Transfer function plot/print automation on test point completion

36 views
Skip to first unread message

Mike Thake

unread,
Feb 13, 2012, 11:49:31 AM2/13/12
to IADS
Hi,

I am relatively new to IADS and need some help with a customer
request. They would like the ability at the end of a test point
(during real-time testing) to take the data from a parameter, create a
transfer function plot of the parameter during the entire test point
and print it. I know I can do this by creating a stripchart of the
parameter, manually selecting the endpoints, right-clicking >
“Compute” > “Phase & Magnitude”, adding my excitation parameter,
adjusting the properties, and then printing the individual plot. This
works well, but they would like it done for about 20 parameters and
want it done within seconds of the test point completion…I am not that
fast. I would like to somehow automate this process (or maybe there
is something already in place that I have yet to find).

This is what I am thinking might work, but please chime in if you have
other thoughts…

1) Add the excitation per parameter in the ParameterDefaults area.
2) Create a DataGroup of my response parameters of interest.
3) Create an action to create the start and stop of the test point of
interest.
4) Create a VB script that will create all 20 transfer function plots
of the full time of the test point, adjust their properties for
optimal viewing, and place them neatly onto a few new Analysis
Windows. (This part is my downfall because I know very little about VB
scripts)
5) Create an action button to print the new analysis windows (or can I
print individual plots without manually right-clicking?) with a header
bar of “Point in the sky” information.
6) Once the test point completes, I would go to “IadsLogs” in the
Dashboard, click the “Test Points” tab, highlight my test point of
interest, click the “Export/Process Selected Rows” button, choose my
Data Group, and select my VB script as the Export Type (not sure what
I would do about the “Output Directory” though), then print my
windows. (Or I could have it execute automatically at test point
completion?).

I am curious if this is the easiest and most efficient solution and
how to go about starting such a VB script (i.e. what the IADS calls
are to create a FR plot, add parameters, change plot properties, etc.)

As a side note, I know when I right-click the stripchart, there is an
option for “Data Group Compute” but the “Phase & Magnitude” option is
not available. Maybe knowing how to use this would speed up my
process?

Also, the parameters discussed here would probably be derived…not sure
if that makes a difference.

Thanks,
Mike

James Bretz

unread,
Feb 13, 2012, 3:48:01 PM2/13/12
to IADS
Hi Mike,

In theory, you could spend the time an effort to build a VB script, but I
don't think it's needed. Also, I honestly don't think the Iads automation
(VB script) interface is setup to perform fixed block frequency computations
in batch mode (yet). You can surely create all of the displays, attach the
parameters, and set some plot attributes, but the interface just isn't there
to execute the fixed blocked computation (compute over a given time frame)
as of yet.

As far as the "Data Group Compute" option, it looks like only the PSD and
Autospectrum options have been implemented... and the FRP based options are
still in the works. In other words, it isn't anything you can execute at
this point. I just added an item to our request database to look into this
issue. Regardless, the datagroup computation (if it was active) probably
wouldn't work for you either. As the plots are generated, they are tiled
atop one another. so it wouldn't be very optimal for printing.

If you want to do this in the real time arena, I think the most
straightforward way is to:

1) Put all of the 20 Phase/Mag plots on an AnalysisWindow
2) Attach all the specific parameters and setup all the plot attributes as
desired
3) Set the TestPoint Action on Start = 'Reset Frequency Averaging'
4) Set the TestPoint Actions on Stop = 'Print AnalysisWindow'

When you run in real-time, the window should reset all the FRPs at the start
of the testpoint and then print at the end. Another option would also be to
"Pause AnalysisWindow" on TestPoint stop, examine all the plots, and then
optionally print the AW by hand.

