Hantek 6022BE 20MHz USB DSO

[RichardK]

Latest Binary: http://jmp.sh/v/TpJwJVuWFtNDFZJMovL8

What's New:
-----------------------------------------------------------
1. Fixed a bug where Interactive Cursors out of sync with AC Coupled waveforms.
2. Full support for viewing stopped waveforms at different timebases and voltage divisions.
3. Cursors now work on Zoomed waveforms .
4. Removed Scale warning for zoomed waveforms (Proper scaling now implemented).
5. Added Reset Zoom menu item and Toolbar Button.
6. Added Fullscreen, Hide Toolbar and Hide Sidebar buttons to Toolbar.
7. Moved Start & Stop acquire buttons closer to the left side of the Toolbar.
8. Various other bugs fixed, performance improvements, code cleanup.

[Plummerman]

Hello Rick,

      I just joined this forum and see that folks are evidently writing software for the HANTEK PC Oscilloscope...  this is encouraging.

      I have a HANTEK 6022BE and want to download the trace data.  So far, I've been successful (sort of) at getting the comma separated variable file... the problem is that there's so much data that it overwhelms my spreadsheet (Open Office Calc).  It appears that every sample is exported.... I've used network analyzers, spectrum analyzers, and other scopes with similarly high sample rates; however the downloaded data is much, much less, like 3K data points instead of 500K - some models have allowed me to set the number of data points (AGILENT for example).  Another problem is that the corresponding x-axis data (time scale) is not exported.  Furthermore, the trace data appears all in the same columns (making extremely long columns) rather than appearing in sequential columns.  I could move this around... except the spreadsheet program tends to choke on the large data set... so forget using a formula to calculate the time base...  This is a problem for me, because I've retired an no longer have access to all the wonderfully expensive equipment I once did.  So... have any of my issues been addressed with software, OEM, or after-market?  Where could I find more information on it?

[Rick Law]

Latest Binary: http://jmp.sh/v/TpJwJVuWFtNDFZJMovL8

What's New:
-----------------------------------------------------------
1. Fixed a bug where Interactive Cursors out of sync with AC Coupled waveforms.
2. Full support for viewing stopped waveforms at different timebases and voltage divisions.
3. Cursors now work on Zoomed waveforms .
4. Removed Scale warning for zoomed waveforms (Proper scaling now implemented).
5. Added Reset Zoom menu item and Toolbar Button.
6. Added Fullscreen, Hide Toolbar and Hide Sidebar buttons to Toolbar.
7. Moved Start & Stop acquire buttons closer to the left side of the Toolbar.
8. Various other bugs fixed, performance improvements, code cleanup.

Thanks RichardK!  I will look forward to testing it tomorrow.

[Rick Law]

Hello Rick,

      I just joined this forum and see that folks are evidently writing software for the HANTEK PC Oscilloscope...  this is encouraging.

      I have a HANTEK 6022BE and want to download the trace data.  So far, I've been successful (sort of) at getting the comma separated variable file... the problem is that there's so much data that it overwhelms my spreadsheet (Open Office Calc).  It appears that every sample is exported.... I've used network analyzers, spectrum analyzers, and other scopes with similarly high sample rates; however the downloaded data is much, much less, like 3K data points instead of 500K - some models have allowed me to set the number of data points (AGILENT for example).  Another problem is that the corresponding x-axis data (time scale) is not exported.  Furthermore, the trace data appears all in the same columns (making extremely long columns) rather than appearing in sequential columns.  I could move this around... except the spreadsheet program tends to choke on the large data set... so forget using a formula to calculate the time base...  This is a problem for me, because I've retired an no longer have access to all the wonderfully expensive equipment I once did.  So... have any of my issues been addressed with software, OEM, or after-market?  Where could I find more information on it?

Plummerman,

The number of data points collected is entirely controlled by time per division.  Once you set time per division (say 10nS/div), that setting alone will dictate the sampling speed and number of data points collected.

