Aliexpress Radioberry - Issues

[st...@radio-kits.co.uk]

Recieved my Radioberry module from Aliexpress and appear to have some issues.

If I alter the power level on TX the control isn't monotonic, using SparkSDR for example as I reduce below a value of 255 the power suddenly increases back to near full output, PIHPSDR does similar. Interestingly SDR console doesn't. I believe the TX gain control internal to the AD9866 has a faulty "bit", Simon is probably not using the attenuators.

The AD9866 runs much hotter than that in the HL2, I guess this is due to using the internal TX amplifier - when it gets very hot (it does have a heat sink fitted) you see a mass of digital noise across the spectrum at only about 50dB down on carrier.

The AD9866 had been poorly reworked, I cleaned it up but no improvement, I couldn't see if the paddle had been soldered so did it just in case.

Just checking these issues aren't normal before I replace the AD9866 and make a claim on the supplier - I guess that won't get me anywhere though! I'm suspecting the AD9866 are "pulls", do others have signs of rework - has TX been tested on other Aliexpress parts?

Steve G6ALU

[Steve Drury]

That should read "If I alter the power level on TX the control isn't monotonic, using SparkSDR for example as I reduce below a value of 224 the power suddenly increases back to near full output".

I note the price has increased substantially... perhaps they have run out of second user chips!

Steve G6ALU

--
You received this message because you are subscribed to a topic in the Google Groups "Radioberry" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/radioberry/rUA8bocApNg/unsubscribe.
To unsubscribe from this group and all its topics, send an email to radioberry+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/radioberry/2ece5692-e70c-4910-9896-5b43d0a8670an%40googlegroups.com.

Virus-free. www.avg.com

[G4ZAL]

Hi Steve,

I also have an Aliexpress RB2v4 - got mine a couple of weeks ago (61GBP delivered), but been on holiday so not much testing done.

Ran mine up on SparkSDR and it seems to reduce power nicely over the range from 255 to 10 with no noticeable increases/jumps.

Did the same with Quisk and similar result.

I currently have nothing capable of checking the pico-power from the RB2, but do see a small defection on my analogue PWR meter set to 0.5W full scale.

I drove my RB2 on SSB into a dummy load and monitored my signal on my Anan 200D and the signal sounds and 'looks' clean, same result monitoring on my HL2.

This was all done on 20M - perhaps I'll go through the rest of the bands...

Maybe I just got lucky?!

I have to say, I'm quite impressed so far and look forward to tinkering some more - small linear/filters etc

Nigel  G4ZAL

[pa3gsb]

Steve

Did you have SPI disabled?

73 Johan

PA3GSB

Op zaterdag 19 juni 2021 om 10:59:15 UTC+2 schreef st...@radio-kits.co.uk:

[Steve Drury]

Hi Johan,

Yes SPI is disabled - the interface screen is setup as per the Wiki instructions, I did note that whilst the install script was running quite a few warnings flashed past. As it was a fresh PI OS install, I used the dev version, I hope this was correct?

I will try another install on a spare SD card just to check nothing "strange" happened.

Thanks for a great product, on a previous thread I read that using the internal TX amplifier caused spuri when on 10m, I've checked on the analyser and can't see anything to be concerned about - if the chip gets very hot I did notice massive digital noise. If the chip is good a small fan should sort that issue - the Pi needs one in any case.

The intent is to build it into a stand alone radio, that was the original intent for the HL2 but the Radioberry is more appropriate.

Steve G6ALU

--
You received this message because you are subscribed to a topic in the Google Groups "Radioberry" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/radioberry/rUA8bocApNg/unsubscribe.
To unsubscribe from this group and all its topics, send an email to radioberry+...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/radioberry/3c88484d-a8d6-42b1-96e1-af1f985351d8n%40googlegroups.com.

Virus-free. www.avg.com

[Steve Drury]

Hi Nigel,

Thanks for coming back to me - can you see any rework (hand soldering) to the SMD parts on your board? All components on mine were re-flowed except for the AD9866.

Trying a fresh install and we will go from there - I will be able to change the chip if necessary - built the HL2 from scratch - but cost of a replacement chip (£36 inc VAT) negates the cost save buying from China in the first place!

