AC voltage math channel

Ask any questions about using PicoScope Automotive software here.
Post Reply
steevegt
TwoWaves
TwoWaves
Posts: 70
Joined: Tue May 12, 2015 10:05 pm

AC voltage math channel

Post by steevegt » Mon Jun 12, 2017 2:43 pm

Hi,

Is there a way of creating a math channel that will output the AC voltage of a signal?
The only way I know how to do it is by creating a measurement and use rules to get the value between them. Not practical and not graphical as I wanted...

My need for this at the moment is to graph the position value reported by a sensor that outputs an AC signal, with amplitude proportional to the position.

Thanks
Attachments
ac_signal_sensor.PNG

steevegt
TwoWaves
TwoWaves
Posts: 70
Joined: Tue May 12, 2015 10:05 pm

Re: AC voltage math channel

Post by steevegt » Fri Sep 07, 2018 11:00 am

After all this time with no response I can assume that there is no way to do this in the software.

Can I suggest a feature request for this? Or no one thinks this is a useful feature?

Thanks
Steeve

Steve Smith
Pico Staff Member
Pico Staff Member
Posts: 597
Joined: Sun Aug 25, 2013 7:22 am

Re: AC voltage math channel

Post by Steve Smith » Mon Sep 10, 2018 3:07 pm

Hello Steeve, thank you for the post and the timely reminder, sorry for the mega late feedback

Could you post the psdata file here as I would like to try a number of maths options to this waveform?

I can see how such a feature would be useful as from your description, amplitude is relative to position and to graph this in a single buffer would be helpful.

We are therefore looking to graph the RMS of this waveform?

I have taken a quick look at Deep Measure topic17711.html with a donor waveform and used Excel to graph Peak to Peak and Peak voltages.
Deep Measure.jpg
Deep Measure
I know this is not the answer at present but I guess the graphs above are what you are looking to obtain in PicoScope?

Thinking this through further, for now, applying Deep Measure will at least allow you to jump to the the cycles of interest with the greatest amplitude and hopefully assist you with engine position.

I hope this helps in the interim

Take care.....Steve

steevegt
TwoWaves
TwoWaves
Posts: 70
Joined: Tue May 12, 2015 10:05 pm

Re: AC voltage math channel

Post by steevegt » Thu Sep 13, 2018 9:45 am

Thank you again Steve for your reply.

I would have to look for the psdata file, it has been some time, not sure where it is, or if I still have it.
If I could not find it, I will see if I could capture it again.

Those graphs are similar to what I want to achieve, yes.

Long story short, this capture was made from a Bosch distributor fuel injection pump, an electronically controlled one.
The collar position is controlled with a solenoid, and it has a feedback sensor that outputs the position. According to the documentation this is an angle sensor, a potentiometer. But according to the measurements I made it does not look like a potentiometer at all. I may be wrong...
The thing I know it that the output from the sensor is an AC signal, with the AC voltage value being proportional to the colar position. So if we want to analyse this sensor output with the scope graphically we have to take into account the AC voltage value, instead of the voltage signal itself. Makes sense?
It is very similar as analysing the signal from a "digital" MAF sensor.

RMS I think it will be ok. Like I said the software allows me to calculate AC RMS value for the capture or for a window of time, using calculated measurements, but unfortunately there is no AC math function for this.

Until now this is the only sensor that I remember to came across that has an AC output signal. Maybe there are or not more like this, not sure... Nevertheless this shouldn't be hard to add to the math channels options I guess, and it will be great for analysing this type of sensors.

Thanks
Steeve

Steve Smith
Pico Staff Member
Pico Staff Member
Posts: 597
Joined: Sun Aug 25, 2013 7:22 am

Re: AC voltage math channel

Post by Steve Smith » Thu Sep 20, 2018 12:43 pm

Hello Steeve, sorry for the late reply

I have to a say a big thank you to Martyn, Gerry and the technical support team for this one.

One thing is for certain “Maths is cool”
MATHS IS COOL.jpg
Maths is cool
The following forum, post has this covered https://www.picotech.com/support/topic2 ... ml#p101421 but I will add a note to help here with the formula.

Assuming your signal is present on channel A, the formula required is “sqrt(integral((A)^2)/T)”

Using our Guided Test for an inductive Crankshaft Sensor (to obtain an example waveform) I have applied the above math channel to channel A in order to graph the RMS voltage of the AC Signal
AC MATHS .jpg
AC RMS Math channel
I am far from a mathematician but the additional information that “maths” can reveal using PicoScope is staggering.

Above we graph the RMS of the AC voltage and with a little help from the scaling feature, we can clearly see the increase in voltage relative to the position of the collar.

So how do we arrive at this graphed RMS value using the formula above?

First we need to “square” (A^2) the AC voltage which has the effect of inverting the negative portion of our waveform (Full Wave Rectification)

The "integral A" of the formula then adds all "areas" under the curve together, which creates a sloping line with an inclination increasing with time

Then we divide “A” by the sweep time across the screen "/T" to level the line calculated using the integral

Finally, the “Sqrt A” will calculate the square root of the above level line to reveal the RMS value

I hope the image below will help.
2018-09-20_06-41-11.jpg
Formula application
Gerry has indicated a number of advisories in his forum post too but to summarise

The start of the math channel can be ignored as the software math function is dividing by zero, hence the confusion and oscillations at the start of the math channel before it climbs to the calculated value of approx. 570 mV RMS

Where using a trigger, the pre-trigger should be set to 0%, (with the yellow trigger diamond positioned on the far left of the screen), and be greater than zero, so that Time is always positive, and the waveform is meaningful.

There needs to be enough cycles of the signal waveform, to allow the Math Channel waveform to converge on a single value. If using measurements, only rulers placed on the converged DC portion of the Math Channel will give correct measurement values.

The RMS Math Channel will not converge quickly enough when zooming in, unless you're zooming in near the start or you CAN zoom in vertically if you use all of the horizontal axis.

The y-axis scaling of a Math Channel is not automatically proportional to the input channel/channels being used, so you may need to adjust it to match the Input Channel scaling.

The psdata file containing the above math channel can be found below:
20180918-0001.psdata
psdata file with AC RMS math
(1.56 MiB) Downloaded 24 times
Thank you again support team, I hope this helps, take care…….Steve

Post Reply