Look at a few pages, you will see a reply containing time/div verses data points and sample speed that I posted.

Rick

[Plummerman]

Hello Rick,

     Thanks for the prompt reply.  My application is for calibrating small wind turbines (< 3kW) output characteristics.  BTW, I have X100 probes because the output voltage in the operating regions is 50V to 300V.  At any rate, the time per division is on the order of milliseconds, which at a 48MSa/sec would yield 1 million samples or more for 10 divisions (1 screen).  As I recall on my first test runs,  I was capturing something less than this (130k data points or more), so it would seem the scope software was adjusting the sample rate somehow.... but that's still way, way, too much (way over-sampled).  Does the OpenHantek software have a more intelligent control over the sample rate than the OEM software?

[Matchless]

RichardK, you are doing outstanding work on this!
I played with that Zoom function a bit and it is really looking very nice.

Please just ignore these questions if they are irrelevant:

Do you intend activating the Calibration function at a later stage?
Is it possible to change the 3 large rotary knobs to a better HD that looks more "quality" in line with your software?
Is it possible or bad practice to allow for clicking on the relevant rotary knob to activate the mouse scroll wheel, where at the moment one has to click the drop down box voltage?

Thanks again for what you are doing!

[frenky]

Richard is doing amazing work with the code so I suggest that we do the modified UI elements so he won't have to.
If you send me some photos of desired buttons I can draw them in Inkscape and upload here.
So Richard would only have to replace the original files...

I have already found one nice button (the big one on the top):

[frenky]

I have now noticed that the buttons in the software are not images but coded graphic elements (circles, dots) so change to rotating image would not be so trivial...
IMHO it's best to wait with UI changes (or requests) until the software is completely done...

[Matchless]

Hi frenky,
              I agree that cosmetics are low on the list for sure, but once the software is complete and working,  that will definitely also make it look good as well!
Unfortunately that is not my field at all, so there is not much that I can contribute to this except some ideas.
I hope I am not breaking any rule, but here is a PCscope with a very good looking GUI and this sort of caused me to ask about the rotary knobs:
xxx.zeitnitz.de/Christian/scope_en   (xxx = www)
Thanks for your offer!

[Rick Law]

Hello Rick,

     Thanks for the prompt reply.  My application is for calibrating small wind turbines (< 3kW) output characteristics.  BTW, I have X100 probes because the output voltage in the operating regions is 50V to 300V.  At any rate, the time per division is on the order of milliseconds, which at a 48MSa/sec would yield 1 million samples or more for 10 divisions (1 screen).  As I recall on my first test runs,  I was capturing something less than this (130k data points or more), so it would seem the scope software was adjusting the sample rate somehow.... but that's still way, way, too much (way over-sampled).  Does the OpenHantek software have a more intelligent control over the sample rate than the OEM software?

Plummerman,

Hantek software (and perhaps hardware) for 6022 capture rates and capture size are automatically set by your time-per-division.  It has not other way of controlling it.

If you want 48MHz capture rate, the most you are going to capture is 1016 data points.
For 1M data points, you have to slow down to 50ms/div. See this table that I put together:
------------------
At 50ms (1Mhz) it starts acquiring at full 1M samples (1,047,552 for each channel total 2x1047552)

At 20ms (1Mhz) it is at 523264 samples each Channel
At 10ms (1Mhz)= 523264
At 5ms (1Mhz)= 523264

At 1ms (1MHz) = 130048
At 100us (1Mhz) = 130048
At 20us (4Mhz) = 130048
At 10us (8Mhz) = 130048
At 5us (16Mhz) = 130048

At 2us (48Mhz) = 1016
Anything faster than 2us, it is still at 48MHz at 1016 samples each channel (2032 total)
------------------

So, your signal determines what minimum capture rate you need.  You said your signal is in the 50-500v range, but that doesn't say how fast you want/need to observe it.

