Hello Minto,
Nice to read about your progress!
Am 07.08.2011 18:06, schrieb Minto Witteveen:
Hi Stefan,
I noticed I wrote ALC, but obviously this must be AGC….
I installed Speclab, and did some tests with the 879. With the ACG off
I adjusted the gain just to the threshold of (audible) distortion of
the DCF39 signal (500 Hz filter) and read-out the level in Speclab.
Then I dialed to 137.700 and read out the (average) noise level. This
turned out to be a 40 dB lower, however I could hear NO bandnoise at
this setting of the manual AGC… only the receiver noise.
So DCF-39 would just be 40 dB above the noise? This is much to low. At
least 30 dB would be missed i estimate.
But what was the FFT setting "width of one FFT bin" in the FFT register
card?
From this I inferred two things:
A. DCF39 is more than 40 dB above the noise (but how much I cannot
measure with this method)
B. The 879 is a really lousy receiver (But I already knew that)
But: there is also good news! I did see (but not hear) your CQ at 17:40
HR CET, see attached screenshot. My first RX on 137!
Congrats to your first LF reception! And it shows me that my
calls are not only for my pleasure :-) So here we may better estimate
how many dBs are missed. Looks like you choosed a relative broad BW for
the FFT settings. 488 (some use 366 mHz) mHz would be a better choice.
And i assume you don't use a noise blanker so far. This will improve
thing too!
So at least the receiver + software works.
In the screenshot you can see your first CQ recorded with the FT897. At
17:32 I switched to the TS130. At first glance the signal looks better
on the TS130. Also notice that this trace is a bit off-frequency and
drifts visibly. The 130 has an analog VFO, and the frequency readout is
in 100 Hz steps. The drift might also be attributable to the SA612
mixer I use (now with a 10 MHz xtal)
This proves that at least QRSS3 QSO’s will be possible, so I have no
excuse not finishing my PA and variometer.
Yes! :-)
How do you go on improving the RX system? Ah, i have an idea: Display
the whole band, as broad as the SSB filter permits. Use a slow
scrolling spectrogram and make your local noise visible. Normally local
QRM sources disappear and appear from time to time. And they could be
band limited.
You should see the band noise ideally and its minimum just after the
local sunrise and a few hours later... If that all is on an equal
level, it must be covered by local noise!
Try to run the PA0RDT on batteries and your TRX and PC as well
(netbook). Do not change the antenna position and see how and if the
noise is changing and so on... :-)
73, Stefan/DK7FC
Regards,
Minto pa3bca
--------------------------------------------------------------------------
Ceterum censeo Carthaginem delendam esse
-----Original Message----- From: Stefan Schäfer
Sent: Sunday, August 07, 2011 11:56
To: [email protected]
Subject: Re: LF: Re: CW sked on 137 friday morning?
Hello Minto,
Am 07.08.2011 00:10, schrieb Minto Witteveen:
Just a question: How do you use speclab to
measure the signal level of for instance DCF39 AND the noisefloor.. By
measuring both and comparing?
I don't use a commercial HF TRX for LF but a down-converter (LF-125
kHz)
so i am not limited to a SSB filter bandwidth. I can display 125...149
kHz when running the soundcard at 48 kS/s. So a slight difference here.
But generally you must switch off the AGC and ALC and noise blanking
and
all that affects the output AF level. Then you can measure the absolute
levels, e.g. -10 dB for DCF and -80 dB for the band noise and -90 dB
for
the RX noise.
Just read it out on the right side.
I did a audio recording during the QSOs. My loop was beaming N/S, so
not
optimally for DCF-39. Anyway it was 85 dB S/N in 488 mHz (i use 488 mHz
for QRSS-3). So this is an example how to simply read out the SNR:
http://dl.dropbox.com/u/19882028/LF/DCF39%20in%20488%20mHz.png
Anyway there could be 10 dB more sensitivity on my RX to have best
reception in a quiet sunday morning in winther...
- but then I would have to disable the ALC of
the FT817?.
Not sure but i would disable anything ;-)
I can try /p sometime, go out in the fields
far away from man-made noise, with the miniwhip, a 12 meter fiberglass
pole, a 7Ah SLA, the transverter and the 817.
If you are outside in a flat field, you maybe do not need to lift the
antenna to 12m! 3m could be reasonable. Yes, beeing outside will at
least show you if its all caused by local QRM...
73, Stefan/DK7FC
Regards, Minto pa3bca
--------------------------------------------------------------------------
Ceterum censeo Carthaginem delendam esse
-----Original Message----- From: Stefan Schäfer
Sent: Saturday, August 06, 2011 16:51
To: [email protected]
Subject: Re: LF: Re: CW sked on 137 friday morning?
Hello Minto,
Am 05.08.2011 23:59, schrieb Minto Witteveen:
Stefan (et al),
I did listen this morning on 136.8 to see if I could copy you or any
other station, but the results were disappointing: I heard nothing.
Thanks for trying anyway.
Well, i was in QSO mit PA0A who must be close to you, so something must
be wrong...
I have an FT817 and an FT879 as possible
receivers. Both are not known for their superb receive quality but with
a 5 pole filter + additional amplification I got them to work quite
reasonable on 500 KHz.
Both rigs are very bad on LF. Very insensitive and prone to heavy
intermodulation, especially at night.
So this weekend I build a quick-and-dirty 4 MHz uptransverter with a
SA612. The Yeasu’s are much more sensitive at these frequencies. To
keep strong LF and MF stations from causing intermodulation had to add
7-pole Chebishev filter between a PA0RDT miniwhip and the transverter.
For details see:
http://www.pa3bca.nl/index.php?option=com_content&view=article&id=23&Itemid=17
Looks good, that page.
This morning (when there was almost no
man-made QRM form TV’s etc) DCF39 was S9+30dB on the 897. On (4000+)
136.8 the noise was S 5 with the 2.4 KHz SSB filter, and S1 with the
500 Hz filter.
Now I know that the FT’s S-meter is not very reliable, especially from
S1-S9. But between S9 and S9+40 it is “reasonably” accurate.
If you use SpecLab for receiving, you directly get the signal level in
dB (i.e. dB below full scale of the ADC). Then you can measure the
DCF39
level and the noise floor level. Meanwhile i am getting DCF39 at 80 dB
above noise in 488 mHz with the RX loop.
As far as I could find out DCF39 generates
50 KW ERP. If Stefan generates 0.5 Watt ERP (My guestimate)
Should be a few dB more ;-)
and the distance to my QTH is approximately
the same, I would expect Stefan to be 50 dB weaker. But 50 dB below
DCF39 would still be above the noise floor? So why didn’t I hear him
(or for that matter, PA0A who is closer and has 1 Watt ERP).
Maybe the S meter is worse than you think, maybe.
Before I finish my 1 kW TX I first have to
make sure that I will be able to copy any other station.... :-)
I had the same problems. But i immediately received DF6NM audible. So
if
you don't even receive PA0A, there seems to be lack of sensitivity.
What happens with the noise floor if you disconnect the PA0RDT? How
many
dB does the noise floor decrease?
Can you do /p tests?
Don't expect to much activity on LF. But there are new stations coming
on the band and you could be one of them and help making the band alie
again.
CU there :-)
73, Stefan/DK7FC
|
|