Also very impressed with it, certainly encourages me to incorporate it into a stand-alone radio, PIHPSDR works nicely on a 7" screen as does SparkSDR, not loaded Quisk on it yet.

Steve G6ALU

To view this discussion on the web visit https://groups.google.com/d/msgid/radioberry/fc2d625f-56ea-4906-bf78-785c76127269n%40googlegroups.com.

[pa3gsb]

Hi Steve

The dev version is fine.

I did not read your post carefull enough... when the temperature is too high you are seeing these spurs... i have not seen these spurs also during long time use.

I have built your 20W PA... board designed by VU3TLD ; see http://www.pa3gsb.nl/2021/03/23/radioberry-tx-amplifiers/

working great!

Have fun in making a standalone radio.

73 Johan

PA3GSB

Op zaterdag 19 juni 2021 om 16:54:30 UTC+2 schreef Steve Drury:

[Steve Drury]

To the group this time...

For the stand-alone I was thinking of a 100W PA using MRF101, last time I played with them they died on me, not even any smoke! Will stick with 20w and 2 x  RD16HHf1, never had any fail on me.

Did a fresh reload onto another SD card but same issue - by the time I had it loaded there were many spurs, disappear with a squirt of freezer - time to talk to China!

Will let you know how the sand-alone goes.... looking forward to it.

Just bought an FTdx10, have hardly used it - find it so boring with just one receiver, I like to run Spark with 4!

Steve G6ALU

To view this discussion on the web visit https://groups.google.com/d/msgid/radioberry/ac74046f-e116-431f-8055-6080e8f0f990n%40googlegroups.com.

[Bob A. Booey]

MRF-101 has slot of gain, easy to overdrive.

73,

Robert, WA2T 

Sent from my iPhone

On Jun 19, 2021, at 12:20 PM, Steve Drury <st...@radio-kits.co.uk> wrote:



You received this message because you are subscribed to the Google Groups "Radioberry" group.
To unsubscribe from this group and stop receiving emails from it, send an email to radioberry+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/radioberry/22135669-8b66-7194-213e-b1034f824903%40radio-kits.co.uk.

[G4ZAL]

Hi Steve,

My board doesn't appear to have been reflowed, although the SMA soldering is poor at best! (pics attached).

C256 was missing when I got it from China.

I do temporarily have a small fan blowing across my Pi/RB2 at the moment and don't see any issues - it does get warm, but not excessively.

If building pihpsdr on the same R-Pi, read the 3rd post down by Ujjwal, here

https://groups.google.com/g/radioberry/c/nSGwiAdv-NQ

I run Quisk and SparkSDR on my Windows PC and connect over TCP/IP to the RB2/R-Pi.

Am just playing with piHPSDR on the host R-Pi as well - seems OK at first glance.

I also have an HL2 and most of the parts to build this linear amp 'add on' designed for an HL2 (using MRF120).  I have one of EA3IGT's PCB's for my build.

https://github.com/ea3igt/HL2-PA70/

Maybe it will also be useable with the RB2 and can 'follow' the I2C band change commands from the RB2?

I may switch the amp build to the RB2 as I already have a Hardrock 50 mated with the HL2 when needed.

Would also like to make my RB2 into a stand-alone/portable all-in-one rig with 7in screen.

Nigel G4ZAL

[Steve Drury]

Plot taken when warm - frequency is arbitrary, happens across the band but severity dependent on attenuator setting.

I've opened up a dispute with the seller.

Steve G6ALU

--
You received this message because you are subscribed to a topic in the Google Groups "Radioberry" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/radioberry/rUA8bocApNg/unsubscribe.
To unsubscribe from this group and all its topics, send an email to radioberry+...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/radioberry/0fdca1a7-c1b6-4ed1-8fa5-d06ddad8cae6n%40googlegroups.com.

Virus-free. www.avg.com

[Yado-san]

QTC!!

  who got Radioberry module or 50MHz LPF chip(F401,F402) from Aliexpress.

My LPF chip was a fake! 

  at RF-OUT filter F401,F402. (50MHz LPF: NFL21SP506X1C3D)

Attached is the output spectrum and filter characteristics chart compared to the real one.

p.s.

