My Digilite test and aligment with Tutioune

[F6DZP]

Hello,

Firstly many thanks to all of you that have spend time for this wonderful tool.
I have finished 2 days ago to build my Digilite board set on 437Mhz, 1MS - (I like very much this value using H264 coding with my others modulators)
The LO is the Ultram board set on 437 Mhz with its MMIC and the attenuator on Digilite Board.
I have a PVR150 in my PC and started to study it just before.

When I put the power to the Digilite board and used Tutioune to monitor the signal, I see very quickly the constellations and could tweak the trimmer to obtain 18db MER.
It's a really good value, knowing that DVB demodulator have much trouble to decode when Symbol rate is low.

digilte 437 -1MS HP8657B -3dBm Tutioune lite.jpg (238.71 KiB) Viewed 10686 times

Using Tutioune I can monitor constellations and MER and IQ level , Offset and Imbalance.
So no need of a spectrum analyser when all go well. I think Constellation and MER are one of the best tool to adjust all parameters. (But if you have more serious problem a spectrum analyser can help)

As I had seen several post that say that we must remove the ultram MMIC and the Digilite attenuator and also put the Ultram in a metal box.
I have done that and now the result is no so good : the 4 constellations clouds are now placed in a losange figure and MER is 14dB.

digilte 437 -1MS Tutioune without MMIC.jpg (491.51 KiB) Viewed 10686 times

I have tried also my HP8657B generator and Digilite board without input attenuator and I have observed:
HP RF level : -7dBm –6dBm –5dBm –4dBm –3dBm –2dBm –1dBm 0dBm 1dBm 2dBm
MER : 16dB 17dB 17.5dB 18dB 18dB 18dB 18dB 17.5dB 17dB 16dB

So in my situation, without attenuator, better values are with 437Mhz input from –4dBm to –1dBm

I will go back and put again the MMIC and attenuator on Digilite Board, results were the best.

Many things to say also about software, I will make a separate post.

digilte 437 -1MS HP8657B -3dBm Tutioune with video.jpg (296.67 KiB) Viewed 10686 times

Many many thanks, I ‘m very happy with that little magic board, and was very surprised that it worked as soon as I plugged 12v on it.

Jean Pierre F6DZP

[F6DZP]

I have also made some measures with my SEFRAM 7856:

digilte 437 -1MS HP8657B -3dBm spectre.jpg (76.3 KiB) Viewed 10685 times

digilte 437 -1MS HP8657B -3dBm MER et VBER.jpg (84.85 KiB) Viewed 10685 times

digilte 437 -1MS HP8657B -3dBm constellations.jpg (64.55 KiB) Viewed 10685 times

Jean Pierre F6DZP

[F6DZP]

all my test in real time are on my Tioune Monitor:
http://www.vivadatv.org/tutioune.php?om ... ation_id=1

Jean Pierre F6DZP

[G4EWJ]

I wonder if your results would be the same on 23cm? The datasheet recommends -10dBm LO into the modulator, but others have found that increasing it to -6dBm reduces the noise. I am running with -6dBm at the input to the balun.

Brian

Tutioune2.JPG (293.92 KiB) Viewed 10658 times

[F6DZP]

RF values I gave were those on the 8657B output setup, not measured at Board input.
Sure I can have 2 or 3 dB lost with not very good connection.
so We can suppose that the real input was about -6dB
Anyway, choosing very well this level enhances the output signal and we can observe that with Tutioune.

I see you have a better MER, I suppose that it's because you are running 4000 kS.
Have you try to measure MER when set to 1000 kS as I do?

I really like 1000 kS with H264 coding. I hope we can very soon use TS files instead of PS files.
Have you tried to make a Dump in a file of the TS obtained with the tuner graph?
If you do that and Digilite transmit accept TS file you have instantaneously a DATV repeater.

Jean Pierre F6DZP

[G4EWJ]

I only have the filter components for SR4000 at the moment, but I will try to test SR1000 at some time.

The UDP version of DL Transmit is working well with TS input. I need to fix a bug in the PS processing and then I will put it out for beta testing. I have constructed a graph with the KNC-One filter and a UDP sender filter and can receive a DigiLite transmission with the KNC-One PCI card on 1249MHz and re-transmit it via DigiLite on 1280MHz.

Brian

[F6DZP]

I would be very glad to test UDP mode
When I choose UDP in Diglite transmit, I can't find any possibilities to tell which Multicast address , port and NIC I use.
How can we setup these 3 values?
Multicast address and port can be fixed if you want, be I need to know which values to use.
NIC must be set on each computer , taken between all values that are available.

Last question:
When I create my TS, must I set the rate to exactly the good value, creating by myself stuffing packets : 1382352 for 1000 kS FEC3/4
or can I use any value under this value and your software add by itself stuffing packets?

Jean Pierre F6DZP

[G4EWJ]

I'll let you know when UDP TS is ready.

Here's a snapshot of a future version of DL Config. It has multiple profiles for different settings and bands, configurable PCR PID, extra test modes for toggling sidebands automatically and the UDP and TS input modes and IP / port settings.

I have only written software for a TS with exact bit rate. Adding padding would be more difficult. I would have to adjust all the PCRs, which would mean decoding the PAT and PMT to find where the PCRs are.

Brian

DLCF150B.JPG (92.06 KiB) Viewed 10598 times

[F6DZP]

Thank you for this important new future version.
I you want, I can do beta test.

I have not seen the UDP Multicast IP setup.
I suppose in your example that 192.168.55.234 is the Network IP used by the sender, we have the UDP port 64802, but no UDP IP?

"I have only written software for a TS with exact bit rate"
OK, there is no problem to set the good TS value when we create it.

I am impatient to try this version.

Jean Pierre F6DZP

[G4EWJ]

I am still learning my way around UDP and this level of networking, so I have just read a bit about multicasting. I thought that point to point UDP transmission would be all that we need for DATV. When would multicast be used?

Brian