►
From YouTube: ORI FPGA Meetup 23 May 2023
Description
Thank you to everyone helping to do ambitious open source digital radio work!
A
A
Thank
you
everybody.
This
is
the
May
23rd
meeting
of
open
research,
institute's
fpga
stand
up
and
office
hours
and
what
we
do
here
is
we
talk
about
what
we've
done,
what
we
have
planned
to
do
over
the
next
bit.
Generally,
we
do
these
weekly,
so
it's
so
we
can.
A
You
know
looking
over
the
weekend
or
more
in
review
and
what
we
have
planned
over
the
next
week
if
we
need
any
resources
in
order
to
do
our
volunteer
work
and
if
we
have
encountered
any
roadblocks
and
might
need
to
call
in
people
to
to
help
us
out,
and
so
that's
the
format
of
the
meeting,
so
let's
go
ahead
and
start
with
James
at
remote
lab
South.
Thank
you
very
much
lots
of
things
going
on
there
in
terms
of
building
and
infrastructure.
The
floor
is
yours.
B
We've
been
doing
a
lot
of
infrastructure
projects
done
here
at
roomb
South
in
terms
of
remote
lab
projects
themselves,
I'm
going
to
be
speaking
with
you
Paul
after
this
meeting
about
some
of
the
deployments
that
we're
going
to
be
doing,
but
otherwise
not
too
much
major
to
report.
Just
a
lot
of
infrastructure
being
set
up
a
lot
more
in
the
way
of
permanent
installations
being
prepared.
A
C
Okay,
we've
been
changing
around
the
hardware
for
fpga
development
on
remote
lab
West.
Here
I
think
we've
reported
on
some
of
this
already
we've
got
a
zcu
102
with
an
adrv
9009,
no
9002
radio
and
the
zc706
has
the
adrv
9009.
Is
that
right?
It's
really
hard
to
keep
these
things
straight.
C
There
are
approximately
three
ways,
or
at
least
three
ways
to
get
one
of
these
rigs
on
the
air
and
doing
something
interesting.
The
easy
way
is
to
use
a
pre-built
hyper
Linux
distro
from
ADI
the
manufacturer
of
the
radio
we've
done
that
on
both
of
the
stations
and
it
works
and
we're
able
to
command
it
with
the
adi's
application
called
the
Iowa
oscilloscope,
and
so
the
hardware
is
there
and
working
and
we're
able
to
do
some
stuff
with
it.
C
Another
important
way
that
we
need
to
get
working
is
using
the
Matlab
support,
so
we
can
run
simulations
directly
to
the
hardware
and
out
through
the
radio.
This
turns
out
to
be
a
little
more
difficult,
not
quite
as
plug
and
play
as
it
ought
to
be,
and
Michelle's
been
taking
the
lead
on
struggling
with
that
we're
making
progress.
C
Getting
close,
I
think
to
being
able
to
to
do
some
of
that
with
Matlab
and
the
Third
Way,
which
we
haven't
tried
yet
on
either
of
the
new
radio
configurations,
is
to
build
with
pedal,
Linux
and
and
have
a
fully
custom
set
up,
which
may
be
what
we
need
in
the
longer
run,
to
use
fpga
designs
on
the
on
the
system
that
straight
from
HDL,
instead
of
with
HDL
code
or
under
Matlab.
So
work
is
going
on
on
this.
C
The
only
blocker
here
is
that
the
documentation
is
marginal
and
the
software
is
fragile
and
we
don't
really
know
what
we're
doing.
But
unless
you
have
a
solution
for
one
of
those
things,
then
we
have
no
immediate
needs.
A
Yeah,
thank
you.
Thank
you
and
well
said
it
sums
it
up
quite
well.
Thank
you.
So
much
for
all
of
your
very
hard
work.
I
think
we
can
see
a
lot
of
it
in
the
background
over
your
right.
Shoulder
on
the
left
of
our
screen
is
the
opulent
voice
station
and
a
lot
of
the
RF
Equipment,
and
then.
D
A
That
that
stuff
and
and
that's
that's
moving
forward
as
well
and
we'll
be
in
space
soon.
So
we're
very
happy
about
that
and
then
over
your
left,
shoulder
with
that
orange
screen.
In
the
background,
that
is
where
all
of
this
work
for
the
fpga
development
stations
has
has
happened.
