ADALM-PLUTO SDR

[G0MJW]

These have started shipping. I ordered mine so long ago I had forgotten until UPS arrived with a parcel. They cost $99 and are based on the AD9369. They are intended as and are nice learning devices that run off a USB port. Frequency coverage is supposed to be 325MHz to 3.8GHz at 20MS 12 bit but those of us who have them have had success from 70MHz to 6 GHz after changing some of the parameters.

Has anyone else been looking at these and done any measurements? Naturally, the first thing I did was to take it apart and take some pictures.

Mike

[hendorog]

Very interesting that they can go from 70-6GHz. I thought the chip AD9363 was 325MHz to 3.8GHz.

Can to explain how you managed that one? It's definitely worth getting one of these if it can still be made to do that.

[G0MJW]

See this tweet from Alexandru Csete https://twitter.com/csete/status/897872430236094464 which pointed me at this post https://wiki.analog.com/university/tools/pluto/users/customizing which shouldn't work but of course we all immediately tried it.

Quote:

There were some early PlutoSDR devices which use the AD9364, which is nearly identical to the AD9363 used the production builds. If you have one of the AD9364 based PlutoSDR devices, it's a quick matter of using the U-Boot's fw_printenv and fw_setenv commands to get that device's larger tuning range (70-6000 MHz) and larger bandwidth (56MHz).

From your favorite serial application, just open a serial connection to the PlutoSDR. The username is root and the password is analog.

This will be the default (based on the AD9363):

# fw_printenv attr_name
## Error: "attr_name" not defined
# fw_printenv attr_val
## Error: "attr_val" not defined
#

To change things to the AD9364 configuration:

# fw_setenv attr_name compatible
# fw_setenv attr_val “ad9364”
# pluto_reboot reset

Unquote.

And there it was. I didn't have one of the AD9364 units, at least, that's not what it says on the chips, but it now seems to cover the wider frequency range. Perhaps the spec is not as good, or maybe there are holes in coverage, or maybe the AD9363 is a AD9364 that just didn't quite meet spec, or perhaps AD just want to be nice to us knowing in the long run it will help sales.

Mike

[hendorog]

Thanks, it seems certainly worth a try. I just ordered one from Digikey - NZD $148 which is about USD 108 at the moment.

Its worth noting that the AD9363 and AD9364 chips are $198  and $312 NZD on Digikey respectively so the PLUTO module is a great deal - even if the hack doesn't work!

[borjam]

I just ordered one

And nice anchovies!

[CopperCone]

Does this work with any popular software?

What is the probability of intercept on signals? I can't seem to find a spec sheet.

[hendorog]

Yes:
MATLAB®, Simulink® support
GNU Radio sink and source blocks
libiio, a C, C++, C#, and Python API

http://www.analog.com/en/design-center/evaluation-hardware-and-software/evaluation-boards-kits/adalm-pluto.html#eb-overview

[hendorog]

Hmm, this recent comment implies the hack no longer works:

http://www.rtl-sdr.com/adalm-pluto-sdr-hack-tune-70-mhz-to-6-ghz-and-gqrx-install/
"I bought two of these from Digikey and I have carried out the AD9364 ‘hack’ on both units. GQRX refuses to tune below 324.999 and the GNURadio source and sink don’t work either. The units remain tuned to the last good freq. Pity because I was keen to try it on 2m."

I would not be surprised if some early chips were just 9364's in disguise and the later ones which are coming through now have been 'refined' to remove that capability.

[CopperCone]

Do you think it was 'refined' to restrict the capabilities from a 'military' or 'business is war' prospective, or if some engineer simply wanted to reduce error ? (i.e. advanced analysis revealed that there is some kind of measurement error with the 'hack' that voided specifications?

It's not like AD is a test equipment manufacturer, they have a vested interest in selling as much as those modules as possible.

https://finance.yahoo.com/quote/ADI/

[hendorog]

Do you think it was 'refined' to restrict the capabilities from a 'military' or 'business is war' prospective, or if some engineer simply wanted to reduce error ? (i.e. advanced analysis revealed that there is some kind of measurement error with the 'hack' that voided specifications?

