CQ Datv82

In this issue
Editorial............................................................. 2
News & World Roundup......................................4
5.8 GHz Transverter........................................... 7

Simple Video Testcard...................................... 11
Micro Corner ASCII Speech Module................ 14
Epson Programmable Oscillators..................... 17
2 Watt 2.4 GHz DVBT Transmitter................... 20
Grass Valley Mixer Conversions Part 15......... 23
One from the Vault........................................... 26
Information......................................................29
The CQDATV editors gratefully acknowledge

all those authors that have contributed
articles for this free magazine.
Production Team
Ian Pawson G8IQU Trevor Brown G8CJS
Terry Mowles VK5TM Jim Andrews KH6HTV
Contributing Authors
Jim Andrews KH6HTV Trevor Brown G8CJS
John Gebuhr WB0CMC John Hudson G3RFL
Rovert Nworb Mike Stevens G7GTN
CQ-DATV 82 - April 2020 Page 1

Editorial Move over commodore the ESP 8266 is here and the new
project has been costed out at under £5.
Welcome to issue 82 of our electronic ATV magazine.
We find ourselves in very serious and troubled times. Our John Gebuhr, WB0CMC is looking at Epson programmable
thoughts go out to all our readers around the world, and their oscillators. The problem is he needed one to work at 27MHz
loved ones. so he could multichannel his new ATV transmitters, but too
many artefacts were confusing the PLL. John was not put off
We hope CQDATV 82 supplies a little escape from these and has some solutions.
problems and brings a little enjoyment into the lock down
that most of the world is now under. What have we got to Trevor G8CJS has produced another chapter in his GVG mixer
distract you? well in this issue? modifications and unfortunately has not yet got it producing
the MIDI data that was discussed in the last issue. The
Jim Andrews KH6HTV has written an ATV handbook. You can Arduino has not made it through from Mike G7GTN. Only a
get your free copy from Jim’s Web site (see News item for pause in the development and Trevor is using the time to
details). Jim has also found time to visit his backburner recap the project and explain where he hopes to go with it.
projects and revive and finish his microwave transverter. This
has put him on 5.8GHz’s but not with FM, but with DVBT. No issue of CQDATV would be complete without “One from
The output power +25dBm and a receiver sensitivity of the vault”. This is a look back at issue 25 and answers the
96dBm. Jim has also revived a 2.4 GHz 2watt transmitter question of why there are so many codec’s.
again for DVBT. Jim describes it as the BEAST, but Jim is
good at taming beasts. The PA used to be 10w, but it’s now I hope this issue has spread a little light into your lives at this
down to 2.5 watts delivering DVBT signals and can be driven difficult time and that by the time CQDATV 83 is published
directly with the Hi Des 320E modulator. that things are starting to improve.
John Hudson G3RFL has found his way back to the shack and From everyone at CQDATV we hope that you and your loved
has made the decision that his old test card generator is ones are safe from this horrific virus.
getting a little long in the tooth, with some of the parts now
becoming obsolete (we were struggling too John). The Take care the CQDATV editorial team
solution is to go back to the drawing board and design a new
one. This one can store a Jpeg picture. John has sent in a
family photo stored in this new design and is working on
software so that we can all follow suite. Thanks John, can’t
wait to try it with the Test Card Maker software. Please note: articles in this magazine are provided
with absolutely no warranty whatsoever; neither the
Mike Stevens G7GTN has been looking at electronic speech. contributors nor CQDATV accept any responsibility or
Some of you may remember the SPO265, well Mike is on the liability for loss or damage resulting from readers
case and takes us back through earlier incarnations, including choosing to apply this content to theirs or others
the commodore 64 and SAM Speech. computers and equipment.

TV Amateur is a German Language ATV Magazine.
It is published 4 times a year.
Please note the website is currently offline.
News and World Round-up Advances in robotics, IP and image sensors over the past
decade have lead to expanding markets for PTZ (pantilt
Guide To PTZ Cameras eBook Is Now Available zoom) cameras. These workhorses have moved beyond
Produced by the editors of TV Technology. traditional surveillance to inside studios and sporting arenas,
giving new and unique points of view for viewers.
In our new ebook on PTZ cameras, we take a look at their
potential uses on the set and in the concert venue as well as
take a look at how AI will enhance PTZ production.
Learn more in the latest free Guide to PTZ Cameras eBook.
Read it free now click here.
New ATV Handbook
Back in 2017, Tom O’Hara, W6ORG, was

writing the chapter 32 on ATV for the
ARRL handbook. He asked me for some
input on DTV to include in his chapter.
Tom’s chapter in the 2018 handbook was
31 pages in length and covered well the
subject.
This past fall, Bill, AB0DH, loaned me his 2019 ARRL

Handbook to look at. I was dismayed by the extremely brief
writeup given in it to ATV. I expressed my displeasure at the
Boulder Friday morning ham breakfast. Larry, K0PYX, made
the suggestion – “Jim, if you are that displeased, why don’t
you write your own ATV Handbook?” Well, I took up Larry’s
challenge. I have just finished writing it. I have entitled it
“ATV HANDBOOK an Introduction to Amateur TV”.
I have just posted it on my web site: www.kh6htv.com as my