Since you are setting it "in the order of mS/division", I assume that is the range you are observing the characteristic you want to see.  In that range, you are pretty much 130048 points.  You can adjust the sampling rate (see table above) but not the length until you go down to 2uS for 1016 points.

However, if at that time-per-division is fast enough but you capture more than you need, you can trim the data with a quick and dirty program.  If you need more capture data than the scope is capable of (at your needed minimum capture rate), than it would be time for another scope.

Hope this helps.

Rick

[RichardK]

Do you intend activating the Calibration function at a later stage?

It has been implemented for quite some time.

Is it possible to change the 3 large rotary knobs to a better HD that looks more "quality" in line with your software?

Anything is possible, but right now I'm focusing on getting core features implemented. I don't see it being very hard to replace the generic knobs with color coded graphical ones, but at a later date.

Is it possible or bad practice to allow for clicking on the relevant rotary knob to activate the mouse scroll wheel, where at the moment one has to click the drop down box voltage?

I'll look into this, but it's low priority for the moment as I'm trying to get Measurements implemented now... Thanks for the suggestions though, all are appreciated.

[Matchless]

RichardK,
              Thanks, I must have missed that. At the moment if I run Calibration on the Hantek software it executes with no further feedback at all. If I run it on Open 6022be it executes and then gives Cannot calibrate: Calibration failed.
I may have missed something. Is there any specific steps to follow, except having it on for a while and having the probes connected to ground?
Thanks

[RichardK]

RichardK,
              Thanks, I must have missed that. At the moment if I run Calibration on the Hantek software it executes with no further feedback at all. If I run it on Open 6022be it executes and then gives Cannot calibrate: Calibration failed.
I may have missed something. Is there any specific steps to follow, except having it on for a while and having the probes connected to ground?
Thanks

I followed the SDK on Calibration, but the SDK documentation might be wrong or there could be a bug in my code, I'll check it out.

[Rick Law]

RichardK,
              Thanks, I must have missed that. At the moment if I run Calibration on the Hantek software it executes with no further feedback at all. If I run it on Open 6022be it executes and then gives Cannot calibrate: Calibration failed.
I may have missed something. Is there any specific steps to follow, except having it on for a while and having the probes connected to ground?
Thanks

Matchless ,

I cannot reproduce your issue -- I think you might have forgot to ground the probes first.

Rick

[Rick Law]

Latest Binary: http://jmp.sh/v/TpJwJVuWFtNDFZJMovL8

What's New:
-----------------------------------------------------------
1. Fixed a bug where Interactive Cursors out of sync with AC Coupled waveforms.
2. Full support for viewing stopped waveforms at different timebases and voltage divisions.
3. Cursors now work on Zoomed waveforms .
4. Removed Scale warning for zoomed waveforms (Proper scaling now implemented).
5. Added Reset Zoom menu item and Toolbar Button.
6. Added Fullscreen, Hide Toolbar and Hide Sidebar buttons to Toolbar.
7. Moved Start & Stop acquire buttons closer to the left side of the Toolbar.
8. Various other bugs fixed, performance improvements, code cleanup.

RichardK,

All the above fixed were tested without problem.  Three minor issue plus a suggestion:

Issue 1: When trigger is not triggered, confusing wave form...

This bug is on NORMAL trigger and doesn't show in automatic trigger.   In earlier release, you said only automatic trigger is working. I'm not sure that NORMAL trigger is "done" or work in progress.  Anyhow, you want to know this anyway.

When trigger point is not within wave-range, the waves are still shown but deformed.  The volt displayed is inaccurate.  (Stock software draws nothing when trigger point is not within wave-range, so it would not be misleading or confusing.)

It may be best to remove the visible trace so as it would not be confusing when normal trigger is not triggered.

To see the confusing trace:
1 use scope’s build-in reference
2 time=1ms/div
3 volt ch1 2v/div
4 volt ch2 1v/div
5 Trigger Edge/Normal/Ch1/+
6 DC coupling for both
7 Trigger point at about 1v/div
Now drag the trigger point up and down a bit to just above the wave form, you can see the wave form CH2 is shown and jumps around, but volt is not accurate.
8 Now set CH1 to AC coupling, both wave forms are shown with inaccurate volts.

