InfiniiVision 1000 X-Series Oscilloscope DSOX1102G - Review

Table of contents

RoadTest: InfiniiVision 1000 X-Series Oscilloscope DSOX1102G

Author: michaelwylie

Creation date:

Evaluation Type: Independent Products

Did you receive all parts the manufacturer stated would be included in the package?: True

What other parts do you consider comparable to this product?: null

What were the biggest problems encountered?: Lack of documentation. Buggy firmware?

Detailed Review:

UPDATE

 

Although it may not seem it, this RoadTest has been exhausting. It consisted of documenting, and proving to Keysight that there was a bug in the Frequency Response Analyzer (FRA) feature of the DSOX1102G. I never received a detailed explanation, but the bug was related to a time-delay in the system. The bug has been fixed in the latest firmware release and the issue I had with the configurability of the number of sample points being hard-coded has also been addressed. The number of points per decade of the analyzer can now be set between 10 and 50 in 10 point increments  I wonder if they read my review before making that decision? Although I was suppose to use the analyzer to measure the frequency response of a few filters, that never happened. Again, instead, I ended up documenting the failure of the analyzer to correctly determine the phase shift in a simple 10 kohm resistive divider. Previously I mentioned my surprise at the inability to configure the FRA, but I also mentioned that one cannot download the data from the FRA. You can save a screenshot of the response, but you can't seem to get a *.csv of the actual data. Pictures are nice, but engineers need data. To make this FRA fully functional they need to add the data download option, in my opinion. I'm sorry this review doesn't seem detailed enough; I admit it seems quite short, but the length of this review doesn't really reflect the work that went into it

 

Here are the old and new FRA results where you can clearly see the phase glitch has been resolved:

 

Note, the scale for the phase measurement has changed because the glitch doesn't make the chart autoscale the upper limit to 60 degrees.

 

Previously I gave this device a 19/60 points for the review. I will be increasing this number, but I haven't decided what the final value should be. I feel like a good effort was made to fix the problems, but ultimately if I had purchased this device I would have boxed it up and sent it back as soon as I discovered the FRA bugs and limitations. Again. this review focuses solely on the FRA ability of the scope, as I outlined below with my RoadTest application.

 

 

Original Review (Which was published before completion)

 

Let me start by saying, I mainly applied for this RoadTest because I wanted to test the Frequency Response Analyzer (FRA) of the new 1000X series scopes. My review is solely of that feature. In fact, my RoadTest application was only about 1 page long, and the focus was this single paragraph:

 

    “My idea for testing this unit is rather simple: I would like to have fun with the DSOX1102G by using its Frequency Response Analyzer to test both passive and active filters. I'm thinking some basic LR, CR, and LCR passive filters and a few active filters such as Butterworth and Chebyshev. I'd like to compare my hand measurements to the automated measurements. I have a few projects where I would utilize the oscilloscope as well, one of which is troubleshooting a continually rebooting Samsung cell phone. I'm trying to keep this application short enough such that you will see practical, achievable goals for this RoadTest. If you are concerned about the quality of the review, simply check the previous reviews I have completed from the list above.

 

THIS REVIEW FOCUSES ON THE FRA FEATURE OF THE DSOX1000 SERIES

Once I started using the FRA I soon realized a few things. Firstly, this feature of the 1000X is not only poorly documented, it’s practically undocumented. I had to email Keysight to get info and the response I got was:

 

    “First, you asked about available documentation.  There isn’t yet a lot of documentation specific to the DSOX1000’s FRA feature.  Additional application notes are coming.  For now, page 71 in the User’s Guide describes how to set the different parameters for the measurement, such as the input / output channels, amplitude, load setting, etc.”

 

So, the documentation is coming. I read the manual before even applying for the RoadTest and I just assumed I missed some documentation, but no, that single page in the manual is all the documentation available for the FRA. I’ve reproduced that single page below:

 

 

 

Fortunately, the scope itself shows you how to make the connections; that is, how to hook up the source, Channel 1, and Channel 2. A whole Scope Month video was dedicated to the FRA and it was even shown on the kick-off video. I know, because I watched all the videos. I wasn’t getting satisfactory results with my FRA, so the agent from Keysight suggested I hook up a resistive divider and measure its frequency response.

 

Measuring a Resistive Divider

I used the topology of Figure 1 for the resistive divider. I did not use a breadboard, instead I used screw terminals to connect the two resistors together as shown in Figure 2. Scope probes were set to 1x attenuation.

Figure 1: Resistive Divider Setup. R1=R2=100k.

 

Figure 2: Implementation of the Resistive Divider.

 

Figure 3 shows the response from the FRA for the circuit of Figure 1 with R1=R2=100k.

 

Figure 3: Frequency Response of the Resistive Divider. Gain is the blue curve and Phase is the Orangish curve.

 

 

As always we ask "What do we expect for a result?". I expect the gain to be a flat - 3 dB - 6 dB (Voltage, not power), until at a higher frequency where some parasitic effects may cause some deviation. I have no idea how the gain curve ever got higher than 0dB in this test ... it's a passive circuit. Thank you for pointing out I had reversed my input and output leads. Of Course! The probes may be influencing the result; I've contacted Keysight support. Let's assume the gain plot is my fault, and look into the phase response. I expect a flat line at 0 degrees for the phase until at higher frequencies where parasitics cause a deviation. The phase has two very wrong points.

 

I took a video of the analyzer running and grabbed the waveforms from the two points of interest. Figure 4 and 5 show the waveforms for frequencies of 158 kHz and 400 kHz. Yellow is the input and green is the output. The measurements taken at 158 kHz and 400 kHz are clearly wrong. For 158 kHz, the reported phase shift is 80°, but it is clearly closer to 30°. At 400 kHz, the reported phase shift is -5.34°, but it is clearly closer to 45°.

 

Figure 4: FRA waveforms for f=158 kHz. Reported phase is 80°. Visual inspection shows the phase is closer to 30°.

 

Figure 5: FRA waveforms for f=398 kHz. Reported phase is -5.34°. Visual inspection shows the phase is closer to 45°.

 

So, there's a bug or glitch occurring, which isn't a big deal IF they fix it. After seeing the above plots I thought I could go back and reduce the effect of the outliers by adjusting the frequency steps of the test. This was the worst part of the RoadTest because YOU CAN'T ADJUST ANYTHING EXCEPT THE AMPLITUDE FOR THIS TEST!!! Are you kidding me!?!?! After talking with Keysight, I found out that YOU HAVE TO PURCHASE THE NEXT MODEL UP to get the ability to change the frequency steps. Once again, are you kidding me? Don't made a big deal about the frequency analyzer if it doesn't have basic functionality like changing the frequency steps. Call it a script or something else, because it sure isn't an Analyzer. I am so disappointed, because I was really looking forward to this feature. To find out it's a crippled script being called an analyzer that you can only really use in a more expensive scope just outraged me, again, because Keysight made such a big deal about the feature. Note, again. The frequency step issue has been resolved with a firmware update.

Anonymous
  • Great work identifying and documenting the FRA bug. I recently got the same scope, and I was having similar issues with incorrect phase readings. When I first got the scope, I looked at the firmware version, saw that it matched the one online (1.10), and assumed that it was up-to-date. After reading this review, I investigated further and noticed that the build on my scope was version 1.10, but with a release date from 2016. Downloading and installing the latest (3/2018) build of the version 1.10 firmware from Keysight's website fixed the issue.

  • Nice work, and well done to identify a bug.

     

    I'd hope that Keysight recognise your efforts and send you something else to check ...

     

    Cheers

    Mark

  • It's good to hear the Keysight were responsive and came back with fixes for the issues you discovered with the FRA. That is encouraging as a lot of companies would leave you high and dry with no solution other than return for a refund. I agree, we engineers need data, if you can't extract the results in any way other than a picture that's a shame and seems a bit of an oversight. Maybe they'll fix that once they read your updated review....

     

    So, I thought I would have a look at how my 3000X compares on this front as I have the DSOX3PWR bundle which I believed should include this frequency analysis functionality.... So I search the menus in the Power module but cant find the one labelled "Control Loop Response (Bode)" anywhere..... A small amount of googling later and downloading a large application note, it looks like this is available in the 3000T, 4000X and 6000X but it would appear not in my 3000X! So it would seem that they can add this to the 1000X but don't make it available at all in the ever so more expensive 3000X series, I'd have to trade in for a 3000T to get it!

     

    There's a Keysight guy who hangs out on the EEVBlog forum so I might have to go on there and have a whinge to him!

  • Yes, I've received the new firmware and it does fix the bug in the Frequency Response Analyzer. This has been an exhausting RoadTest - I will be updating the review in the next few days.

  • After talking with Keysight, I found out that YOU HAVE TO PURCHASE THE NEXT MODEL UP to get the ability to change the frequency steps.

    There has been a firmware patch released that allows you to change the frequency steps on the DSOX1000 series. It allows users to set the FRA points-per-decade to 10, 20, 30, 40, or 50 instead of just 10.

    http://bit.ly/Patch1000XFRA

     

    Details as posted by Daniel from Keysight over on EEVBlog.

    http://www.eevblog.com/forum/testgear/keysight-1000-x-series-oscilloscope-fra-patch-coming/msg1223184/#msg1223184

  • Michael,

     

    Very interesting accidental post.  I haven't tried using the FRA feature yet as I have been concentrating on bus decode and triggering investigations.

    I am curious now.  I will find time to reproduce your experiment and add the results to my review.

     

    Mark A

  • Ah yes, I noticed this while testing last night as well. The gain should be -6 dB. I'm going to check my pictures and videos to see if the probes are swapped.

  • "I expect the gain to be a flat -3 dB, until at a higher frequency where some parasitic effects may cause some deviation. I have no idea how the gain curve ever got higher than 0dB in this test ... it's a passive circuit."

    I'd expect -6dB. -3dB would be a voltage gain of 0.707. Are you used to working with power measurements?

     

    The curves in that first plot are upside down (as Cosmin says, the probe capacitance on the output will dominate and take the gain and phase curves down rather than up), so the most obvious explanation would be that the probes were swapped over (turning the expected low-frequency 6dB attenuation into a 6dB gain and sending the phase off in the wrong direction higher up).

     

    The instrument does seem to have some real problems measuring the phase, though.

  • I greatly appreciate your honesty.

     

    I completely agree with your observation on needing the next model up to get the useful features.

    If the device said it could support your FRA needs and then does not, then we call that "bait and switch!"

     

    I look forward to seeing your testing within the limits of the system as delivered.

     

    You might use shorter test leads.  Once you get into the higher RF frequencies, everything looks like an antenna and there are very interesting effects.

     

    We had to deal with these issue in the old days when today's digital frequencies were considered extreme RF.  Yes it dates me, but we learned a lot about analog circuit response to parasitic and harmonic creation.

     

    DAB