As for the spurious problem during 24.9MHz TX,

I inserted a BPF between RadioBerry2/RF-OUT and Pre-Ampand/RF-IN, the problem went away completely. 

-> temporarily used the QRP-Labs 12m BPF.

73GL

Yado-san

jg1twp

2021年6月20日日曜日 5:29:08 UTC+9 Steve Drury:

[Steve Drury]

Hi Yado-San,

I didn't realise the problem was with 24.9MHz TX.

With an unmodified Radioberry - 24.9MHz TX output of 15dBm I am seeing a spur @ -40dBm 28.664MHz, do you see a worse figure? Was this what was identified with the HL1? I guess it could be worse on some Radioberry.

By modifying the circuit so RF is taken from pins 51 and 52 (Iout_P- and Iout_P+) you get 3dBm of clean RF - I did this by removing C414 and FB401, replacing C414 with a short. Cutting track from pin 46 and 50 and linking to pins 51 and 52.

See attached plots:

JPG3 - 24.990MHz TX showing 28.66MHz spur
JPG4 - Two tone, -80dB ref to PEP
JPG5 - Both before and after modification I had a spur at 13.41MHz (about -63dBm)

I will go with this modification as its easier to achieve another 10dB gain than switch in additional filters.

My filters appear to be OK

Steve G6ALU

On 20/06/2021 08:29, Yado-san wrote:

QTC!!

  who got Radioberry module or 50MHz LPF chip(F401,F402) from Aliexpress.

My LPF chip was a fake! 

  at RF-OUT filter F401,F402. (50MHz LPF: NFL21SP506X1C3D)

Attached is the output spectrum and filter characteristics chart compared to the real one.

p.s.

As for the spurious problem during 24.9MHz TX,

I inserted a BPF between RadioBerry2/RF-OUT and Pre-Ampand/RF-IN, the problem went away completely. 

-> temporarily used the QRP-Labs 12m BPF.

73GL

Yado-san

jg1twp

2021年6月20日日曜日 5:29:08 UTC+9 Steve Drury:

Plot taken when warm - frequency is arbitrary, happens across the band but severity dependent on attenuator setting.

I've opened up a dispute with the seller.

Steve G6ALU

On 19/06/2021 18:54, G4ZAL wrote:

Hi Steve,

My board doesn't appear to have been reflowed, although the SMA soldering is poor at best! (pics attached).

C256 was missing when I got it from China.

I do temporarily have a small fan blowing across my Pi/RB2 at the moment and don't see any issues - it does get warm, but not excessively.

If building pihpsdr on the same R-Pi, read the 3rd post down by Ujjwal, here

https://groups.google.com/g/radioberry/c/nSGwiAdv-NQ

I run Quisk and SparkSDR on my Windows PC and connect over TCP/IP to the RB2/R-Pi.

Am just playing with piHPSDR on the host R-Pi as well - seems OK at first glance.

I also have an HL2 and most of the parts to build this linear amp 'add on' designed for an HL2 (using MRF120).  I have one of EA3IGT's PCB's for my build.

https://github.com/ea3igt/HL2-PA70/

Maybe it will also be useable with the RB2 and can 'follow' the I2C band change commands from the RB2?

I may switch the amp build to the RB2 as I already have a Hardrock 50 mated with the HL2 when needed.

Would also like to make my RB2 into a stand-alone/portable all-in-one rig with 7in screen.

Nigel G4ZAL

--
You received this message because you are subscribed to a topic in the Google Groups "Radioberry" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/radioberry/rUA8bocApNg/unsubscribe.
To unsubscribe from this group and all its topics, send an email to radioberry+...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/radioberry/0fdca1a7-c1b6-4ed1-8fa5-d06ddad8cae6n%40googlegroups.com.

Virus-free. www.avg.com

--
You received this message because you are subscribed to a topic in the Google Groups "Radioberry" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/radioberry/rUA8bocApNg/unsubscribe.
To unsubscribe from this group and all its topics, send an email to radioberry+...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/radioberry/54f3c392-6ded-4a23-b501-9e5344a2b9bcn%40googlegroups.com.

