Five best breast pumps of 2023 - A complete review

Having problems ? let us know the details here
Post Reply
Gerry
PICO STAFF
PICO STAFF
Posts: 1144
Joined: Mon Aug 11, 2014 11:14 am

Re: Capture a dynamic time measurement between two different channels

Post by Gerry »

Hi khaldrogo,

Sorry for the delayed response (I've been on leave).

You didn't give any more detail about the pulses, so I'll assume that they are the same or similar, and one is always leading the other.

What you can do is position the trigger near to the start of the capture, and then select a Timebase that is just large enough to capture the 2 pulses when they are the furthest apart that they will ever be. Then create a Math Channel of A+B and add a measurement of Cycle Time, I have created an example data file, with a screenshot showing that the measurements, both manually and automatically (dynamically), are close enough for agreement, below:
Leading edge difference measurement for non-overlapping pulses.png
(Note that the data file was too large to post with all of the captures that make up the statistics, so it only has one capture. However, you can see the statistics for the other captures in the image file)

Unfortunately, this will only work if the pulses don't overlap. For pulses that overlap you will need to use the Math Channel abs(B-C) with a High pulse Width measurement, as shown in the data file and image below:
Leading edge difference measurement for overlapping pulses.png
So, you can have both Math Channels active and one will always be giving you the correct value.

Regards,

Gerry
Gerry
Technical Specialist

Gerry
PICO STAFF
PICO STAFF
Posts: 1144
Joined: Mon Aug 11, 2014 11:14 am

Re: Capture a dynamic time measurement between two different channels

Post by Gerry »

Hi khaldrogo,,

One thing I forgot to mention in my previous post was that you need to be able to guarantee isolated single pulses on each channel for the duration of the captures. If there is more than one pulse on a channel, the Math Channels will not work correctly (as you will have a mixture of cycle times for one and a mixture of pulse widths for the other).

If you can't isolate just one pulse on each channel, for example, where pulses are periodic (instead of single shot) and one period is a multiple of the other, then there's an alternative method that you can use to get automatic (dynamic) measurements, and that is use Deep Measure to perform the measurements. As Deep Measure only measures one channel of pulses relative to the Trigger point, you need to trick Deep Measure into referencing the Trigger point on the other channel. I have posted an example data file and screenshot for Periodic pulses on both channels (zoomed in to the first edge for the Start Time measurement) below:
Measurement of Periodic 1st and 2nd pulse timing difference.png

You can see that the manual measurement (for the difference between the leading edge and the trigger at the start of the other channel pulses leading edge) when made using the measurement ruler, matches the Start Time measurement from Deep Measure. Notice also that, as Deep Measure is a completely separate function to the other automatic Measurements, you won't get measurement statistics for them.
Note also that, the individual rows of Deep measurements represent the other pulses in the same capture, not pulses from different captures, as demonstrated in the example below. In the example, the capture is of pulses with just one period on one channel, and a single shot pulse on the other, (which would corrupt the Math Channels if they had been used) which only results in one row of Deep Measurements.
Measurement of single shot1st and periodic 2nd pulse timing difference.png

You can create your own statistics, but you will have to export CSV files of the Deep Measurements for every capture. For instance, if you're using a repeat/rapid trigger, Picoscope 6 will just dump all of the CSV files for each capture into a folder when it does the export. So, to make it more automatic, you could setup the import of just the Start Time measurement for the first row of every CSV file into in a CSV reading, Mathematical tool such as Excel, Matlab, Python, Wolfram, etc, to be processed by your own previously created Measurement Statistic functions.

If, on the other hand you would prefer to just look through the start times of the other captures rather than setting up a tool to automatically import the start times and create the statistics for them then you can just scroll through the other deep Measurements using the Buffer Overview icon (the one that looks like a compass).

Note that Deep Measure won't work in the examples I gave in my previous post (i.e. where there is only one pulse captured on each channel) because it requires Periodic waveforms.

Regards,

Gerry
Gerry
Technical Specialist

Gerry
PICO STAFF
PICO STAFF
Posts: 1144
Joined: Mon Aug 11, 2014 11:14 am

Re: Capture a dynamic time measurement between two different channels

Post by Gerry »

11/06/21 UPDATE
A recent customer query has highlighted some information that needs to be added here.

In the previous posts it was assumed that the measurements would be between square shaped pulses. Of course, it could be possible that these are rounded pulses (e.g. due to the filtering effect of the transmission, or generation) or even single cycles of a rounded waveform. If the purpose of the measurement was to establish, for instance, the Time-of-flight of a signal sent on Channel A and received on Channel B, then there could be some insertion loss by the time the signal is received. I have created an example below, which shows this:
Time-of-flight measurement.psdata
(1.84 MiB) Downloaded 184 times
Time-of-flight measurement.png

In the data file and image you can see that I have assumed a rounded pulse will be sent (on Channel A in blue) and received (on Channel B in red) with a reduction in amplitude of the received signal, compared to the sent signal. This makes it difficult to use the same reference point for the measurement of the pulses (i.e. the midpoint of the pulse on Channel A is not the same voltage level as the midpoint on pulse B). However, to get around that problem you can create a Math Channel that normalizes both pulses to the same level (i.e. normalized Channel A in green, and normalized Channel B in orange) which enables zero volts to be used as the reference point for measuring the start of both pulses. So in this instance the time-of-flight is 2.2605ms.

So in this case the previous Math Channels would need to be modified accordingly

Regards,

Gerry
Gerry
Technical Specialist

Post Reply