It's not like AD is a test equipment manufacturer, they have a vested interest in selling as much as those modules as possible.

https://finance.yahoo.com/quote/ADI/

Of course not - and this is complete speculation on my part of course based on a single comment on a forum of someone saying it doesn't work for them...

I just think they always intended to remove the AD9364 functionality from the AD9363 chip once production ramped up. They could get 2 SKU's to market for the price of one and then refined the 9363 to match its specs.

[CopperCone]

hmm, do you think you can replace the chip?

seems unlikely that the code would 'ask' the chip but you never know

[hendorog]

hmm, do you think you can replace the chip?

seems unlikely that the code would 'ask' the chip but you never know

I'm guessing yes as the pinout is almost identical. But the price of the chip is so high I wouldn't bother.

Will wait and see what happens when it arrives.

[CopperCone]

Based on that post, it seems that whatever command is used to reconfigure the device was blocked...  though there is some contention on the website if its not just user error in doing the 'hack'

Or does it perform some kind of self test to see if the span settings are reasonable? Perhaps if you run the transmitter without a load there is crosstalk specification between TX and RX that is guaranteed by design as a self check.

I assume that " but it’s not guaranteed to give the full tuning range and bandwidth for every single unit." means that there is weird behavior, not software detected lockouts..

[borjam]

Curiously, the possibility of expanding the frequency range is mentioned on the Mathworks website.

https://es.mathworks.com/hardware-support/adalm-pluto-radio.html

[CopperCone]

does anyone know what USB speed this device uses?

[RoGeorge]

USB 2.0
https://wiki.analog.com/university/tools/pluto/users/name?rev=1487174293

[hendorog]

Mine has just arrived from Digikey. Need to download stuff I assume and will post an update once I've had a crack at the hack.

[borjam]

Mine has just arrived from Farnell. Seems to be out of stock everywhere.

So far I am looking at documentation and checking what I need in order to install GnuRadio on Mac OS X properly.

For now I have plugged it, updated firmware to the latest version (0.22) and tried two hacks. Enabling the AD9364 hack
in order to increase the tuning range, and something else I found on plutosdr.com: enabling a second CPU core.

dmesg shows this after doing the AD9364 hack:

Code: [Select]

cf_axi_adc 79020000.cf-ad9361-lpc: ADI AIM (10.00.b) at 0x79020000 mapped to 0xe0938000, probed ADC AD9364 as MASTER

Not sure wether it really detected an AD9364 or it just assumes it thanks to the fw environment variables I set.

As for the dual core hack shown here: https://www.plutosdr.com/viewtopic.php?f=7&t=11

It worked for me:

Code: [Select]

# cat /proc/cpuinfo
processor : 0
model name : ARMv7 Processor rev 0 (v7l)
BogoMIPS : 666.66
Features : half thumb fastmult vfp edsp neon vfpv3 tls vfpd32
CPU implementer : 0x41
CPU architecture: 7
CPU variant : 0x3
CPU part : 0xc09
CPU revision : 0

processor : 1
model name : ARMv7 Processor rev 0 (v7l)
BogoMIPS : 666.66
Features : half thumb fastmult vfp edsp neon vfpv3 tls vfpd32
CPU implementer : 0x41
CPU architecture: 7
CPU variant : 0x3
CPU part : 0xc09
CPU revision : 0

Hardware : Xilinx Zynq Platform
Revision : 0003
Serial : 0000000000000000

So much to explore so far, for now I have just done blind "script kiddying"

[hendorog]

Bit of an update from me:- it does seem to tune ok after the hack. I could see the FM BCB on the waterfall.

A few tips and references:

The easiest way to do an initial test and get it to do something  is to install the windows IIO oscilloscope app.
https://wiki.analog.com/resources/tools-software/linux-software/iio_oscilloscope

The easiest path by far to get GNU Radio going with the drivers is to use a linux box. Virtual box is fine.