[pa3gsb]

Steve,

Adding an additional resistor could give you again some 10db . according to measurements done by Claudio, https://www.qsl.net/in3otd/ham_radio/Hermes-Lite/TX_out.html#specout33

If you adding your 20W amp you have gain enough.

Please let us know if this helps.

73 Johan

PA3GSB

Op zondag 20 juni 2021 om 17:06:12 UTC+2 schreef Steve Drury:

[Yado-san]

Hi Steve,

These were measured with my Radioberry2.

My spur values ranged from -50.7 dB to -53.5 dB,  with Radioberry2_CL016 and CL025.

Your readings are a little better (-55dB =Tx 15dBm - spur -40dBm), but it may just be a measurement error. I'm using RIGOL DSA815.

I knew about this spurious problem from this topic.

 https://groups.google.com/g/radioberry/c/eK0Y6Do7NR0

  https://github.com/softerhardware/Hermes-Lite/wiki/Spectrum-12M#spurs-at-nfs-plus-or-minus-mftx

  -> AD9866-IAMP generated nonharmonic spurs.

The following 12m and 10m are likely to be a problem , and cannot be removed by LPF.

  24.9MHz spur 29.1MHz (=2*76.8-5*24.9)  

  28.1MHz spur 13.1MHz (=2*76.8-3*28.1)

I think it would be difficult to add an RF amplifier OPA2677 with TxDAC output like Hermes Lite2. So I thought of inserting a BPF.

But I saw your changes and the measurement waveform. It looks nice. 

 (I'm not shure the spur at 13.41MHz/-62.9dB. JPG5)

I will try your modifications too.

Tnx!

Yado-san

jg1twp

2021年6月21日月曜日 0:06:12 UTC+9 Steve Drury:

[Steve Drury]

Johan, Yado-san,

Thanks for that, I hadn't seen it. Loading the TxDac output with 47R (I didn't have a small 33R) reduced the TX output too much but increasing this to 100R across the TxDac output made a significant difference to the spur. I initially tried a 33R between each output and ground but this reduced output to ~ -9dBm.

Original design 15dBm output - 24.990MHz TX, spur at ~ 28.66MHz 55dB below carrier
100R between TxDac pins gave 13dBm output - 24.990MHz TX, spur at ~ 28.66MHz 64dB below carrier

I did a test with a new PA configuration - Radioberry..4dB attenuator..single RD06HHF1 in class A..two RD16HHF1 in class ab. Running the first stage in class A gives more gain, interstage matching was guessed so drive could probably be further reduced but in that state I only needed 70% drive from SDRConsole to achieve 20W output, couldn't use other software due to the faulty AD9866.

As drive was reduced the 28.66MHz spur was even further reduced, at 20W out the worse is that spur at ~13MHz. Will test gain once I get the new chip, as there is plenty of drive I guess the 100R could be reduced further to give a greater improvement.

Now to work out a reliable way to add a resistor between pins 51 and 52 :)

Steve G6ALU

To view this discussion on the web visit https://groups.google.com/d/msgid/radioberry/bf9e5f14-365a-469f-9424-53364399824fn%40googlegroups.com.

[Ronald Nicholson]

Hi,

I ordered one of the Aliexpress Radioberry boards.  

Are there any critical areas to inspect for poor soldering and,/or other assembly issues, before I try connecting the board to my Raspberry Pi 4 and powering it up?

I also have a Makerfabs built Hermes Lite 2v9, and have successfully operated it from my Pi 4, M1 Mac, and iPad, using several different software applications, including stuff that I've written (such as hl2_tcp on GitHub).  Are there any differences I need to watch out for when running or developing software for Radioberry?

Thanks and 73,

Ron

n6ywu

[Radioberry]

  Hi Ron,

Just plugin. Visual inspection is always good of course.

The wiki and release page gives information.:

https://github.com/pa3gsb/Radioberry-2.x/wiki

https://github.com/pa3gsb/Radioberry-2.x/releases

I do also own a HL2 but i do not have a list of differences....but some notes..

The radioberry started a long time ago with a RB1 (double layer PCB) based on the ideas from HL-1  ..... after gaining knowledge a 4 layer PCB was designed.

