Author Topic: Hantek 6022BE 20MHz USB DSO  (Read 578885 times)

0 Members and 5 Guests are viewing this topic.

Offline RobertH

  • Contributor
  • Posts: 11
  • Country: de
    • monochromatic
Re: Hantek 6022BE 20MHz USB DSO
« Reply #300 on: February 12, 2014, 11:03:44 pm »
Wow, that's a fast response  :-+

It solves the problem to some extend:

Down to 500mV/DIV things are working properly ... with 200mV/DIV and 100mv/DIV actual resolution is 400mv/DIV (that's strange!), 50mV/DIV gives you 200mV/DIV and 20mV/DIV appears to be working correctly again ...

The peak-peak readout is somewhat erratic too:

To make things even worse: when switching to x10 (only the probe, thus reducing the signal to 10%) all results are as they are supposed to be ...

While summing the above things up I did some additional testing. The main result: disregard all after good morning!

Your update fixed my issues! I've reproduced the R11's flaws on two other computers, so there might be some changes in hardware, which led to this behaviour.

The rest seems to be a clipping issue, as there are several clipping conditions (@x1):

a) Vin exceeding 500mV
b) Vin exceeding 5*V/DIV (if Vin is above 500mV)
c) Vin exceeding 5V (due to hardware limitations: internal /10 & 500mV max for a/d converter)

So clipping occurs at the following settings vs. voltages:

V/DIV    Vclip (x1) Vclip(x10) Vclip (x100)
20mV    500mV    n/a            n/a
50mV    500mV    n/a            n/a
100mV  500mV    n/a            n/a
200mV  1V          5V              n/a
500mV  2.5V       5V              n/a
1V         5V          5V              n/a
2V         5V*        10V            50V
5V         5V*        25V            50V
10V       5V*        50V            50V
20V       n/a         50V*         100V
50V       n/a         50V*         250V
100V     n/a         50V*         500V
200V     n/a         n/a            500V*
500V     n/a         n/a            500V*
1kV       n/a         n/a            500V*
(values marked with * are running into hardware limitation)

So, as a conclusion: all settings above 1V times probe attenuation run into the hardware limit and all settings above 100mV times probe attenuation will hit the software limit (assuming it is in fact a software limit).

For AC sources the software limit (again, if it isn't, please correct me!) is quite reasonable, as it exceeds the 8 DIVs on the scope. For DC sources this limit is somewhat low, as it leaves up to 3 DIVs unusable ...

So, as a conclusion: the three highest settings are to some extend obsolete, as they will run into the hardware limit of 5V max @ BNC. If (I know, too many if's) the 5*V/DIV setting is a software limit (with some doubts, as this applies to the mfg's software too), I'd suggest removing it, at least for DC measurement ...

My 6022BE is now decorated with a big sticker: "Max usable range: 5V * probe attenuation"

Best regards and, once again, thanks for the fast respone

Robert
Sorry, I'm not a signature, I'm just tidying up this mess ...
 

Offline Rick Law

  • Super Contributor
  • ***
  • Posts: 2654
  • Country: us
Re: Hantek 6022BE 20MHz USB DSO
« Reply #301 on: February 13, 2014, 03:50:54 am »
Latest Binary: http://jmp.sh/v/pgnwMU9slvz6OIK8YhKr
....

RichardK, great work, I owe you a few.  See below.  Also, there are two new issues - likely just one as they look related.

+1. Added the ability to Import & Export Calibration data to file (Utility->Zero Offset Calibration)
>>   I assume the export/input is to allow zero-reset without having to wait for the system warm up and recal it to what the normal zero-offset would be.   Do you plan to re-import settings at start up?

+2. Added Clear option to all Channel sub-menus which allows you to clear the data on that
   channel to zero.
>> Tested AOK

+3. Added Scroll Bars to Vertical Controls when they are out of view.
for (i=0;i<=1000000;i++ ) { say: This is wonderful!  Works beautifully!  I owe you one.}   In other words, thanks a million plus one!

+4. Fixed a bug when undocking-redocking the Toolbar which would cause Scope display to flicker,  and or reduced performance.
>> Fixed work, I can no longer trigger that flicker playing around with toolbar.

+5. Fixed a bug where Math and Reference Cursor Measurements would be incorrect in certain
   situations.
>> Math cursor is very stable, tested under different volts and freq.  See two other issues below with math trace.

