What sort of form factor would people want (base/mobile/portable)?
What sort of money would people be prepared to pay ?
The parts are slowly coming together, we have the PI for MPEG4 encoding/decoding
The Pi camera for video capture, Element14/Wolfson are about to release a proper sound board
for the PI. There is the 9 inch HDMI PI screen on Kickstarter. So what we are left with is providing the
RF front end and the signal processor.
I had a look on the Analog Devices website and they produce front end chips designed for BPL / ADSL
which could be used to produce a transceiver with an IF of around 21 Mhz, which could then be mixed
up to the final frequency.
I am currently thinking about using OQPSK (Offset QPSK). I have an old TI appnote that describes their
implementation for one of their DSP chips. At these narrow bandwidths it might be better to use a DSP
rather than an FPGA.
I know the VHF consultation has been delayed but I think we should have a plan ready to present to OFCOM
as to what we are going to do with these new frequencies. Probably we will need to knock up a couple of
prototypes based on commercial EVMS so we can lay claim to the frequencies as soon as the NoVs are
available. Producing a low cost system can then follow.
- Charles G4GUO
Also we have the problem of working with SMD multipin chips and people's ability to deal with such devices let alone see what they are doing or indeed have a steady enough hand to do any such soldering. So ready made sub assemblies are going to be a requirement, and some people have difficulty drilling all the right sized holes in a box so is this a serious kit? i.e one with all the parts ready to assemble and wire up? Such as a Comtech controller and box.
There are a lot of amateur bands that are often not used much, but we suffer with band planning where a contiguous section of band never seems to be available, a for instance of this is 70cm, where we have narrow band activity at the bottom then packet, then repeaters then space communications then more repeaters, hang on where is the clear space for ATV? Plus of course the primary user.
At present 70cm is our main Dx band as it can go round hill tops better than any of the higher frequencies and is less attenuated by local trees than 23cm and up. We are often limited here to exceptional conditions and temerature inversion to get long distance contacts, and 175kM is the typical limitation.
F2 layer propogation comes in at 28MHz, sometimes as low as 21MHz, but certainly more regularly on 50MHz, 70MHz and 144MHz and the question comes round as to what we would suggest as the technology for these bands. More immediately is the offer of more 2metres, and we need to have enough clear space without interference to make a contact possible. So what do we need to employ, how much will it cost and what bandwidth can we squeeze it into, and what tools can we use to line up antennas, before a decode is successful.
During the Analogue TV days I had a Tv with the european channels E1 to 4 covering 45 to 65 MHz. During the summer months Iceland was first in the morning, then as the day progressed Sverige, then Poland, finishing up with some Italian then Spanish, and if lucky Africa having seen a Zimbabwe test card. That space is largely vacant. Of course BBC1 405 lines on 45MHz from Alexandra Palace was seen in Australia.
On 4 metres I have easily worked North Africa at 30 over 9, well enough for a DATV signal.
So the question is to define a system that can be used on the lower bands without taking up too much space and then get asome frequency allocations sorted out.
(As seen in Radcom!)
The RSGB's VHF manager has some ideas of the type of thing he would like to see on 147 MHz
and the way to achieve it. Currently his thinking is towards 500 KHz channels and some form
of COFDM. They are hoping to have a demo system available for the RSGB thing in the Autumn.
For transmit / receive the use of SDRs and GNURadio is being considered. At these fairly slow
data rates a fast PC should be adequate and you could probably get away with a Realtek Dongle
for receive or maybe a HackRF for transmit/receive.
I will probably use my USRP2 which has been sitting idle for a number of years.
For the video encoding a fast PC means we can use something more efficient than MPEG2.
Google's VP9 open source codec comes to mind although it is not designed for Video Conferencing at the
moment but that is in the works. It has a similar compression efficiency to HEVC (about 1/4 the bitrate
for the same video S/N as MPEG2).
So my line up is looking like a 2m 60w power brick (run at about 10w), a USRP2 SDR, GNURadio, GStreamer
a fast PC and an awful lot of filtering. Probability of me actually working anyone on the band is virtually zero
because of my location. But at least I will be able to wear the tee-shirt
- Charles G4GUO
I think it maybe already with internet gateways & TV repeater streaming etc.G4GUO wrote:Who said anything about Windows?
One day DATV will just be another downloadable Android/iOS app on your smart-phone.
I tend to agree with Mike's thought train about PC's I have to admit.
Unfortunately the 147 MHz allocation may be gone in a couple of years time and
it is simply so much faster to develop for the PC.
ARM devices are becoming faster and faster, we already have the MK808 which is a dual core
and recently I bought an MK802IV which is a quadcore device which has Linux support and Rikomagic
http://www.rikomagic.co.uk/are coming out with an even faster TV gizmo in July.
DVB-S (QPSK) on 1255MHz or 437 MHz can be done at low SR : 250 kS/s using a digilite modulator and a RaspberryPi for H264 encoding.
Tutioune is used for receiving with S2-3200 card or S2-1600 card.
Several tests have been made by me and Evariste F5OEO this summer.
sample can be seen here : http://www.vivadatv.org/viewtopic.php?f=72&t=332
I think I will reach soon 100 kS/s.
next test will be tests using DVB-S2 up to 32APSK modulation.
Jean Pierre F6DZP