Started on RPI-2 ...... now the target is RPI-4 using a kernel driver and linux service system; also made a variant gateware based on the HL-2.

The latest development is a replacement of the RPI... by a juice board.

Please look around and have fun!

73 Johan

PA3GSB

Op zondag 4 juli 2021 om 21:15:55 UTC+2 schreef ron.ni...@gmail.com:

[Ronald Nicholson]

My AURSINC Radioberry PCB from Aliexpress does not have a version number printed on the silkscreen.  

Is there an easy way to determine by visual inspection the version?

The PCB does not appear to have any re-work in the vicinity of C256?  Do I need to remove C256?

What do I need to check to determine whether my PCB has a short in this area?

Should there be any solder visible in the large PCB thru-holes below the AD9866 and the FPGA?  

My board does not appear to have any.  The thru-holes look clean.

Does this need to be fixed, and if so how?  (amount of solder, safe soldering temperature, etc.)

Thanks and 73,

Ron

n6ywu

[Radioberry]

Ron

Can you make a good high resolution photo of the top and bottom of the PCB.

73 Johan

PA3GSB

Op vrijdag 9 juli 2021 om 23:01:32 UTC+2 schreef ron.ni...@gmail.com:

[Ronald Nicholson]

Hi Johan,

Do these 2 photo's have enough resolution to determine the PCB revision?

http://www.burntwhiskers.com/rhn/files/IMG_1466.jpg 

http://www.burntwhiskers.com/rhn/files/IMG_1467.jpg 

I installed the software manually, using commands from :

https://raw.githubusercontent.com/pa3gsb/Radioberry-2.x/master/SBC/rpi-4/releases/stable/radioberry_install.sh

Then plugged the PCB into a Pi 4 and powered it up.

Here's a trimmed portion of the startup log:

sudo modinfo radioberry

        version:        0.8

        srcversion:     D797CE1A2D0B5B2005EDF7C

        vermagic:       5.10.17-v7l+ SMP mod_unload modversions ARMv7 p2v8

        ...

sudo radioberry

        Supports 4 receivers and 1 transmitter.

        Build version: 2021.01.17

        Radioberry gateware version 255-255.

        Radioberry firmware initialisation succesfully executed.

        Your SBC device is missing the following driver: '/dev/i2c-1'

        ...

hl2_tcp -d

        discover ip = 10.0.1.123 : 0x0A00017B

        MAC Address: 00.01.02.03.04.05

        Protocol: 6

        Gateware Version: 255

Attempts to read 1 slice of Rx data at 48ksps from UDP port 1024 with my test code seems to be returning data at a much higher rate than 48k.

The 9866 also seems to be running too hot to touch, even in idle, so I powered off the system, until I can figure out how to run more detailed diagnostics.

Thanks and 73,

Ron

n6ywu

[Yado-san]

Hi Ron, Johan

> Your SBC device is missing the following driver: '/dev/i2c-1'

Please enable I2C in Raspberry Pi configuration.

ps. 

 I'm not sure, but... 

  may be better to use the development version instead of the stable version.

  Pics. appears that the logo and version5 of Radioberry have been deliberately removed!

  Not a very positive thing.

73

Yado-san

jg1twp

2021年7月11日日曜日 10:13:18 UTC+9 ron.ni...@gmail.com:

[Radioberry]

Hi Ron

tnx for the nice photos.

C256 is installed at your board.

The board i received from ali; this C is removed.

Under the C256 there is a VIA (CLKOUT1) which can be connected to +3V3 rails....

remove the C256 and remove the short if 

( Manfred DB2OJ; described already the fix  https://groups.google.com/g/radioberry/c/TOTKc9GZpGk/m/TKz84lWwAAAJ)

Also use the development release; it is real easy to use the install scripts as described (https://github.com/pa3gsb/Radioberry-2.x/releases)

You logging shows: Radioberry gateware version 255-255.   This should be 73.0 (using dev release)

Iam not aware of the command hl2_tcp -d   what kind of app are you using??

Hope this helps to get your radio going.

