Max SNR as Indicator of RX Performance?

[VE3MEO]

Re: Re: Re: Re: Max SNR table - the ascent to 40dB

Here are the latest changes to the table.

73, Tom

May 22:
Bob's trusty old AR-88 tube radio hits 32.5 dB!

May 26:
Daniel, DF8UO, gained another dB to 37.8 with his FT757 and is nipping at the heels of simone's FRG-100 for 3rd place.

Originally posted by VE3MEO
It's been quiet as far as new entries or advancements in the table since Mar 20! What gives? Has nobody had a better result? Are there no new members hitting 30+dB SNR? Or is this a subject that has run its course?

73, Tom

[df8uo]

Comparison FT757/Digital world traveller

Hi all,

DF9RB's comparisons motivated me to perform also some comparison tests. I read a lot of the DWT, however missed real practical tests. Find below my results after having a DWT for 2 weeks now.

Today i made a comparison of the DWT and FT757GX, using the VOR signal on 15780kHz.I was using two different antnnas. Antenna 1: Dipole 2x5m Antenna 2: Shortened trap dipole (for 10, 15, 20m).
Antenna 1 gives in general better DRM results. Attached the reults:
DWT(red/blue) 9h00-10h00 Antenna 1; 10h00-11h00: Antenna 2
FT757 (green/yellow) 9h00-10h00 Antenna 2; 10h00-11h00: Antenna 1
Signal of VOR on the FT757 was S5-S9 with antenna 2 and S8 to S9+ with antenna 1. The comparison showed me, that my FT757 can handle adjacent channel interference much better than the DWT. On 15785 a partly quite strong signal was present (probably Radio Galei Tzahal from Israel). The AM filter of the FT757 is a little to small for DRM, which for these cases has also advantages. From all the experiments up to know i could see that SNR on my FT757 under similar conditions is 3 to 10 dB better
than the DWT. If i compare weight and size of the DWT and my FT757, this small device is quite good.
The most critical point for the DWT is the usage of provisorical wire antennas. Here i will do some further experiments. Using dipole antennas are performing quite well.

Some examples for the DWT and antenna 2 you can find here:
<p><a href="http://www.drmrx.org/forum/showthread.php?s=&amp;postid=18302#post18111">ORF 6155: </a><br>
<a href="http://www.drmrx.org/forum/showthread.php?s=&amp;postid=18304#post18304">DW 7390</a>: <br>
<a href="http://www.drmrx.org/forum/showthread.php?s=&amp;postid=18287#post18123">BBC 7320</a> <br>
<a href="http://www.drmrx.org/forum/showthread.php?s=&amp;postid=18321#post18306">Migh ty 5830</a> <br>
<a href="http://www.drmrx.org/forum/showthread.php?s=&amp;postid=18321#post18299">Migh ty 7430</a> </p>

More to come.

73, Daniel

[hsseppo]

Hi Tom and all,

Building a DRM-receiver has been very educational for me. Especially measuring the specs like the mixer 1dB compression point and the intercept point. I used two-tone technique to measure the third-order distortion. My AD831 mixer compression point was +(8-9) dBm (abt. 1.6 Vp-p) and IP3 about +(15-18) dBm. Using the two-tone method it was easy to see that at level 0dBm 3rd order output was -25 dBm and at level -10 dBm 3rd order was -55 dBm. At level +4dBm 3rd order was only -16 dBm and at +8dBm input level the distance of 3rd order was about 10 dB.

When RF input increases 10dB 3rd order increases 30dB!
It looks like the 3rd order distortion is a very important factor for DRM performance. My Sony ICF-2001D receiver has about 35dB distance of 3rd order, maybe this is the reason for the low maximum 28dB SNR and not the PLL phase noise.

As I have a AD8099 preamplifier (44dB gain) feeding the AD831 mixer, the input RF levels of my receiver are: -36dBm to -126 dBm (10 mVp-p to 0.3 uVp-p).

Next I measured the maximum SNR, feeding dream -t to SA612 mixer and got 50 dB SNR.