We
call
that
the
fpga
table,
because
it
allows
us
access
to
the
VMS
that
run
all
of
this
stuff
and,
like
Paul,
said,
there's
a
zcu.
The
Z
stands
for
I,
don't
know
what,
but
it
sounds
like
xilinx.
A
So
the
zcu
102
is
a
xilinx
Ultra
scale,
plus
Dev
board.
It's
paired
with
a
adrb
9002,
which
is
an
amazing
chip,
so
this
upgrade
was
really
good.
Thank
you
so
much
to
the
Neptune
team
for
identifying
this
chip
for
us.
It
works
really
well
with
the
102.
The
9000
2
is
a
mobile,
targeted,
Advanced
zero.
If
system
on
chip
from
Analog
Devices,
it's
12
kilohertz
up
to
40,
megahertz
transmit
bandwidth.
A
So
it's
perfect
for
the
the
work
that
we're
doing
with
Neptune,
which
is
aimed
towards
drone
data
links
that
will
handle
first
person,
view
video
and
also
Aerospace
and
possibly
CIS
lunar
work.
The
other
station
that's
accessed
from
the
orange
glowing
rectangle
above
your
left
shoulder
is
the
zcu
706,
which
is
a
7000
series
ultrascale.
Oh,
no
sorry,
7000,
series,
xilinx
chip,
a
big
Workhorse
of
their
lineup
and
the
the
radio
card
that's
attached
to
that
is
the
adrb
9009,
and
this
is
the
true
you
know
progression
from
the
9371.
A
So
it's
a
very
solid
Broadband
capable
system
on
chip.
This
is
designed
not
for
mobile
applications
like
the
9002,
but
it's
designed
for
infrastructure,
repeaters,
transceivers
transponders
and
that's
what
we're
targeting
now
for
our
open
source,
heo
and
Geo
transponder,
which
can
also
be
used
for
terrestrial.
A
But
it's
fdma
up
and
tdma
down
the
down
link
is
dbbs2
with
our
custom,
IP
and
the
Uplink
is
frequency
division.
Multiple
access,
the
native
digital
mode,
is
opulent
voice.
Our
high
fidelity
voice
and
data
data
protocol
and
all
the
multiplexing
is
handled
in
the
9009
and
then
transmitted
down
with
dbbs2
S2X.
A
The
other
station
that
we
have
here
in
the
lab
is
Pluto
SDR
with
the
JTAG
soldered
in
so
that
you
can
access
it
directly
over
the
web
and
do
experiments.
This
has
been
very,
very
helpful
to
the
DTV
community,
so
there's
been
some
big
steps
forward
and
we've
published
code
in
several
places
with
the
station
and
it's
still
there
still
available.
So
that's
the
big
changes
that
we've
done
and
we're
we
are.
Oh,
yes,
there's
a
there's
one
there's
the
Pluto.
We
have
multiple
varieties
of
those
scattered
around
the
one.
A
That's
in
the
lab
that
has
all
of
its
guts
exposed.
You
can
see
it's
undies
and
all
of
the
JTAG
stuff
is
quite
quite
quite
good.
It's,
but
it's
it's
out
of
its
case,
so
yeah,
the
and
and,
as
Paul
said,
there's
three
different
ways
to
to
handle
these
systems.
A
So
in
general
you
need
an
fpga,
you
need
a
processor
and
you
need
a
radio
chip
and
the
combinations
that
we
have
are
all
you
know
very
modern
and
featured
and
and
and
are
complicated
and
require
a
pretty
significant
and
steep
learning
curve.
A
There's
there's
three
different
ecosystems
of
Linux
builds
to
Target
the
processor
running
the
hard
processor,
like
it's
an
actual
dedicated
processor
in
the
zinc
on
all
of
these
platforms,
and
you
know
the
Linux
build
that
targets
them
can
either
be
petal
Linux,
which
is
from
the
the
xylex
universe,
and
we
know
how
to
do
that
and
we've
done
this.
We've
used
this
in
the
past.
There's
the
math
Works
build
root
version
of
Linux.
A
This
is
from
Matlab
and
enables
HDL
coder
GPU
coder,
all
sorts
of
amazing
things
that
they
have
that
we
have
access
to
because
we
have
license
for
all
of
the
toolboxes.
So
we're
looking
for
some
benefits
for
for
working
through
all
of
the
build
problems
with
the
max
Works
math
Works
build
route.
That's
what
we've
been
working
on
over
the
past
four
days,
pretty
much
solid
and
we
have
had
success.
