[Freetel-codec2] LoRa Modulation

[Steve]

Someone asked me if we could use LoRa modulation on UHF. I said, I don't
know, let me look.

LoRa is a Chirp Spread Spectrum mode (CSS). Well, the FCC allows Spread
Spectrum on UHF I seemed to recall. Let me check.

97.3(c)(8): *SS. Spread Spectrum emissions using bandwidth-expansion
modulation emissions having designators with A, C, D, F, G, H, J or R as
the first symbol; X as the second symbol; X as the third symbol.*

Boy, this doesn't look good. I find that LoRa has an emission code of X1D.
X because no other emission mode works for spread spectrum.

Wait a minute. If no other mode works, why did they list all those. The
second and third emission symbols are mysterious, as it is obviously one
channel of data.

Good Lord. Looks like we can't do spread spectrum on UHF like they said we
could. How about AXX emission mode. AM spread spectrum, hee.

This isn't very digital voice related, but I always thought frequency
hopping on UHF would make a good voice modem.

I'm thinking about sending in a petition:

https://docs.google.com/document/d/15n1LYDJhhdqRon7j5lM3n2rFlOgoqrk-kCenKtx9YLo/edit?usp=sharing

Have fun,

Steve

[Daniel Mundall]

Hi Steve,

I think that this is a great idea, we've been playing with LoRa for a few
years now with great results. And now with some SDR libraries showing up it
stands to be an interesting option for UHF Voice/messaging.

73's
Daniel

Daniel Mundall

Post by Steve
Someone asked me if we could use LoRa modulation on UHF. I said, I don't
know, let me look.
LoRa is a Chirp Spread Spectrum mode (CSS). Well, the FCC allows Spread
Spectrum on UHF I seemed to recall. Let me check.
97.3(c)(8): *SS. Spread Spectrum emissions using bandwidth-expansion
modulation emissions having designators with A, C, D, F, G, H, J or R as
the first symbol; X as the second symbol; X as the third symbol.*
Boy, this doesn't look good. I find that LoRa has an emission code of X1D.
X because no other emission mode works for spread spectrum.
Wait a minute. If no other mode works, why did they list all those. The
second and third emission symbols are mysterious, as it is obviously one
channel of data.
Good Lord. Looks like we can't do spread spectrum on UHF like they said we
could. How about AXX emission mode. AM spread spectrum, hee.
This isn't very digital voice related, but I always thought frequency
hopping on UHF would make a good voice modem.
https://docs.google.com/document/d/15n1LYDJhhdqRon7j5lM3n2rFlOgoq
rk-kCenKtx9YLo/edit?usp=sharing
Have fun,
Steve
------------------------------------------------------------
------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2

[David Rowe]

Lora is also quite popular for telemetry in High Altitude Ballooning
circles, which my local club (http://www.areg.org) is involved in.

As it's patented and a locked down chip set I'm quite interested in open
source alternatives. Our FSK modem (same one as used for some FreeDV
modes) is allowing us to get much higher bit rates than Lora, e.g. 100
kbit/s with a very powerful LDPC code.

We have managed to send HD images over 100km paths using just 50mW:

http://www.rowetel.com/?p=5344

At 100 kbit/s with a LNA and $20 SDR dongle the MDS is a rather low
-114dBm - which would make analog FM on a HT scratchy.

As a next step I'd like to develop custom radio hardware, like a $5
radio board, but fully open and with our high quality modems (many
closed chip-set modems under perform).

Cheers,

David

Post by Daniel Mundall
Hi Steve,
I think that this is a great idea, we've been playing with LoRa for a
few years now with great results. And now with some SDR libraries
showing up it stands to be an interesting option for UHF Voice/messaging.
73's
Daniel
Daniel Mundall
Someone asked me if we could use LoRa modulation on UHF. I said, I
don't know, let me look.
LoRa is a Chirp Spread Spectrum mode (CSS). Well, the FCC allows
Spread Spectrum on UHF I seemed to recall. Let me check.
97.3(c)(8): /SS. Spread Spectrum emissions using bandwidth-expansion
modulation emissions having designators with A, C, D, F, G, H, J or
R as the first symbol; X as the second symbol; X as the third symbol./
Boy, this doesn't look good. I find that LoRa has an emission code
of X1D. X because no other emission mode works for spread spectrum.
Wait a minute. If no other mode works, why did they list all those.
The second and third emission symbols are mysterious, as it is
obviously one channel of data.
Good Lord. Looks like we can't do spread spectrum on UHF like they
said we could. How about AXX emission mode. AM spread spectrum, hee.
This isn't very digital voice related, but I always thought
frequency hopping on UHF would make a good voice modem.
https://docs.google.com/document/d/15n1LYDJhhdqRon7j5lM3n2rFlOgoqrk-kCenKtx9YLo/edit?usp=sharing
<https://docs.google.com/document/d/15n1LYDJhhdqRon7j5lM3n2rFlOgoqrk-kCenKtx9YLo/edit?usp=sharing>
Have fun,
Steve
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
<https://lists.sourceforge.net/lists/listinfo/freetel-codec2>
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2

[Steve]

Post by David Rowe
As a next step I'd like to develop custom radio hardware, like a $5
radio board, but fully open and with our high quality modems (many
closed chip-set modems under perform).