>>>>
Dream
Software Version 1.2.5cvs
Starttime (UTC) 2005-06-30 11:34:32
Frequency 10012 kHz
Latitude
Longitude
Label Dream Test
Bitrate 20.96 kbps
Mode B
Bandwidth 10 kHz

MINUTE SNR SYNC AUDIO TYPE
0000 50 150 1500/10 0
0001 50 150 1500/10 0
0002 50 150 1500/10 0
0003 50 150 1500/10 0
0004 50 150 1500/10 0
0005 51 150 1500/10 0
0006 51 150 1500/10 0
0007 51 150 1500/10 0
0008 51 150 1500/10 0
0009 51 150 1500/10 0
0010 51 150 1500/10 0
0011 50 150 1500/10 0

SNR min: 49.0, max: 51.4

CRC:
<<<<


It was also a lucky day for me, as I got my record 37.6 dB SNR from 15780 kHz (Voice of Russia)

http://www.drmrx.org/forum/showthread.php?s=&postid=18509#post18509

Best Regards,

Seppo

[Rene Schmitz]

Hi Seppo and all,

Congrats to your archeivement! 51dB even in the lab, thats something!

Not difficult to imagine what happens when the mixer is overdriven, instead of two tones you have the 100+ carriers intermodulating each other. I have made similar observations with my mods for the Elektor RX. Most of which focussed around getting more gain but also more linearity out of the IF stage and the final mixer. (If anyone is interested, the IF amp I built has been published here
) The original crummy FET-mixer of the RX was later changed to an AD831, but finally again to a ring mixer. While I'd clearly favour the AD, now that I made the comparison, its probably staying that way for a while.

Cheers,
René

[VE3MEO]

Re: Max SNR table - the ascent to 40dB

Here are the latest changes to the table.

73, Tom

June 30:
Two big moves for Seppo (hsseppo) using his revised modular receiver:
a) 51.4 dB in closed circuit test (I included this because it indicates nearly ideal conditions that might occasionally be realisable from a real system comprising a powerful transmitter and perfect propagation! If anyone achieves a higher test result, it will replace this one as a point of reference.

b) 37.6 dB with this receiver pulling in VoR - 5th place among actual reception reports.

July 2:
Fritz Wue's (DK6QI) new Sat-Schneider DRT1 radio achieved 37 dB - 6th place and 2.6 dB better than his previous best with his FT840

[VE3MEO]

Max SNR table - the ascent to 40dB

Here are the latest changes to the table.

73, Tom

July 18:
Fritz Wue (DK6QI) vaulted from 6th to 3rd place with his new Sat-Schneider DRT1 radio achieving 39.3 dB, 2.3 dB better than his previous best! Now he is just 0.7 dB away from the rarified atmosphere at 40 dB and above where only two have ascended so far under real (non-lab) conditions.

[df9rb]

Hi all,

I built the up-mixer described by hsseppo and Seppo sent me a CD with the DRM testsignals (again thanks to Seppo). First of all: if you build this mixer also do not connect the capacitor at the input between the input and the potentiometer. The correct placement is between pin 1 and the potentiometer! I think pin 1 must be floating as pin 2. I did not get any modulation at the output with the original schematic.
Here first results:
audio fed directly to the sound card: SNR 55 dB
DRM signal via the up-mixer to my TS850: SNR 44 dB
The SNR via the up-mixer is very level sensitive above a SNR of 35 dB. Now I will try to optimize the level chain (up-mixer, manual Gain control of the TS850, level to the 12 kHz mixer, input level of the soundcard). Perhaps I can also achieve the 50 dB boarder.

73s Bernd, DF9RB

[df9rb]

Hi all,

I made some additional tests with the DRM-signal directly fed to my TS850 via an up-mixer. Here the results:
- activated Bandpass Filter in Dream always deliveres the best SNR
- Modified Metrics has no influence
- Number of Iterations has no influence
- default settings of Channel Estimations and Time Sync are the best
- Input Level -10 dB is the best (even though this is displayed in red)
- the 6 kHz Filter!!! is the best even though the Input PSD is strongly attenuated at the edges
- it is better to adjust higher amplification of the soundcard input and reduce the level at the 12 kHz down-mixer.
- high SNR is very level sensitive
- the maximum SNR of the TS850 is 44.5 dB varying 1 dB even under this stable "lab"-conditions

I also connected the up-mixer (12 to 467 kHz) output to the down mixer (467 to 12 kHz) input to determine the possible SNR of the mixer itself. This don t work!. At the up-mixer output the carrier, upper and lower sideband is delivered. These signals modulate the down-mixer and no decodeable DRM signal is available anymore.