Issue 2: Zoom...
Zoom works really nice - particularly the reset zoom.  It lacks just one thing.  One doesn't know how much is zoomed, and reading the screen can easily confuse one into reading a wrong number.

You need either some magnification indicator (1X, 2.3X) somewhere on screen, and/or resize the grid.  With the grid unchanged but zoomed, one can easily be confused and it is hard to discern the true value of the Horizontal or Vertical.

Issue 3: Full Screen
This works nice too!.  It took me a bit of trying to realize that right-mouse button give me the option to re-window it.  Perhaps you should put a tiny icon at the corner to re-window.  First few times, I close the app and restart to regain window-ed mode.  After a few tries, I want to see if there is a way and I tried all kinds of keys - eventually got to trying the other mouse button.

A small icon in the corner would be a nice touch.


Suggestion:
Earlier point Mark_O made about when one screen of 260 screens of data is shown, it is hard to see where the visible screen is relative to the start of data.  I pointed out that your mini-wave at top-middle of screen and the T points at exactly the start position of data.

A further improvement would make it even more informative - may be on the left of the mini-wave, show x/130048 where x is the data point the screen trace begin and 130048 is however many data points is collected or available for scrolling.  Is that possible?

(Not a request, just blue-skying)  For super-super showoff, you can even have an "auto-advance" and a fast/slow button as well.  So when auto-advance is engaged, the wave start scrolling from start of saved data forward so the Wave moves right to left (like a movie moving in time).  Pressing the fast/slow button increase/decrease the play speed.  That would be super super awesome way of inspecting the saved data points.

Hey Richard, this is great work!  Thanks!

Rick

[RichardK]

Issue 1: When trigger is not triggered, confusing wave form...

This bug is on NORMAL trigger and doesn't show in automatic trigger.   In earlier release, you said only automatic trigger is working. I'm not sure that NORMAL trigger is "done" or work in progress.  Anyhow, you want to know this anyway.

When trigger point is not within wave-range, the waves are still shown but deformed.  The volt displayed is inaccurate.  (Stock software draws nothing when trigger point is not within wave-range, so it would not be misleading or confusing.)

It may be best to remove the visible trace so as it would not be confusing when normal trigger is not triggered.

To see the confusing trace:
1 use scope’s build-in reference
2 time=1ms/div
3 volt ch1 2v/div
4 volt ch2 1v/div
5 Trigger Edge/Normal/Ch1/+
6 DC coupling for both
7 Trigger point at about 1v/div
Now drag the trigger point up and down a bit to just above the wave form, you can see the wave form CH2 is shown and jumps around, but volt is not accurate.
8 Now set CH1 to AC coupling, both wave forms are shown with inaccurate volts.

Yeah some graphical aspects of triggering are not finished, like what you brought up, how the waveform should not be updating when the trigger is outside wave range.

Issue 2: Zoom...
Zoom works really nice - particularly the reset zoom.  It lacks just one thing.  One doesn't know how much is zoomed, and reading the screen can easily confuse one into reading a wrong number.

You need either some magnification indicator (1X, 2.3X) somewhere on screen, and/or resize the grid.  With the grid unchanged but zoomed, one can easily be confused and it is hard to discern the true value of the Horizontal or Vertical.

There is a zoom indication in the Main form Caption when zoomed in or out, as for zooming the grid, the SDK DrawGrid doesn't support zooming the grid so I would have to make my own DrawGrid function or reverse engineer theirs.

Issue 3: Full Screen
This works nice too!.  It took me a bit of trying to realize that right-mouse button give me the option to re-window it.  Perhaps you should put a tiny icon at the corner to re-window.  First few times, I close the app and restart to regain window-ed mode.  After a few tries, I want to see if there is a way and I tried all kinds of keys - eventually got to trying the other mouse button.