73 Johan

PA3GSB

Op zondag 11 juli 2021 om 03:13:18 UTC+2 schreef ron.ni...@gmail.com:

[Ronald Nicholson]

Hi Johan,

I removed C256, but my PCB version appears to not have a via underneath C256. 

See: http://www.burntwhiskers.com/rhn/files/IMG_1469.jpg

Does this mean my AURSINC Radioberry PCB is an older version?

Also, when installing the use the development release,

should I select 1 = CL016 , or 2 = CL025 , for the .rbf firmware?

hl2_tcp is my HL2 RTL-SDR rtl_tcp transcoder. 

See: https://github.com/hotpaw2/hl2_tcp 

hl2_tcp -d broadcasts Metis discovery UDP packets that I use to test connectivity.

IIRC, the 9th byte in the HL2 UDP reply packet should include the FPGA gateware version.

Thanks and 73,

Ron

n6ywu

http://www.nicholson.com/rhn/

[pa3gsb]

Ron

The clock via at the top is there:

Seem that the via in the built from ali  AURSINC is moved??

You check if the clock from the AD9866 is present at both pins;  pin 53 and pin 69.

The radioberry can run with a cyclone CL106 cheaper variant and the CL025 cyclone FPGA.... i assume the ali provider places the CL025 so the option depends what you have bought?

Look in the kern.log (see other posts about gateware loading).

Hope this helps,

Johan 

PA3GSB

Op zondag 11 juli 2021 om 17:37:02 UTC+2 schreef ron.ni...@gmail.com:

[Ronald Nicholson]

Hi Johan,

Success.  

I enabled I2C, installed your development release, and configured it for CL025.

hl2_tcp -d    

now reports: Gateware Version: 73

My iPhone SDR app now receives signals from the Aliexpress AURSINC Radioberry on my Raspberry Pi 4.

Very nice.

Thanks and 73,

Ron

n6ywu

[Ronald Nicholson]

The AD9866 on my Aliexpress Radioberry board seems to run hotter than the AD9866 in my Hermes Lite 2 when running in receive-only mode.  Hot enough that the Radioberry occasionally stops working (returns lots of noise instead of signal data).  I have to shut down power to the Pi 4 to let the board cool down before it starts working again.  I'm pretty sure that my SDR software is setting the MOX bit in the Protocol 1 packets to zero, so the Radioberry board should not be heating up from trying to transmit power.

Is there some difference in the firmware driver or the FPGA gateware that could be causing more ADC power consumption on a Radioberry board than in an HL2?  Would more solder in the thru-hole under the chip help?  Or could this power consumption difference just be due to manufacturing variation in the batch of AD9866 chips that was used for assembly?  A bad batch of chips or something?

73,

Ron

n6ywu

[Bob A. Booey]

Not soldered (or not soldered well) AD9866 exposed pad?

You received this message because you are subscribed to the Google Groups "Radioberry" group.

To unsubscribe from this group and stop receiving emails from it, send an email to radioberry+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/radioberry/bb843ba5-b0d4-4e26-ac90-671f22713978n%40googlegroups.com.

[softerh...@gmail.com]

Hi Ron,

The gateware is pretty much the same, but the gateware does configure the AD9866 from a table at power on. Some of these configuration settings may be different for the Radioberry. For example, the Radioberry uses the IAMP for extra TX power directly from the AD9866. This may increase standing current and power even during RX only.

The HL2 has a larger PCB. The PCB plays a significant role in heat dissipation. Also, the HL2 can dissipate heat via the enclosure rails.

The quality of the AD9866 used in AliExpress builds is not guaranteed. If you search for AD9866 on AliExpress, you will get many hits. I've bought and tested some of these parts before. The experts at the local assembly house refused to use some of these parts for replacement as they found hard-to-notice cracks and damage to the case. This damage may degrade the die to package thermal pad heat transfer. An early HL2 used one of this AD9866s and always ran hot, even though the thermal pad was well soldered to the PCB.

Early versions of the AD9866 may not have been as power efficient. There is a date code on the IC package. It is easy to find older ICs, even 10 to 15 years old, on the Chinese gray or second market.

For the HL2, we buy the AD9866 directly from the Analog devices or other reputable sources. This is out of control for a board which just appears on AliExpress.

73,

Steve

kf7o