Next I will test my Sangean

Bernd, DF9RB

[df9rb]

Hi all,

now I determined the max SNR of my Sangean ATS909. The picture shows 4 modi:
- supplied by batteries, manual gain control - SNR 28.5 dB
- supplied by batteries, AGC - SNR 22 dB
- supplied by the original Sangean power supply, manual - SNR 22 dB
- Power supply, AGC - SNR 20.5 dB
Till the end of the diagramm again with batteries and manually adjusted
Every measurement was done at - 10 dB input level to the soundcard.

Bernd, DF9RB

[FritzWue]

- the 6 kHz Filter!!! is the best even though the Input PSD is strongly attenuated at the edges

Hi Bernd,

found the same here when modifying my Yaesu FT-840.
The narrow CFW455HT filter with -6dB/6kHz gave best results, especially under adjacent interference conditions.

[hsseppo]

Hi,

as there has been some questions about my upconverter, I include the schematics on this thread.
To get a high SNR source I made a CD using the following method:
My soundcard is Terratec 24/96 EWX and I have Steinberg WaveLab Lite software. So I used Dream -t with sound out to EWX 24/96 wave audio and WaveLab Lite was recording 12kHz audio on disk. Then I made 30 minute recording and burned adio.wav on CD-rom. I used a standard CD-player to feed my up-converter. When CD output is connected direct to PC audio input I get 61.8 dB SNR. If I use EWX 24/96 S/PDIF my SNR is 66.0 dB

73s Seppo.

Thanks to Bernd, he noticed my mistake in the dream-up-converter. Now it should be ok.

[df9rb]

Hi,

I tested the DRT1 from Sat-Schneider with a signal from an upconverter and the DRM-CD from hsseppo.
The DRT1 delivers an amazing SNR of 47 dB!
(the varying of the SNR are caused by some alignment trials)

Bernd, DF9RB

[VE3MEO]

Re: Max SNR table - the ascent to 40dB AND BEYOND!

Here are the latest changes to the table.

73, Tom

Aug 15:
Fritz Wue (DK6QI) made 39.9 dB with his DRT1 radio, a 0.6dB gain from July 18. more

Aug 23:
Fritz breaks through 40 dB with a bang! 41 dB peak SNR. more

Sep 9: a BIG day!
Fritz hits 41.9 dB from DW on 7265 kHz... more and more

Bernd (DF9RB) blazed into first place with 43.6 dB from VoR on 15780 kHz on his DRT1! more.

Looks like the DRT1 is the highest SNR commercial receiver on the chart.

[VE3MEO]

Re: Re: Max SNR table - the ascent to 40dB AND BEYOND!

Here are the latest changes to the table.

73, Tom

Oct 3: Daniel, DF8UO, joins the rare ranks of the off-air 40dBers, bringing the count to 5 reported!

The last two months have seen dramatic results - thanks to the DRT1 in many cases, perhaps general learning and optimisation at both senders and receivers, and maybe propagation. We have seen hsseppo do a loopback test attaining 51.4 dB so it's possible that we will see some advance yet on DF9RB's 43.6dB off-air.

[VE3MEO]

Re: Max SNR table - the ascent to 40dB AND BEYOND!

Here are the latest changes to the table.

73, Tom

Oct 14: Carsten (carknue) advanced two positions in the table with a new best of 36.3dB with his AOR7030. He also achieved 32dB with his new Digital World Traveller from Coding Technologies and Degen DE31 active loop antenna.

With the surpassing of the 40dB mark, Carsten has revised the ceiling of the graphical scale on his excellent DRMcalc software to 44dB. I'll bet that's surpassed this winter! DF9RB has come within a 1/2 dB.