A small icon in the corner would be a nice touch.

That's not too hard, and I was thinking maybe binding the escape key to return to windowed mode also.

Suggestion:
Earlier point Mark_O made about when one screen of 260 screens of data is shown, it is hard to see where the visible screen is relative to the start of data.  I pointed out that your mini-wave at top-middle of screen and the T points at exactly the start position of data.

The 'T' actually points to the center of view, and the green area exactly represents the visible area (the waveform in view) in respect to the rest of the data.

Also, check out View->Debug Information... Is that similar to what you are talking about?

[Rick Law]

...
Issue 2: Zoom...
Zoom works really nice - particularly the reset zoom.  It lacks just one thing.  One doesn't know how much is zoomed, and reading the screen can easily confuse one into reading a wrong number.

You need either some magnification indicator (1X, 2.3X) somewhere on screen, and/or resize the grid.  With the grid unchanged but zoomed, one can easily be confused and it is hard to discern the true value of the Horizontal or Vertical.

There is a zoom indication in the Main form Caption when zoomed in or out, as for zooming the grid, the SDK DrawGrid doesn't support zooming the grid so I would have to make my own DrawGrid function or reverse engineer theirs.

Ah...  I totally missed the window frame name changed to have the word zoom and actual scale following - and - the second 1/2 of the 3rd line has the information I am looking for.  If that is on screen, I think it would help prevent misreading the scale.

I don't think the grid needs to be redraw - just some indication that the wave is zoomed.  The window frame is easy to miss.  Perhaps the 1/2 of the 3rd line in debug window can be brought to the wave area?  Or something more "in your face" to make it hard to miss.

...
Suggestion:
Earlier point Mark_O made about when one screen of 260 screens of data is shown, it is hard to see where the visible screen is relative to the start of data.  I pointed out that your mini-wave at top-middle of screen and the T points at exactly the start position of data.

The 'T' actually points to the center of view, and the green area exactly represents the visible area (the waveform in view) in respect to the rest of the data.

Also, check out View->Debug Information... Is that similar to what you are talking about?

The first number of debug line 1 (DataSize) and the first two numbers of debug info line 3 (ViewStart and ViewSize).  That 3 numbers shown as [ ViewSize @ ViewStart/DataSize] to the left or the right of the mini-wave would be great.  If it can't always be shown, I suppose once scrolling starts (ie: no longer at "default view") or when trigger is paused would be good too.  Then as one scroll left or right, one knows exactly where one is relative to the the exported data.   I think this is exactly what would address Mark_O's issue and a feature I would love too!

With the stock software, I can do a little of that by saving the wave form as a reference and as a TXT - To find things with stock, I have to look at the reference graphically and look at the exported text to hunt.  But that is so difficult to use and so troublesome to scroll.  This is so much more superior.

Hey, thanks, Richard.  This is getting nicer and nicer.

[Matchless]

RichardK,
              Thanks, I must have missed that. At the moment if I run Calibration on the Hantek software it executes with no further feedback at all. If I run it on Open 6022be it executes and then gives Cannot calibrate: Calibration failed.
I may have missed something. Is there any specific steps to follow, except having it on for a while and having the probes connected to ground?
Thanks

Matchless ,

I cannot reproduce your issue -- I think you might have forgot to ground the probes first.

Rick

Rick,
      They were grounded as the Calibration software calls for on the screen. I also removed them completely as the Help calls for. Neither method works.
This is quite peculiar I must say!

[Matchless]

Quote

   

Issue 3: Full Screen
    This works nice too!.  It took me a bit of trying to realize that right-mouse button give me the option to re-window it.  Perhaps you should put a tiny icon at the corner to re-window.  First few times, I close the app and restart to regain window-ed mode.  After a few tries, I want to see if there is a way and I tried all kinds of keys - eventually got to trying the other mouse button.

 A small icon in the corner would be a nice touch.