6. Fixed several form resizing flicker issues.
>> Rock stable!

7. Modified the Measure Output window to have a horizontal layout (Separate Tree View for each   Measurement Channel) which better utilizes the available horizontal space.
for (i=1;i<=1000000;i++ ) { say: Wonderful!  This is sweet.}

8. Added preliminary support for a Probe Compensation Auto Setup (See Auto Button on Toolbar).
>> not tested, can't find the silly little screw driver

9. Fixed a few minor Measurement bugs.
>> So far, so good, I have not seen any issues.

10. Renamed some Measurement items in the Measurement Window (Positive/Negative now in parentheses
   after the Measurement name).
>> Works, no issue

11. Swapped positions of "Toggle Controls" & "Toggle Measure" buttons on the Toolbar
>> Wonderful

12. Added little 'X' buttons to Measure, Output and Control Windows which hides them when clicked.
>> very nice touch

13. Fixed a bug where resizing the form would not rescale the Channel Waveform Position Levers.

14. Made Zooming In and Out more granular (Halved zoom-rate)
>> Tested AOK.

15. Fixed a bug where Voltage Division might not properly set for CH1 & CH2 in certain situations.

RobertH: Hopefully #15 fixes your issues.
....

New Issues 1:
1. Math is changing where it should not be (setting dependent, described below).

When both Ch1 and Ch2 are measuring the same thing, the two waves match as expected.  Math trace (A+B) should look exactly like  Ch1 and Ch2 with exact same frequency, different ampltitue, and same 0 crossing.  By window resize and mouse movement describe below, it could get into a mode where it shifts left/right a little (that is what caught my attention).  The math trace begin rolling left and right and keep rolling.  Once that starts, it can get to a point the rolling worsens to like an untriggered wave form. 

Settings:
Both Ch1 and Ch2 hooked to the same spot, 100KHz, Square,  +-3.5V peak to peak, math=A+B
Set to default 2V/div, 5uS.  Leave trigger at program start up default.
Note:Square, Sine, doesn't matter, v/div doesn't matter, just frequency and time/div matters.  I know I don't get it at the low 1KHz reference wave. I am not sure at what frequency it starts. At 100KHz (or 90KHz), I know the bug will hit.

Steps to trigger the uncontrolled scrolling of math waveform:  (Next issue related to this, don't change a thing yet after you reproduce this bug)
1. I started with fresh reboot, signal all set up (100KHz, Square, +-3.5v)
2. Start the program, change time/div to 5uS, math A+B
3. Size the window smallest (both verticle and horizontal) and move it to top left for sreen-space.  LET GO of the button before next step.
4. Enlarge the window both verticle and horizontal about 1.5x size
5. Start the mouse in the trace-drawing area and move the mouse left and right very very rapidly about 10 times.(mouse movement within trace area, moved perhaps 1/2 trace area width)
6. It may start after a few (<5) such rapid mouse movement left and right, if not go back to step 3.  I can usually get the trace scrolling  going back to step 3 less than 5 times.
7. Once you get it to roll, move moment appears irrelevant.  Sizing window smaller stops the roll, sizing it big starts it again.  It became uncontrollable after a few small-large cycle..
8. Do not change a thing yet.  Look at the next issue.

New Issues 2: probably related to Issue 1

(Once you hit issue 1) Trigger is default setting.  Top right display of trigger voltage update delayed (by > 5 sec) Moving the update trigger above wave, the trigger voltage should update.  The new display of trigger voltage you added on lower left updates immediately. The trigger voltage on top right (stock software trigger voltage display) does not update immediately.  The more you go "resize-small (Vert&Horiz), release button, size big (V&H)," the slower the the update gets.  I have waited for >10 seconds for it to update.  (The trigger voltage change can be within or beyond wave-voltage.  I used the trigger pointer tab to change trigger voltage in case there is another way I don't know)

With your added display on lower right already showing it immediately, the delayed stock-display of trigger voltage will merely confuse the user.  If this is difficult to fix, perhaps just don't display the stock trigger voltage update.
« Last Edit: February 13, 2014, 04:01:28 am by Rick Law »
 

Offline Rick Law

  • Super Contributor
  • ***
  • Posts: 2654
  • Country: us
Re: Hantek 6022BE 20MHz USB DSO
« Reply #302 on: February 13, 2014, 04:02:45 am »
Wow, that's a fast response  :-+

It solves the problem to some extend:

Down to 500mV/DIV things are working properly ... with 200mV/DIV and 100mv/DIV actual resolution is 400mv/DIV (that's strange!), 50mV/DIV gives you 200mV/DIV and 20mV/DIV appears to be working correctly again ...

