Capturing serial output with Rigol 1054Z - seeking some guidance

[Grigs]

I have some experience with analog scopes, but I have a new digital scope that I don't fully understand all the features of. (Rigol DS1054z)

I recently wanted to capture some serial output from my PC to a radio I had. The communication between the two wasn't working, and I decided I wanted to see both RX and TX during the attempted handshake. So, I managed to get a probe on both RX and TX, channels 1 & 4, set the vertical to 5 volts, the horizontal to 5 ms. The pulse of data was rather quick. I could see it flash across the screen, but that was about it.

In short, my question is, how do I better capture this so I can view it?

I did manage to record it, I kind of got lucky. I just happened to push record at the right time. My recording duration was rather low (5 periods of 5 ms I think?) I am not entirely sure what determines my recording length, I believe it's my sampling depth?

I was looking at storage options for waveforms and traces. I saved some files for both, but I couldn't view them and I was not entirely sure storage for either was what I wanted to do here.

Anyways, back to the original question, what is the best way to view this output, which happens very fast? Below is a screen shot of what I managed to capture.

[aveekbh]

In short, my question is, how do I better capture this so I can view it?
...
I did manage to record it, I kind of got lucky. I just happened to push record at the right time. My recording duration was rather low (5 periods of 5 ms I think?) I am not entirely sure what determines my recording length, I believe it's my sampling depth?

You need to set up the triggering correctly. I see that you have pattern trigger on channel 2, which is not active. Try setting up a level trigger on the falling edge of channel 1.

Another technique I use with unknown signals is roll mode (use the horizontal menu to select it). You can see the waveforms as they come in (no triggering). Then you can stop the scope and zoom in to see more detail and set up a suitable trigger.

The record duration is simply time/division (5 ms) x 12 (number of horizontal divisions) = 60 ms.

I was looking at storage options for waveforms and traces. I saved some files for both, but I couldn't view them and I was not entirely sure storage for either was what I wanted to do here.

CSV is the most flexible way to save the trace data. You can also save the setup information along with the CSV, it goes into a text file for handy reference.

[Grigs]

You were right, I did need to setup the triggering more correctly. My eyes are still getting adjusted to everything the output is telling me, I didn't even notice I was triggering on channel 2. I managed to capture it with how you suggested, I used the roll mode to kind of see what was going on, and then set a level trigger on channel 1 for the falling edge. One thing I noticed too, was that I had the sweep on auto. Changing it to normal seemed to also help.

Roll mode:


Adjusted trigger:


Thanks for the info on recording duration, that makes more sense.

I am going to play with the storage options some more. I assume there are a number of applications that would accept a CSV file for viewing this data. Is there a recommended free utility that works well?

Thanks again.

[leppie]

Both channel 1 and 4 is the same. Either there is a short or you connected both probes to the same pin.

[AG6QR]

For capturing single events, turn the sweep to "single" instead of "normal" or "auto".  Single will start recording when a trigger happens, and then it will stop, and disable further triggering.  You'll have to push the "run/stop" button if you want to try to capture another try.  But the nice thing about single sweep is, your captured sample is safely frozen unless you DO push run/stop again.

Be sure to turn on long memory if you want to capture the longest time period with the most detail.

You can adjust sample rate to trade off the recording time for detail.  I've only got an old DS1052E, so I'm not sure if the details will be the same for you, but on mine, if you go into the horizontal menu, you can see the sample rate as you turn the horizontal scale knob.  You want to capture at least a few samples for each bit.  Note that the scope captures samples with much more horizontal resolution than it displays on the screen during recording.  You can set the scale to get the desired sample rate, then capture the sample, and you may not be able to see much detail.  But after the sample is captured, you can use the horizontal scale and position knobs to zoom in and pan around and see more detail.

You can save a .wfm file to the scope's internal storage, or to a USB drive.  There are various programs to view .wfm files, though the scope itself is often good enough.

Regarding the input and output showing the same thing, that's probably a side effect of the way many radios communicate with a computer for uploading/downloading frequency lists.  They use a single line for both sending and receiving.  It's got weak pull-ups on each end, and each end has a strong pull-down, along with a buffer that's connected to a UART.  If you're looking at the RS-232 side of things, you've got a separate lines for TXD and RXD.  Everything set on TXD will show up on RXD, but things received from the other end will only show up on RXD.  Those interested in more details can Google for "Icom CI-V interface" (a similar idea is used by many radios, not just Icom).

[aveekbh]

Be sure to turn on long memory if you want to capture the longest time period with the most detail.

You'll find this in Acquire > Mem. Depth on the 1054. Default is auto, but often it's quite shallow, so I prefer to set it manually.

Aside - I don't know why auto doesn't just use the maximum memory possible, like on Agilent Keysight. Perhaps this would come at the cost of waveform update rate.

[AG6QR]

Aside - I don't know why auto doesn't just use the maximum memory possible, like on Agilent Keysight. Perhaps this would come at the cost of waveform update rate.

I'm not sure either, but I also suspect waveform update rate has something to do with it.  However, in "single" sweep mode, waveform update rate isn't much of an issue.  At least when the sweep is set to single, it seems to me that the default should be to have long memory on.  At least I've never come up with a reason to leave long memory off in single sweep mode.

I also wish the scope would display the total recording time in a more straightforward way, rather than showing the sample rate and making you do the math to convert that into a recording time, based on the memory depth.  Sometimes I know the duration of what I want to capture, and I just want to do so with the maximum possible resolution.

[Grigs]

Regarding the input and output showing the same thing, that's probably a side effect of the way many radios communicate with a computer for uploading/downloading frequency lists.  They use a single line for both sending and receiving.  It's got weak pull-ups on each end, and each end has a strong pull-down, along with a buffer that's connected to a UART.  If you're looking at the RS-232 side of things, you've got a separate lines for TXD and RXD.  Everything set on TXD will show up on RXD, but things received from the other end will only show up on RXD.  Those interested in more details can Google for "Icom CI-V interface" (a similar idea is used by many radios, not just Icom).

You hit the nail on the head here. I am trying to troubleshoot my IC-7000, which appears to just have stopped communicating. It worked fine for a couple years, and for whatever reason has stopped. I see an attempted negotiation by the PC, but nothing RX from the transceiver. I suppose I could just have the software misconfigured, but at this point I've tried a couple PCs, and a couple of cables. I'm still poking at it. If there is any interest, I can post some photos of what I am probing, what I've tried, etc.

Thanks to both of you on the information. AG6QR, many thanks, that clarified a lot for me on general usage that I wasn't aware of. A huge help, both of you. -K1TCP