A
The
next
step,
for
both
of
these
is
to
incorporate
the
device
tree
for
the
radio
card
so
that
we
can
fully
Target
the
the
stations
in
Matlab
and
use
the
full
force
of
Matlab
in
order
to
develop
both
Neptune
and
high
Faria
code
bases
and
the
third
ecosystem,
as
Paul
mentioned,
is
the
Kuiper
build
that's
directly
from
Analog
Devices,
and
this
is
a
Linux
build
that
you
can
just
download
and
get
or
build
on
your
own
from
Analog
Devices,
and
this
all
of
these
builds
give
you
IO-
and
this
is
industrial
input,
output,
library
from
Analog
Devices
to
that
allows
you
to
order
your
radio
card
around
and
all
of
these
use
the
HDL
reference
design
from
Analog
Devices.
A
So
these
three
different
ecosystems
all
have
the
same
guts.
They
use
iao
for
the
API.
They
use
the
HDL.
The
hardware
descriptive
language.
This
is
the
this-
is
the
fpga
design
from
Analog
Devices
for
their
radio
cards.
They
all
use
xilinx.
So
that's
the
the
common
foundation
for
all
of
these.
The
upper
levels
of
petal,
Linux
MathWorks,
build
root
and
Kuiper
I'll,
give
you
slightly
different
interfaces.
Different
field,
different
different
and
they're
different,
so
they're,
not
interchangeable,
except
that
I
can
pull
out
an
SD
card.
A
I
can
pull
out
the
petal
Linux
SD
card
and
put
in
a
MathWorks
build
route.
If
we
need
to
use
HDL,
coder
I
can
pull
that
out
and
put
in
Kuiper
if
we
simply
just
need
to
do
some
basic
operations
with
the
stations.
So
we've
had
some
huge
step
forward
here
with
new
hardware
and
and
realigning
ourselves
with.
Essentially
what
the
market
is
chosen,
which
is
the
9002
and
9009,
and
it
is
kind
of
a
shame
that
the
9371
did
not
get
a
lot
of
traction
in
the
marketplace.
A
That
was
a
bet
on
our
part
three
years
ago
and
I
don't
think
it
paid
off,
but
it
was
a
good
bet
and
the
9009
has
all
of
the
features
and
more
of
the
9371.,
so
we'll
be
selling
off
the
old
gear,
and
it's
all
good
gear
and
people
will
probably
be
very
happy
with
it.
We'll
put
in
some
some
bonuses
and
some
swag
along
with
it,
but
we'll
be
firmly
launching
ourselves
forward
to
the
9000
series
from
Analog
Devices.
A
So
that's
what's
going
on
in
remote
labs,
this
will
affect
remote,
Labs
South
a
little
bit
if
they
have
any
sort
of
remaining
equipment.
That's
radio
equipment,
I
think
the
only
thing
that
might
affect
remote
lab
South
is
that
they
have
a
zcu
106
rather
than
the
zco102,
and
the
zcu
106
has
a
smaller
fpga,
but
it
does
have
video
Hardware
on
board
a
video
Codec
HDMI
stuff,
which
is
desirable.
However,
the
zcu
106
is
not
well
supported
in
the
whole
Analog
Devices
in
Matlab
ecosystem.
A
So
it's
a
it
has
more
features
a
smaller
fpga.
We
may
want
to
go
ahead
and
sell
the
zcu
106.
That
is
still
I,
think
it's
still
in
the
box
at
remote,
lab,
South
and
and
either
and
then
go
ahead
and
pick
the
zcu
102,
and
that
would
give
us
the
most
flexibility
and
also
would
would
tie
us
together
to
have
the
same
and
be
able
to
support
each
other.
A
A
Okay,
that's
the
main
part
of
of
my
report.
I'm
gonna
pause
here
for
any
questions
or
comments
or
anything
that
anybody
needs
to
know
from
me.
So
go
ahead.
A
A
A
We
already
have
a
lot
of
of
existing
IP
for
this,
that
we
need
to
get
up
and
running
and
end
and
tested
and-
and
it
looks
like-
maybe
the
preference
for
that
set
of
teams
is
is
pedal.
Linux
is
something
that's
that's
less
tied
into
to
Matlab.
A
I.
Think
Matlab
could
probably
give
us
a
big
leverage
here,
but
if,
but
if
we,
if,
if
the
team
really
kind
of
prefers
to
to
just,
do
it
themselves
to
not
use
HDL
encoder,
then
we
need
to
go
with
whatever
the
team
prefers.