The peak-peak readout is somewhat erratic too:
...
Robert

Robert, I will give it a try to reproduce your issue, see if I can somehow get it to replicate on my system the next couple of days.  I will keep you posted.

Rick
 

Offline Mark_O

  • Frequent Contributor
  • **
  • Posts: 937
  • Country: us
Re: Hantek 6022BE 20MHz USB DSO
« Reply #303 on: February 13, 2014, 05:42:14 am »
With slower settings (i.e. 500ms/DIV) I get an updated screen every 5 seconds. Is that the standard output?

Hi, Robert.

If you multiply it out, 500ms/div x 10 divs = 5 sec.  It has to capture the data, before it can display it.  ;)

Quote
If yes: I'd love to have a "continuous drawing mode" - either real-time samples on the right edge of the scope and the plot moving to the left of, if more feasible, continuous drawing of samples from left to right with "overwriting" on the next pass.

I can see where that would be handy.  However, it would be difficult to do.  The reason being that the interface with the Hantek DLL sets up the acquisition parameters, then the driver funnels the data directly to the memory buffers.  There's no incremental interactivity (query/response, or periodic notify callbacks).

The reason I say difficult, and not the expected impossible, is because it would be possible to prefill the buffer (say, with FF's).  And have a second task running that monitored the high-water mark (as the FF's got overridden by incoming data).  Then update the display incrementally.  I'm not sure that's something Richard would want to tackle at this stage of the development.

Of the two display methods you suggested, the fill-then-overwrite would be lower overhead, than trying to blit a chunk of screen space while scrolling the whole area (on top of the grid pattern).  Some or all of this may be completely impossible, using the DisplayDLL that Hantek has provided.  I'm not familiar with it's constraints, but I doubt it's overly fancy.  So it may effectively be wishful thinking.
 

Offline Matchless

  • Regular Contributor
  • *
  • Posts: 119
  • Country: za
Re: Hantek 6022BE 20MHz USB DSO
« Reply #304 on: February 13, 2014, 07:24:38 am »
What an improvement!!!

RichardK, just one small thing I picked up when running the Autoset for scope compensation adjustment, the V/Div goes to 100mV on the OEM Hantek and Open6022be goes to 1.00V.

Saying you are doing excellent work is an understatement!!!
Regards
Matchless
 

Offline RobertH

  • Contributor
  • Posts: 11
  • Country: de
    • monochromatic
Re: Hantek 6022BE 20MHz USB DSO
« Reply #305 on: February 13, 2014, 09:38:21 am »
Wow, that's a fast response  :-+

It solves the problem to some extend:

Down to 500mV/DIV things are working properly ... with 200mV/DIV and 100mv/DIV actual resolution is 400mv/DIV (that's strange!), 50mV/DIV gives you 200mV/DIV and 20mV/DIV appears to be working correctly again ...

The peak-peak readout is somewhat erratic too:
...
Robert

Robert, I will give it a try to reproduce your issue, see if I can somehow get it to replicate on my system the next couple of days.  I will keep you posted.

Rick

Rick,

don't waste the time - that was clipping that hit me  |O

Robert
Sorry, I'm not a signature, I'm just tidying up this mess ...
 

Offline Rick Law

  • Super Contributor
  • ***
  • Posts: 2654
  • Country: us
Re: Hantek 6022BE 20MHz USB DSO
« Reply #306 on: February 13, 2014, 04:25:52 pm »
Wow, that's a fast response  :-+

It solves the problem to some extend:

Down to 500mV/DIV things are working properly ... with 200mV/DIV and 100mv/DIV actual resolution is 400mv/DIV (that's strange!), 50mV/DIV gives you 200mV/DIV and 20mV/DIV appears to be working correctly again ...

The peak-peak readout is somewhat erratic too:
...
Robert

Robert, I will give it a try to reproduce your issue, see if I can somehow get it to replicate on my system the next couple of days.  I will keep you posted.

Rick

Rick,

don't waste the time - that was clipping that hit me  |O

Robert

OK, good you found the problem.
 

Offline RichardK

  • Regular Contributor
  • *
  • Posts: 157
Re: Hantek 6022BE 20MHz USB DSO
« Reply #307 on: February 13, 2014, 05:01:45 pm »
What an improvement!!!

RichardK, just one small thing I picked up when running the Autoset for scope compensation adjustment, the V/Div goes to 100mV on the OEM Hantek and Open6022be goes to 1.00V.

