Analog Devices ADALM2000

[RobRob]

Mouser sells the Digilent breakout/breadboard that's compatible with the ADALM2000 for $19.95. It plugs into all the output pins like the Digilent scope adapter.

[BFX]

Hi guys,

board presented by friend of mine mrjoda was my idea and he drawn the first version of PCB.
Why? I don't like original Digilent board because if you attach standard bnc scope probes it's huge and often fell down.
My idea was something light with inputs at front. And hey! There are probes for those crappy small Chinese scopes


I was surprised by quality and decision was easy.
And here is result (little bit different from mrjoda drawn)

There is additional button for monitoring output of first generator output by ch.1 of oscilloscope. (together with 6db attu is bode plot really simple )

There is missing one ninigi connector for PSU which I forgot to order.
Next step is design nice 3D printed case.
I'm so happy with the result, that's exactly I expected.

[BFX]

I forgot Last picture 

[Kosmic]

Discovered something interesting today. I always assumed that the output of the signal generator was calibrated for 50\$\Omega\$ load. Looking at the schematic kind of confirm that assumption.



Testing with a multimeter I was observing the right voltage. Ex setting 1V pp I was reading around 0.354V RMS. But a multimeter normally 1M\$\Omega\$ not 50\$\Omega\$. So I did some more testing with my oscilloscope and it look like the default calibration (nothing connected on the signal generator port) is for 1M\$\Omega\$ load.

Now the interesting part is that if you have a 50\$\Omega\$ load (terminator or other) connected to the signal generator port during calibration, the signal generator will now be calibrated for a 50\$\Omega\$ load.

[mrjoda]

here is my board. If you want, send me pm i have 4 pcs left. I will send it for postage + 1€ for board.

[mrjoda]

so, that was quick. Spare PCBs are already reserved 

[BFX]

Discovered something interesting today. I always assumed that the output of the signal generator was calibrated for 50\$\Omega\$ load. Looking at the schematic kind of confirm that assumption.



Testing with a multimeter I was observing the right voltage. Ex setting 1V pp I was reading around 0.354V RMS. But a multimeter normally 1M\$\Omega\$ not 50\$\Omega\$. So I did some more testing with my oscilloscope and it look like the default calibration (nothing connected on the signal generator port) is for 1M\$\Omega\$ load.

Now the interesting part is that if you have a 50\$\Omega\$ load (terminator or other) connected to the signal generator port during calibration, the signal generator will now be calibrated for a 50\$\Omega\$ load.

During my testing of transformers I found out that it is 50 ohm termination and therefore I've put those 6db attu on the board.
Without attenuation reading was so unstable and useless.

And Bode100 also has an attenuator on the output. (Poorly documented  )

[Ben321]

I'm considering buying a ADALM2000 myself. It has an analog-to-digital sample rate of 100MSPS, which is really impressive, as it means I should get a full 50MHz bandwidth, great for capturing a few seconds of 50MHz wide RF spectrum for later analysis (100MSPS with 12bit bitdepth padded to 2 bytes per sample gives you 200MBytes per second, so a huge file but useful for analysis). Just like a professional high speed digitizer except only a fraction of the price. However the specs sheet at https://media.digikey.com/pdf/Data%20Sheets/Analog%20Devices%20PDFs/ADALM2000_Web.pdf is telling me I only get 5MHz bandwidth out of it. Apparantly they decided to incorporate an internal 5MHz LPF in it between the input pins and the ADC. A different spec sheet for the same device at a different website https://docs.rs-online.com/fb1d/0900766b815d846d.pdf says the bandwidth is 25MHz (better than 5MHz for sure, but not as good as 50MHz). What's the actual specs? Did the specs actually change? Did Analog Devices actually modify the manufacturing process for the ADALM2000, so there actually exist 2 different variants of the ADALM2000 (each with different specs), but with the exact same product name?

Also, is it possible to bypass the internal LPF (maybe soldering a wire across a component), to get the full 50MHz bandwidth that's possible with a 100MSPS sample rate?