application note, AN55. My new book is 39 pages in length. I
broke it into the following chapters covering: Basics of TV
Video Signal, TV Video Sources, Amateur TV bands, TV Signal
Quality Reporting, TV Modulation Methods, TV Receivers, TV
Transmitters, TV Antennas, TV Propagation, TV Repeaters, TV
in ARES, and a list of ATV Suppliers.
73 de Jim Andrews, KH6HTV
BATC and Corona The 2021 WIA Conference will take place in Hobart.
In order to support the worldwide amateur The WIA board would like to encourage all operators to
radio community during the COVID19 virus practice social separation, wash and dry your hands regularly
pandemic, the BATC is offering free use of and promptly seek advice if you feel you are affected by
its video streaming service and chat facility COVID19.
to any radio club or group of radio
amateurs. One event that would have gone over the 100 'guest rule' as
to take place just north of Brisbane in VK4 so REDFEST 2020
has been cancelled.
This will enable clubs to hold talks, presentations amd virtual
meetings with HD video and audio streaming and a "Greetings from the Redcliffe & Districts Radio Club VK4RC,
membership chat window for real time feedback and right here in south East QLD. Robert Thomson VK4TFN here,
discussion. with a REDFEST 2020 update.
To enable this, we are offering free oneyear “cyber This is indeed a sad turn of events. However, the Executive
membership” to any radio club or group of radio amateurs. has had no choice but to cancel REDFEST 2020, due to the
current public health concerns.
Full details of how to apply are here on the BATC Wiki.
https://wiki.batc.org.uk/Radio_Club_membership_2020n Next year, there will be a REDFEST 2021, circumstances
permitting.
Australian (Ham Radio) Corona Casulties
Along with the many other cancellations of events throughout
No news presentation seem complete these days without the country where there are a number of people likely to
some mention of the COVID19, the Novel Coronavirus that attend them Amateur Radio New South Wales has done
has driven panic and fear into the worlds economic and social likewise and cancelled the scheduled Sunday March the 29th
engines. Trash and Treasure event along with the Experimenters Group
lecture which would have been held later in the day.
And this broadcast is no exception
Other ARNSW activities of members at the Dural site are
The health and well being of the Amateur cohort and our under review. Interested parties are reminded to listen to the
members is of the highest priority. Out of an abundance of Sunday VK2WI News sessions for updates.
caution, consideration for the health and wellbeing of our
members, and in an effort to deliver certainty to the Amateur Notice of cancellation of The Urunga Radio Convention
community (not to mention a federal government mandate),
the board of the WIA met last Wednesday and endorsed the The Urunga Radio Convention Committee advises that the
recommendation of the 2020 WIA Conference Organising Field Days to be held at Easter weekend, 11 and 12 of April
committee to cancel this event. 2020 has been cancelled due to the restrictions and
conditions placed upon the Convention by the various
government authorities in the light of COVID19 pandemic.
It is the first Urunga Convention that has failed in 72 years

and it is a personal blow to the organising committee.
We wish to thank all our supporters and we will return at
Easter in 2021.
VK3 EMDRC Committee after due consideration, have

decided to cancel/postpone 2020 EMDRC HamFest on Sunday
29 March.
VK3 GippsTech 2020 has been cancelled
VK6 Bob VK6POP in a phone hookup has told how PerthTech

at Bassendean is to be cancelled.
VK7 Meet the Voice organiser Dani VK7FREQ also has made
the decision to cancel the Meet the Voice gathering.
ZL For our Kiwi friends your Conference 2020 is cancelled.

NZART Council has made the decision to cancel the
Conference in Shantytown this year. This is due to the
COVID19 situation and taking on board all the advice for
doing so.
Source: Excerpts from VK National News Broadcast

MARCH 22 2020
This is your free ATV magazine.

Please consider contributing an article!

5.8 GHz Transverter
Written by Jim Andrews, KH6HTV
Reprinted from Boulder Amateur Television Club TV
Repeater’s REPEATER March, 2020
Don, N0YE, had previously built over the years, several

homebrew, 10 GHz, SSB rigs. He loaned these out for some
initial 10 GHz, DVBT, outings this past fall.
I got interested in what I might be able to throw together for

a microwave DTV rig. My junk box is pretty sparse for 10 GHz
stuff, but I did have an assortment of misc. C band (48 GHz)
components. So, I decided to try to build a 5.8 GHz
transverter for DTV. My initial plan was was to use my HiDes
Back in August, shortly after my ATV Repeater article HV320E modulator and HV120 receiver as the IF transceiver
appeared in QST, I got an email from Fumio, KA0RUZ, in and to use the new Analog Devices microwave frequency
Japan. He described their microwave DXpreditions where synthesizers as my LO for up/down converting. I discussed
they achieved a distance of 287 km on 5 GHz using the these AD synthesizers in the October newsletter, # 21. I took
Japanese equivalent of DVBT (ISDBT). Fumio also included the old 2.4 GHz transverter I had built a couple of years ago
a link to You Tube to see them in action on 5 GHz. This and gutted it for parts and rebuilt it for 5.8 GHz service. This
immediately triggered an interest here in Boulder among past fall, then Don, N0YE, and I took to the field to test it
several ATVers to get busy pushing DVBT to the higher out. Don set up his rig on Flagstaff Mtn. and I set up mine in
microwave bands, beyond 23 cm. See the Sept. newsletters, my back yard where I had a direct lineofsight path to
issue # 19 & 20, Oct. #22. Flagstaff. Don immediately reported seeing my DVBT signal.
Then he proceeded to transmit to me. NOTHING ! We were
both running similar rf power levels and similar antennas, but
I was not able to see anything. Bummer ! – Back to the lab.
The project then ended up on the back burner for awhile as I