Saying you are doing excellent work is an understatement!!!

It varies from scope to scope because the stock software is scanning horizontal and vertical scale and looking for some condition that it finds suitable. I'm not quite sure what exactly that condition is yet, so the Auto button is "dumb" right now and when I am able to figure out what it is exactly they are doing I'll make it do the same.
 

Offline RichardK

  • Regular Contributor
  • *
  • Posts: 157
Re: Hantek 6022BE 20MHz USB DSO
« Reply #308 on: February 13, 2014, 05:16:07 pm »

+1. Added the ability to Import & Export Calibration data to file (Utility->Zero Offset Calibration)
>>   I assume the export/input is to allow zero-reset without having to wait for the system warm up and recall it to what the normal zero-offset would be.   Do you plan to re-import settings at start up?

It was easy to implement and I figured someone might find a use for it so why not... I don't plan on adding any start-up import, I am already grabbing the stored calibration data from the scope at start-up, so I'll leave the usage of this feature to the imagination of the end-user.

Quote
New Issues 1:
1. Math is changing where it should not be (setting dependent, described below).

When both Ch1 and Ch2 are measuring the same thing, the two waves match as expected.  Math trace (A+B) should look exactly like  Ch1 and Ch2 with exact same frequency, different ampltitue, and same 0 crossing.  By window resize and mouse movement describe below, it could get into a mode where it shifts left/right a little (that is what caught my attention).  The math trace begin rolling left and right and keep rolling.  Once that starts, it can get to a point the rolling worsens to like an untriggered wave form.

Not sure what could be causing this but I'll be investigating it today.

Quote
New Issues 2: probably related to Issue 1

(Once you hit issue 1) Trigger is default setting.  Top right display of trigger voltage update delayed (by > 5 sec) Moving the update trigger above wave, the trigger voltage should update.  The new display of trigger voltage you added on lower left updates immediately. The trigger voltage on top right (stock software trigger voltage display) does not update immediately.  The more you go "resize-small (Vert&Horiz), release button, size big (V&H)," the slower the the update gets.  I have waited for >10 seconds for it to update.  (The trigger voltage change can be within or beyond wave-voltage.  I used the trigger pointer tab to change trigger voltage in case there is another way I don't know)

Was able to slightly reproduce this and it might have been caused by that portion's draw event having a much lower priority and when the draw queue is choked full of draw wave events (which take some time) it tends to sit on the back burner for a while. Since rendering the trigger level takes very little time I gave that particular draw event a higher priority than draw wave events and it's much more snappy now.
 

Offline RichardK

  • Regular Contributor
  • *
  • Posts: 157
Re: Hantek 6022BE 20MHz USB DSO
« Reply #309 on: February 14, 2014, 04:42:17 am »
I was going to skip the 13th build, but it's the 13th of February and I fixed a few issues, so I might as well get it out of the way... Lucky Build #13: http://jmp.sh/r4BjH8a

What's New:
-----------------------------------------------------------
1. Various Performance enhancements.
2. Reduced tendency for Math channel to roll.
3. Implemented FFT Fundamental frequency rendering.
4. Fixed a glitch in FFT Rendering.

Not a whole lot changed, but this should clear up the latest issues and get #13 out of the way *knock on wood*.
« Last Edit: February 14, 2014, 02:18:00 pm by RichardK »
 

Offline Mark_O

  • Frequent Contributor
  • **
  • Posts: 937
  • Country: us
Re: Hantek 6022BE 20MHz USB DSO
« Reply #310 on: February 14, 2014, 05:01:42 am »
I was going to skip the 13th build, but it's the 13th of February and I fixed a few issues, so I might as well get it out of the way... Lucky Build #13: http://jmp.sh/JEjvatW

Hmm.  "File Not Found."  That string looks suspiciously short?  Maybe it's just me.
 

Offline Rick Law

  • Super Contributor
  • ***
  • Posts: 2654
  • Country: us
Re: Hantek 6022BE 20MHz USB DSO
« Reply #311 on: February 14, 2014, 05:12:25 am »
I was going to skip the 13th build, but it's the 13th of February and I fixed a few issues, so I might as well get it out of the way... Lucky Build #13: http://jmp.sh/JEjvatW

Hmm.  "File Not Found."  That string looks suspiciously short?  Maybe it's just me.

