EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: tokmag on January 08, 2017, 02:10:36 pm
-
Hi,
I've recently bought my first oscilloscope, a Rigol MSO1074z-s so I am a very beginner in this field. I am making my first steps learning to use and discover its features. I've encountered with a strange behavior of the logic analyzer of the scope and I would like to ask for some advice or tips on it.
I was just wanted to try out the serial decoding so I've set up a microcontroller to send out the "Test" string over and over, with some delay, on a SPI port. I've connected the logic analyzer's test leads to the SPI lines and set up a decoder for those signals. For some reason on the MOSI line instead of correct bit signal form I had only some bursts and the serial decoder was unable to decode it. Firstly I thought that the problem is in my program but after verifying everything, I did not find any issue with my code. I tried to connect an analog test lead to the same MOSI line and there the bit pulses were appearing correctly and using this line as MOSI, the serial decoder was successfully decoding my data.
On the attached screenshot (LA_bursts.png) this issue is visible. On the upper part of the screen the bursts are visible on D0 input, labeled as MOSI. The CH1 lead is connected to the same point on my circuit, this is visible on the yellow waveform. I've set up two decoders. The upper one uses the D0 as MOSI input, the lower one uses the CH1 as MOSI.
Do you have any idea why this is happening? What I am doing wrong?
Thank you in advance!
PS1: At some point this started to work (LA_correct.png), but after turning the scope off and on after a short brake, the problem came back.
PS2: What I've tried:
- move the cables, clips, every connector to exclude loose connection.
- change the time base which automatically changes the sampling rate of decoder. On this scope I cannot change this directly. Any way, I think this problem is not coming from decoder.
- change the baud rate of the SPI from my code.
- change the trigger from MISO line to CS line.
- moved the MISO line to other input of logic analyzer. This actually solved the problem. This moves me toward the conclusion that my D0 line is not working all the time. It is strange because the other day I was able to do some decoding on RS232 using the same D0 line and worked perfectly for hours without a glitch.
By the way my firmware version is 00.04.03.SP2. I don't know yet if there is a newer version.
-
Hi tokmag,
I don't have an MSO1074Z but maybe you set up a too picky logic level threshold?
Just my 2 cent...
Best,
0xfede
-
Hi 0xfede,
Thank you very much for your response, and sorry for late reaction... I had lot of other stuff to do.
I tried what you've suggested and adjusted the threshold levels but unfortunately this didn't solve the issue.
More and more tend to think that this is a hardware issue in my unit. I've tried to run a probe calibration for the logic analyzer and that failed for multiple times. So I think I had to contact the seller. However I would still appreciate any other idea.
-
If a different line works then it seems D0 is broken but you better check if D0 works again. After all it can be a bad contact in the grabber, probe or the plug into the scope.
-
Comparing the "bad" with the "good" screenshot, one can see that in the "bad" setup, the spikes on the MOSI line occur exactly where the positive edges are in the correct MOSI signal. So this might indeed be a bad connection, where the MOSI signal is only capacitively coupled.
You mentioned that you have wiggled the cables and connectors without success. I guess the connection problem could also be a bad solder joint inside the scope -- maybe between the logic probe connector and the main board? It may be worth a try to gently press or bend the connector from the outside while measuring, and check whether that makes a difference. (If it does, of course it's time to contact the seller and ask for a replacement or warranty repair.)
-
Thank you very much for your responses. I hope I will get some time at the weekend and will try your suggestions. I will get back with the results.
-
There is no reason that you cannot connect multiple logic analyser inputs to the MOSI line at the same time. See if they all show the same thing.
I agree with bastler; the problem looks like an open in the D0 signal path and capacitive coupling is showing the rising edges.
-
Yes, actually this was my thought as well today and I made a fast setup. I've connected all of the LA lines to that microcontroller. I hooked up the probes from D0-D7 range to the MOSI line and the D8-D14 to the clock line. The remained D15 is hooked to CS and I used as trigger. The reason of this separation is that I did not have enough space on the breadboard to hook up all probes on the same line.
The results are on the attached screenshot. It seems that I have more dead lines D0, D5 and D10.
I've also tried to gently push the connector on the scope and at the other end of the cable, as ebastler mentioned, but nothing has changed.
-
Looks like the inputs are actually capturing a rising-edge trigger pulse or some similar derivative of the signal, rather than the signal itself. If there's nothing in the scope setup that could cause this then I'd say it's broken...
-
I concur: D0, D5 and D10 are dead so it is time to pack the scope into it's box and let it meet it's maker ;)
-
Thank you for all of you for responses. I will contact the seller and let see how this will be handled.