worked on other projects. I intermittently went back to it
trying different combinations of IF frequencies, different local
oscillators, mixers, etc. I concentrated on the 5.8 GHz
receiver sensitivity. What I discovered was very “eye
opening”. After many false starts, I realized that a major
issue was poor phase noise in the Analog Devices frequency For my receiver, I found an old WatkinsJohnson mixer from
synthesizers. While they worked fine for receiving high level my junk box worked well. I used a DownEast Microwave,
DVBT signals, when one really got down to the 80 90dBm model L5ULNA, as my preamplifier. Testing it on an HP noise
range, NO Picture ! The poor phase noise destroyed the figure meter, I found the DEMLNA had 15dB of gain and a
signal to noise ratio. Another eye opener was the discovery 1.1dB noise figure. With this receiver, if I only used the WJ
that some diode mixers also were crummy performers for mixer, the DVBT sensitivity was 92dBm. Adding the LNA,
really weak signals. the sensitivity was improved to 96dBm.
So after many hours ( really days & weeks ) of trying various For the transmitter chain, I used the new HMC219N mixer
combinations, I finally hit upon a workable solution. It is from China. It was a very poor performer for a receiving
shown in the above block diagram and also the detailed mixer, so I used it instead in the transmitter chain. The
schematic on a following page. I scrapped out the idea of amplifiers used in the transmitter chain were from my
using the Analog Devices synthesizers. I am now using a low microwave junk box. The driver was an Avantec AMT8052.
noise, Frequency West brick oscillator as my LO (see Oct. The final amp was an Amplica 6535CSL. Using a mixer one
newsletter #23). Because it was a gift from Don, N0YE, there gets as the output, the desired sideband, in this case the
was no choice of the LO frequency. I had to take what I was lower sideband, plus the undesired upper sideband and also
given. The LO crystaled to work on 6.0924 GHz. Thus to some leakage of the LO frequency. I thus needed to filter out
operate on 5.678 GHz, my IF frequency had to be 414.4 MHz. the LO and USB.
I accomplished this using a pair of simple tee notch filters.
They were SMA tees with a short piece of opencircuited, RG
174 coax cable on the third arm. I used my Wiltron network
analyser to fine tune these filters. I made the coax initially
too long and then using wire cutters to carefully trim the coax
length to put the notches on the desired frequencies. The
notches were about 26 to 28dB in depth and I had less than
1dB loss at the transmitter frequency of 5678 MHz.
An SMA coax relay (again from the junk box) was used as the
antenna switch. I controlled the timing of the turn on / turn
off of the various amplifiers to avoid transmitting back into
the receiver. I used LM2941 low dropout voltage regulators.
I added an RC circuit to the enable pin on these regulators
to slow their turnon.
In operation, I leave the HiDes receiver powered up all the

time. I only power up the HiDes modulator when I want to
transmit. There is sufficient rf leakage to the receiver when
transmitting that I am able to use the receiver to monitor the
transmitted video.
So, how well does it work? The transmitter is definitely NOT

high power. It is a milliQRP rig. For DVBT, with the
modulator’s rf drive power adjusted so the spectrum skirts
breakpoints are set to 30dB, the output power is a
whopping +14 dBm ! The max. saturated output from the
Amplica amplifier is +25dBm. The receiver is quite good
(finally ! ) with a sensitivity of 96dBm when tested with
“Normal” digital parameters (1080P, 5/6 FEC, etc.). The final
acid test was to go out in the field and exchanging pictures
with Don, N0YE. The following photos are proof that it really
works !
I set up my rig at the Boulder 911/EOC near the Boulder

airport looking south toward’s N0YE’s QTH. Don lives on a
high ridge line on the south side of town. We have a good
lineofsight path between the two locations. The path
distance was 7.4 km. We both used +23dBi dish antennas.
Don was transmitting +10dBm with 4dB coax cable loss. I
was transmitting +14dBm with 0.3dB cable loss. I was able
to receive Don’s picture, as shown in the photo.
The received power was measured to be 83dBm with a s/n

of 14dB. Radio Mobile computer program predicted that I
would get 79dBm from Don.
Don was not so lucky. He did not receive my signal. Don said
he was having issues with his HV110 receiver locking up. His
receiver stopped receiving even his own transmitted signal a
few minutes before I started transmitting to him. Rebooting
several times didn’t solve the problem until much later back
in the ham shack on the work bench.
Whew ! – you know this microwave DVBT stuff was a lot

more work and required an awful lot more parts than doing
this with the elcheapo 5.8 GHz, FMTV gear ! ! !

Teletext input if required.
Simple Video Testcard There was also page
Written by John Hudson, G3RFL selector for stepping
through the preset test
I often have a requirement cards on an external push
for a simple video pattern button. The RGB and sync
generator and for a long time signals fed a modern AD724
I used my old favourite, my surface mount RGB to PAL
electronic test card that used encoder mounted under the
a MACH or Lattice field PCB.
programmable array. It
originally appeared in CQ The problem was It was not
DATV 8. easy to programme the
MACH and the EPROM. I
It used two programmed chips the MACH and a 27C512 made many of these units
EPROM. The logic was programmed into the 44 pin MACH. and the programming alone
This was shown has a block diagram rather than the more took about 2 hours.
conventional circuit diagram which would show all the TTL
gates of the array. This had several counters such as Pixel, This unit now getting a little dated and at least one of the
Character Column, Sub Character, Line and Character Row. chips (EXO3) has become obsolete. I thought it might be
The main clock was at 8MHz and was sub divided down to 4, time to go back to the drawing board and come up with a
2 and 1 MHz the next section was the EPROM multiplexer unit new design. I settled for a PIC design, yes it would mean
having a latch and 11 x 4 multiplexer that controlled the writing some fresh code, but into everyone’s life a little rain
Column, Line, Row Character and Palette. Also, in the MACH must fall.
was the bit pattern generator that had an 8 x 2 multiplexer, 4
D type latches a further 8way multiplexer and a 250 nano The new unit still uses the same PAL/NTSC coder, the AD724
second digital delay line. as this produces some pleasing results, is not too expensive
and still available from several sources although now the
The final part of the MACH was a broadcasters have gone wide screen PAL is technically
Colour/ Control latch which obsolete, but many of us still live in the world of yellow, red
switched between the and white phono plugs, but it might be time to stock pile
background/Foreground Palettes, some of the coder chips, before they also start to disappear.
this had another 6 x 2way
multiplexer and 6 D type latches I think the AD724 is my preferred choice for all future
with extra logic for control, plus PAL/NTSC projects. I also like the ability to use subcarrier
field reset with interlace. Then we xtals or 4 times subcarrier xtals by changing the logic level
come to the video output stage on pin 12. I have both in my junk box.
buffering the RGB and supplying a
The new design is at present on a home etched PCB In the new design the PIC is the only custom programmed
chip, while in the old design I had to programme individual
details into an EPROM and batch programme the MACH chips
which thankfully were all the same.
The new testcard is easier to programme than the old one, I

