Oscilloscope Bode Plot Speed Comparison

[2N3055]

Had a bit of time.

Keysight MSOX3104T.

200Hz - 10MHz
101 points per decade (boils down to 478 points).
ONLY ONE (1)  channel analysis (Siglent can have 3 simultaneously).
4 minutes, 34 seconds

So in order to measure 3 band active X-over 13 minutes, 30 seconds roughly with probes reconnecting overhead between 3 runs...
It shows how everything is relative, depending what you do...

Some comments: Keysight uses some different method. With Siglent's and Hexamer FRA approach measuring at lower frequencies takes more time at lower frequencies and gets faster as it goes to higher ones. With MSOX3104T, it is almost same speed all the time... As if almost like it uses standard phase/amplitude measurements on just few periods...

[nctnico]

More likely the algorithm is more optimised. You can pre-filter / mix a signal to decimate it to a dataset with a fixed length and thus make the time needed for analysis constant.

[Martin72]

All this test gear but no stopwatch or enough disk space for a screenshot?

You say that...

[2N3055]

More likely the algorithm is more optimised. You can pre-filter / mix a signal to decimate it to a dataset with a fixed length and thus make the time needed for analysis constant.

Yeah I get what you're saying, but it shows time domain data, and it is always a screen with exactly 4 periods shown. At every frequency..
Could be just display..

[Martin72]

Hi,

Today the tests with 2504X-HD and 1104X-E.
Same settings as before because we want to compare apples with apples.
To my surprise the bodeplot on the 1104X-E allows three outputs just like the 2000 series.
So I test the 2014X-plus, 2504X-HD and 1104X-E.
And the "winner" is...The little 1104X-E. 
Nearly the half of time for all situations and :
Bode stopping and Bode ending is executed immediately, unlike the large models.
Here the times, later a little table:
SDS2504X-HD:
1. 6min54sec
2. 8min33sec
3. 10min11sec
SDS1104X-E:
1. 3min39sec
2. 4min43sec.
3. 5min47sec

Would like to see other results, ideally with the demoboard or similar filter*
And with R&S RTB, SDS5000X, SDS6000A or similar model..

Martin

*) will examine the board, then present the circuit (it´s passive)

[KungFuJosh]

Today the tests with 2504X-HD and 1104X-E.

Did you use the same AWG for all of the tests, or did the models with internal AWGs get the internal AWGs used?

[Martin72]

Hi,

As described in the first post, I also use an external generator for the 2000 models because the 1104 does not have an internal one.
At work, with the 2104X+, it was a SDG1062X, here at home a SDG2122X.
That should not play a role for the speed.
And as I'm writing this, I could have done without the external ones on the 2000s, because their internal generators should also run via usb (internal).

[Martin72]

will examine the board, then present the circuit (it´s passive)

Under the (wonderful) andonstar microscope you can see three components for the filter, a coil, cap, resistor.
Measuring them with my mastech LCR tweezer, the Resistor is 470ohm, the coil 118µH (I guess 100µH)a and the cap varies between 1µF and 1nF depending on the testfrequency.
Not to forget the jumpers on the board for termination:
50 ohm to ground, 75 ohm terminal in to the filter.
Because of the "unknown" cap I run a simulation, 100nF should be the right value..
The simulation is close enough to say these are the values you could take for the filter ( Lowpass: Signal after the coil, bandpass: signal after the cap.) when you don´t have the demoboard(99.99% of all contributors here I guess.. )

Martin

[M Harris]

Keysight MXR608A, about 1.2s per point.

[Performa01]

Keysight MXR608A, about 1.2s per point.

So for ~500 points its a total of 10 minutes.

[mawyatt]

Not to confound this thread, which is about DSO Bode Plot relative speed with a given DUT,  but to update our issues with the Analog Discovery 2 we mentioned earlier. The site for the operational software for the AD 2 was fixed, and we were able to download such.

We were able to get the AD 2 to perform with the mentioned 10.5KHz Peltz Oscillator under Injection Locking conditions. To get meaningful displays required narrow bandwidth, multiple waveform cycles and averaging, all settable parameters like the PicoScope FRA utilizes, and also slow as expected. Once these parameters were honed in the AD 2 performed admirably, actually better than expected

Keep in mind we are asking these measurements to be performed on a difficult DUT. Where the large "noise" or "interference" is considered, as this "noise" will get very close to the measurement frequency, then actually becomes the measurement frequency under locked conditions, then releases after lock is lost and the measurement frequency moves away when sweeping upwards.

Would be nice if the knowledgable folks at Siglent (and others) would offer some user selectable parameters which have a direct influence on the measurements at hand, such as those offered with the PicoScope FRA and the Analog Discovery 2 

Best,

[teddychn]

Here is the comparison of 61-points Bode test between AD2, SDS2074X Plus and SDS2104X HD. The generator signal amplitude was fixed at 500mV.

AD2 : 17.6 Sec
SDS2074X Plus :  ~ 177 Sec
SDS2104X HD :   ~ 177 Sec

[tautech]

Here is the comparison of Bode test between AD2, SDS2074X Plus and SDS2104X HD. The generator signal amplitude was 500mV.

AD2 : 17.6 Sec
SDS2074X Plus :  ~ 177 Sec
SDS2104X HD :   ~ 177 Sec

Use of the blue Print/Save button for screenshots on the front panel preserves the menu in use showing your settings. 

[techneut]

And here is the SDS5034X:
1 9.05,29
2 11.03.93
3 Timing error, will do it again

[teddychn]

Sorry, I haven't noticed the testing criteria and posted my previous test data for the SMPS loop gain test.

Here is the test result of AD2 with a very similar configuration:

Bandbass : 21.5 Sec
Bypass :    21.1 Sec