That's not too hard, and I was thinking maybe binding the escape key to return to windowed mode also.

I vote for the Esc key! Its a fairly accepted way.
Thanks!

[Rick Law]

RichardK,
              Thanks, I must have missed that. At the moment if I run Calibration on the Hantek software it executes with no further feedback at all. If I run it on Open 6022be it executes and then gives Cannot calibrate: Calibration failed.
I may have missed something. Is there any specific steps to follow, except having it on for a while and having the probes connected to ground?
Thanks

Matchless ,

I cannot reproduce your issue -- I think you might have forgot to ground the probes first.

Rick

Rick,
      They were grounded as the Calibration software calls for on the screen. I also removed them completely as the Help calls for. Neither method works.
This is quite peculiar I must say!

Geeze - RichardK, you are right.  I totally missed that on Matchless' post.  Sorry, Matchless.

Time to consider an upgrade to my reading glasses.  (Gotta blame it on something else.)

[Matchless]

Geeze - RichardK, you are right.  I totally missed that on Matchless' post.  Sorry, Matchless.

Time to consider an upgrade to my reading glasses.  (Gotta blame it on something else.)

No problem Rick! 

Hopefully my comments/thoughts are useful to Richard in some way. I think I will just leave my noise concern on the back burner for a while, so as not to distract from the project which is more important to us anyway.
It's quite fascinating to watch this roll out!

[Plummerman]

However, if at that time-per-division is fast enough but you capture more than you need, you can trim the data with a quick and dirty program.  If you need more capture data than the scope is capable of (at your needed minimum capture rate), than it would be time for another scope.

Hope this helps.

Rick

Hello Rick,

     Thanks again for taking the time to respond.  Since I'm looking at some pretty low frequency stuff (alternating current of a wind turbine), my time scales will be in the milliseconds range.  I don't know of a spreadsheet program that won't simply choke on 130K to 500K (or more) data points.  As for a "quick and dirty program", I'm not much of a code guru, and attempts at using a formula to lookup and extract a smaller data set simply sent Open Office into la-la land.  What about the screen memory/display manipulations in the HANTEK and/or OpenHANTEK software?  Is every data point being shown in the display, or is there some extrapolation from the sampled data set?  I'm sure hoping there is some way to get a more manageable data chunk out of this thing..., or perhaps, as folks reading this forum are working on the Open HANTEK code, they will think about and implement ways to export manageable data sets? 

[Mark_O]

I don't know of a spreadsheet program that won't simply choke on 130K to 500K (or more) data points.  As for a "quick and dirty program", I'm not much of a code guru, and attempts at using a formula to lookup and extract a smaller data set simply sent Open Office into la-la land.

I get the impression you're looking at CSV files?  If so, and you can use a text-editor, just open the file, select a few thousand lines (up to 60k should be fine), and save them out.  No code guru (or programming) required.

perhaps, as folks reading this forum are working on the Open HANTEK code, they will think about and implement ways to export manageable data sets?

Good suggestion.

[Rick Law]

However, if at that time-per-division is fast enough but you capture more than you need, you can trim the data with a quick and dirty program.  If you need more capture data than the scope is capable of (at your needed minimum capture rate), than it would be time for another scope.

Hope this helps.

Rick

Hello Rick,

     Thanks again for taking the time to respond.  Since I'm looking at some pretty low frequency stuff (alternating current of a wind turbine), my time scales will be in the milliseconds range.  I don't know of a spreadsheet program that won't simply choke on 130K to 500K (or more) data points.  As for a "quick and dirty program", I'm not much of a code guru, and attempts at using a formula to lookup and extract a smaller data set simply sent Open Office into la-la land.  What about the screen memory/display manipulations in the HANTEK and/or OpenHANTEK software?  Is every data point being shown in the display, or is there some extrapolation from the sampled data set?  I'm sure hoping there is some way to get a more manageable data chunk out of this thing..., or perhaps, as folks reading this forum are working on the Open HANTEK code, they will think about and implement ways to export manageable data sets?