[Kosmic]

I'm considering buying a ADALM2000 myself. It has an analog-to-digital sample rate of 100MSPS, which is really impressive, as it means I should get a full 50MHz bandwidth, great for capturing a few seconds of 50MHz wide RF spectrum for later analysis (100MSPS with 12bit bitdepth padded to 2 bytes per sample gives you 200MBytes per second, so a huge file but useful for analysis). Just like a professional high speed digitizer except only a fraction of the price. However the specs sheet at https://media.digikey.com/pdf/Data%20Sheets/Analog%20Devices%20PDFs/ADALM2000_Web.pdf is telling me I only get 5MHz bandwidth out of it. Apparantly they decided to incorporate an internal 5MHz LPF in it between the input pins and the ADC. A different spec sheet for the same device at a different website https://docs.rs-online.com/fb1d/0900766b815d846d.pdf says the bandwidth is 25MHz (better than 5MHz for sure, but not as good as 50MHz). What's the actual specs? Did the specs actually change? Did Analog Devices actually modify the manufacturing process for the ADALM2000, so there actually exist 2 different variants of the ADALM2000 (each with different specs), but with the exact same product name?

Also, is it possible to bypass the internal LPF (maybe soldering a wire across a component), to get the full 50MHz bandwidth that's possible with a 100MSPS sample rate?

Schematics are over here: https://wiki.analog.com/university/tools/m2k/devs/hardware

There is some minor differences between Rev A and D on the input.

I remember testing the spectrum analyzer functionality up to 50Mhz. I might be mistaken though. Could do more tests tonight if you want.

[Ben321]

I'm considering buying a ADALM2000 myself. It has an analog-to-digital sample rate of 100MSPS, which is really impressive, as it means I should get a full 50MHz bandwidth, great for capturing a few seconds of 50MHz wide RF spectrum for later analysis (100MSPS with 12bit bitdepth padded to 2 bytes per sample gives you 200MBytes per second, so a huge file but useful for analysis). Just like a professional high speed digitizer except only a fraction of the price. However the specs sheet at https://media.digikey.com/pdf/Data%20Sheets/Analog%20Devices%20PDFs/ADALM2000_Web.pdf is telling me I only get 5MHz bandwidth out of it. Apparantly they decided to incorporate an internal 5MHz LPF in it between the input pins and the ADC. A different spec sheet for the same device at a different website https://docs.rs-online.com/fb1d/0900766b815d846d.pdf says the bandwidth is 25MHz (better than 5MHz for sure, but not as good as 50MHz). What's the actual specs? Did the specs actually change? Did Analog Devices actually modify the manufacturing process for the ADALM2000, so there actually exist 2 different variants of the ADALM2000 (each with different specs), but with the exact same product name?

Also, is it possible to bypass the internal LPF (maybe soldering a wire across a component), to get the full 50MHz bandwidth that's possible with a 100MSPS sample rate?

Schematics are over here: https://wiki.analog.com/university/tools/m2k/devs/hardware

There is some minor differences between Rev A and D on the input.

I remember testing the spectrum analyzer functionality up to 50Mhz. I might be mistaken though. Could do more tests tonight if you want.

Thanks. That would be great if you could check that. Also, does the device's analog input support continuous streaming (sends all samples to the PC)? Or does it only support block transfer (sends a block of samples and then there may be some delay before the next block is sent)? I know that the official Scopy software has an osciloscope mode, that only supports block transfer mode (each block is sent in response to a trigger, or free run, but no guarenty that there will be no delay between blocks). However, I also know that the ADALM2000 has an SDK for you to design your own software, so if the hardware permits a continuous streaming mode for analog input, then I should be able to write my own software to use that streaming mode.

[Kosmic]

So I tested at different point (1, 10, 20, 30, 40 and 49Mhz) at 0dBm. I initially wanted to use the max hold feature but it stop working after some time  My source was a calibrated IFR2025 with a accuracy of +-0.8dBm.

the Spectrum Analyzer app was sweeping from 1Hz to 50MHz with a 762.94Hz resolution bandwidth, gain mode low, averaging 16 samples.

