Would you please advice me on choosing usb scope?

[z]

I came to Melbourne from Turkey 6 months ago and could not bring my oscilloscope or any other instruments with me.
I want to buy new scope. I am undecided on which scope to buy.

Would you please advice me on choosing between the following scopes?

First one;
Picoscope 3405D https://www.picotech.com/oscilloscope/3000/picoscope-3000-oscilloscope-specifications

The other;
Keysight u2701A http://literature.cdn.keysight.com/litweb/pdf/5990-9537EN.pdf?id=2150237

I used to have an analog scopes.  In past years, I bought digital oscilloscope but later undestood the I made wrong choice.

Could you please advice on what kind of digital oscilloscope would be the better choice?

Yes 3405D four channel. But this is not very importand for me.

[Eric_S]

What are you going to use it for, what will you be measuring?

[z]

I am interested by analog, digital and power electronics.
The 40Khz pwm signal-induced power circuits are very important to me.
However, there are situations where I have to see a very detailed square wave or sine wave 20Mhz signals.

[Andreas]

Hello,

if power = mains voltage you should consider that you need a special probe (not the standard 10:1) on both scopes to measure voltages out of the +/-200V range.
Both scopes have +/-20V as maximum (1:1) input range.

With best regards

Andreas

[MrW0lf]

Pico 4444 is interesting differential scope for power stuff (12/14bit), but 20MHz is max and even that not with all probes. 3405D seems reasonable, but only 8bit. However 3000 is only Pico where can have 4ch analog + 16ch MSO. For MSO part I have come up with a way to generate up to 16 bit extra analog channel "tunneling" it thru digitals, or more lower bit ones:
https://www.eevblog.com/forum/testgear/what-you-do-with-math-channels-on-oscilloscope/msg1130997/#msg1130997
Speed will limited by MSO part sampling rate and how good external hardware is (I used Arduinos).
If choose between 3405D and 3404D MSO then latter will be cheaper but probably display -3dB at approx 100MHz if anything like other Picos. On the other hand... 3405D probably 130...140MHz.

Edit: Noticed big difference in timebase accuracy: 3*0[5|6]D have 25x better timebase accuracy +-50ppm vs +-2ppm.

As for U2701A, I looked one time at their software at it seems more geared towards using with (large) array of other instrumentation in modular system, than standalone scope application. Also it seems to lack: ETS, signal gen, protocol decoders. Also no mention of wfm/s rate. Pico is max 100kwfm/s in "persistence" mode, 1.6Mwfm/s in 10kwfms burst in segmented mode. FFT is 1.2kpts on Keysight vs n*2Mpts on Pico.

For other info heres my still uncomplete review of 2000:
https://www.eevblog.com/forum/testgear/picoscope-2000/

Seems that 2000 is out of luck regarding new analytical feature called "Deep measure", described here in context of automotive scopes:
https://www.picoauto.com/support/topic17711.html
Otherwise most should be similar.

BTW for anything high energy I also use differential probe. For low energy worked with 100x on HV but must keep it DC coupled at all times when DUT voltage > scope tolerance (+/-20V)! AC coupling capacitors charge to full average DC value at the tip of the *x probe without attenuation.
In automotive world they use these also:
https://www.picoauto.com/products/test-leads/101-attenuator
But "This attenuator must not be used to measure mains voltages, secondary ignition or other hazardous voltages.". So active diff probe still seems best idea...

[z]

USB3.0 is more important feature for oscilloscopes?

[MrW0lf]

USB3.0 is more important feature for oscilloscopes?

Pico has 2 main ack modes. Default I call "analytical". There it drags full record over USB to PC. They have done some optimizations but this is still heavily subject to USB bw. So if have large memory of 256M and actually use it often this may become important.
If using SDK 31MSa/s (2.0) vs 125MSa/s (3.0) streaming.
Second mode is high wfm/s "persistence" where stuff seems pre-rendered on scope and USB should have relatively less effect. Otherwise this mode has limited analytical functionality and I do not use it often.