have developed software which will take a Jpeg file and
convert it to PIC code, which can then be programmed into
the unit. The new generator also produces a VITS (Vertical
Interval Test Signal) on lines 17,18,330,331. The results are
very encouraging, with an enlarged colour pallet, which is
composed of 5 bit red, 5 bit blue and 5 bit green, or 32 levels
or shades of all three primary colours if you prefer to look at
it that way. So, 32 x 32 x 32 = 32,768 colours if I have done
my sums correct.
This is considerably more than the old Testcard Generator, so

how does it look, if you programme it with a picture, the
answer is remarkably good, no sign of digital puddling and it
Some of the components are also mounted on the really pushes the AD724 with the enlarged colour pallet which
underside was not produced by the previous design.
I am hoping that this design will work with the Test Card
Maker software from the previous issue of CQDATV so that
you can design your own testcard and then convert your
design into PIC code and programme it into the new unit. A
pity the Test Card Maker seems to have disappeared or I
would be onto the author to add a menu choice to export a
design to G3RFL format file. Ian has put the programme on
the CQDATV download site, so all is not lost, just it is frozen
to future enhancements without the source code.
This is a giant leap for me as I dread to think of the number

of hours I spent working out QTH details and converting them
to hex to programme the old design and that is excluding the
large callsign letters at the top of the old testcard.
My thanks to Dave G3ZGZ for all his work with the pictures,
time to build a new Testcard Generator Dave and I won’t
need to programme it for you (once I get the Jpeg converter
software running).
The full AD 724 data sheet can be found at

https://www.analog.com/media/en/technical
Jpeg picture stored in the PIC of a group of family documentation/datasheets/AD724.pdf
friends
I have put a hex file (PIC_TCG5. hex) on the CQDATV site

that will allow you to programme the PIC with this same
picture and I will make available the programme to convert
your own pictures into PIC code, it is at the moment being
revised to include some additional features, but in the
meantime you can test your hardware.
The output is 625 line and the PAL coder which is PAL/NTSC
is hard wired into PAL by the grounding of pin 1.
The data sheet also insists on C6 to block the DC, which is a

pity, but I suspect it’s a casualty of single rail working.

Micro Corner - ESP8266 Based ASCII Enter SAM Speech
Speech Module Software Automatic Mouth shortened to SAM was originally
developed to run on the Commodore 64 computer and is a
Written by Mike Stevens G7GTN fully software implemented algorithm requiring no additional
hardware for speech generation.
For some of our television shack projects a requirement

might be that of having some speech output available.
Back in 1981 we had devices such as the SPO256AL2 speech

allophone processor from General Instruments. This was a
fully custom device that had the data loaded in a parallel
fashion. The speech phrases are made up in segments called
“allophones” to create words that we might understand.
The quality of output was always a little dubious. So we had a

good measure of tweaking and adding in little delays in our
built up strings. I still have an original SPO256AL2 chip
alongside two other custom chips stored in my secure vault.
Instead of making use of this device and the requirement of

using all I/O pins on available micro controllers I did some
searching to see what other options existed in this Century. I Hardware
found that the author of Annex RDS Basic also created a
custom project called “SAM SPEAKER” that allows text to be The main speech module requires an ESP8266 module that is
spoken via 115200 baud rate pure ASCII serial. That seemed loaded with the custom speech firmware. I used a wemos D1
like a good idea to escape now quite rare speech devices mini alongside a PAM8306 amplifier module as the custom
where the possibility of buying fakes will be higher than back speech module. Using my keypad project as a simple test
when these were actually available and under manufacture. driver to send the custom text as serial data the hardware
ended up as being simple to implement.
Basic Control Software
Using the free Basic for Ardunio a software serial port on

Digital I/O pin 11 was implemented to run at 115200 and
simple text strings sent with the following command format
on a keypad button press.
A handshake signal is available on the speech processor

GPIO0 (D3) that is logic HIGH when the module is waiting for
data and goes LOW when the speech is being processed or
spoken. The analog A0 pin is monitored for this logic level
change and used to indicate the status on the first line
(position 15) of the LCD. Use is made of an internal resistor
to always pull this pin high.
As long as your projects processor is able to send data at the

required baud rate then the control could not be much
simpler.

Follow the link to obtain the custom ESP8266 firmware and They both make use of a Software Serial library that allows
following the instructions to install this on to your chosen the use of a different custom set of I/O pins to send our
module. Since we cannot make use of any additional I/O pins defined messages. Both are available from a single ZIP file
on the ESP8266 using a WeMos D1 Mini made sense as a called SPEECH from the download page.
quite compact dedicated platform. image
Options and quick testing Conclusion
With the ESP8266 module flashed and connected a terminal We still have very 1980’s robotic style speech and so this
set to 115200 can be used to look at available options and might fit in more with personal shack projects where the
also to send text for testing purposes. speech is not actually put on air with repeaters for example.
Using MP3 style SD based modules would of course provide a
better option in that use case. You can check out the quality
of the generated speech
https://simulationcorner.net/index.php?page=sam before
you decide whether to flash an ESP8266 Module. You can also
easily experiment by sending data using your favourite
terminal software via the USB Connection. An added bonus
which has not been covered here is the ability to also control
the speech processor via a network connection.
The cost of parts should be under £5.00 if you use suppliers
such as eBay or similar sources.
I have also created the exact same project in Ardunio C++ if
you prefer this software IDE and programming language. References
https://sites.google.com/site/annexwifi/projects/sam
speakernetwork
https://simulationcorner.net/index.php?page=sam
https://www.smbaker.com/counterfeitfakejustplainbad
sp0256aal2chips
https://en.wikipedia.org/wiki/General_Instrument
https://www.b4x.com/b4r.html

Epson Programmable Oscillators
Written by John Gebuhr, WB0CMC

Repeater’s REPEATER www.kh6htv.com
After seeing the article on the Epson programmable crystal

oscillators I did some checking to see if they would work to
multichannel my new ATV transmitters. They don’t at the 27
MHz range the exciter uses. Too many artifacts that confuse
the PLL in the chip I’m using in it. After some experimenting I
found it takes a fairly pure signal to make it work right. My
siggen will drive it and works over the entire ATV portion of
the band. I then tried some at the 100 MHz range and tried
quadrupling them and driving the amp/modulator directly.
Here is the result using 3 of the most common frequencies.
A three channel frequency quadrupler using a step recovery
diode and a buffer amp. Circuit below.

