it will never give 
the amount of phase coherence as ...

(1) A "good" soundcard (like yours, presumably) where ADC + DAC use 
truly identical clocks,

(2) A hardware solution as suggested (and already implemented) by Andy 
G4JNT.
The DDS doesn't need to be an IC in TSSOP (or even worse, BGA) case, a 
resistor ladder network as a simple DAC, plus some smart machine code 
does the job. Just clock the PIC (or similar micro) with the 10 MHz from 
the OCXO, GPS unit, or whatever you have available.

Don't get me wrong, if the reference-locked soundcard-based solution 
works, there's no reason why it shouldn't be used. But if it doesn't 
work, instead of spending time and money to test different soundcards, 
try a simple hardware-based solution.
For higher frequencies, there are very affordable ARM-based 
microcontrollers available, which are able to multiply the externally 
provided "crystal" clock to frequencies in the range of a few hundred 
MHz, which can then be divided down through a programmable digital 
timer/counter with a very good precision (used for PWM and motor 
control). After initialisation, the CPU itself is not involved in the 
generation of the PWM signal. The good news is that some of these 
Cortex-M based microcontrollers are available as homebrew-friendly 
development boards ("LPCXpresso" starter kits).
If there was a bit more time to spend, I would definitely give this a 
try since I use similar (but larger) microcontrollers at the QRL.

Cheers,
   Wolf.


--------------030306060007080502050104
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit

<html>
  <head>
    <meta content="text/html; charset=ISO-8859-1"
      http-equiv="Content-Type">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    Am 25.04.2012 18:24, schrieb Stefan Sch&auml;fer:
    <blockquote cite="mid:4F98254C.9050300@iup.uni-heidelberg.de"
      type="cite">
      <meta content="text/html; charset=ISO-8859-1"
        http-equiv="Content-Type">
      Hi Wolf,<br>
      <br>
      Am 25.04.2012 18:02, schrieb wolf_dl4yhf:
      <blockquote cite="mid:4F981FFA.20201@freenet.de" type="cite">
        <meta content="text/html; charset=ISO-8859-1"
          http-equiv="Content-Type">
        <br>
        Much easier and more reliable to build a dedicated hardware,
        clocked by
        a 10 MHz reference, using a PLL, DDS, or similar. <br>
      </blockquote>
      Really easier?<br>
    </blockquote>
    Definitely easier than building a control loop (Regelkreis) which
    involves the soundcard's input, output, and a bunch of
    system-dependent latencies. Ok, the loop doesn't need to be fast,
    but it will never give the amount of phase coherence as ...<br>
    <br>
    (1) A "good" soundcard (like yours, presumably) where ADC + DAC use
    truly identical clocks,<br>
    <br>
    (2) A hardware solution as suggested (and already implemented) by
    Andy G4JNT. <br>
    The DDS doesn't need to be an IC in TSSOP (or even worse, BGA) case,
    a resistor ladder network as a simple DAC, plus some smart machine
    code does the job. Just clock the PIC (or similar micro) with the 10
    MHz from the OCXO, GPS unit, or whatever you have available. <br>
    <br>
    Don't get me wrong, if the reference-locked soundcard-based solution
    works, there's no reason why it shouldn't be used. But if it doesn't
    work, instead of spending time and money to test different
    soundcards, try a simple hardware-based solution.<br>
    For higher frequencies, there are very affordable ARM-based
    microcontrollers available, which are able to multiply the
    externally provided "crystal" clock to frequencies in the range of a
    few hundred MHz, which can then be divided down through a
    programmable digital timer/counter with a very good precision (used
    for PWM and motor control). After initialisation, the CPU itself is
    not involved in the generation of the PWM signal. The good news is
    that some of these Cortex-M based microcontrollers are available as
    homebrew-friendly development boards ("LPCXpresso" starter kits).<br>
    If there was a bit more time to spend, I would definitely give this
    a try since I use similar (but larger) microcontrollers at the QRL.<br>
    <br>
    Cheers,<br>
    &nbsp; Wolf.<br>
    <br>
  </body>
</html>

--------------030306060007080502050104--