ADALM2000 vs Analog Discovery 2 vs PicoScope 2205 (2010my)

[MrW0lf]

Gentlemen, arm your triggers!



It's going to be long hard race. Pico is old model but spec wise very similar to current 2205A and will serve as software demonstrator. First test soon.

[MrW0lf]

First things first. Analog bandwidth and general usage of FFT function. Analog Discovery 2 BNC adapter board was used on both BNC-less scopes.

Sine sweep 0-50MHz. Calibrated each scope for 0 at 100kHz. 50Ω signal gen (Siglent SDG2122X), 50Ω inline terminator.

ADALM2000: 0dBV in, -3dB@27MHz



ADALM2000: -20dBV in, -3dB@28MHz



Software is a bit primitive but did not take long to get result.

AD2: 0dBV in, -3dB@30.9MHz



AD2: -20dBV in, -3dB@>50MHz



Software is very good for the task, but huge freq. response difference depending in input level. Need to investigate it further sometime. Edit: Turns out it has 2 ranges. Look later post.

PS2205: PicoScope 5 software: 0dBV in, -2.98dB@21.17MHz (+-2V range)

Had to use PS5 because it allows sweep up to 50MHz (PS6 limited to 25MHz on given scope). GUI very outdated cannot change vertical zoom (at least did not find obvious way).



PS2205: PicoScope 6 software: 0dBV in, -3dB@18.8MHz (+-2V range)



PS2205: PicoScope 6 software: -20dBV in, -3dB@21.95MHz (+-200mV range)



Since PS2205 has actual relay switched ranges there are some differences also. Note this is obsolete scope and current one is probably more consistent, at least my PS2408B is. It took loooong time to get the GUI visual right because have to fiddle with zoom factor and offset % instead of just entering needed values eg it does not offer spectrum analyzer like control.

Winner: Analog Discovery 2, but very large difference in freq. response depending on amplitude. ADALM2000 is most consistent in that regard.

[MrW0lf]

Step response test to shed more light on response vs amplitude differences. Source is negative going square with 30ps rise, amplitude can be varied. For starters would be nice to center amplitude around zero with DC offset but some scopes do not have it. Can be still emulated by various methods and test software & features at once.

ADALM2000:

±50 mV: RT=~10ns BW=~35MHz (RT calculation broken)
±1 V: RT=11.2ns BW=31.3MHz



As can see response is indeed consistent. But many problems some little some bigger. Maybe have not discovered all the right ways to use software but still...
- seemingly no DC offset emulation, but can do with math channel
- to create reference must first save to CSV and then reload
- did not found averaging option
- no oversampling/ETS
- full +-1 sample jitter around trigger point (+-10ns @ 100MHz)
- zoom is very nasty to use with mouse, seems touch oriented, in general I do not like all those soft and fuzzy zooms,  need numerical input for precise x10 or whatever
- many little weirdness, software needs polishing for UX
- no Sinc interpolation (min timebase whopping 100ns/div)

AD2:

±50 mV: RT=8.8ns BW=39.7MHz
±1 V: RT=9.9ns BW=35.4MHz



Big difference in response indeed. GUI is again good for the task, can even create function that calculates bandwidth from RT on the fly. There is oversampling that will give 1.6GHz effective rate but from past experience I know that it can have problems related to underlying RTS trigger jitter. No issues this time however.

PS2205:

+-100mV: RT=18.8ns BW=18.6MHz
+-2V: RT=19.7ns BW=17.8MHz



Response performance worse than with sine test, maybe due to using only small portion of ADC range (no DC offset). Software setup quite easy. 4GHz ETS (equivalent time sampling) gives extremely solid relative timing accuracy for this class of scope.

Winners: AD2/PS2205 because each one has strong points. ADALM2000 clearly inferior in everything but consistent response with differing amplitudes - software clearly needs more work.

[MrW0lf]

Differences in AD2 bw measurements were suspiciously large so revisited from the perspective of ranges. Turns out AD2 and ADALM2000 both have 2 physical ranges. Many numbers are calculated/measured so take with a grain of salt.

ADALM2000 (12bit ADC)

+-2.5V (~+-3.023V), 1mV/div - 500mV/div
Resolution ~1.476mV

+-25V, (~+-31.818V) 1V/div - 10V/div
Resolution ~15.536mV

Cannot find a way to force SA to use lower range, it works only in upper one so this maybe explain why all so linear... If someone knows how to change it please let me know. Currently it's basic resolution is pretty non-existent if compare to others.

AD2 (14bit ADC)

+-2.5V (+-2.743V), 100uV/div - 500mV/div
Resolution ~0.335mV

+-25V (+-30.162V), 1V/div - 5V/div
Resolution ~3.682mV

Abs maximum ratings (full scale) can be found from Constant measurements:
Spectrum Analyzer / Measurements / Constant / FS
At FS it will openly distort, but fun with bandwidth starts much earlier. Here test for same 5Vpp under +-2.5V (max low range) and +-5V (min high range). One must be very aware of this so can pick either max bandwidth or max resolution.



PS2205 (8bit ADC)

±50mV, ±100mV, ±200mV, ±500mV, ±1V, ±2V, ±5V, ±10V, ±20V physical ranges, checked 2 of them:

±50mV (~+-58.24mV) (10mV/div)
Resolution ~0.455mV

±20V (~+-23.2V) (4V/div)
Resolution ~181.3mV

Calculated numbers vary depending on amplitude so possibly some post-correction is going on.

So turns out at low end Pico and AD2 have similar resolution. ADALM2000 fairly limited and SA functionality badly needs access to low range. AD2 bw differences are huge. Pico lacks dynamic range. No perfect scope in sight.

[jsfotografie]

thx for the nice test i went with the ADALM2000 because of the price MEH now i found that the claimed THD measurment in the Advertisment is not included in the Scopy software? https://www.analog.com/en/design-center/evaluation-hardware-and-software/evaluation-boards-kits/adalm2000.html i also didnt find the sweep generator? which version of scopy you use?

cheers and thx janosch