Top of board showing the 3 RF inductors Foil & component side with the 3 Epson chips
The 2 pins are ground and +5V, the 3 are channel select and The harmonics don’t make it through the transmitter nor do
the SMB is output of about +8dBm. Swaged standoffs not the subs and fundamental. A second hair pin and 210 pf cap
yet installed. Connected to the amplifier in the transmitter I can be added right at the output for more attenuation of
noted that the power dropped slightly from 439.25 to 421 25 other than the desired 4th harmonic. All top inductors are
by about a watt, probably due to the pass band of the amp. #22AWG on 1/8 inch diameter air core. The highest harmonic
It is specified from 440470 MHz so not surprising. The other than shown below is the 14th @ about 30dBc. I think
pedestal level also changed somewhat but video didn’t with some retuning modifications it should work fine for a
change noticeably. 1.2GHz source. Likewise on 902928.
The series LC on the input of U1 is 3rd harmonic trap. The X13 are available from Digikey and are ordered to frequency.
hairpin and trimmer cap is the pass band tuning. It will tune 421.25= 105.3125. 427.25= 106.8125, 434= 108.5 and
the 5th harmonic @ 530 MHz or so but at 430 the 5th is 439.25=109.8125.
about 20 dBc. Other harmonics are even lower. With no The 2 trimmer caps I got from Mouser: P# 768JZ300, 530pf
tuning the harmonics are significant to at least 1.5 GHz. Nice & 768JZ100. 210 pf. The step recovery diode is also from
comb generator, huh? Mouser: P# 937MAVRO4476912790T.
0 1GHz with 434 near center 427.25 near center
Since its original publication, John says, I discovered to

late for the article that the bypass for the 3 oscillators is
recommended to be .1mfd. I used .0047 which at 100 MHz
should be fine. I swapped them out today and the .1s did
improve the noise floor by about 5 or 6 dB. Don’t know
how I missed that but I did. I don’t think it’s a big deal but
worth mentioning.
0 1GHz with 439.25 near center

2 Watt 2.4 GHz DVB-T Transmitter
Written by Jim Andrews, KH6HTV
Repeater’s REPEATER March, 2020
Transmitter after “Gutting”
Several years ago, Bill, K0RZ, gave me a surplus ShowTime amplifier, the metering circuits, and the necessary, big,
MDDS, analog TV transmitter. It was used on the commercial heavy, linear dc power supplies and one very noisy cooling
MDDS, 2.52.7 GHz band. It was built by EMCEE and was fan. It was still big, bulky and heavy. So would it work ? Read
their model TTS10HS and labelled to work on channel H3. It on.
was found to contain two identical transmitters, one Visual
and the other Aural. Each transmitter was rated for 10 watts. I have just completed some evaluation tests on the “Beast”. I
It was a BIG Beast with two very noisy cooling fans. I didn’t used a HiDes, model HV320E, DVBT modulator as my DTV
know what to do with it for a long time, so it sat gathering signal source. I used 2393 MHz as the test frequency. To
dust on a shelf in my garage for an extended period of time. measure the output power, I used a Narda, 30dB, 50 W
attenuator and my HP432 power meter with an rms
When I eventually became interested in trying out the 2.4 responding, HP thermistor head. To look at the output
GHz ham band for DVBT, I decided to see what I could do spectrum, I lashed together a downconverter consisting of a
with this transmitter. Opening the box, I found a whole lot of VariL, 3 GHz mixer, a frequency synthesized Local Oscillator
stuff. Most of it was unusable for my purposes. So, I set to 2.0GHz, +7dBm and the IF went to my Rigol DSA815
proceeded to gut out most of it. Included was pulling out the spectrum analyser (0.11500MHz). Setting the Rigol to 393
second, Aural transmitter, which I then gave to Don, N0YE. MHz allowed me to thus look at the 2393 MHz spectrum.
All that was left was the one driver amplifier, the final 10 watt
Output Directional Coupler
It was set for “Normal”, amateur digital parameters with 6

Driver & 10 W Final Amplifier MHz BW, 1080P resolution, QPSK, 5/6 FEC, etc. I used a step
attenuator on the output of the HV320E to carefully set the
All tests were at the desired 13cm operating frequency of input rf drive level. I monitored the resultant spectrum on the
2393 MHz. I first ran a CW test of the power curve of the Rigol. I increased the rf drive upwards until the spectrum
amplifier and found the small signal gain to be 57.5dB. The shoulder breakpoints (measured at ± 200 kHz beyond the
1dB gain compression, Pout(1dB) occurred at +38.5dBm = band edges) hit 30dB. I then used the HP power meter to
7.1 Watts. measured the output power. I found it to be +33 dBm = 2
Watts (rms). At this level, the front panel power meter
The max. saturated power output was +39.7dBm = 9.3 indicated 50%. This amplifier will work well as a 2.4 GHz
Watts. I then ran tests for digital TV. I used the HV320E AfterBurner and can be driven directly by the HiDes HV
modulator. 320E modulator.
Front Panel Metering