The -3dB point seem to be at around 20Mhz. At 49MHz its something like -9dB off.

[Kosmic]

I'm considering buying a ADALM2000 myself. It has an analog-to-digital sample rate of 100MSPS, which is really impressive, as it means I should get a full 50MHz bandwidth, great for capturing a few seconds of 50MHz wide RF spectrum for later analysis (100MSPS with 12bit bitdepth padded to 2 bytes per sample gives you 200MBytes per second, so a huge file but useful for analysis). Just like a professional high speed digitizer except only a fraction of the price. However the specs sheet at https://media.digikey.com/pdf/Data%20Sheets/Analog%20Devices%20PDFs/ADALM2000_Web.pdf is telling me I only get 5MHz bandwidth out of it. Apparantly they decided to incorporate an internal 5MHz LPF in it between the input pins and the ADC. A different spec sheet for the same device at a different website https://docs.rs-online.com/fb1d/0900766b815d846d.pdf says the bandwidth is 25MHz (better than 5MHz for sure, but not as good as 50MHz). What's the actual specs? Did the specs actually change? Did Analog Devices actually modify the manufacturing process for the ADALM2000, so there actually exist 2 different variants of the ADALM2000 (each with different specs), but with the exact same product name?

Also, is it possible to bypass the internal LPF (maybe soldering a wire across a component), to get the full 50MHz bandwidth that's possible with a 100MSPS sample rate?

Schematics are over here: https://wiki.analog.com/university/tools/m2k/devs/hardware

There is some minor differences between Rev A and D on the input.

I remember testing the spectrum analyzer functionality up to 50Mhz. I might be mistaken though. Could do more tests tonight if you want.

Thanks. That would be great if you could check that. Also, does the device's analog input support continuous streaming (sends all samples to the PC)? Or does it only support block transfer (sends a block of samples and then there may be some delay before the next block is sent)? I know that the official Scopy software has an osciloscope mode, that only supports block transfer mode (each block is sent in response to a trigger, or free run, but no guarenty that there will be no delay between blocks). However, I also know that the ADALM2000 has an SDK for you to design your own software, so if the hardware permits a continuous streaming mode for analog input, then I should be able to write my own software to use that streaming mode.

There's a C++/C#/Python library you can use.

https://wiki.analog.com/university/tools/m2k/libm2k/libm2k#download
https://github.com/analogdevicesinc/libm2k

Looked quickly at the API, and yes you can stream.

[Kosmic]

Hmmm this is new, they have a driver for Labview now

https://wiki.analog.com/university/tools/m2k/labview

[Ben321]

I'm considering buying a ADALM2000 myself. It has an analog-to-digital sample rate of 100MSPS, which is really impressive, as it means I should get a full 50MHz bandwidth, great for capturing a few seconds of 50MHz wide RF spectrum for later analysis (100MSPS with 12bit bitdepth padded to 2 bytes per sample gives you 200MBytes per second, so a huge file but useful for analysis). Just like a professional high speed digitizer except only a fraction of the price. However the specs sheet at https://media.digikey.com/pdf/Data%20Sheets/Analog%20Devices%20PDFs/ADALM2000_Web.pdf is telling me I only get 5MHz bandwidth out of it. Apparantly they decided to incorporate an internal 5MHz LPF in it between the input pins and the ADC. A different spec sheet for the same device at a different website https://docs.rs-online.com/fb1d/0900766b815d846d.pdf says the bandwidth is 25MHz (better than 5MHz for sure, but not as good as 50MHz). What's the actual specs? Did the specs actually change? Did Analog Devices actually modify the manufacturing process for the ADALM2000, so there actually exist 2 different variants of the ADALM2000 (each with different specs), but with the exact same product name?

