LarsB

Thanks, Attila, copying a part of the capture to a new tab is really helpful.

Duh, just noticed the Cursor view, which answers my first question.

Thanks for the explanation! I figured you really need to set a trigger to capture what you want.

Is it possible to mark a point in time in the plot, e.g., by a (colored) vertical line? There are labels, but they don't move with the capture, so I'm not even sure what they're for. Alternatively, is it possible in the Event View (or anywhere else) to search for something after a given point in time, e.g., the previous search, or the current position in the capture on screen, or a numeric time value? I'm having a hard time analyzing the capture, and manually copying and comparing times such as "0.00211153" is quite cumbersome.

What message do I have to wait for? If I wait for the box turn green and the message "Recording", then wait 1 seconds before I press stop, I only have captured a few ms. Do I have to wait for "Recording Trig'd ...", even though I don't have a trigger set? Or is there a limit to the sample rate when recording that WaveForms isn't aware of?

When I try to record (16 bits, 128 Mi samples at 50 MHz), WaveForms doesn't capture anything if I start the recording and then press stop after a second. Why is there a delay before recording starts? WaveForms states it's "Recording Prefill"?! Can I avoid that? Trigger is set to None.

I see. As I said, not a big deal. :-)

Forgot to mention: It's version 3.18.1 on Linux.

Great, I didn't know that! In my version, it says "swap event navigation", but it works. Very minor nitpick: It also reverses the arrow symbol, which is still confusing, so maybe keep the arrow's direction while swapping navigation? Or maybe it's already like that in your version?

I have three signals in Logic: If I drag TRIG up below A0 then the dragged signal loses its name: Or is there a different way to reorder signals?

Oh yes, I didn't notice that all data is shown if I leave the entry field blank. Perfect! And the bidirectional navigation will be really helpful, thanks!

This is just a minor issue, and maybe my expectations are wrong, but: When I click on one of those little arrows at the side of the signal grid, e.g., I get the previous data segment, i.e., the signals scroll to the right (might makes the signal segment at the right edge disappear), which is consistent with the right-facing arrow. But my expectation is that clicking this arrow on the right side of the grid would show the data that follows after the edge, i.e., more. In other words, the grid should scroll left. Please compare this behavior with virtually any image gallery on the web, and you'll find that clicking the right arrow shows the next image, not the previous. Compared to this de-facto standard, Logic has it backwards. Am I the only one who thinks this way?

With WaveForms Logic, it seems that I have to set the number of samples upfront. Is there a technical reason for this? Is it possible to start and stop the measurement with a click each instead? Also, why does the sample rate change when zooming in and out of the signal graph? Shouldn't it be fixed after the measurement?

I'm using the logic analyzer of WaveForms to measure two buses. The buses are clocked by the same signal. WaveShows shows the bus value at each clock edge, like it should: Sometimes, however, I find it easier to work with a list, like in the Data view. Unfortunately, the Data view lists every measurement: (Also, while clicking on a row in the Data view positions the Signal Grid, why is the other way around not supported?) Is there a way to get a view of bus values only? Or in general a list of a protocol values?

Oh, I just noticed that increasing the sample rate by a lot (that UI again ...) makes those slopes appear upright when zoomed appropriately. I guess that is what I'm seeing in the help sample diagram. I guess it's OK then in practice.

I'm using the Logic Analyzer tool in WaveForms. Is there a way to draw straight lines instead of those slanted signal lines? Even the in-program help uses straight (for lack of a better name) lines: I couldn't find an option in WaveForms.

Oh, finally! At last this is looking as it should be. I'll try with all 25 signals later, but at least I'm confident now. Thanks for your suggestions, Attila.

Yes, I connected one group wire of the Discovery to ground. I didn't even know that you need to configure Supplies, but here is the current setting: Voltage is correct, threshold is fine, pull-down. I wasn't sure about the strength of the pull-down, so I retried without it, but I still get only 0. I'll try the beta, though.

To further investigate, I tried to capture only the clock signal by both Discovery and Saleae at the same time. For that, I soldered an extra wire to the signal pin, and connected Discovery to the pin and Saleae to the wire. But while Saleae did record activity, I only got 0s in WaveForms: Here's the workspace with capture: newtry.dwf3work I even tried both DIN 0 and DIN 1 in case a wire was bad, but I got the same result. Also note that the number of samples captured indicate a recording time of about 0.5 seconds, which is shorter that the actual recording time of 1-2 seconds. The data the Saleae captured: saleae.csv I'm not sure what to make of this. To me, it seems that my Digital Discovery is broken. Can you suggest anything else to try? Is a crucial setting wrong?

Here is the decoded bus data for Saleae, i.e., the hex values that should be displayed in the event viewer for that bus with clock D15: saleae-bus-data.txt

Obviously, you have to add attachments to the post, so let me try again. default.dwf3logic Oh and now I'm no longer allowed to upload a 9 MB file. Please get it here: http://endlos.net/saleae-csv.zip

With my new Digital Discovery logic analyzer, I captured the bus of an electronic device (see attached project). There are two potential clocks: D15 is the clock for the functionality I want to examine, and D16 is the general READ ACTIVE clock. For some reason, the saved project shows D16, but doesn't allow D16 as clock for the bus. In the Event viewer, the bus with clock D15 shows seemingly random data, although almost all data should be in the h4000 to h5FFF range. Also, there are awfully few entries. Likewise, D14 shows very little activity. When I capture the exact same operation with my old Saleae, I get plenty of activity in the above range, with the equivalent clock of D15 (see attached CSV file). Sample rate of the Saleae was 16 MS/s. Why the discrepancy? I triple checked that the connections between DD and probes are correct, on two different days. Also the probes are correct, and in any case, the DD capture and the Saleae capture was done with the same probes, within one hours. Right now I seems that either the DD or WaveForms is broken. If needed, I can happily provide additional data.

Thank you for your answers, Attila, that clears up all of my questions. The only issue remaining is the TRIG signal one. TRIG is HIGH for at least 150 ns and LOW for at least 200 ns, so at 20 MHz or even 100 MHz it should be detected. But I'll do further testing before reporting back.

My first capture also isn't quite what I expected. Look at the TRIG signal in that screenshot: I zoomed into every blip in TRIG, but didn't find more than two HIGHs each. Also, the number of events is only 184. Compare that with the Saleae capture, showing the same signal (for a new run): You can see that there is a lot of activity going on for slightly more than 3 us. Why is WaveForms not showing that? Do I have to configure something special for DIOs >= 24? Despite that capture problem, WaveForms really has some annoyances. If you click on one of the signals to the left, the entire line gets highlighted, but it's impossible to get rid of the highlighting -- at least one line will always be highlighted, distracting from the capture for that signal. While you can save your project, why is WaveForms not automatically loading the most recent project when starting WaveForms? The zoom in effect is not that great. Whereas Saleae immediately shows detailed signals when zooming, WaveForms blurs the picture while zooming, making searching for signals much more cumbersome. And I still don't understand why scrolling and zooming affects both Samples and Rate settings.

I've just noticed "list of values on the bus" is covered by the Events view.