Grass Valley Mixer Conversions - Part 15 I have written a script in Annex BASIC that has enabled me
to map out the addresses of the lamp’s buttons and pots.
Written by Trevor Brown, G8CJS The script requires the ESP to be flashed with the BASIC
interpreter, there is a lot of help available on the Annex site.
Recap The buttons now control all the lamps without the crate, so
The Grass Valley mixer project we have a standalone panel. The software is on the CQDATV
revolves around three basic download site.
models the GVG100, GVG 110
and GVG 1000. The mixer Vmix
appeared in the 80’s and
consists of two parts, a crate The next stage is to connect this panel up to Vmix which is a
which is available in two software package that will run on a PC and allows mixing of
options, component and nonsynchronous sources. Vmix far outshines the original
composite (various standards PAL and NTSC included). Both GVG crate with digital moves, electronic captions and the
crates and all three panels are interchangeable, unless you ability to manage clips as VT inserts, providing they are on
know differently. the hard drive of the PC. The only downside is that
conventional external cameras need an interface or capture
The crates are all a little dated in that they will only mix cards to interface to the host PC. Webcams can be brought
synchronous sources, the control panels are connected to the up as vision sources. Vmix software is not free, but there is a
crate by a 15way lead using D connectors. The two D free demo version which controls a limited number of
connectors are connected one to one, thanks Heiner Juers for sources.
that information.
The mixer was designed so that it could be run by pushing

buttons on the panel as you would expect, or by a RS 422
link for remote operation, eg an edit controller. This was a 9
pin interface that was very popular at the time and allowed
other kit to be interconnected eg RS 422 sound mixers. The
protocol for this is often called P2 or Sony protocol.
In this series of articles, we have discarded the crate and

concentrated on the panel alone. Powered it with +14 and
+10 from an external power source, removed the internal
microprocessor and fitted a dongle in its place. The dongle is
based on an ESP micro with the addition of 5 external 8bit
ports (PCF 8574). The port chips are I2c controlled and are
interfaced to the lamps, buttons TBar and analogue pots.

There are paid versions that enable more sources and there The Panel history
is a 60day option that enables the full version of the
software to be evaluated. I also wanted to get this project up and running before all
these control panels disappear, they were very popular in the
Vmix control screen broadcast industry for several reasons. The number one
being their RS 422 control. This was possible because of the
I have downloaded the 60day full package and looked at micro control which could not be added to older AB mixers.
externally controlling this software, with the GVG panel. Vmix
can be controlled by a PC keyboard, a mouse or a MIDI TV production has moved a long way from the traditional
interface. The ESP BASIC does not support MIDI commands, model of put the VT in record, cue the studio and punch
so Mike G7GTN stepped in and interfaced the software to an around on the vision mixer, which was how TV production
Arduino, which can deliver the necessary commands (see last functioned in the 60’s with a perhaps a little scene to scene
month’s issue). The micro Mike programmed will also talk I2c editing if required.
so the plan is that it can be located down stream of my ESP
micro and translate I2C commands into MIDI. By the 70’s TV production was becoming more film orientated
with more sophisticated editing, sound dubbing, some could
The video interface even argue that by the time three machine editing arrived,
that enabled takes to be mixed in the edit suite that perhaps
Mike and I live at different ends of the country and the it had features above film editing. This was still slow with its
Arduino is now in the postal service having left Mike, but not neg cutting, AB rolls and print the mixes in the film
yet arrived with me. These are extraordinary times that were processing lab. With the emergence of single camera shoots
not with us when we started this project and there is a lot using camcorder’s in the 80’s, the two were almost on a level,
more to worry about, rather than getting a micro to me. It using similar production techniques. The ace in the hole was
does leave us with a problem as CQDATV is going to press VT mixes happened in the edit suite and the GVG 100 was
and perhaps some of our readers are wondering how we have the key.
progressed. The other problem is that I downloaded the
Vimix 60day package two magazines ago and soon it will The larger UK equipment manufactures were a little slow to
stop working. I don’t expect it to drop to the free limited respond to these changes and often had the old 60’s model of
source version and from memory some of the features such TV production as their equipment target. Many of the ITV
as chroma key don’t work on the free version. I want to studios engineered their own solutions. Grass Valley changed
deliver more than the cut and mix version as the pots on the that and the GVG100 was at the head of edit suite mixing as
GVG panel are all delivering digital values via the I2c Bus. P2 protocol equipment appeared. The Studio galleries also
Yes, we have worries about latency, having two micro’s daisy took to GVG products (using the larger mixers). The push
chained, we will find out. Other potential problems such as buttons that could be customised with overhead projector
the digital steps of the GVG pots, Tbars and joysticks may be gels to create legends within the illuminated buttons, became
too large for some applications. The problems you think an industry winner. Gone were the china graphs pencils used
about are often not the problems that bite. to mark up the sources.

The future
I worry industry may have junked this pleasing control panel,

with the move out of analogue TV. Ross did produce a digital
crate that was compatible with the GVG panel, but I have
never had my hands on one to evaluate so am unsure of the
market penetration. The digital nature of the panel, with very
few electrolytic caps, marked for future replacement, all the
chips are socket mounted, so low maintenance, (fingers
crossed). There are also changes on the net where the lamps
have been changed to LEDs and the 14volt rail eliminated,
something I plan to do later.
Big Brother
GVG produced much larger mixers, some a little too large for
the average shack or workshop. They all have the same look,
use the same push buttons. It’s a long time since I tinkered One of GVG 100’s big brother, no I have no intention
with the innards of these. Are the i2c changes compatible of reverse engineering one for your ham shack
with the larger range of mixers? I am unable to report.
Star Wars https://www.vmix.com/
The larger mixer will always be remembered for the Star https://sites.google.com/site/annexwifi/home
Wars film where Darth Vader used one to blow up Princess
Leia’s home planet, (Alderaan). George Lucas has never https://www.reddit.com/r/MechanicalKeyboards/comments/3l
released that interface, perhaps that’s for the best. tscc/grass_valley_video_switcher_in_star_wars_1977/
If you see a panel at a reasonable price it might be worth a https://www.youtube.com/watch?v=l7l4MY9u7pA&feature=y

