►
From YouTube: ORI FPGA Office Hours 21 March 2023
Description
0:10 Small talk
2:10 Meeting called to order
23:00 Caroushell Photos from @burningman
A
A
A
For
for
all
one
of
the
people
listening,
this
is
the
carousel.
This
is
a
burning
man,
art
project,
that
I
did
the
lights,
for
is
that
late
evening,
before
it
got
completely
dark
picture.
If
I
move
out
of
the
way
here,
you
can
see,
the
central
column
is
the
well,
it's
a
central
column.
These
vertical
lines
of
lights
are
animated.
These
are
boxes
that
are
back
illuminated
single
color
per
per
panel
and
they
stayed
the
same
all
area
and
the
whole
thing
rotated
under
human
power
until
it
didn't.
A
B
So
I'll
include
some
some
photographs
of
that
at
the
end,
all
right,
it's
10.
welcome
to
open
research,
institutes,
fpga
standup,
it's
the
21st
of
March
2023
and
let's
go
ahead
and
just
kick
it
off.
Let's,
let's
go
to
James
first,
so
that
he
can
give
us
report
and
if
he
has
to
leave,
then
he
can
go.
A
B
A
B
B
D
B
Oh
fantastic,
that's
really
good
news.
Yeah
Pilots!
For
those
of
you
that
don't
know,
pilot
signals
are
signals
that
come
out
and
just
like
a
lighthouse.
They
they
help
the
receiver
receive.
So
the
pilot
signals
are
powerful,
very
simple
signals
that
are
at
a
regular
interval,
and
this
helps
in
several
ways.
So
it
helps
you
both
in
frequency
and
time.
Pilot
signals
are
optional
and
dbb
s2s2x,
and
we
include
them
to
make
our
receivers
have
an
easier
time
of
doing
stuff.
So
getting
the
pilots
working
in
the
transmitter
is
a
good
deal.
B
C
Yep
all
taken
care
of
here
doing
fairly
well
at
room
at
lab.
South
we've
been
continuing
some
of
our
outdoor
work
with
the
weather,
committing
though
this
next
week
we're
expecting
to
see
a
bit
more
rain,
but
even
with
that,
we
should
be
doing
fairly
well
we're
working
more
in
inside
this
week
because
of
the
weather
and
I'm
also
going
to
be
a
bit
less
active,
because
this
is
finals
week
for
me,
but
otherwise
we're
doing
fairly
well
down
here
at
remote,
lab
South
and
getting
more
preparations
done.
B
B
B
Okay,
so
you
have
three
things
to
report
and
and
then
a
little
bit
about
our
our
upcoming
edition
of
the
newsletter.
So
we
we
now
have
a
newsletter,
it's
at
least
monthly.
We
try.
Sometimes
we
we
send
out
a
special
edition
and
the
next
newsletter
will
come
out
on
our
favorite
holiday,
which
is
April
fools.
Oh,
my
gosh,
okay,
so
I
know
I've,
told
people
all
along
that
our
favorite
holiday
is
Halloween,
but
no,
it's
really
April
fools
so
be
aware.
B
There
may
be
some
pranking
in
the
newsletter
when
it
comes
out.
So
here
we
are
at
our
stand-up
and
there's
there's
three
things
that
we
know
about.
We
have
an
issue
with
the
dvb
fpga
that
was
raised
by
David
Horowitz.
We
have
a
big
update
for
lab
equipment
and
and
some
some
protocol
development
and
sort
of
a
publishing
schedule
for
Neptune,
and
we
have
some
some
progress
on
opulent
voice.
B
So
David
wrote
to
us
through
GitHub
issues.
He
says
he
has
an
issue
sending
data
to
the
encoder.
This
is
our
dbbs2
encoder
from
swato
via
axi
stream
interface
and
he's
just
not
getting
it
to
send
out
data,
and
he
has
some
stream
connections
issue.
The
excise
stream
is
not
active.
He
wants
to
share
his
design
now
he's
using
this
I
believe
on
the
rfsoc.
B
This
is
a
new
radio
frequency
Communications
chip
from
xilinx.