Hope this helps... and just FYI, derived parameters or not... doesn't really
change the situation,
Jim

--------------------------------------------------
From: "Mike Thake" <mpt...@gmail.com>
Sent: Monday, February 13, 2012 8:49 AM
To: "IADS" <ia...@googlegroups.com>
Subject: [IADS] Transfer function plot/print automation on test point
completion

> Hi,
>
> I am relatively new to IADS and need some help with a customer
> request. They would like the ability at the end of a test point
> (during real-time testing) to take the data from a parameter, create a
> transfer function plot of the parameter during the entire test point
> and print it. I know I can do this by creating a stripchart of the
> parameter, manually selecting the endpoints, right-clicking >

> �Compute� > �Phase & Magnitude�, adding my excitation parameter,


> adjusting the properties, and then printing the individual plot. This
> works well, but they would like it done for about 20 parameters and

> want it done within seconds of the test point completion�I am not that


> fast. I would like to somehow automate this process (or maybe there
> is something already in place that I have yet to find).
>
> This is what I am thinking might work, but please chime in if you have

> other thoughts�


>
> 1) Add the excitation per parameter in the ParameterDefaults area.
> 2) Create a DataGroup of my response parameters of interest.
> 3) Create an action to create the start and stop of the test point of
> interest.
> 4) Create a VB script that will create all 20 transfer function plots
> of the full time of the test point, adjust their properties for
> optimal viewing, and place them neatly onto a few new Analysis
> Windows. (This part is my downfall because I know very little about VB
> scripts)
> 5) Create an action button to print the new analysis windows (or can I
> print individual plots without manually right-clicking?) with a header

> bar of �Point in the sky� information.
> 6) Once the test point completes, I would go to �IadsLogs� in the
> Dashboard, click the �Test Points� tab, highlight my test point of
> interest, click the �Export/Process Selected Rows� button, choose my


> Data Group, and select my VB script as the Export Type (not sure what

> I would do about the �Output Directory� though), then print my


> windows. (Or I could have it execute automatically at test point
> completion?).
>
> I am curious if this is the easiest and most efficient solution and
> how to go about starting such a VB script (i.e. what the IADS calls
> are to create a FR plot, add parameters, change plot properties, etc.)
>
> As a side note, I know when I right-click the stripchart, there is an

> option for �Data Group Compute� but the �Phase & Magnitude� option is


> not available. Maybe knowing how to use this would speed up my
> process?
>

> Also, the parameters discussed here would probably be derived�not sure


> if that makes a difference.
>
> Thanks,
> Mike
>

> --
> You received this message because you are subscribed to the Google Groups
> "IADS" group.
> To post to this group, send email to ia...@googlegroups.com.
> To unsubscribe from this group, send email to
> iads+uns...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/iads?hl=en.
>

John Bretz

unread,
Feb 13, 2012, 5:43:30 PM2/13/12
to IADS
"> 1) Put all of the 20 Phase/Mag plots on an AnalysisWindow"

20 FRPs in one 8 1/2 by 11 seems kinda tough to see. Do we have the
ability to print the individually?

John

James Bretz

unread,
Feb 13, 2012, 6:51:30 PM2/13/12
to IADS
The TestPoint actions don't have the ability to print individual plots....
but through the script interface, yes. If need be, the Actions on Stop could
launch a script to print each plot (and flood the printer lol).

Also, let me correct my last post. The DataGroup compute for the frequency
plots actually overlays all the results on a single display... so I assume
if we had the FRP capability, it would do the same. Again, not really what
is needed IMO... perhaps if we had an extensive legend and they were using a
color printer.

Jim

--------------------------------------------------
From: "John Bretz" <johnw...@roadrunner.com>
Sent: Monday, February 13, 2012 2:43 PM
To: "IADS" <ia...@googlegroups.com>
Subject: [IADS] Re: Transfer function plot/print automation on test point
completion

> "> 1) Put all of the 20 Phase/Mag plots on an AnalysisWindow"

Mike Thake

unread,
Feb 14, 2012, 12:56:58 PM2/14/12
to IADS
Great, thank you! You guys are saving me lots of wasted time.

John - I could just set up 5 AW with 4 FRP's each, or something to
that effect and create a button to print them all.

Jim - As for the real-time method you suggested, it really wouldn't be
a "fixed block" computation would it? I would just get up to 5 blocks
frequency averaged...but if I made the FFT block size larger it would
increase the window of total data I could consume for the FRP,
correct? I will need the full window of the test point in the FRP.

Mike

John Bretz

unread,
Feb 14, 2012, 1:29:20 PM2/14/12
to IADS
Mike,

Great idea (5 windows)!

> Jim - As for the real-time method you suggested, it really wouldn't be
> a "fixed block" computation would it?  I would just get up to 5 blocks
> frequency averaged...but if I made the FFT block size larger it would
> increase the window of total data I could consume for the FRP,
> correct?  I will need the full window of the test point in the FRP.

If you set the averaging to "0", a running average is computed from
the start to the stop time. The larger block size actually hinders
confidence in the data by decreasing the number of fft's in the
average (from what I have read). You have to balance frequency
resolution against confidence. The Chirp Z Transform may also be a
tool that would help if your freq range of interest is rather narrow
(like 0-100 given a sample rate of 1K), because you can use much
smaller blocksizes and get good resolution along with many averaged
fft's. Set overlap small unless your CPU bogs down.

John

James Bretz

unread,
Feb 14, 2012, 1:49:49 PM2/14/12
to IADS
Mike,

> As for the real-time method you suggested, it really wouldn't be
> a "fixed block" computation would it? I would just get up to 5 blocks
> frequency averaged...but if I made the FFT block size larger it would
> increase the window of total data I could consume for the FRP,
> correct? I will need the full window of the test point in the FRP.

As John mentioned, set the number of averaging blocks to "0" which denotes
infinite averaging. If you use infinite averaging and also reset the
averaging at the beginning of the TestPoint (via our friendly Actions on
Start) then it should be mathematically equivalent to performing a fixed
block computation over that time range.

John, please correct me if I'm wrong.

If you want 5 windows you'll need to have 5 monitors. The window has to be
visible to print properly. In other words, I don't think you can set the
"Run while Minimized" option on the Aw and it compute and print in the
background. It will compute properly for sure, but I'm guessing that when it
goes to print you'll get a blank page (but if you want you can test this
theory).

Your best bet is to use the "DisplayFolder". Create 5 tabs and put 4 FRPs
per tab. Once that is complete, activate the "Multi Print Tab Displays"
option on the right click menu of the IrigTime panel (on the Dashboard).
Make sure this AW is visible during the test (secondary monitor).

Of course, I suggest testing things out before going live ;)
Jim

--------------------------------------------------
From: "John Bretz" <johnw...@roadrunner.com>

Sent: Tuesday, February 14, 2012 10:29 AM


To: "IADS" <ia...@googlegroups.com>
Subject: [IADS] Re: Transfer function plot/print automation on test point
completion

> Mike,

Mike Thake

unread,
Feb 14, 2012, 3:02:19 PM2/14/12
to IADS
Jim, John,

> As John mentioned, set the number of averaging blocks to "0" which denotes
> infinite averaging. If you use infinite averaging and also reset the
> averaging at the beginning of the TestPoint (via our friendly Actions on
> Start) then it should be mathematically equivalent to performing a fixed
> block computation over that time range.

Really useful information, guys. I wasn't aware that 0 blocks =
infinite averaging and of the TestPoint Actions. In theory, this
should solve all my problems (whew, no VB Scripting).

I will test this out in the coming days and let you know how it played
out.

> Your best bet is to use the "DisplayFolder". Create 5 tabs and put 4 FRPs
> per tab.  Once that is complete, activate the "Multi Print Tab Displays"
> option on the right click menu of the IrigTime panel (on the Dashboard).
> Make sure this AW is visible during the test (secondary monitor).

This sounds like a more elegant way to do the multi-plot anyway. Good
idea.

Thanks,
Mike
Reply all
Reply to author
Forward
0 new messages