Disk images for virtualbox and vmware can be downloaded from here:
http://www.osboxes.org/ubuntu/

Edit: I'm using Ubuntu 16.10 for the following. I've tried 17.10 but found the gnu-radio compile fails.
There is an alternative approach here for 17.04 and newer: https://wiki.analog.com/resources/tools-software/linux-software/gnuradio

Once linux is up:
Install pip:

Code: [Select]

sudo apt-get install python-pip

Use pybomb to install gnu radio - instructions came from here: https://www.gnuradio.org/blog/pybombs-the-what-the-how-and-the-why/

Code: [Select]

sudo pip install pybombs
pybombs recipes add gr-recipes git+https://github.com/gnuradio/gr-recipes.git
mkdir prefix/
pybombs prefix init -a default prefix/default/ -R gnuradio-default

Edit the gr-iio pybomb recipe to install the latest code instead of a tag - instructions came from here https://www.reddit.com/r/RTLSDR/comments/6tnjba/getting_started_with_adalmpluto_and_gnu_radio/

Code: [Select]

sed -i 's/gitrev.*/gitbranch: master/' ~/.pybombs/recipes/gr-recipes/gr-iio.lwr

Now use pybomb to install gr-iio

Code: [Select]

pybombs install gr-iio

Run Gnu Radio

Code: [Select]

source ~/prefix/default/setup_env.sh
gnuradio-companion

Load and run the attached gnuradio sample file

Note the source URL which worked for me: ip:192.168.2.1
The one in the docs didn't work (ip:pluto.local) - perhaps I missed something.

[RoGeorge]

As found on Internet, one of the ADALM-PLUTO designers talks about the board using a custom Zynq 7010, with only one core. Maybe they switched to the usual 2 core Zynq now.

Regarding the other hack, for the frequency range, reporting the chip as being an AD9364 is one thing, but it is not clear enough for me, because I don't have a board:

Can you tune the Rx/Tx in the 70-6000 MHz range after applying the frequency hack, please?
Can you, let's say, receive an FM radio station after applying the frequency hack, or see the spectrum of a 5 GHz Wi-Fi router?

[hendorog]

As found on Internet, one of the ADALM-PLUTO designers talks about the board using a custom Zynq 7010, with only one core. Maybe they switched to the usual 2 core Zynq now.

Regarding the other hack, for the frequency range, reporting the chip as being an AD9364 is one thing, but it is not clear enough for me, because I don't have a board:

Can you tune the Rx/Tx in the 70-6000 MHz range after applying the frequency hack, please?
Can you, let's say, receive an FM radio station after applying the frequency hack, or see the spectrum of a 5 GHz Wi-Fi router?

Yes I read that single core comment too. However the thing definately has two cores, I've tried it and get same results as borjam.

I mentioned in my last comment that it tunes to the FM broadcast band, I guess you didn't read that bit - check out that attached image.

[CopperCone]

did you measure its probability of intercept?

I may do so today using my tektronics 284 picosecond pulser or my wavetek pulse generator.

I guess i hook it up to a counter and an antenna and see how many intercepts there is over a period of time for various wave periods , not sure of the proper procedure

then calculate it over a course of a few hours

[hendorog]

did you measure its probability of intercept?

I may do so today using my tektronics 284 picosecond pulser or my wavetek pulse generator.

I guess i hook it up to a counter and an antenna and see how many intercepts there is over a period of time for various wave periods , not sure of the proper procedure

then calculate it over a course of a few hours

Nope, and I don't know much about that apart from what I read about SignalHounds implementation.

I guess the USB 2 will be a limitation it so some sort of processing on the device would be required.

[G0MJW]

did you measure its probability of intercept?

I may do so today using my tektronics 284 picosecond pulser or my wavetek pulse generator.

I guess i hook it up to a counter and an antenna and see how many intercepts there is over a period of time for various wave periods , not sure of the proper procedure

then calculate it over a course of a few hours

Don't take this the wrong way but what are you talking about? And maybe also why is it important?

[CopperCone]

Ok, I repaired my pulse generator. I think I can determine probability of intercept by comparing my waterfall graph to a counter hooked up to a pulse generator with specific delay between short pulses.

Is there a way to some how 'count' the waterfall with GNU Radio?