I temper my feelings about closed designs, when I realize that a lot of CPU
and Floating Point designs are closed. Some have no qualms about using
an ARM CPU and then stress over the modem.

Regardless, I won't get into that. I do think Hams are missing the boat with
the FM blinders at VHF and above. I'd really like to do do a digital I/Q
modulator interface. USB3 or better, connected to a 12 watt transmitter.

Separate antenna jacks for transmit/receive. SDR receiver, with 16-bit I/Q
channels.

Full duplex, with four transmit channels, two of which can be changing
frequency, while the others are transmitting. (To enable Frequency Hopping,
using (B/Q)-PSK or FSK).

I'd even go $30 :-)

[Andreas Weller]

Post by David Rowe
As a next step I'd like to develop custom radio hardware, like a $5
radio board, but fully open and with our high quality modems (many
closed chip-set modems under perform).

You may have a look into analog devices upcoming ADALM Pluto platform:
https://wiki.analog.com/university/tools/pluto
http://www.digikey.com/product-detail/en/analog-devices-inc/ADALM-PLUTO/ADALM-PLUTO-ND/6624230

AFAIK it includes FPGA and ARM processor that allows Linux to run directly
on the device. So SDR will run native on the device without an "external"
PC...


Regards,
Andreas Weller

[glen english]

My 2c
You can do alot with a STM32F4/F7. There is plenty of blood in that stone.

MUCH more than most people think. A good programmer can get plenty out.
be smart about cache , now. Writing code that works on PC AND STM32 is a
bit of a drag I think if you can let go of that chain.

really.

My 8-32 channel commercial SDRs use FPGA coupled to a single STM32F4/F7
that does all demod , mod, filtering, signalling, modems..x8 , x16 ...

I actually used to have the entire SDR in FPGA but the software effort
to tweak modems and the demods was too high, hence the stm32 via 80
Mbps SPI interface..... The next programming effort step is NEON in the
A5. I prefer NEON over SSE...

g

[Dana Myers]

Post by glen english
My 2c
You can do alot with a STM32F4/F7. There is plenty of blood in that stone.
MUCH more than most people think. A good programmer can get plenty out.
be smart about cache , now. Writing code that works on PC AND STM32 is a
bit of a drag I think if you can let go of that chain.
really.
My 8-32 channel commercial SDRs use FPGA coupled to a single STM32F4/F7
that does all demod , mod, filtering, signalling, modems..x8 , x16 ...
I actually used to have the entire SDR in FPGA but the software effort
to tweak modems and the demods was too high, hence the stm32 via 80
Mbps SPI interface..... The next programming effort step is NEON in the
A5. I prefer NEON over SSE...

+1

73,
Dana K6JQ

[David Rowe]

Hello,

Over the past few weeks I've been adding "instrumentation" to the FreeDV
GUI program, to help me understand how well it is working over real and
simulated channels. It's all checked into SVN BTW, if anyone else wants
to play.

One feature simulates an interfering carrier (like a birdie), something
I have seen on the air. Attached is a plot of a carrier in the middle
of one of the 700C carriers, but about 10dB higher. The upper RH plot
is a rolling plot of bit errors for each carrier. You can see one
carrier is really messed up, lots of bit errors. The average bit error
rate is about 1%, which is where FreeDV 700C starts to become difficult
to understand.

These bit errors would not be randomly distributed, but would affect one
part of the codec all the time. For example the pitch might be
consistently wrong, or part of the speech spectrum.

I found that as long as the interfering carrier is below the FreeDV
carrier, the effect on bit error rate is negligible.

Take away: tune away from any interfering carriers that poke above the
FreeDV signal carriers. Placing the interfering tones between FreeDV
carriers is another possibility, e.g. a 50Hz shift of the tx signal.

Cheers,

David

[David Rowe]

Here are a couple of wave files that can be used to crowd-source the
testing and development of 700C:

i) http://rowetel.com/downloads/codec2/700c_tune/sine_analog_700c.wav

ii)
http://rowetel.com/downloads/codec2/700c_tune/sine_analog_testframes700c.wav

Each file consists of three signals, a 1000Hz sine wave, analog, and
700C. (i) contains Codec 2 700C vocoder frames, (ii) contains fixed
test frames useful for measuring bit error rate.

The three signals are at the same RMS power, so we can use them to
compare SSB versus 700C objectively. The sine wave at the start can be
used to measure the actual SNR.

The idea is:

i) The Tx station plays these files through their SSB tx.

ii) the Rx station records them and various SNRs and channel conditions,
e.g. high SNR, low SNR, fading. They can be recorded using FreeDV, or
any wave recorder. Using FreeDV lets you monitor the received signal
yourself, e.g. measure BER in testframe mode (Tools- Options menu), or
monitor the speech quality.

Pls use 8kHz, mono, 16-bit integer wave files.

iii) email recorded wave files to me.

I'll be tx-ing these files on 7.177 MHz in .au over the next few days
and weeks and gathering some results on 700C performance.

Thanks,

David

[Stuart Longland]

Post by David Rowe
http://www.rowetel.com/?p=5344
We were receiving 115 kbit/s data on just 50mW of tx power at ranges of over 100km.

Okay, silly question, how wide, approximately, was the transmitted
signal? 115kbps is bloody good for 70cm FSK, giving D-Star 23cm digital
data (which runs at 128kbps) a run for its money!

I suspect it is wider than the standard 25kHz FM channel, hence the need
for a "specialised" dongle (okay, bog standard digital TV receiver
dongle based around the infamous Realtek chipset).

It might be that the discriminator output available via the 9600 baud
packet radio jacks isn't up to the task… but even ¼ of this data rate
would be a brilliant alternative to the Bell-203 modulation we're used to.

Regards,

[Brady O'Brien]

The way our modem is set up, we're using 2*Rb bandwidth (not to mention
guard bands and things like that). So for 115 kbit/s, the FSK in our repo
is using 230KHz. We're also not going through an analog FM demod, so a 9600
packet radio port probably won't be up to the task anyway. The FMFSK modem
used for 2400B would probably be better there.

Post by David Rowe

Post by David Rowe
http://www.rowetel.com/?p=5344
We were receiving 115 kbit/s data on just 50mW of tx power at ranges of

over 100km.
Okay, silly question, how wide, approximately, was the transmitted
signal? 115kbps is bloody good for 70cm FSK, giving D-Star 23cm digital
data (which runs at 128kbps) a run for its money!
I suspect it is wider than the standard 25kHz FM channel, hence the need
for a "specialised" dongle (okay, bog standard digital TV receiver
dongle based around the infamous Realtek chipset).
It might be that the discriminator output available via the 9600 baud
packet radio jacks isn't up to the task
 but even Πof this data rate
would be a brilliant alternative to the Bell-203 modulation we're used to.
Regards,
--
Stuart Longland (aka Redhatter, VK4MSL)
I haven't lost my mind...
...it's backed up on a tape somewhere.
------------------------------------------------------------
------------------
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2

[David Rowe]

100's of kHz wide. For FSK you need about the symbol rate spacing
between tones, so for 2FSK at 115 kbit/s, the tones are spaced at
115kHz-ish, plus a bit more, think we use 400 kHz separation in practice.

Yes we only managed these fine results through very careful engineering,
and deliberately avoiding the common trap or re-using legacy FM radio kit.

Cheers,

David

Post by Stuart Longland

Post by David Rowe
http://www.rowetel.com/?p=5344
We were receiving 115 kbit/s data on just 50mW of tx power at ranges of over 100km.

Okay, silly question, how wide, approximately, was the transmitted
signal? 115kbps is bloody good for 70cm FSK, giving D-Star 23cm digital
data (which runs at 128kbps) a run for its money!
I suspect it is wider than the standard 25kHz FM channel, hence the need
for a "specialised" dongle (okay, bog standard digital TV receiver
dongle based around the infamous Realtek chipset).
It might be that the discriminator output available via the 9600 baud
packet radio jacks isn't up to the task… but even ¼ of this data rate
would be a brilliant alternative to the Bell-203 modulation we're used to.
Regards,

[David Ranch]

Hello David,

I did read your blog entry, the associated WeNet presentation, and the
Wenet github page. It all sounds fascinating but it was very high level
on the TX side (just highly brushed over). I'd love to see a higher
throughput solution out there and I have questions like:

- Which FSK TX board were you using? It looks like it's based on a
RFM98W RoRA board which might be soldered onto another Rpi HATT board
(hard to tell from the 2016_11_wenet_presentation.pdf but maybe that's
the FSK modem sitting on top of another Hab Supplies board?
- This design seemed to be uni-directional but do you think it would
be hard to make bidirectional? The current design seems to take a
picture, apply FEC, etc. and then send in batch.
- Did you ever measure the current draw when transmitting?
- There is very little detail on the antenna you used. Did you just
use a dipole? Maybe more of a panel design?
- Thanks for all the work in publishing the details at
https://github.com/projecthorus/wenet

--David
KI6ZHD

Post by David Rowe
100's of kHz wide. For FSK you need about the symbol rate spacing
between tones, so for 2FSK at 115 kbit/s, the tones are spaced at
115kHz-ish, plus a bit more, think we use 400 kHz separation in practice.
Yes we only managed these fine results through very careful engineering,
and deliberately avoiding the common trap or re-using legacy FM radio kit.
Cheers,
David

[Dana Myers]

Post by David Ranch
Hello David,
I did read your blog entry, the associated WeNet presentation, and the Wenet github page. It all sounds fascinating but it was
very high level on the TX side (just highly brushed over). I'd love to see a higher throughput solution out there and I have
- Which FSK TX board were you using? It looks like it's based on a RFM98W RoRA board which might be soldered onto another
Rpi HATT board (hard to tell from the 2016_11_wenet_presentation.pdf but maybe that's the FSK modem sitting on top of another
Hab Supplies board?

David's question leads me to a couple of questions.

The TX description says:

"the transmit side is designed to run on a Raspberry Pi, and the
UART (/dev/ttyAMA0) is used to modulate a RFM98W (yes, a LoRa module) in
direct-asynchronous mode. "


From this, I gather that data is applied to the DIO pin asynchronously, bypassing
the packet handler, transmit BT shaping and is in non-LoRa FSK mode. Correct?

This probably means there's no need to pay the premium price for the LoRa radio
(Semtech gets a healthy fee for the LoRa IP) and that a conventional FSK radio (like
the RFM69HW) could be equally used? (I believe it supports the same continuous
mode parameters as the LoRa radios).

I also gather that the coding-gain/throughput trade-off of the LoRa modem makes
it un-appealing for SSDV here.

73,
Dana K6JQ

[David Rowe]

Hi Dana,

Post by Dana Myers
From this, I gather that data is applied to the DIO pin asynchronously, bypassing
the packet handler, transmit BT shaping and is in non-LoRa FSK mode. Correct?

Yes - it's just being used as a FSK tx. Any FSK tx will do.

- David

[David Rowe]

Hi David,

I'll let Mark VK5QI field these questions - as he developed all the
aspects covered by your questions below.

I should point out that Wenet is a joint effort of many people. While I
did some modem and integration work, Mark developed the payload software
and hardware, Bill VK5DSP designed the LDPC code, and Brady KC9TPA
ported the FSK Modem to C. The Wenet payloads are launched and tracked
as part of http///www.areg.org.au club High Altitude Balloon (HAB)
activities.

Cheers,

David

Post by David Ranch
Hello David,
I did read your blog entry, the associated WeNet presentation, and the
Wenet github page. It all sounds fascinating but it was very high level
on the TX side (just highly brushed over). I'd love to see a higher
- Which FSK TX board were you using? It looks like it's based on a
RFM98W RoRA board which might be soldered onto another Rpi HATT board
(hard to tell from the 2016_11_wenet_presentation.pdf but maybe that's
the FSK modem sitting on top of another Hab Supplies board?
- This design seemed to be uni-directional but do you think it would
be hard to make bidirectional? The current design seems to take a
picture, apply FEC, etc. and then send in batch.
- Did you ever measure the current draw when transmitting?
- There is very little detail on the antenna you used. Did you just
use a dipole? Maybe more of a panel design?
- Thanks for all the work in publishing the details at
https://github.com/projecthorus/wenet
--David
KI6ZHD

Post by David Rowe
100's of kHz wide. For FSK you need about the symbol rate spacing
between tones, so for 2FSK at 115 kbit/s, the tones are spaced at
115kHz-ish, plus a bit more, think we use 400 kHz separation in practice.
Yes we only managed these fine results through very careful engineering,
and deliberately avoiding the common trap or re-using legacy FM radio kit.
Cheers,
David

------------------------------------------------------------------------------
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2

[Mark Jessop]

Radio Module: Yes, it's a RFM98W on a UpuTronics shield. TX data straight
from the UART into DIO2, bypassing all the packet handling stuff. I have
been using LoRa modules for other balloon telemetry (command and control),
so I just modified some of the shields I already had to add the line
between the RPi's TXD and the RFM98W's DIO2 line.

I did actually start out using RFM22B modules, as I wasn't initially aware
the RFM98W's could do the direct-async thing. They have the same footprint
as the RFM98W modules, hence I still had them soldered down to a LoRa
shield. I think that's whats shown in one of the presentation photos.

I'm only using these kinds of modules because it was the easiest way to
transmit FSK at 115kbaud at short notice. If anyone has any better
alternatives I'm all ears! I suspect moving to something like 4FSK is going
to require me to offload packet buffering and modulation onto something
that actually does operate in realtime.

- Mark

Post by David Rowe
Hi David,
I'll let Mark VK5QI field these questions - as he developed all the
aspects covered by your questions below.
I should point out that Wenet is a joint effort of many people. While I
did some modem and integration work, Mark developed the payload software
and hardware, Bill VK5DSP designed the LDPC code, and Brady KC9TPA
ported the FSK Modem to C. The Wenet payloads are launched and tracked
as part of http///www.areg.org.au club High Altitude Balloon (HAB)
activities.
Cheers,
David

Post by David Ranch
Hello David,
I did read your blog entry, the associated WeNet presentation, and the
Wenet github page. It all sounds fascinating but it was very high level
on the TX side (just highly brushed over). I'd love to see a higher
- Which FSK TX board were you using? It looks like it's based on a
RFM98W RoRA board which might be soldered onto another Rpi HATT board
(hard to tell from the 2016_11_wenet_presentation.pdf but maybe that's
the FSK modem sitting on top of another Hab Supplies board?
- This design seemed to be uni-directional but do you think it would
be hard to make bidirectional? The current design seems to take a
picture, apply FEC, etc. and then send in batch.
- Did you ever measure the current draw when transmitting?
- There is very little detail on the antenna you used. Did you just
use a dipole? Maybe more of a panel design?
- Thanks for all the work in publishing the details at
https://github.com/projecthorus/wenet
--David
KI6ZHD

Post by David Rowe
100's of kHz wide. For FSK you need about the symbol rate spacing
between tones, so for 2FSK at 115 kbit/s, the tones are spaced at
115kHz-ish, plus a bit more, think we use 400 kHz separation in

practice.

Post by David Ranch

Post by David Rowe
Yes we only managed these fine results through very careful engineering,
and deliberately avoiding the common trap or re-using legacy FM radio

kit.

Post by David Ranch

Post by David Rowe
Cheers,
David

------------------------------------------------------------

------------------

Post by David Ranch
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2

------------------------------------------------------------
------------------
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2

[David Ranch]

Hello Mark,

I was curious if anyone was going to mention 4FSK as it sure sounds like
the better route to go. I'd love to see something running 115.2k but I
don't know about 100s of Khz wide signals. :-) Maybe something like
the Moeino boards with an on-board arduino would be a decent solution
(run to completion style)

https://lowpowerlab.com/shop/category/60


--David
KI6ZHD

[glen english]

------------------------------------------------------------------------------
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford

[Steve]

Just doing a little web browsing and I found this neat little 27 dBm TX
circuit. Needs some filtering of course, but otherwise pretty simple with a
MMIC chip. If I read that right, it needs a 2 dBm drive.

It's a MSA-0886 (Agilent, Avagotech) or newer MAR‐8ASM+ Mini-Circuits

There's a company selling kits, but at $15 the markup seems high. Should be
able to lay that out easy right on the data board.


​

​

[glen english]

------------------------------------------------------------------------------
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford

[Brady O'Brien]

LSI is Avago is Broadcom now. That setup is literally just the datasheet
recommended biasing configuration, heh. There are a couple of cheaper
recommended replacements for the MSA-0886 in an obsolescence notice I
found. http://www.farnell.com/datasheets/1635257.pdf

Post by Steve
Just doing a little web browsing and I found this neat little 27 dBm TX
circuit. Needs some filtering of course, but otherwise pretty simple with a
MMIC chip. If I read that right, it needs a 2 dBm drive.
It's a MSA-0886 (Agilent, Avagotech) or newer MAR‐8ASM+ Mini-Circuits
There's a company selling kits, but at $15 the markup seems high. Should
be able to lay that out easy right on the data board.
​
​
------------------------------------------------------------
------------------
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2

[Steve]

I tried to buy one on digi-key. obsolete, ha. No wonder he's selling them
for $15...

Yea, I noticed the diagram similarity. I've bought stuff from mini-circuits
before, but it was a pain.

Post by Brady O'Brien
LSI is Avago is Broadcom now. That setup is literally just the datasheet
recommended biasing configuration, heh. There are a couple of cheaper
recommended replacements for the MSA-0886 in an obsolescence notice I
found. http://www.farnell.com/datasheets/1635257.pdf

Post by Steve
Just doing a little web browsing and I found this neat little 27 dBm TX
circuit. Needs some filtering of course, but otherwise pretty simple with a
MMIC chip. If I read that right, it needs a 2 dBm drive.
It's a MSA-0886 (Agilent, Avagotech) or newer MAR‐8ASM+ Mini-Circuits
There's a company selling kits, but at $15 the markup seems high. Should
be able to lay that out easy right on the data board.
​
​
------------------------------------------------------------
------------------
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2

------------------------------------------------------------
------------------
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2

[glen english]

------------------------------------------------------------------------------
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford

[Steve]

The replacement AVT-53663 is pretty nice though. 5V 20 dB gain at 44 cents.

[Brady O'Brien]

The AVT-53663 has 20dB of gain, but a P1dB of only 15.1 dBm, which is about
35mW.

Post by Steve
The replacement AVT-53663 is pretty nice though. 5V 20 dB gain at 44 cents.
------------------------------------------------------------
------------------
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2

[glen english]

------------------------------------------------------------------------------
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford

[glen english]

And don't buy that kit !

why?

because the design is awful. it is complete rubbish.
You cant use construction techniques like that, those audio style RF
chokes etc with a 5GHz MMIC...

the chances are, into some loads it will oscillate all over the place

stay clear of it ! rubbish like this buyer beware.

This guy has some decent kits :

http://www.minikits.com.au/electronic-kits/rf-amplifiers

nothing state of art but all stable and of reasonable design
post to USA no problem.

everything from +20- +23 MMIC based
http://www.minikits.com.au/electronic-kits/rf-amplifiers/rf-wideband

to BRICK outputs with proper chip components

http://www.minikits.com.au/electronic-kits/rf-amplifiers/rf-high-power

and low pass filters

http://www.minikits.com.au/electronic-kits/filter-kits

like this
http://www.minikits.com.au/electronic-kits/filter-kits/vhf-lowpass-filters


not that other rubbish.

[Steve]

Thanks Glenn, I'm cheap, so $15 is a lot of ice cream!

I was looking for a way to amplify VHF cheaply, and I figured the MMIC
was competitive with just a transistor.

[glen english]

you wont find 0.5W MMICs with 50 in out (unless it is a 32V GaN one)
not really, there are some but you have to hang some matching and know
how around them
the Triquint TQP7M9102 or 9103
from Mouser
yes, they are speced from 400 meg upward but they match fine at 100 megs.

Now, most of this stuff is for GHz, so what you need to do for the RF
supply choke for these low freqs like 150 MHz is put something like
20-50nH in series with your 600-800nH choke.

by itself the 600nH choke will probably have enough strays to cause the
things to oscillate. but you need something that big for the <300MHz.
with the 22nH ish first series from the MMIC, this pretty much sorts
out the GHz stability. couple of 30 gauge turns on a 1/32 drill bit etc

AH322 Triquint (mouser supply) is another goodie

otherwise
I would suggest Mini Circuits GALI84 (250mW full smoke)




g

Post by Steve
Thanks Glenn, I'm cheap, so $15 is a lot of ice cream!
I was looking for a way to amplify VHF cheaply, and I figured the MMIC
was competitive with just a transistor.
------------------------------------------------------------------------------
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2

[Dave]

Steve:
No way will you w33 27.5 dBm. The MSA0886 data sheetshows a max
output of 12.5 dBm. The 3rd order intercept point is 27 dBm.

Dave - WB6DHW
<http://wb6dhw.com>

Post by Steve
Just doing a little web browsing and I found this neat little 27 dBm
TX circuit. Needs some filtering of course, but otherwise pretty
simple with a MMIC chip. If I read that right, it needs a 2 dBm drive.
It's a MSA-0886 (Agilent, Avagotech) or newer MAR‐8ASM+ Mini-Circuits
There's a company selling kits, but at $15 the markup seems high.
Should be able to lay that out easy right on the data board.
​
​
------------------------------------------------------------------------------
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2

[Steve]

I better stop mis-reading data sheets.

No way will you w33 27.5 dBm. The MSA0886 data sheetshows a max output
of 12.5 dBm. The 3rd order intercept point is 27 dBm.
Dave - WB6DHW
<http://wb6dhw.com> <http://wb6dhw.com>

[David Rowe]

One project I'd like to work on is custom RF hardware for open source
telemetry - a tx module would be a good start. If anyone would like to
work on this with me, pls contact me. We've proven that we can meet or
exceed the performance of the commercial chips sets with open source -
next step is to come up with competitive hardware.

Another challenge is the interface to the RPi - currently we are using a
UART with start/stop bits - which is effectively 2/10 bits wasted.

Cheers,

David

Post by Mark Jessop
Radio Module: Yes, it's a RFM98W on a UpuTronics shield. TX data
straight from the UART into DIO2, bypassing all the packet handling
stuff. I have been using LoRa modules for other balloon telemetry
(command and control), so I just modified some of the shields I already
had to add the line between the RPi's TXD and the RFM98W's DIO2 line.
I did actually start out using RFM22B modules, as I wasn't initially
aware the RFM98W's could do the direct-async thing. They have the same
footprint as the RFM98W modules, hence I still had them soldered down to
a LoRa shield. I think that's whats shown in one of the presentation photos.
I'm only using these kinds of modules because it was the easiest way to
transmit FSK at 115kbaud at short notice. If anyone has any better
alternatives I'm all ears! I suspect moving to something like 4FSK is
going to require me to offload packet buffering and modulation onto
something that actually does operate in realtime.
- Mark
Hi David,
I'll let Mark VK5QI field these questions - as he developed all the
aspects covered by your questions below.
I should point out that Wenet is a joint effort of many people. While I
did some modem and integration work, Mark developed the payload software
and hardware, Bill VK5DSP designed the LDPC code, and Brady KC9TPA
ported the FSK Modem to C. The Wenet payloads are launched and tracked
as part of http///www.areg.org.au <http://www.areg.org.au> club High
Altitude Balloon (HAB)
activities.
Cheers,
David

Post by David Ranch
Hello David,
I did read your blog entry, the associated WeNet presentation, and the
Wenet github page. It all sounds fascinating but it was very high

level

Post by David Ranch
on the TX side (just highly brushed over). I'd love to see a higher
- Which FSK TX board were you using? It looks like it's based on a
RFM98W RoRA board which might be soldered onto another Rpi HATT board
(hard to tell from the 2016_11_wenet_presentation.pdf but maybe that's
the FSK modem sitting on top of another Hab Supplies board?
- This design seemed to be uni-directional but do you think it

would

Post by David Ranch
be hard to make bidirectional? The current design seems to take a
picture, apply FEC, etc. and then send in batch.
- Did you ever measure the current draw when transmitting?
- There is very little detail on the antenna you used. Did you

just

Post by David Ranch
use a dipole? Maybe more of a panel design?
- Thanks for all the work in publishing the details at
https://github.com/projecthorus/wenet

<https://github.com/projecthorus/wenet>

Post by David Ranch
--David
KI6ZHD

Post by David Rowe
100's of kHz wide. For FSK you need about the symbol rate spacing
between tones, so for 2FSK at 115 kbit/s, the tones are spaced at
115kHz-ish, plus a bit more, think we use 400 kHz separation in

practice.

Post by David Ranch

Post by David Rowe
Yes we only managed these fine results through very careful

engineering,

Post by David Ranch

Post by David Rowe
and deliberately avoiding the common trap or re-using legacy FM

radio kit.

Post by David Ranch

Post by David Rowe
Cheers,
David

------------------------------------------------------------------------------

Post by David Ranch
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up

for an

Post by David Ranch
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2

<https://lists.sourceforge.net/lists/listinfo/freetel-codec2>
------------------------------------------------------------------------------
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
<https://lists.sourceforge.net/lists/listinfo/freetel-codec2>
------------------------------------------------------------------------------
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2

[glen english]

Stuart, nice work.

what block size is the LDPC code for said project ? They are my 1st love.

g

[Mark Jessop]

Tone spacing is around 120 kHz. Occupied bandwidth is a lot more :-)

The code is a (2064,516) code (r=4/5). 2064 being 256 bytes of packet data
+ 2 byte checksum. Yes, I could probably get rid of the checksum, it's
still there for legacy reasons (and so we could compare PER of coded and
uncoded).

- Mark VK5QI

Post by glen english
Stuart, nice work.
what block size is the LDPC code for said project ? They are my 1st love.
g
------------------------------------------------------------
------------------
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2

[glen english]

------------------------------------------------------------------------------
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford

[M5AKA]

Steve wrote: "Looks like we can't do spread spectrum on UHF like they said we could"
Are you perhaps confusing the separate terms "SS" and "Spread Spectrum" ? They have different meanings in Part 97.
The SS definition you quoted defines the limited sub-set of Spread Spectrum emissions which CANNOT be used below 222 MHz but can be used above that frequency.

Those Spread Spectrum emissions which are NOT "SS" are permitted on all bands subject to § 97.309

73 Trevor M5AKA



On Monday, 27 February 2017, 22:01, Steve <***@gmail.com> wrote:


Someone asked me if we could use LoRa modulation on UHF. I said, I don't know, let me look.
LoRa is a Chirp Spread Spectrum mode (CSS). Well, the FCC allows Spread Spectrum on UHF I seemed to recall. Let me check. 97.3(c)(8):SS. Spread Spectrum emissions using bandwidth-expansion modulationemissions having designators with A, C, D, F, G, H, J or R as thefirst symbol; X as the second symbol; X as the third symbol.Boy, this doesn't look good. I find that LoRa has an emission code of X1D. X because no other emission mode works for spread spectrum.Wait a minute. If no other mode works, why did they list all those. The second and third emission symbols are mysterious, as it is obviously one channel of data.Good Lord. Looks like we can't do spread spectrum on UHF like they said we could. How about AXX emission mode. AM spread spectrum, hee.This isn't very digital voice related, but I always thought frequency hopping on UHF would make a good voice modem.I'm thinking about sending in a petition:https://docs.google.com/document/d/15n1LYDJhhdqRon7j5lM3n2rFlOgoqrk-kCenKtx9YLo/edit?usp=sharing
Have fun,Steve

[Dana Myers]

I think you might be mis-reading the relevant rules here. I suggest you give
the ARRL Lab a call and sort it out.

73,
Dana K6JQ

P.S. I'd be shocked if you could come up with a compatible implementation that
doesn't trample on Semtech's patents, which doesn't prevent experimentation, but
draws a sharp line at any kind of commercial utilization.

P.P.S. I am not an attorney despite having been accused of that several times.

Post by M5AKA
Someone asked me if we could use LoRa modulation on UHF. I said, I don't know, let me look.
LoRa is a Chirp Spread Spectrum mode (CSS). Well, the FCC allows Spread Spectrum on UHF I seemed to recall. Let me check.
97.3(c)(8): /SS. Spread Spectrum emissions using bandwidth-expansion modulation emissions having designators with A, C, D, F, G,
H, J or R as the first symbol; X as the second symbol; X as the third symbol./
Boy, this doesn't look good. I find that LoRa has an emission code of X1D. X because no other emission mode works for spread
spectrum.
Wait a minute. If no other mode works, why did they list all those. The second and third emission symbols are mysterious, as it
is obviously one channel of data.
Good Lord. Looks like we can't do spread spectrum on UHF like they said we could. How about AXX emission mode. AM spread
spectrum, hee.
This isn't very digital voice related, but I always thought frequency hopping on UHF would make a good voice modem.
https://docs.google.com/document/d/15n1LYDJhhdqRon7j5lM3n2rFlOgoqrk-kCenKtx9YLo/edit?usp=sharing
Have fun,
Steve
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2

[Bruce Perens]

LoRa has modulation designator DXX for our purposes. Pretty much everything
fits in there. If you are in doubt, you can just ask for an STA.

Thanks

Bruce

Post by Dana Myers
I think you might be mis-reading the relevant rules here. I suggest you give
the ARRL Lab a call and sort it out.
73,
Dana K6JQ
P.S. I'd be shocked if you could come up with a compatible implementation that
doesn't trample on Semtech's patents, which doesn't prevent
experimentation, but
draws a sharp line at any kind of commercial utilization.
P.P.S. I am not an attorney despite having been accused of that several times.
Someone asked me if we could use LoRa modulation on UHF. I said, I don't know, let me look.
LoRa is a Chirp Spread Spectrum mode (CSS). Well, the FCC allows Spread
Spectrum on UHF I seemed to recall. Let me check.
97.3(c)(8): *SS. Spread Spectrum emissions using bandwidth-expansion
modulation emissions having designators with A, C, D, F, G, H, J or R as
the first symbol; X as the second symbol; X as the third symbol.*
Boy, this doesn't look good. I find that LoRa has an emission code of X1D.
X because no other emission mode works for spread spectrum.
Wait a minute. If no other mode works, why did they list all those. The
second and third emission symbols are mysterious, as it is obviously one
channel of data.
Good Lord. Looks like we can't do spread spectrum on UHF like they said we
could. How about AXX emission mode. AM spread spectrum, hee.
This isn't very digital voice related, but I always thought frequency
hopping on UHF would make a good voice modem.
https://docs.google.com/document/d/15n1LYDJhhdqRon7j5lM3n2rFlOgoq
rk-kCenKtx9YLo/edit?usp=sharing
Have fun,
Steve
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
_______________________________________________
------------------------------------------------------------
------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2