I think perhaps it is a site problem.  I saw this PR13 post from RichardK a few hours ago on my laptop.  Approx 1 hr ago,  I got on my PC to download - and I did not find this PR13 post from RichardK.  I was beginning to wonder if I was imagining it or perhaps he changed his mind and removed it somehow.  I came back to my laptop after having some snack, this post from RichardK reappeared.

I was really wondering what was in my dinner that caused such imagination.  I am glad I see your post.  That tells me there must be something wrong on the site or Google Chrome cache is not working right showing me an old page.

Well, I am sure Richard K will post a new link again shortly.

Rick
 

Offline RichardK

  • Regular Contributor
  • *
  • Posts: 157
Re: Hantek 6022BE 20MHz USB DSO
« Reply #312 on: February 14, 2014, 02:16:30 pm »
Sorry about that guys, Jumpshare gave me the wrong URL...  :-//

http://jmp.sh/r4BjH8a
 

Offline RichardK

  • Regular Contributor
  • *
  • Posts: 157
Re: Hantek 6022BE 20MHz USB DSO
« Reply #313 on: February 14, 2014, 02:22:21 pm »
I think perhaps it is a site problem.  I saw this PR13 post from RichardK a few hours ago on my laptop.  Approx 1 hr ago,  I got on my PC to download - and I did not find this PR13 post from RichardK.  I was beginning to wonder if I was imagining it or perhaps he changed his mind and removed it somehow.  I came back to my laptop after having some snack, this post from RichardK reappeared.

I was really wondering what was in my dinner that caused such imagination.  I am glad I see your post.  That tells me there must be something wrong on the site or Google Chrome cache is not working right showing me an old page.

Both, I found another issue and noticed nobody had Downloaded it yet so I removed it and rebuilt it, reuploaded with new fix and one other new feature...

But... Lucky #13 strikes back and Jumpshare gives me a dud URL (it was /invalid or something) when I'm making the post, so I copy it again and finally get a normal looking one (but I didn't test it, DOH!) and of course it's a DUD too  :palm:

Well time to un-bury my house, friggen snow  :rant:
 

Offline Rick Law

  • Super Contributor
  • ***
  • Posts: 2654
  • Country: us
Re: Hantek 6022BE 20MHz USB DSO
« Reply #314 on: February 15, 2014, 02:25:05 am »
I was going to skip the 13th build, but it's the 13th of February and I fixed a few issues, so I might as well get it out of the way... Lucky Build #13: http://jmp.sh/r4BjH8a

What's New:
-----------------------------------------------------------
1. Various Performance enhancements.
2. Reduced tendency for Math channel to roll.
3. Implemented FFT Fundamental frequency rendering.
4. Fixed a glitch in FFT Rendering.

Not a whole lot changed, but this should clear up the latest issues and get #13 out of the way *knock on wood*.

1. Various Performance enhancements.
It does seem more responsive now

2. Reduced tendency for Math channel to roll.
Not fully fixed.  Issue#1 (recap below) still triggers a roll.  When I posted originally, I said 100KHz down to 90 KHz I can get it to roll.  I experimented more and I get it to roll at 10KHz at 5uS/division.  (A,B at the same spot on signal gen, Square wave, +- 2 to 2.5V).  The rolling stop at slower time/division.  At 50uS, I did not get it to roll.
Even as slow as 5KHz, I can also see rolling.  But 5KHz at 5uS/div, so little of the wave is visible determining rolling (or not) becomes difficult.
Unrelated Suggestion
Doing tests makes me switch time/div a lot and I notice the drop down shows just a few choices (7+current selection), so I ended up scrolling around a lot.  This may be an inconvenience during normal use as well.   Since that drop down is so high up on the screen, offering more choices will not impact screen real-estate.  Can you make the drop down a little longer so it show more time/div choices?  Only if it is easy - this one is polishing.  Offer more choices for this drop down menu may make it more user-friendly.  It would also look more polished to get rid of some of the digits after the decimal since they are always zero anyway.

3. Implemented FFT Fundamental frequency rendering.
4. Fixed a glitch in FFT Rendering.
Not sure how to test 3&4.  So I did not test these two much.

Issue#2 is certainly fixed.  The stock trigger voltage display now updates immediately.  The conflicting trigger-voltage display no longer exist to confuse the user.

Recap of Issue #1 so you don't have to hunt down the original post down reproduction info:
New Issues 1:
1. Math is changing where it should not be (setting dependent, described below).