It's
a
very
similar
development
platform
to
the
adrb
9371
that
we
use,
but
it's
it's
got
a
lot
more
Communications
support
on
it.
So
he's
he's
got
a
similar
development
board,
but
it's
not
quite
the
same
and
he's
asking
for
help.
This
was
about
five
days
ago,
so
I
haven't
seen
any
action
on
this,
but
I'm
pretty
sure
that
there's
Communications
going
on
so
I'd
like
to
elevate
this
up
and
put
it
out
there
I'll.
B
Let
you
know
that
on
our
GitHub
at
at
our
at
our
GitHub
site
for
dvb
fpga,
that
we
have
this
issue
and
let's
figure
out
the
problem
with
something,
would
help
with
or
have
control
over
another,
and
also
there
also
is
a
recent
change.
So
swatches
updated
the
light.
Fury
I
think
that's
light
Fury
and
not
fitty
bit
the
plot
design
and,
as
Everest
mentioned,
that
there's
progress
on
the
pilots,
which
is
huge.
So
all
of
this
is
a
big
deal.
Thank
you
so
much
for
everybody.
B
That's
working
so
hard
to
make
this
a
reality.
This
is
a
complicated,
big
digital
thing.
It's
a
great
design.
We
really
like
this
protocol
and
we
really
want
to
enable
open
source
use
of
a
dbb,
S2
S2X
and
to
get
just
a
broader
interest
in
these
sort
of
digital
Communications
protocols.
Amateur
radio
and
citizen
scientists
and
hobbyists
really
deserve
to
to
have
modern
techniques
and
we're
here
to
try
to
help
that
every
way
we
that
we
can
all
right.
The
next
thing
I
have
to
talk
about
is
opulent
voice.
B
This
is
our
Uplink
protocol,
so
this
is
a
as
you
might
might
guess,
and
or
know
that
the
the
opulent
is
a
riff
on
Opus,
the
Opus
codec.
So
what
we
do
is
we
take
the
Opus
open
source
voice
codec
and
we
use
it
as
sort
of
the
the
base
design
for
the
for
the
Uplink.
The
University
of
Puerto
Rico
is
going
to
use
it
in
their
upcoming
sounding
rocket
launch
at
Wallops
and
or
for
Roxette,
and
we're
very
happy
about
that.
B
So
our
our
method
here
is
to
use
Matlab
and
simulink
use
HDL
coder
toolbox,
which
we
have
through
our
startup
license
to
produce
open
source,
clearly
readable
via
verilog
image,
deal
implementations
of
of
this
for
both
the
receive
and
transmit
and
in
the
upper
part
of
your
screen.
You
can
see
a
very
colorful
like
circular
mandala
thing.
Well,
that's
just
one
of
the
visualizations
from
modeling
this
in
Matlab,
and
this
is
a
whole
bunch
of
different
Channel
models
with
increasing
or
actually
decreasing
SNR.
B
So,
as
the
noise
gets
higher
and
higher,
the
signal
gets
more
and
more
kind
of
out
of
control.
Here's
what
it
looks
like
in
Matlab
right
now
for
opulent
voice
upper
left
is
the
modulated
signal.
This
is
what
it
looks
like
it's
about:
got
about
a
54,
kilohertz
bandwidth
right
now
for
the
the
high
definition
voice
for
Erie
FSK,
it's
minimum
frequency
shift
keying.
This
is
what
it
looks
like
when
you
modulate
it
with
some
of
the
system.
Objects
in
Matlab.
The
middle
bottom
is
what
it
looks
like
when
you
transmit
it
through
channels.
B
So
you
take
the
signal
you
transmit
it
through
your
channel
and
when
you
transmit
it
over
the
air,
you
may
have
lots
of
noise
or
not
a
lot
of
noise,
and
what
we've
done
is
modeled
that
so
you
can
see
that
on
the
top,
the
noise
is
actually
higher
than
the
signal
and
towards
the
bottom.
The
noise
barely
touches
the
signal,
and
so
what
we
do
is
we
have
a
complete,
a
collection
of
different
channels
that
we
send
our
signal
through
to
try
to
see
like
what's
going
on
with
it
on
the
upper
right.
B
This
is
what
happens
after
you
use
the
receive
filter.
So
we
used
a
match,
filter
approach
and
we
are
you
know
calming
down
the
noise.
You
can
see
that
it's,
you
know,
sort
of
the
expected
result
and
in
some
cases
you
get
a
signal
emerging
out
of
the
noise
and
in
other
cases
it's
still
kind
of
buried,
and
we
have
bit
error
rate
curves.
B
What
we're
trying
to
do
is
make
absolutely
sure
that
we
understand
all
of
the
theory
that
matches
up
with
our
implementation,
we're
going
to
take
this
implementation
of
Matlab
we're
going
to
take
it
to
simulink
we're
going
to
take
it
to
digital
coder
and
then
we're
going
to
publish
bhdil
vhl,
slash,
verilog
code,
open
source
for
this
type
of
communications
protocol,
and
it
does
fill
a
niche
that
you
know
fills
a
need
that
doesn't
really
exist
right
now.
We
have
a
high
definition
voice.
It
starts
out
at
16
kilowatt
per
second
Opus
voice.
B
Codec
can
be
used
on
70,
centimeters
and
above
and
can
you
can
do
data
or
voice?
The
protocol
in
is
built
to
handle
either
Voice
or
data
without
having
to
switch
to
a
second
packet
mode?
So
we
would
like
this
to
be
very,
very
user
friendly
and
to
to
kind
of
fill
that
fill
that
need.
It
also
has
authentication
and
authorization
sort
of
built
in
so
this
is
in
the
protocol,
so
that
you
can
use
those
sorts
of
functions.
B
And
finally,
Neptune
is
a
drone
related
project
that
we
are
pushing
forward
and
and
working
hard
to
make
happen.
It's
a
ofdm
orthod,
orthogonal
frequency
division,
multiplexing
protocol,
it's
an
open
protocol
and,
if
you're
familiar
with,
maybe
cellular
ofdm
you'll
see
a
lot
of
similarities.
With
this
we
take
out
like
the
things
that
work.
B
So
we're
we're
looking
towards,
like
CC,
ccsds
and
NASA
guidance,
to
make
sure
that
ofdm
in
space
will
will
will
still
work,
that
we
may
have
a
solution
for
some
of
the
problems
that
are
being
posed
by
by
the
by
the
missions,
particularly
lunar
missions
and
deep
space
missions.
So
we're
we're
ready
here,
but
but
the
immediate
need
and
the
immediate
market
and
application
is
for
for
drones.
That
means
that
we'll
get
a
lot
of
attention
towards
lightweight
low
power
efficient,
there's
a
there's,
an
emphasis
on
on
video.
B
So
those
are
the
sorts
of
bandwidths
that
we're
looking
at
for
this
particular
protocol,
it's
called
Flex
link
and
will
should
be
published
very
soon.
So
what
we'll?
What
you'll
see?
First,
is
the
physical
layer
followed
by
the
data
link
layer?
What
we've
done
in
order
to
prepare
for
supporting
this
project
is
to
purchase
an
adrb
9002.
B
This
is
Analog
Devices
chipset,
the
9002
and
the
adrb
is
The
Analog
Devices
radioverse.
This
is
their
Dev
board.
So
it's
it's
designed
to
show
off
the
9002
chip
and
to
give
it
lots
and
lots
of
peripherals
you
plug
it
into
a
baseboard
that
has
a
powerful
fpga
so
that
you
can
run
it.
The
adrb
9002
comes
in
two
variants,
so
it's
not
like
our
93
71.
That
is
the
entire
frequency
range
of
the
the
underlying
chip
it
breaks
it
out
into
two
boards.
B
One
of
them
is,
like
I,
think
it's
300
megahertz
to
three
gigahertz.
The
other
is
three
gig
to
six
gigs,
since
five
gig
is
where
the
drones
are
at
we've,
we've
selected,
at
least
for
now
at
least
one
the
version
that
will
support
five
gigahertz.