Also, is it possible to bypass the internal LPF (maybe soldering a wire across a component), to get the full 50MHz bandwidth that's possible with a 100MSPS sample rate?

Schematics are over here: https://wiki.analog.com/university/tools/m2k/devs/hardware

There is some minor differences between Rev A and D on the input.

I remember testing the spectrum analyzer functionality up to 50Mhz. I might be mistaken though. Could do more tests tonight if you want.

Thanks. That would be great if you could check that. Also, does the device's analog input support continuous streaming (sends all samples to the PC)? Or does it only support block transfer (sends a block of samples and then there may be some delay before the next block is sent)? I know that the official Scopy software has an osciloscope mode, that only supports block transfer mode (each block is sent in response to a trigger, or free run, but no guarenty that there will be no delay between blocks). However, I also know that the ADALM2000 has an SDK for you to design your own software, so if the hardware permits a continuous streaming mode for analog input, then I should be able to write my own software to use that streaming mode.

There's a C++/C#/Python library you can use.

https://wiki.analog.com/university/tools/m2k/libm2k/libm2k#download
https://github.com/analogdevicesinc/libm2k

Looked quickly at the API, and yes you can stream.

Is it better to use libm2k for this? or is it better to use libiio for this? Part of installation of the ADALM2000 in the first place is installation of libiio on my PC. The only problem I have with using libiio right now is knowing what folder it put libiio in when it installed it. I don't see a libiio folder in either Program Files or Program Files (x86). Of course if libm2k is the preferred method for writing software for controlling the ADALM2000, then I will use that instead of libiio.

[Kosmic]

I'm considering buying a ADALM2000 myself. It has an analog-to-digital sample rate of 100MSPS, which is really impressive, as it means I should get a full 50MHz bandwidth, great for capturing a few seconds of 50MHz wide RF spectrum for later analysis (100MSPS with 12bit bitdepth padded to 2 bytes per sample gives you 200MBytes per second, so a huge file but useful for analysis). Just like a professional high speed digitizer except only a fraction of the price. However the specs sheet at https://media.digikey.com/pdf/Data%20Sheets/Analog%20Devices%20PDFs/ADALM2000_Web.pdf is telling me I only get 5MHz bandwidth out of it. Apparantly they decided to incorporate an internal 5MHz LPF in it between the input pins and the ADC. A different spec sheet for the same device at a different website https://docs.rs-online.com/fb1d/0900766b815d846d.pdf says the bandwidth is 25MHz (better than 5MHz for sure, but not as good as 50MHz). What's the actual specs? Did the specs actually change? Did Analog Devices actually modify the manufacturing process for the ADALM2000, so there actually exist 2 different variants of the ADALM2000 (each with different specs), but with the exact same product name?

Also, is it possible to bypass the internal LPF (maybe soldering a wire across a component), to get the full 50MHz bandwidth that's possible with a 100MSPS sample rate?

Schematics are over here: https://wiki.analog.com/university/tools/m2k/devs/hardware

There is some minor differences between Rev A and D on the input.

I remember testing the spectrum analyzer functionality up to 50Mhz. I might be mistaken though. Could do more tests tonight if you want.

Thanks. That would be great if you could check that. Also, does the device's analog input support continuous streaming (sends all samples to the PC)? Or does it only support block transfer (sends a block of samples and then there may be some delay before the next block is sent)? I know that the official Scopy software has an osciloscope mode, that only supports block transfer mode (each block is sent in response to a trigger, or free run, but no guarenty that there will be no delay between blocks). However, I also know that the ADALM2000 has an SDK for you to design your own software, so if the hardware permits a continuous streaming mode for analog input, then I should be able to write my own software to use that streaming mode.

There's a C++/C#/Python library you can use.

https://wiki.analog.com/university/tools/m2k/libm2k/libm2k#download
https://github.com/analogdevicesinc/libm2k

Looked quickly at the API, and yes you can stream.

