Saleae continues to get better. Please keep up the good work!
Still no method to save/load configuration and default is failing.
Purpose: Different fpga configs might have different signal assignments to pins, so would be nice to load a config with all windows in the correct order, size, and correct names to avoid having to go somewhere else to find the right settings and then type in the config again and drag channels around.
Even worse, I have two Saleae’s attach to the same system and they share the same config, sometimes. Port 0 on one does not have the same definition as port 0 on the other!
When I set up the pro16 on one config then quit, when I next started up, I had the default config of digital+analog ports 0+1 with default names.
(It loaded the pro8 by default, but the pro16 didn’t have the correct config either when I switched to it)
Revenge of uncorrelated digital and analog waveforms
I have one of the ports (line0) with an analog and digital waveform.
When I am zoomed out, so that I can see ~1s across the screen, the analog and digital waveforms start at different types, with the analog approximately 50ms later than the digital.
As I zoom in, the relative time between the start/stop shrinks. The on-screen distance also shrinks. At 1.2ms on the whole screen, the difference between the two is 10ųs.
Continuing to zoom in, at 150ųs across the screen, the gap is 200 ns, maybe.
All weird, but it gets weirder. When I bring up the measuring tool the analog waveforms snaps to the correct place (e.g. aligned with the trigger). When I dismiss the measuring tool, it dealigns.
Actually, any time the side panel is open (e.g. analyzer, config), it aligns. With the side panel closed it dealigns. Actually at different zoom levels it doesn’t necessarily align. It always moves, but doesn’t always align.
If it matters, Pro8 with 0/4 on analog at 50 MS/s and 0-7 at 125 MS/s. Manchester decoding on ports 0 and 4. Trigger on port 0 (rising edge).
This problem persisted after a program restart.
Signal analyzer display fail
I have Manchester decoding set up. When I look at the digital waveform, it properly decodes the values. However, when I look at it on the analyzer, I don’t see any values underneath DATA (value/start/duration).
Config: manchester port 0, 1000000, negative in 1, 17 bits, MSB 1st, 0 ignore, 25% tolerance, show in protocol result table, NO stream to terminal
Same config for port 4 (which was not seeing any traffic this test).
I tried clicking on data table–no change. I tried clicking on pro terminal, nothing.
I enabled stream to terminal, then click on pro terminal, and got some values that look vaguely correct; but still nothing in the data table.
I would love to see #1 as well. It sounds like a good idea. Have you added it as a feature request in their system?
I can confirm #2. I thought it was just me but I’ve seen this exact thing. I captured an analog signal plus a digital signal made from it via comparators. The digital signal LEADS the analog signal in Logic even though obviously the digital signal is in reality lagging slightly since it goes through conditioning and the comparator. This is kind of spooky looking but I can overlook it and tell myself not to read into the lack of time sync between the two.
Still no method to save/load configuration and default is failing
True, it’s high on our todo list! (and on the feature requests board)
I believe that we’ll add that in a few weeks. Sorry about that…
Revenge of uncorrelated digital and analog waveforms
@markgarrison is looking into this. We’ll keep you posted.
Signal analyzer display fail
The data table currently supports I2C, SPI, Async Serial, as a temporary state while we’re upgrading its infrastructure. A new version with support for all analyzers will be released next week. if you want to share with us a capture with Manchester data, we’ll be able to verify that it works beforehand.
@sethnobaka and @Collin, I’m struggling to reproduce this, could either of you send a saved capture where you saw this?
@sethnobaka I’d like to make sure I’m reproducing your scenario as closely as possible. when you said 0/4 channels, did you mean channels 0 and 4, not 0-4, right?
About how much time do you think passes between when you press start and when the trigger is found?
What value do you have set for “memory buffer size”?
What is your setting for “capture duration after trigger”?
Do you have “trim pre-trigger data to last X s” enabled? if so, how many seconds?
Do you have the glitch filter enabled?
Also, try saving the capture and reloading it. Does the problem persist? If not, please send us a video of the problem.
Also, please send us your machine ID, so we can look up error reports from your system. The front-end of the application will report even non-crashing unhandled errors, and it’s possible an unhandled error is causing this problem. The machine ID can be found by opening the device switcher from the main menu and clicking the “?” button on the lower left corner to copy to clipboard.
Last question, what operating system are you using?
@Collin, if you can send in any of the above information, that would be very helpful too.
Another fantastic release - you guys just keep on making this software better and better. Absolutely loving it.
I REALLY like the new big fat yellow trigger marking! Fantastic.
Any chance that you could flag every instance where the trigger is true, rather than just the first one? I’d love to be able to record 100s of data, and have lots of yellow trigger markers show up every place the trigger settings were true. (Yes, I’m sure I could do this with a custom protocol analyzer, but this seems like such a simple ask. . . )
We’re planning to add continuous triggering support this quarter (probably in a month). It will allow you to jump back and forth between triggers, view triggers in real-time, and more.
@markgarrison: What is the best way of sharing data? It is a 52MB file. I could email it or upload it somewhere.
The data capture has both the example of the analog/digital mismatch and has some Manchester encoding examples.
Yes, I had the analog data captures running on channels 0 and 4. As seen the example, it was a few seconds between recording start and data start and my machine/environment isn’t the fastest.I have the default capture settings, more or less, so 1second capture after, no trim, no filter, 3GB Memory.
This machine id information is unlikely to be very interesting. The machine normally is not connected to the internet directly. I hope the calibration information is cached! This is a Fedora 25 system. To be upgraded very soon, I hope.
I’ve just PM’ed you a dropbox file request link you can use to send me the capture.
I hope the calibration information is cached!
We’ve added features to improve managing calibration for offline users or folks behind company proxies that block the calibration download. It will be much easier to view the calibration status, and the software will give you a link for you to download the calibration file manually it can’t download it for you, and then let you load the file through the UI.