So
we
got
the
three
gig
to
six
gig
version,
it'll
be
here
in
a
couple
days.
It
will
go,
it
will
fit
onto
the
zcu
106,
the
xilinx
fpga
development
baseboard,
and
we
are
definitely
going
to
use
HDL
coder
for
Matlab
to
implement
this
design.
B
What
this
will
also
give
us
which
I'm
I'm
very
excited
about
is
some
real
experience
with
implementing
a
really
a
pretty
complicated
design
using
this
particular
tool
set,
which
is
industry
standard
and
that
will
allow
us
to
inform
open
source
tools
Developers
as
like
what
do
you
really
need
to
worry
about
what
worked
really
well
at
HDL
code
or
what
was
missing?
What
was
done
well,
what's
what
its
capabilities
are
in
2023
and
I?
Think
that
that
will
be
of
great
benefit
for
all
of
us.
B
So
that's
our
goal,
the
other,
the
other
stuff
that
I
have
to
talk
about.
Okay,
so
we
have
a
class
coming
up,
because
HDL
coder
is
non-trivial,
it's
a
toolbox
for
Matlab
that
you
can't
get.
If
you
have
the
home
license.
So
those
of
you
that
have
the
home
license,
you
might
have
noticed
that
there's
a
couple
of
toolboxes
that
you
simply
can't
get
an
HDL
coder
is
one
of
them.
B
The
reason
that
you
can't
get
it
is
because
Matlab
says
they
have
so
many
calls
for
tech
support
that
if
you
have
the
150
dollar
license,
they
just
they
can't
afford
it.
Okay,
so
I
have
some
opinions
about
this,
but
let's
set
that
aside
fine,
but
we
have
the
HDL
coder
and
we
can
get
some
technical
support,
so
we
also
get
from
from
MathWorks
with
our
license.
We
get
a
custom
class,
and
so
what
we
are
putting
together
is
a
class
that
will
be
virtual
for
people
that
want
to
take
it.
B
Yes,
it
will
cost
the
money
and
it
will
teach
you
digital
digital
Communications,
Digital
radio,
from
an
open
source
perspective.
That
is
what
we're
wanting
to
put
into
this:
how
to
use
these
tools
that
we
have
access
to
through
our
startup
license
for
you
like.
How
do
you
get
to
execute
fancy
radios
with
this
stuff
with
code
that
you
own,
that
you
can
publish
that
then
can
be
used
as
a
baseline
for
other
designs?
So
that's
what
we
would
like,
because
this
is
leverage.
This
is
a
big
step
forward.
B
It
includes
Matlab
and
simulink
and
HDL
coder,
where
we're
looking
at
and
we're
trying
to
finalize
a
draft
of
the
virtual
class
content
today,
send
it
to
MathWorks
get
their
opinions
on
what
they
think
would
work
for
for
a
class,
since
this
is
at
least
four
different
classes,
sort
of
cut
up
mixed
up
and
then
and
then
sort
of
re-imagined
for
our
community
we're
going
to
try
to
schedule
and
carry
out
this
class
as
soon
as
possible
and
and
there's
lots
of
benefits.
So
that's
that's
the
goal
with
this
particular
class.
B
Finally,
there
are
open
source
options
here,
so
there's
amaranth,
which
is
a
python
based
framework
for
converting
I'm,
not
going
to
say
it
converts
python
directly
into
HDL,
but
it's
that's
kind
of
the
aim
and
goal.
We
have
a
wonderful
article
that
we
will
have
in
our
April
newsletter.
We
have
a
guest
editor.
Thank
you.
Thank
you
very
much
to
Dr
estefas
for
agreeing
to
do
this.
He
is
written
a
fantastic
article
about
Maya.
This
is
his
this
is
he
authored
Maya?
B
This
is
a
a
project
and
what
it
does
is
it's
an
implementation
of
digital
Spectrum
analysis,
a
spectrum
analyzer
on
the
Pluto,
and
it
is
entirely
on
the
fpga
in
the
Pluto
and
it
was
written
with
amaranth.
So
and
then
this
is
a
completely
open
source
tool
chain,
project
product.
This
is
very
exciting.