Plummerman,

They (scope manufacturers) are more likely to work on exporting larger and larger sets instead of smaller data sets.  Deep-memory is a selling point.

130,000 points is not that many.  You could use Microsoft Word to slides the file into 2 or three, and go from there.  Microsoft Word has the capability to go to specific line of text if I remember right, so you can get to 100,000th line or whatever point you want to slice the file.

If you can get a buddie to write you a script or a macro for the word processor you use, you could speed that up.  A macro (recordable bunch of keystroke) can "delete a line, move forward N lines, repeat".  With such a macro, you can just let it run for a while and it will trim down to 1/N of the original size.

Good Luck!
Rick

[Plummerman]

However, if at that time-per-division is fast enough but you capture more than you need, you can trim the data with a quick and dirty program.  If you need more capture data than the scope is capable of (at your needed minimum capture rate), than it would be time for another scope.

Hope this helps.


Rick

Hello Rick,

     Thanks again for taking the time to respond.  Since I'm looking at some pretty low frequency stuff (alternating current of a wind turbine), my time scales will be in the milliseconds range.  I don't know of a spreadsheet program that won't simply choke on 130K to 500K (or more) data points.  As for a "quick and dirty program", I'm not much of a code guru, and attempts at using a formula to lookup and extract a smaller data set simply sent Open Office into la-la land.  What about the screen memory/display manipulations in the HANTEK and/or OpenHANTEK software?  Is every data point being shown in the display, or is there some extrapolation from the sampled data set?  I'm sure hoping there is some way to get a more manageable data chunk out of this thing..., or perhaps, as folks reading this forum are working on the Open HANTEK code, they will think about and implement ways to export manageable data sets?

Plummerman,

They (scope manufacturers) are more likely to work on exporting larger and larger sets instead of smaller data sets.  Deep-memory is a selling point.

130,000 points is not that many.  You could use Microsoft Word to slides the file into 2 or three, and go from there.  Microsoft Word has the capability to go to specific line of text if I remember right, so you can get to 100,000th line or whatever point you want to slice the file.

If you can get a buddie to write you a script or a macro for the word processor you use, you could speed that up.  A macro (recordable bunch of keystroke) can "delete a line, move forward N lines, repeat".  With such a macro, you can just let it run for a while and it will trim down to 1/N of the original size.

Good Luck!
Rick

Hello Rick,

     I can see a desire for "deep data" in order to catch high speed transients, glitches, and the like; however, an FFT of the trace would yield the highest frequency component.  From there, the Nyquist criteria can be used to determine a maximum useful sample rate, above which, nothing useful is added to having "more data".  A problem with doing this is the resulting variable number of samples with each capture.  My desire is to do repetitive data captures of side by side data in a single file in performing a series of tests for later analysis - rather than having a host of files to deal with.  Excel and OpenOffice Calc choke on just two data channels of 130K data points- and that doesn't even include generating the x-axis data (time of each sample).  Handling a file of 20 or 30, or more such traces would be, well, unthinkable.  I know what a macro is - at one time I wrote quite a few of them in Excel VBA, but that required a "functioning" application...  The Excel  VBA macro recorder helped a lot with my weaknesses at scripting.  As yet, I'm unfamiliar with the Open Office VBA equivalent.

     AGILENT spectrum analyzers employ a user selectable number of data points per trace which may be irrespective of the instrument's sample rate.  This of course, affects the displayed results because of the bandwidth represented for each data point - thus requiring more knowledge and experience on the part of the user for accurate data representations.

      At any rate, I may be able to solve my individual problem with much head banging  , but that doesn't do much for the next guy.  Perhaps young blokes that live to write applications with neat features will benefit from this discussion.  I'm still thinking a bit about some technical stuff - but that's almost from a bygone life now.  The older I get, the less time I want to spend in front of a computer.

Thanks again,

Plummerman