When both Ch1 and Ch2 are measuring the same thing, the two waves match as expected.  Math trace (A+B) should look exactly like  Ch1 and Ch2 with exact same frequency, different ampltitue, and same 0 crossing.  By window resize and mouse movement describe below, it could get into a mode where it shifts left/right a little (that is what caught my attention).  The math trace begin rolling left and right and keep rolling.  Once that starts, it can get to a point the rolling worsens to like an untriggered wave form. 

Settings:
Both Ch1 and Ch2 hooked to the same spot, 100KHz, Square,  +-3.5V peak to peak, math=A+B
Set to default 2V/div, 5uS.  Leave trigger at program start up default.
Note:Square, Sine, doesn't matter, v/div doesn't matter, just frequency and time/div matters.  I know I don't get it at the low 1KHz reference wave. I am not sure at what frequency it starts. At 100KHz (or 90KHz), I know the bug will hit.

Steps to trigger the uncontrolled scrolling of math waveform:  (Next issue related to this, don't change a thing yet after you reproduce this bug)
1. I started with fresh reboot, signal all set up (100KHz, Square, +-3.5v)
2. Start the program, change time/div to 5uS, math A+B
3. Size the window smallest (both verticle and horizontal) and move it to top left for sreen-space.  LET GO of the button before next step.
4. Enlarge the window both verticle and horizontal about 1.5x size
5. Start the mouse in the trace-drawing area and move the mouse left and right very very rapidly about 10 times.(mouse movement within trace area, moved perhaps 1/2 trace area width)
6. It may start after a few (<5) such rapid mouse movement left and right, if not go back to step 3.  I can usually get the trace scrolling  going back to step 3 less than 5 times.
7. Once you get it to roll, move moment appears irrelevant.  Sizing window smaller stops the roll, sizing it big starts it again.  It became uncontrollable after a few small-large cycle..
8. Do not change a thing yet.  Look at the next issue.

New Issues 2: probably related to Issue 1

(Once you hit issue 1) Trigger is default setting.  Top right display of trigger voltage update delayed (by > 5 sec) Moving the update trigger above wave, the trigger voltage should update. 

Thanks again!
Rick
« Last Edit: February 15, 2014, 02:28:47 am by Rick Law »
 

Offline RichardK

  • Regular Contributor
  • *
  • Posts: 157
Re: Hantek 6022BE 20MHz USB DSO
« Reply #315 on: February 15, 2014, 03:14:05 am »
2. Reduced tendency for Math channel to roll.
Not fully fixed.  Issue#1 (recap below) still triggers a roll.  When I posted originally, I said 100KHz down to 90 KHz I can get it to roll.  I experimented more and I get it to roll at 10KHz at 5uS/division.  (A,B at the same spot on signal gen, Square wave, +- 2 to 2.5V).  The rolling stop at slower time/division.  At 50uS, I did not get it to roll.
Even as slow as 5KHz, I can also see rolling.  But 5KHz at 5uS/div, so little of the wave is visible determining rolling (or not) becomes difficult.

I think the rolling is caused by the fact that the Math function takes some time to execute and by the time it's rendered the source waveforms are at a slightly different time, and with the noise the trigger could be in another location, thus causing the offset.

I optimized it as much as I can, so I think in order to squash this issue I'll have to find another approach, possibly compute the math as soon as possible after the raw data is acquired, although I am already doing it pretty much right after CH1 & CH2 are acquired.

Quote
Unrelated Suggestion
Doing tests makes me switch time/div a lot and I notice the drop down shows just a few choices (7+current selection), so I ended up scrolling around a lot.  This may be an inconvenience during normal use as well.   Since that drop down is so high up on the screen, offering more choices will not impact screen real-estate.  Can you make the drop down a little longer so it show more time/div choices?  Only if it is easy - this one is polishing.  Offer more choices for this drop down menu may make it more user-friendly.  It would also look more polished to get rid of some of the digits after the decimal since they are always zero anyway.

Good idea, will be in the next release.
 

Offline Rick Law

  • Super Contributor
  • ***
  • Posts: 2654
  • Country: us
Re: Hantek 6022BE 20MHz USB DSO
« Reply #316 on: February 15, 2014, 03:39:20 am »
2. Reduced tendency for Math channel to roll.
Not fully fixed.  Issue#1 (recap below) still triggers a roll.  When I posted originally, I said 100KHz down to 90 KHz I can get it to roll.  I experimented more and I get it to roll at 10KHz at 5uS/division.  (A,B at the same spot on signal gen, Square wave, +- 2 to 2.5V).  The rolling stop at slower time/division.  At 50uS, I did not get it to roll.
Even as slow as 5KHz, I can also see rolling.  But 5KHz at 5uS/div, so little of the wave is visible determining rolling (or not) becomes difficult.