A
Okay,
so
if
you're
listening
to
this
and
have
an
opinion,
let
me
know
so
we'll
be
we'll
be
looking
at
this
really
closely
over
the
next
week,
so
yeah,
good,
stuff,
I,
think
we've
got
a
number
of
code
bases
that
are
working.
Lots
of
really
good
equipment
made
lots
of
big
steps
forward
in
terms
of
infrastructure
and
support.
We
have
the
agility
to
pick
between
three
different
environments
that
work
for
us.
We've
gotten
a
lot
of
competence
with
using
the
filter
wizards.
They
call.
A
So
for
these
system
on
chips
for
these
radios
they're
very
complicated,
they
have
a
lot
of
different
registers
and
a
lot
of
configurations
and,
like
Paul,
said
some
of
the
documentations,
not
that
complete
or
great
or
clear,
unless
you
know
already
all
that
you
need
to
know,
and
we've
been
slogging
through
this
and
and
as
Communications
engineering
types
and
as
amateur
the
true
amateurs
like
we
really
love
this
stuff.
Then
a
lot
of
this
is
really
familiar
like
we
see
that
there's
an
interpolation
filter.
We
know
what
that
means.
A
We
can
see
all
the
settings
we
can
get
pretty
close,
but
there's
an
awful
lot
of
stuff
that
isn't
really
written
down,
that
you
have
to
hunt
through
forums
or
track
down
the
right
person
at
a
company
to
tell
you
and
that's
just
part
of
the
job,
so
we're
doing
those
jobs
and
we're
we're
trying
to
document
it
as
we
go
so
that
other
people
can
find
it.
So
it's
less
hard
for
anybody
that
comes
after
us
and
we're
at
the
point
where
we're
developing
what's
called
profiles
for
all
of
these
devices.
A
This
is
how
the
Analog
Devices
system
on
chips
work.
You
have
a
a
text
file,
it's
human
readable.
It's
a
text
file
that
you
then
present
to
your
transceiver
chip,
and
this
has
almost
every
setting.
This
has
all
the
filters,
the
filter,
coefficients
all
the
clock
settings
everything
that
it
needs
in
order
to
work
well
at
the
frequency,
bandwidth
and
Etc
that
you've
chosen
for
it
and
sometimes
it
works,
and
then
sometimes
we
found
it
does
not
like.
A
We've
had
situations
where
we
follow
the
instructions
and
the
profile
simply
doesn't
load
onto
the
chip,
because
these
are
computers
too,
and
all
of
the
problems
that
you
have
with
computers,
processors
you
have
with
these
these
complicated
systems.
So
what
we're?
What
we're
trying
to
do
is
allow
people
to
develop
open
source
in
amateur
friendly
code
and
and
get
things
done
in
a
in
an
a
complex
environment
that
is
really
kind
of
aimed
towards
large
commercial
teams.
A
So
this
is
difficult,
but
we've
been
able
to
pull
it
off
so
far
and
we'll
keep
making
progress
and
I
think
we
have
a
better
foundation
with
the
9009
and
9002
than
we
did
before
so
over
the
next
few
weeks
and
at
IMs
2023
in
June,
where
we
have
a
demo
and
at
Defcon
in
in
Las
Vegas
in
in
August,
at
our
showcase
I
think
we'll
be
able
to
to
present
and
to
share
video
evidence
that
that
we're
we're
doing
good,
solid
work
and
moving
the
entire
field
forward.
A
Quite
a
bit
I,
don't
know
of
any
other
amateur
radio
team
working
in
this
realm
at
this
level,
we're
looking
actively
looking
for
them
and
trying
to
to
hook
up
with
them
and
to
help
any
other
organization
or
any
other
team
or
any
other
amateur
radio
Endeavor.
To
take
advantage
of
these
sorts
of
chips,
code
bases
published
work.
A
A
So,
if
you're
listening
to
this-
and
you
are
interested
in
this
sort
of
work
or
if
you
know
of
a
organization
or
club
or
or
group
or
or
anybody
that
we
can
can
get
in
touch
with
that,
that
might
either
benefit
from
this
work
or
be
able
to
contribute
to
make
it
work
better,
get
in
touch
with
us,
and
all
of
that
will
be
at
the
end
of
this
in
and
also
in
the
the
comments
and
description
all
right.