[Lowfer] WE2XGR/2 MFSK8, MFSK16

[Jay Rusgrove]

Rick

The sound cards in use here have been characterised using SpectrumLab>Time
Domain Scope>Presets>Phase Meters>Phasemeter for 800 Hz. An 800 Hz signal
from a GPS referenced HP3325A feeds the line in of the sound card. The
actual sampling rates are determined and stored into Spectrum Lab and other
programs that will accept actual sampling rate inputs. This accurately
calibrates the receive side. The Soundblaster Live cards I'm using on three
different receivers/computers are quite stable over the short and long term.
I typically check them once a month and changes are rarely needed. The main
receiver is a Harris RF-590 which is also GPS referenced. Once a receiving
setup is calibrated it can then be used to verify the accuracy of the
transmitted signal.

Many of the programs allow the user to input the exact sampling rate of the
sound card in use, however, MultiPSK (which I use for MFSK8 and MFSK16) only
allows limited adjustment steps of the sampling rate. With that adjustment I
was able to reduce the transmitted frequency error to about .05 Hz - which I
suspect is probably close enough. I was able to remove the remaining .05 Hz
offset by adjustment of the LO frequency (from another GPS referenced
HP3325A) into the phasing exciter.

Yes, the sound cards support the programs sample rate.

Jay
----- Original Message -----
From: Rick Kunath
To: Discussion of the Lowfer (US, European, & UK) and MedFer bands
Sent: Thursday, January 03, 2008 12:29 PM
Subject: Re: [Lowfer] WE2XGR/2 MFSK8, MFSK16


Jay Rusgrove wrote:

> I conducted local reception tests last night of the two modes - at a
signal
> level about 15 dB above the receiver noise floor.   No antenna was
connected
> and therefore the decoder had no noise or fading to contend with -probably
> close to ideal conditions for the decoder. BTW, the transmitter and
receiver
> frequency accuracy here are about .01 Hz (both TX and RX are GPS
> disciplined).

<Snipped>

> Suppose another worthwhile battery of tests would be to run the various
> digital modes under both very weak and moderate signal levels to determine
> tolerable frequency error. Not all stations on the band will have
> 'laboratory grade' frequency stability, especially during an emergency
> communication period, so some digital modes may not be practical under
those
> conditions. Perhaps John (if he's not burned out from the last tests!) and
I
> can run tests over the near future to get a better handle on this issue.

Were the respective sound cards in the various machines that were
generating and decoding the MFSK checked for sampling rate accuracy,
sampling rate stability over time, and were any needed correction
factors for differences between the expected and actual sound card
sampling rates applied to the correction factor entries in the MFSK
software?

Were the sound cards in use able to natively support the requested MFSK
application's sample rate, or was re-sampling necessary? If so, what was
the re-sampling accuracy set to in the Windows control panel, if these
were Microsoft machines?

Rick Kunath
_______________________________________________
>From the Lowfer mailing list
Send messages to: Lowfer at mailman.qth.net
To sub/unsub visit: http://mailman.qth.net/mailman/listinfo/lowfer