DATV-Express Project – Jan report – boards can be ordered
Posted: Sat Feb 01, 2014 4:27 pm
DATV-Express can be ORDERED!!
It’s been a long, slow 3 year process developing a quality state-of-the-art digital ATV transmitter board. The DATV-Express project team is pleased to announce that the software is now ready for production release and the boards can now be ordered via PayPal on the PURCHASE page of our web site at www.DATV-Express.com. (Please note that you need to be logged in to the web site in order to see the PURCHASE page.)
Charles G4GUO worked throughout January trying to refine the “PCR jitter” design work-around that had stopped the start of sales at the beginning of January. In December the video was freezing after a few minutes because the video time stamps were arriving after the presentation-time-stamps. Slowly, but surely, the timing design was improved and Transport Stream analyzers now say the TS quality is good. Ken W6HHC ran a 12 hour transmission test without any interruptions. DVB-S now works very well.
Charles, G4GUO also discovered a programming math mistake that had broken the DVB-T protocol back in November when some functionality was moved into the FPGA. The bug was fixed and now DVB-T is working again. For DVB-T protocol, at this point 6-7-8 MHz configurations of bandwidth have been reasonably tested. Testing has started on the new ham 2-3-4 MHz bandwidths of DVB-T, but this testing is not completed at this point.
Charles also tried to mathematically improve the nearby alias spurs that occur with DVB-T signals using his original “x1 iFFT math” (spurs are shown in Fig 1 using 2 MHz bandwidth)
Figure 1 - DVB-T 2k mode using 2048 point iFFT with aliases
With a little mathematical and DSP magic, using 4096 point iFFT and using only the centre DSP bins, the others bins filled with zeros….the spurs have disappeared. See Fig 2. The main problem with the “x2 iFFT math” is that it doubles the USB2 traffic/load for IQ stream. This “x2 iFFT math” will work OK for a DVB-T 2 MHz bandwidth signal. But at DVB-T 7 MHz bandwidth….the IQ stream will exceed the capacity of the USB2 interface.
Figure 2 – DVB-T 2k mode using 4096 point iFFT
The main priorities of the project team for the next month will be to focus on the initial sales process, respond to the inevitable tech support e-mails that will come in, and finish building the next batch of production hardware boards.
“full speed ahead”….de Ken W6HHC
It’s been a long, slow 3 year process developing a quality state-of-the-art digital ATV transmitter board. The DATV-Express project team is pleased to announce that the software is now ready for production release and the boards can now be ordered via PayPal on the PURCHASE page of our web site at www.DATV-Express.com. (Please note that you need to be logged in to the web site in order to see the PURCHASE page.)
Charles G4GUO worked throughout January trying to refine the “PCR jitter” design work-around that had stopped the start of sales at the beginning of January. In December the video was freezing after a few minutes because the video time stamps were arriving after the presentation-time-stamps. Slowly, but surely, the timing design was improved and Transport Stream analyzers now say the TS quality is good. Ken W6HHC ran a 12 hour transmission test without any interruptions. DVB-S now works very well.
Charles, G4GUO also discovered a programming math mistake that had broken the DVB-T protocol back in November when some functionality was moved into the FPGA. The bug was fixed and now DVB-T is working again. For DVB-T protocol, at this point 6-7-8 MHz configurations of bandwidth have been reasonably tested. Testing has started on the new ham 2-3-4 MHz bandwidths of DVB-T, but this testing is not completed at this point.
Charles also tried to mathematically improve the nearby alias spurs that occur with DVB-T signals using his original “x1 iFFT math” (spurs are shown in Fig 1 using 2 MHz bandwidth)
Figure 1 - DVB-T 2k mode using 2048 point iFFT with aliases
With a little mathematical and DSP magic, using 4096 point iFFT and using only the centre DSP bins, the others bins filled with zeros….the spurs have disappeared. See Fig 2. The main problem with the “x2 iFFT math” is that it doubles the USB2 traffic/load for IQ stream. This “x2 iFFT math” will work OK for a DVB-T 2 MHz bandwidth signal. But at DVB-T 7 MHz bandwidth….the IQ stream will exceed the capacity of the USB2 interface.
Figure 2 – DVB-T 2k mode using 4096 point iFFT
The main priorities of the project team for the next month will be to focus on the initial sales process, respond to the inevitable tech support e-mails that will come in, and finish building the next batch of production hardware boards.
“full speed ahead”….de Ken W6HHC