punt as I suspect the button and lamp maps I have re outu.be
produced in the previous issues will enable you to press them
into service. I will continue with the software as soon as
Mikes Arduino turns up and report back to you in CQDATV.
If you see a scrap one, I am short of cut button and a couple

of lamps, should they be harvestable, otherwise keep an eye
out on ebay for one, three have appeared since I started this
series of articles.
One from the Vault JPEG is a single layer so it's not got an Alpha layer to define
transparency, PNG does have a layer to define transparency,
Video Codec's and why so many but the file size is larger (but is lossless!). JPEG degrades
Written by Rovert Nworb every time you edit/resize. BMP and Tiff are better files but
First published in issue 25 can be quite large.
If ever there was a recurring Let's start TV pictures from a camera and AVCHD
question in Digital Television it's compression. This compression will get the data rate down to
what is a Codec and why are there about 7 to 10MB/second, so it can be stored on an SD card.
so many different kinds. But at what cost and how does it do it.
Let's start with the name Codec it Let's go back to sampling, it's a little like those children's
stands for COmpressor puzzles where we join up numbered dots to create a picture,
DECompressor, that the easy part, the more detailed the picture the more dots you need. So for
what do they do well in our case PAL think about the subcarrier and a single sine wave cycle
(Television) they convert the TV sampled three times. Chose where you take the samples you
signal into 1 & 0, OK so it's an A are not going to draw a sine wave with three dots.
to D and D to A, no its much
more. If we take a 6MHz video signal and sample it at 3
times subcarrier (4.433 MHz in the UK we get over 13MHZ)
and then convert it to 10 digital bits 130 MB/second that's for
625 PAL if we go for 1080p and 12 bits we are up at
155MB/Second, that's 560Gb to store for 1 hours worth of
video, as for 4K Television well let's not go there. We are at a
bit rate that hard disks struggle with, but what we wanted is
to record video on an SD card.
Television pictures are not the only medium, with digital

problems, that are solved by Codecs.
Audio and still images, with the MP3 being the most familiar
with AAC and AIF not far behind and JPG, PNG and TIFF for
still images. Until fairly recently, the compression algorithm The dots carry binary amplitude information that indicate
used in JPG files was copyrighted, whereas PNG is an open, amplitude, so 8 binary bits would present 256 analogue
public domain format. levels. Still we are going to get a lumpy sine wave picture.
Increase the samples and the analogue levels and the data
These are digital files that have been compacted in different will be coming so fast that we will have storage problems.
ways. The problem is each format has its short comings.
together. This is because there is prediction software built
into the codec that guesses how the pictures are going to
change throughout the GOP (Group of Pictures), but of
course, it never guessed you were going to shake the
camera.
Compression Codec's work best if you do not feed them with

coded pictures. Otherwise you are PAL coding and then digital
coding and you get the worst out of both worlds. But some
DATV transmitters give you very little option as they have
only PAL or SVHS inputs. Also we have moved into a wide
screen world and PAL coding or NTSC for that matter was
never designed for the wide screen world, so pictures are
best sourced as component pictures.
This is ok if you start today, but particularly in the broadcast

world where they want to inflict programmes made in an
Reduce it and we will see ether quantisation puddling or older analogue world that were stored as 12 by 16 in PAL.
artefacts caused by low sample rates. Filtering is essential
and nor something software does well. This can be aspect ratio converted by removing part of the
picture and stretching other parts and the can be converted
Once over that hurdle we come to the next stage: removing to component with high quality comb decoders, but they still
the redundant part of the picture to save on space and suffer and can be found every day on some of the less well
reduce data rates. off digital channels.
This is done by looking at a group of pictures, transmitting Codec's are very much horses for courses. They have specific
the first one and the last one and then the rest of the uses, some are ideal for transmitting pictures and will cope
pictures in the group as the difference between the first and with reduced bandwidth, such as Mpeg 4, little difficult to edit
last picture. with as you do not have a group of pictures to scrub past and
chose edit points, but software can create the illusion of the
Does it work well providing there is not too much movement whole group of pictures being there, but this can be very
in the picture? Yes, the more movement there is in the demanding on your CPU.
picture the more it is likely to not work.
Also, for equipment manufactures, there are copyright issues
In the early days all you had to do was put the camera into as Codec's might be in use everywhere but they were all
record and shake it to give maximum picture change for each developed by somebody and that process might not be in the
frame and your Codec would be in trouble, but we have had public domain which leads manufacturers to implement
considerable improvements and yes, they do now hold various versions of common Codec's.
There are also Transcoders, which will convert between the
various codec formats but these are not necessarily lossless
and can again lead to picture degradation.
Summing up, choosing the right horse for the course. Not
that we always have much choice, avoiding sourcing a picture
that will be in a digital domain from a coded source. The old
computer adage of rubbish in rubbish out still holds and just
because it is digital does not mean it is good.
Republication of CQDATV magazine material is encouraged

as long as source credit is properly given.
Exception: “Reprinted by permission” material must have
the original publisher’s/authors permission.