Is it better to use libm2k for this? or is it better to use libiio for this? Part of installation of the ADALM2000 in the first place is installation of libiio on my PC. The only problem I have with using libiio right now is knowing what folder it put libiio in when it installed it. I don't see a libiio folder in either Program Files or Program Files (x86). Of course if libm2k is the preferred method for writing software for controlling the ADALM2000, then I will use that instead of libiio.

Better to use libm2k since you have specific API for the Adalm2000. libiio is really generic and was specifically created to transport data from the device to the PC. I would not be surprise if libm2k was using libiio under the hood.

Edit: Yes libm2k use libiio.

[Ben321]

I'm considering buying a ADALM2000 myself. It has an analog-to-digital sample rate of 100MSPS, which is really impressive, as it means I should get a full 50MHz bandwidth, great for capturing a few seconds of 50MHz wide RF spectrum for later analysis (100MSPS with 12bit bitdepth padded to 2 bytes per sample gives you 200MBytes per second, so a huge file but useful for analysis). Just like a professional high speed digitizer except only a fraction of the price. However the specs sheet at https://media.digikey.com/pdf/Data%20Sheets/Analog%20Devices%20PDFs/ADALM2000_Web.pdf is telling me I only get 5MHz bandwidth out of it. Apparantly they decided to incorporate an internal 5MHz LPF in it between the input pins and the ADC. A different spec sheet for the same device at a different website https://docs.rs-online.com/fb1d/0900766b815d846d.pdf says the bandwidth is 25MHz (better than 5MHz for sure, but not as good as 50MHz). What's the actual specs? Did the specs actually change? Did Analog Devices actually modify the manufacturing process for the ADALM2000, so there actually exist 2 different variants of the ADALM2000 (each with different specs), but with the exact same product name?

Also, is it possible to bypass the internal LPF (maybe soldering a wire across a component), to get the full 50MHz bandwidth that's possible with a 100MSPS sample rate?

Schematics are over here: https://wiki.analog.com/university/tools/m2k/devs/hardware

There is some minor differences between Rev A and D on the input.

I remember testing the spectrum analyzer functionality up to 50Mhz. I might be mistaken though. Could do more tests tonight if you want.

Thanks. That would be great if you could check that. Also, does the device's analog input support continuous streaming (sends all samples to the PC)? Or does it only support block transfer (sends a block of samples and then there may be some delay before the next block is sent)? I know that the official Scopy software has an osciloscope mode, that only supports block transfer mode (each block is sent in response to a trigger, or free run, but no guarenty that there will be no delay between blocks). However, I also know that the ADALM2000 has an SDK for you to design your own software, so if the hardware permits a continuous streaming mode for analog input, then I should be able to write my own software to use that streaming mode.

There's a C++/C#/Python library you can use.

https://wiki.analog.com/university/tools/m2k/libm2k/libm2k#download
https://github.com/analogdevicesinc/libm2k

Looked quickly at the API, and yes you can stream.

Is it better to use libm2k for this? or is it better to use libiio for this? Part of installation of the ADALM2000 in the first place is installation of libiio on my PC. The only problem I have with using libiio right now is knowing what folder it put libiio in when it installed it. I don't see a libiio folder in either Program Files or Program Files (x86). Of course if libm2k is the preferred method for writing software for controlling the ADALM2000, then I will use that instead of libiio.

Better to use libm2k since you have specific API for the Adalm2000. libiio is really generic and was specifically created to transport data from the device to the PC. I would not be surprise if libm2k was using libiio under the hood.

Edit: Yes libm2k use libiio.

So libm2k can be used with ADALM2000? What about Pluto2000 (an SDR also offered by the same company)? Does the Pluto2000 also work with libm2k, or do I need to use the lower level libiio with that one? Also if libm2k doesn't work with Pluto2000, is there another community created lib similar to libm2k, designed to work with Pluto2000?

[Kosmic]

Interesting, the Adalm2000 now has a Mixed Signal Mode. I'm surprise they are still updating the firmware and software after all this time. Scopy is working a lot better now. Less crash and weird stuff going on.