Page 1 of 1

The DVB-T2 "1.7 MHz" Standard

Posted: Thu Nov 04, 2021 5:10 pm
by G8GKQ
Starting a new thread on this subject as the old one had drifted.

I have been investigating the problem with Patrice's reception of the 1.7 MHz DVB-T2 Paris repeater, and I think that I have the explanation.

During early Knucker testing, I tried to make it receive 1.7 MHz bandwidth DVB-T generated by the Portsdown and I was unsuccessful. On further investigation, it appeared that the tuner clock parameters were being incorrectly set for 1.7 MHz DVB-T/T2, so I corrected then amd the Tuner then worked well with 1.7 MHz bandwidth DVB-T. I included this "correction" in the production software for both Ryde and Portsdown and thought no more of it.

Patrice noticed that the Ryde software did not decode the Paris repeater; eventually, we found that the original "uncorrected" tuner clock parameters would successfully decode that repeater. I was mystified because they were not for 1.7 MHz bandwidth, but for 1.614 MHz bandwidth.

I have now found this table in the DVB-T2 standard:
Clock table.PNG
Clock table.PNG (17.56 KiB) Viewed 9194 times
Comparing the "Elementary Period" values with the Bandwidth values, it appears that the real bandwidth of 1.7 MHz DVB-T2 is actually 1.61443662 MHz. Perhaps it is called 1.7 MHz bandwidth to appeal to the DAB community, whose bandwidth slots are 1.712 MHz wide?

So, if setting up a "1.7 MHz" DVB-T or DVB-T2 transmitter, the actual bandwidth should be set to 1.614 MHz. I have confirmed that setting the Portsdown to transmit 1.614 MHz DVB-T results in a successful decode with the original unmodified software when "1.7 MHz" bandwidth is selected on the receiver.

Can anybody in the industry confirm my findings?

I will restore the original software in the next update of both the Ryde and the Portsdown, and so bring back compatibility with so called "1.7 MHz Bandwidth" transmissions.

Dave, G8GKQ