Information Copyright
The articles contained in this publication remain the copyright
External links of their respective authors and NOT dotMOBI. Any
reproduction of such articles must be approved by the author
If you have an eBook reader that does not have WiFi then of that article.
you will not be able to use the hyperlinks in this publication.
If you have an eBook reader that has WiFi then you will be
able to providing you are in a WiFi zone.
Notice to Contributors
Authors are alone responsible for the content of their articles,
But if you have a Kindle 3G then yes, but only to Amazon, including factual and legal accuracy, and opinions expressed
and there is not a lot of ATV material on their site. by them may not reflect the editorial stance of the
Smart phone reading apps are ok providing that you have a publication. Material submitted to dotMOBI should not
3G data connection. infringe the copyright of other writers or bodies.
Contributions are accepted for publication on this basis alone.
Note: These links will fire up your devices browser and if you dotMOBI publications https://cqdatv.mobi
are using 3G/4G then you will incur data usages charges.
Author Guidelines
Legal Niceties (the small print) CQDATV welcomes contributions from our readers. It does
E&OE. Whilst every care is taken in the production of this not necessarily have to be on ATV, as long as it is of interest
publication, dotMOBI accepts no legal responsibility for the to our readers.
advice, data and opinions expressed. dotMOBI neither
endorses nor is it responsible for the content of Although a formatted article showing the layout can be sent,
advertisements or the activities of those advertisers. No we prefer an unformatted text file of the script, along with
guarantee of accuracy is implied or given for the material annotations of where important images should be placed. All
herein. dotMOBI expressly disclaims all liability to any person images should be identified as Fig 1 etc and sent seperately.
in respect of anything and in respect of the consequences of
anything done or omitted to be done wholly or partly in Images should be in PNG format if possible and the best
reliance upon the whole or any part of this publication. quality available. Do not resize or compress images, we will
As the regulations for the operation of radio frequency do all the rework necessary to publish them.
equipment vary in different countries, readers are advised to
check that building or operating any piece of equipment If you are sending a construction project, please include the
described in dotMOBI will not contravene the rules that apply dimensions of any pcb's and make the pcb image black and
in their own country. white, not greyscale.
All copyrights and trademarks mentioned in this publication
are acknowledged and no infringement of the intellectual CQDATV reserves the right to redraw any schematics and
copyright of others is intended. pcb layouts to meet our standards.
Coming up in CQ-DATV
Is this the latest issue of CQDATV? Click here to go to our
web site to check to see if there is a later edition available.
CQDATV is published on the last day of the month. The cut

off day for submissions/corrections/alterations is 5 days
before the day of publication.
Please note that this

mailing list is only used
to advise interested
people about the
availability of new issues
of CQDATV magazine.
The list is not, and never
will be, shared with any
other organisations.
Want to be notified when issues of CQDATV are published?
Join our mailing list.

CQ Datv82

Uploaded by

Copyright:

Available Formats

CQ Datv82

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

CQ Datv82

Uploaded by

Copyright:

Available Formats

In this issue

5.8 GHz Transverter........................................... 7

The CQ­DATV editors gratefully acknowledge

CQ-DATV 82 - April 2020 Page 1

CQ-DATV 82 - April 2020 Page 2

New ATV Handbook

Back in 2017, Tom O’Hara, W6ORG, was

This past fall, Bill, AB0DH, loaned me his 2019 ARRL

I have just posted it on my web site: www.kh6htv.com as my

It is the first Urunga Convention that has failed in 72 years

VK3 ­ EMDRC Committee after due consideration, have

VK3 ­ GippsTech 2020 has been cancelled

VK6 ­ Bob VK6POP in a phone hookup has told how PerthTech

ZL ­ For our Kiwi friends your Conference 2020 is cancelled.

Source: Excerpts from VK National News Broadcast

This is your free ATV magazine.

CQ-DATV 82 - April 2020 Page 6

Don, N0YE, had previously built over the years, several

I got interested in what I might be able to throw together for

The project then ended up on the back burner for awhile as I

In operation, I leave the Hi­Des receiver powered up all the

So, how well does it work? The transmitter is definitely NOT

I set up my rig at the Boulder 911/EOC near the Boulder

The received power was measured to be ­83dBm with a s/n

Whew ! – you know this microwave DVB­T stuff was a lot

CQ-DATV 82 - April 2020 Page 10

The new testcard is easier to programme than the old one, I

This is considerably more than the old Testcard Generator, so

This is a giant leap for me as I dread to think of the number

The full AD 724 data sheet can be found at

I have put a hex file (PIC_TCG5. hex) on the CQ­DATV site

The data sheet also insists on C6 to block the DC, which is a

CQ-DATV 82 - April 2020 Page 13

For some of our television shack projects a requirement

Back in 1981 we had devices such as the SPO256AL2 speech

The quality of output was always a little dubious. So we had a

Instead of making use of this device and the requirement of

Using the free Basic for Ardunio a software serial port on

A handshake signal is available on the speech processor

As long as your projects processor is able to send data at the

CQ-DATV 82 - April 2020 Page 15

Options and quick testing Conclusion

CQ-DATV 82 - April 2020 Page 16

Reprinted from Boulder Amateur Television Club TV

After seeing the article on the Epson programmable crystal

CQ-DATV 82 - April 2020 Page 17

Since its original publication, John says, I discovered to

0 ­ 1GHz with 439.25 near center

CQ-DATV 82 - April 2020 Page 19

Transmitter ­ after “Gutting”

It was set for “Normal”, amateur digital parameters with 6

CQ-DATV 82 - April 2020 Page 22

The mixer was designed so that it could be run by pushing

In this series of articles, we have discarded the crate and

CQ-DATV 82 - April 2020 Page 23

CQ-DATV 82 - April 2020 Page 24

I worry industry may have junked this pleasing control panel,

Star Wars https://www.vmix.com/

If you see a panel at a reasonable price it might be worth a https://www.youtube.com/watch?v=l7l4MY9u7pA&feature=y

The CQDATV editors gratefully acknowledge

VK3 EMDRC Committee after due consideration, have

VK3 GippsTech 2020 has been cancelled

VK6 Bob VK6POP in a phone hookup has told how PerthTech

ZL For our Kiwi friends your Conference 2020 is cancelled.

In operation, I leave the HiDes receiver powered up all the

The received power was measured to be 83dBm with a s/n

Whew ! – you know this microwave DVBT stuff was a lot

I have put a hex file (PIC_TCG5. hex) on the CQDATV site

0 1GHz with 439.25 near center

Transmitter after “Gutting”

Republication of CQDATV magazine material is encouraged

CQDATV is published on the last day of the month. The cut