I think the rolling is caused by the fact that the Math function takes some time to execute and by the time it's rendered the source waveforms are at a slightly different time, and with the noise the trigger could be in another location, thus causing the offset.

I optimized it as much as I can, so I think in order to squash this issue I'll have to find another approach, possibly compute the math as soon as possible after the raw data is acquired, although I am already doing it pretty much right after CH1 & CH2 are acquired.


I think this is a minor issue - not one that should hold up release since your release represents so many important improvements.  It could be documented in the release write-up that a fast CPU is needed for proper rendering of math function @ faster than 5uS/division.  But of course, it is your call.
 

Offline RichardK

  • Regular Contributor
  • *
  • Posts: 157
Re: Hantek 6022BE 20MHz USB DSO
« Reply #317 on: February 15, 2014, 04:22:24 am »
2. Reduced tendency for Math channel to roll.
Not fully fixed.  Issue#1 (recap below) still triggers a roll.  When I posted originally, I said 100KHz down to 90 KHz I can get it to roll.  I experimented more and I get it to roll at 10KHz at 5uS/division.  (A,B at the same spot on signal gen, Square wave, +- 2 to 2.5V).  The rolling stop at slower time/division.  At 50uS, I did not get it to roll.
Even as slow as 5KHz, I can also see rolling.  But 5KHz at 5uS/div, so little of the wave is visible determining rolling (or not) becomes difficult.

I think the rolling is caused by the fact that the Math function takes some time to execute and by the time it's rendered the source waveforms are at a slightly different time, and with the noise the trigger could be in another location, thus causing the offset.

I optimized it as much as I can, so I think in order to squash this issue I'll have to find another approach, possibly compute the math as soon as possible after the raw data is acquired, although I am already doing it pretty much right after CH1 & CH2 are acquired.


I think this is a minor issue - not one that should hold up release since your release represents so many important improvements.  It could be documented in the release write-up that a fast CPU is needed for proper rendering of math function @ faster than 5uS/division.  But of course, it is your call.

Now that I have looked at the code some more, it might also be caused by the fact that Data Capture is in one thread and Wave Drawing in another, and when the CH1 & CH2 waves are being drawn, a new Data Capture may have occurred in the other thread, BUT before the latest Math has been calculated. So when Math Channel is finally rendered, it's old Math data rendered with new Channel data, and if the new channel data has a different trigger position... Well you know what happens :)

In the next release I'll have a change in the trigger code that won't affect any situation where Channel Data and Math Data are in sync, BUT if they are not in sync it should correct it. 

I also noticed the Interactive Cursors bouncing off the Math trace, and I have just fixed this also.

Edit: Think I found a solution that works! I just had to create a boolean that would delay a raw data capture until the Math channel was rendered... The Math Channel will still shudder when dragging, but that's due to extra Draw Wave events being generated... But if you are just observing, even down to 5uS/DIV the Math channel is lock-step with CH1 & CH2.
« Last Edit: February 15, 2014, 06:08:53 am by RichardK »
 

Offline Matchless

  • Regular Contributor
  • *
  • Posts: 119
  • Country: za
Re: Hantek 6022BE 20MHz USB DSO
« Reply #318 on: February 15, 2014, 08:57:11 am »
RichardK,
               I found a slightly more descriptive note on the function of the Autoset button in the DSO1000 series user manual:
It seems as if its primary function is to scan all the inputs and then automatically adjust the settings for the best display on the screen when the button is pressed. I could get this to work like this on the Hantek6022be 1.0.4 software.

The same function is then just used to properly display the built in 2V P-P wave when used to adjust the probe compensation.

Open6022be does not do this at present as I think you did mention.

