Magnova oscilloscope

[KungFuJosh]

now I need to delete my post 

Or we can wait until the forum decides to change the image again. 

[Tomki]

@Josh : is there a new "driver" for Testcontroller for the Magnova? Where can I find it?
Thanks
Thomas

[HKJ]

@Josh : is there a new "driver" for Testcontroller for the Magnova? Where can I find it?

The latest update is here:
https://www.eevblog.com/forum/testgear/program-that-can-log-from-many-multimeters/msg6226847/#msg6226847

[john74]

question to all Magnova users:

if you could add/change one feature in the firmware....what would it be?

For me the most annoying thing is the current behavior of the cursors.
I have to use them a lot and the fact that they are not linked to the position and voltage scaling
of the signal I am measuring is a pain in the ass.

The zoom function is a bit of a workaround, but still....why is it even implemented like this?

I'm wondering which feature you're missing.

[KungFuJosh]

I agree that the cursors and their label positions need some improvement.

One of the features I'm most waiting for is the windowing ability that they already have planned.

I think what I'd like to see the most though is the ability to switch the detailed measurement window from vertical to horizontal as a user option (but still keep the simple option they have already as well). This will go smoother once the windowing is setup for the rest of the UI.

To take the measurement window a step further, it would also be nice to individually determine which measurements are basic, and which are advanced.

For example, I might want to see full details on Vrms and CNT-Freq, but some options don't need more than the label and number. Since the Magnova can do up to 24 measurements at once, I might want to setup like 3 to 5 primary measurements with details, and then smaller boxes for additional measurements that I don't need extra details from. That way more measurements can fit on screen without needing to scroll.

ETA: I also made the suggestion already that CNT-Freq should have a permanent placement on screen like what's fairly standard on most scopes. Even if the measurement window is disabled, the frequency counter should still have a spot somewhere on the screen.

ETA2: I would also like some dedicated Undo and Redo buttons on the top or bottom toolbars.

Thanks,
Josh

[king2]

Hello!
Can I ask few questions to Magnova developers?

1. When can be expected hystogram capability for measurements?

2. When can be expected PC software or web-based interface? Access from PC to scope, copy-paste of misс info (decoded packets, measurements, and so on) can be very helpful.

3. Can you help me to understand how measurements works?
For example, I have meander signal, with gradually changing pulse width.
I want to see how and how much pulse width changing.
Sounds simple: make measurement for pulse width, zoom pulses to be one per display, then activate trending chart, expecting to see graph of each pulse
But as I understand, it is not possible, not each pulse will be counted, trend chart have dynamic horizontal zoom, averaging data, so I cannot clearly see if, for example, that each third pulse have slightly less width?

Can I do something to get such information (measurement for each pulse) from scope?
For example - we have signal that contains 60 pulses, so I want to get 60 widths measured and saved to file, shown on display as table or as graph.
Main idea is how to get full measurement info from non-repetitive signal, each part of which is important (like it done for decoded packets).
Maybe some software exists to get such information from scope (by csv files, for example), for post-analyze data collected?

Thanks in advance!

[Andre77]

For me the most annoying thing is the current behavior of the cursors.
I have to use them a lot and the fact that they are not linked to the position and voltage scaling
of the signal I am measuring is a pain in the ass.

The zoom function is a bit of a workaround, but still....why is it even implemented like this?

I have good news for you (and others who have already asked for this): With the next Magnova firmware update, the cursors will no longer be tied to screen positions but to the actual time and voltage values of the signal. As a result, when you pan or zoom the signal, the cursors will move along accordingly instead of staying fixed on the screen.

Additionally, cursor positions outside of the currently visible screen area will be supported, which should make working more convenient.

We will also apply the same improvements to the FFT cursors.

[Andre77]

1. When can be expected hystogram capability for measurements?
2. When can be expected PC software or web-based interface? Access from PC to scope, copy-paste of misс info (decoded packets, measurements, and so on) can be very helpful.

I know you asked about both features long time ago already. Since then, we have added many other features, but not these yet - sorry about that.

I can confirm that the web server functionality is now quite high on our priority list. The histogram capability for measurements is planned for a later update.

3. Can you help me to understand how measurements works?
For example, I have meander signal, with gradually changing pulse width.
I want to see how and how much pulse width changing.
Sounds simple: make measurement for pulse width, zoom pulses to be one per display, then activate trending chart, expecting to see graph of each pulse
But as I understand, it is not possible, not each pulse will be counted, trend chart have dynamic horizontal zoom, averaging data, so I cannot clearly see if, for example, that each third pulse have slightly less width?
Can I do something to get such information (measurement for each pulse) from scope?
For example - we have signal that contains 60 pulses, so I want to get 60 widths measured and saved to file, shown on display as table or as graph.
Main idea is how to get full measurement info from non-repetitive signal, each part of which is important (like it done for decoded packets).
Maybe some software exists to get such information from scope (by csv files, for example), for post-analyze data collected?

Some general points:
- The trend chart can either auto-scale or be set to fixed ranges. You can choose what fits better for your analysis.
- The Magnova has a very short trigger re-arm time. Depending on your signal, it can capture many pulses, but of course not always all of them.
- To avoid unnecessary capture time, you can turn off Extended Capture in the Acquire menu.

I recorded a short video where a sequence of pulses with changing width is captured and replayed using the History function. The measurements are shown in a trend chart, and the triangular change in pulse width is clearly visible. This may or may not fit your exact case, depending on your signal details.



Another possible method:
- Take a single shot over a longer time interval that includes all the pulses to be analyzed.
- Use the Search function to find/mark all edges
- Step through the events using the search controls while observing the measurement values in the trend chart.
- Perhaps you need to use the 'gated' measurement to specify the measurement area.

A true “measurement tracking” feature, which automatically records values across the full screen, is already planned for a later update.

As you mentioned, external analysis is also possible. You can export the raw data via USB/LAN or to a USB stick, and then analyze it on a PC (for example with CSV-based tools).

[john74]

I have good news for you (and others who have already asked for this): With the next Magnova firmware update, the cursors will no longer be tied to screen positions but to the actual time and voltage values of the signal.

Cool, sounds awesome....thanks.

With each update the scope gets better and better.
I hope that you are pleased with the sales numbers, to be able to keep the updates running.

P.S. my Keysight MSOX3000 is sold, no need for that anymore 

[john74]

QOL suggestion 1:

when trigger type is set to "decoder", add into the upper right corner
the selected "decoder group" and maybe the "event type".
(same goes for trigger type "pattern")

To be able to visualize the current config
and even have a possibility for changes with a single click.

In all other trigger types thats allready the case (see screenshots).

QOL/feature suggestion 2:

In the "search type" menu use the same button layout as in the "trigger type" menu
and maybe it is possible to add "decoder" as last missing search type item ?!?.

feature suggestion 3:

expand the "save decoder data" to all the data in the history buffer.
And maybe add an option for a compact CSV file, like the table view visualization, instead of a single byte per line)

[FloBX]

Hello John,

thank you for your valuable suggestions:

QOL suggestion 1:

when trigger type is set to "decoder", add into the upper right corner
the selected "decoder group" and maybe the "event type".
(same goes for trigger type "pattern")

To be able to visualize the current config
and even have a possibility for changes with a single click.

In all other trigger types thats allready the case (see screenshots).

Further elaborating the display of trigger type options and extending the existing “easy access” concept to Decoder and potentially Pattern trigger types makes sense.

QOL/feature suggestion 2:

In the "search type" menu use the same button layout as in the "trigger type" menu
and maybe it is possible to add "decoder" as last missing search type item ?!?.

Adding Decode as an additional option for the Search functionality is a legitimate request as well. We already plan to implement this, thereby making Search and the Trigger system complementary overall.

feature suggestion 3:

expand the "save decoder data" to all the data in the history buffer.
And maybe add an option for a compact CSV file, like the table view visualization, instead of a single byte per line)

As you know, general waveform export is already available (in various formats) for all waveforms or selected ranges. We will discuss how we can optimally translate this to decoded waveform data. Extending the advantages of the compact table view to the export of decoded data is a valid point as well.

We will add these items to our list of future improvements.

Best regards,
Florian

[john74]

As you know, general waveform export is already available (in various formats) for all waveforms or selected ranges.

thanks for the reply, and my inputs are only suggestions from my point of view using the scope.

You are right, there is an export of the waveform data...
but then I need an external tool/script to decode this waveforms back into the serial data.
It would be much easier when the scope could do this for me :-)

And most of the time (for digital signals) I am more interested in the content of the decoded data and not so much
in the timing of every databit in the stream....but as I said, from my point of view.

Thanks again.

[kasperle]

question to all Magnova users:

I'm wondering which feature you're missing.

Digital filtering on Math channels would be really great (I think it was mentioned in this thread already), would make them much more useful (diff(), I am looking at you).

And an improvement:

Just today I used the Bode Plot application and got a little bit annoyed by the fact that the cursors can only be scrolled on the actual points, not on the interpolated curve. If you e.g. search for a 3dB point without a priori knowledge, this means that you either have to set lot of points (which takes time, often without necessarily increasing precision) or you might get unlucky and the two neighboring points to the actual 3dB point might be quite far off.
Would be cool if that where to be improved somewhat, the cursor step width could be dynamically adapted to the current X-scale just like with the normal cursors (and the feature to scroll over interpolated points could be a selectable option in the Bode plot menu, leaving it optional for those who like the current behavior). The cursors being touch-enabled (via "grab-handle") would also be great.

[john74]

another question:

does anybody know how to measure the on/off ratio, when the signal has more than one puls?

I tried the "duty-cycle" measurement "between gates" but here only the very first impulse is measured.
I want to know the combined high-time and the low-time between the cursors
(see screenshot) or at least the proportion factor like 30/70.

Is there any other measurement type that I can use instead of "duty-cycle"?

[rf-messkopf]

does anybody know how to measure the on/off ratio, when the signal has more than one puls?

Not directly, but you could use the integration function of the math channels to count the time the pulse is high, and relate that to the total time. So you do the formula intg(CH1 > 0). This requires your pulse is actually negative going to be reliable.

Then you can use tracking cursors on the math channel to measure differences and relate them. For quicksave_1_screen.png, the duty cycle between the two cursors is 16µ / 85µ = 0.188. As a check a square wave with 0.5 duty cycle in quicksave_2_screen.png, we get 45µ / 85µ = 0.53 between the cursors.

[john74]

So you do the formula intg(CH1 > 0).

thanks, that worked just fine

[rf-messkopf]

Looking at this again, I noticed the following thing: On the math channel, the cursors, when set to Tracking, do not exactly track the waveform. See the attachment: The cursors are consistently 1µV below the waveform.

Even though the displayed delta would be correct in this case, I suppose this is a bug?

Edit: This seems to be related to the y-resolution being limited to 1 µV, also for the indication of the Δy-indication on the left, even tough there are three decimals displayed in the encoder section of the screen.

[john74]

The cursors are consistently 1µV below the waveform.

I noticed this behavior as well, the measurement with the cursors are not ideal at the moment.
(especially when you try to measure "big" time deltas...you lose precision)

But, I improved your idea with the math channels to convert the signal ratio directly into [%]
(well 100% = 100V :-)

M1 = intg(CH4<0)
M2 = intg(M1>0)
M3 = (M1/M2) * 100

now M3 ist the negativ Duty-Cycle in [1V = 1%]

[rf-messkopf]

But, I improved your idea with the math channels to convert the signal ratio directly into [%]
(well 100% = 100V :-)

M1 = intg(CH4<0)
M2 = intg(M1>0)
M3 = (M1/M2) * 100

now M3 ist the negativ Duty-Cycle in [1V = 1%]

Cool.   I totally forgot that you can do math on math on the Magnova.

The cursors are consistently 1µV below the waveform.

I noticed this behavior as well, the measurement with the cursors are not ideal at the moment.
(especially when you try to measure "big" time deltas...you lose precision)

Yes, the limited resolution of the cursors should be fixed. Perhaps Batronix could include this as a suggestion for improvement in a future update.

[KungFuJosh]

Yes, the limited resolution of the cursors should be fixed. Perhaps Batronix could include this as a suggestion for improvement in a future update.

They already have plans to vastly improve the cursors in the next fw update:

I have good news for you (and others who have already asked for this): With the next Magnova firmware update, the cursors will no longer be tied to screen positions but to the actual time and voltage values of the signal. As a result, when you pan or zoom the signal, the cursors will move along accordingly instead of staying fixed on the screen.

Additionally, cursor positions outside of the currently visible screen area will be supported, which should make working more convenient.

We will also apply the same improvements to the FFT cursors.

[john74]

If someone needs a starting point for a python script,
which can communicate with the BMO over TCP.

Here are two examples:
main.py (creates a static screenshot from CH-1)
main_stream.py (shows a live stream of all 4 analog channels)

you need python installed and some packages (PyQt5, pyqtgraph)
(you have to set the IP address accordingly)

and be aware, it's made with chatGpt in less than 2h, so ...buggy as hell :-)

[KungFuJosh]

The driver I made for TestController can already control most functions remotely, and also includes a screenshot button.

Thanks,
Josh

[kasperle]

Maybe stupid question: Is it somehow possible to change the vertical scale on a channel without changing the offset?

If I use the scale Encoder, the offset is scaled accordingly. Kinda annoying for low frequency signals with a DC bias...

[jusaca]

Maybe stupid question: Is it somehow possible to change the vertical scale on a channel without changing the offset?

I asked for that functionality as well.
Batronix responded that they liked the idea and would add it to the wishlist. But no promises for any date of implementation:
https://www.eevblog.com/forum/testgear/magnova-oscilloscope/msg6091473/?topicseen#msg6091473

[king2]

One more proposal - make possibility to use USB WiFi adapter to get wireless connectivity with network instead of ethernet RJ-45 socket. Just to minimize wires quantity and add more wiring options.
Maybe this implemented already?