Extract from user manual below:


Quote
To compensate probes
Perform this adjustment to match the characteristics of the probe and the channel input. This should be performed whenever attaching a probe to any input channel the first time.
1. From CH1 menu, set the probe attenuation to 10X (press CH1?Probe?10X).
Set the switch to X10 on the probe and connect it to CH1 of the oscilloscope.
When using the probe hook-tip, insert the tip onto the probe firmly to ensure proper connection.
Attach the probe tip to the probe compensator connector and the reference lead to the ground pin, Select CH1, and then press AUTO.
2. Check the shape of the displayed waveform
3. If necessary, use a non-metallic tool to adjust the trimmer capacitor of the probe for the flattest square wave being displayed on the oscilloscope.
4. Repeat if necessary.
WARNNING: To avoid electric shock while using the probe, be sure the perfection of the insulated cable, and do not touch the metallic portions of the probe head while it is connected with a voltage source

To display a signal automatically
The oscilloscope has an Auto feature that automatically sets up the Oscilloscope to best display the input signal. Using Auto requires signals with a frequency greater than or equal to 50 Hz and a duty cycle greater than 1%. Press the AUTO button, the oscilloscope turns on and scales all channels that have signals applied, and it selects a time base range based on the trigger source.
The trigger source selected is the lowest numbered input that has a signal applied. The oscilloscope has two-channels input. Connect a signal to the CH1 input.
1.Connect a signal to the oscilloscope.
2.Press AUTO. The oscilloscope may change the current setting to display the signal. It will automatically adjust the vertical and horizontal scaling, the trigger coupling, type, position, slope, level, and mode settings.
Regards
Matchless
 

Offline Matchless

  • Regular Contributor
  • *
  • Posts: 119
  • Country: za
Re: Hantek 6022BE 20MHz USB DSO
« Reply #319 on: February 15, 2014, 04:48:23 pm »
Finally sat down and started editing the user manual in readiness for all the changes and improvements Richardk is adding. This is basically the original user manual with the changes:

Edit taken down: It was not showing a virus from this side, but others were getting warnings! See new post about 9 posts further on and Jumpshare new link.

I have just done a first edit up to page 12. Mostly this can really only be done when the software is complete, but starting now may be useful to somebody. If anyone wants to help with this please just enter in red or PM me the new text or upload somewhere and I will update the master as we go on.
« Last Edit: February 15, 2014, 07:44:59 pm by Matchless »
Regards
Matchless
 

Online Fraser

  • Super Contributor
  • ***
  • Posts: 8412
  • Country: gb
Re: Hantek 6022BE 20MHz USB DSO
« Reply #320 on: February 15, 2014, 05:40:33 pm »
Warning !

That download site just tried to infect my PC with a Trojan. I have rebuilt my PC today and the first download was this one. Be careful. I will attach a screen shot from MS Secuirty Essentials.

« Last Edit: February 15, 2014, 05:48:58 pm by Aurora »
 

Offline RichardK

  • Regular Contributor
  • *
  • Posts: 157
Re: Hantek 6022BE 20MHz USB DSO
« Reply #321 on: February 15, 2014, 05:44:12 pm »
Warning ![/size

That download site just tried to infect my PC with a Trojan. I have rebuilt my PC today and teh first download was this one. Be careful. I will attach a screen shot from MS Secuirty Essentials.

That's Media Fire for you, they will let anyone advertise on their site and they don't screen for malicious advertisements.
 

Offline Rick Law

  • Super Contributor
  • ***
  • Posts: 2654
  • Country: us
Re: Hantek 6022BE 20MHz USB DSO
« Reply #322 on: February 15, 2014, 05:46:17 pm »
Warning ![/size

That download site just tried to infect my PC with a Trojan. I have rebuilt my PC today and teh first download was this one. Be careful. I will attach a screen shot from MS Secuirty Essentials.

Please clarify - are you are referring to the Document Mr. Matchless put on mediafire.com? or are you referring to the RichardK's binary download site?

My desktop also gone to hell (hung during Windows start up) this morning and I had a hard time getting on.  I plan to restore my machine after lunch too.
 

Offline RichardK

  • Regular Contributor
  • *
  • Posts: 157
Re: Hantek 6022BE 20MHz USB DSO
« Reply #323 on: February 15, 2014, 05:47:25 pm »
My desktop also gone to hell (hung during Windows start up) this morning and I had a hard time getting on.  I plan to restore my machine after lunch too.

Sounds like a job for CCleaner.
 

Online Fraser

  • Super Contributor
  • ***
  • Posts: 8412
  • Country: gb
Re: Hantek 6022BE 20MHz USB DSO
« Reply #324 on: February 15, 2014, 05:53:31 pm »
To be clear, the SITE that matchless has uploaded the edited manual to is the one with Virus issues.

I selected the manual download and it gave a file not found error and my AV went off saying that it was defending against an attack. The manual file never got downloaded, so the virus Payload came from the site when I pressed the 'Download' button (next to the manual details). NOT an advert or one of those cr*p bogus download buttons often found on sites.
« Last Edit: February 15, 2014, 05:55:13 pm by Aurora »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf