Hantek 6022BE 20MHz USB DSO

[Rick Law]

Richard,

The Open 6022 front end you wrote is quite an accomplishment.  It has a nice feel to it.  Allow me to point out a few things where I found trying:

1. The start up seems to have a bit of delay.  A few times, I re-clicked and ended up starting multiple instances.

2. The interactive measurement is quite nice.  It however escapes me on how to catch the peak, min, or max – points that I cannot exactly put a cursor on.

3. Wave form for Ch2 is weird.  I have Ch1 on a TTL square wave and Ch2 on the output part of the circuit which transform the TTL square wave transformed to Sine wave.  The two waves are same frequency and phase.  With the native software, the two wave forms are locked together as they should (since one is input and one is output of the same wave source).  With Open6022 however, the Ch2 wave keeps rolling left/right as if it is not related to the Ch1 wave.  I tried from low KHz to 4MHz using the UB1308S function generator.

4. I was confused for a while as only one set of vertical control is visible only.  Until I found the drop down to change from Ch1 to Ch2, I was for a moment very confused by not able to deal with Ch2.  It was rather inconvenient to drop down to switch to control “the other channel” and back again.  I think keeping both Ch1 and Ch2 vertical control knobs visible would be nice.  Perhaps use the same check box: when wave for Ch? Is shown, the vertical controls are also shown.

I particularly like these new abilities you provided:
1. AC couple via software, love this.  Too many circuits at “points of interest”, I had an extra capacitor to no where merely to provide a hook up point for the scope so as I can see how things are working.  With this AC coupling, it makes taping into any point less cumbersome.
2. Change color – something so simple yet it can make the software seem so much more friendly.
3. Looks like the top “mini wave window” shows what part of the captured data is being displayed as I drag the wave forms left and right.  I like that.  It does not seem fully functional yet since the T like pointer doesn’t seem to move consistently with my dragging.  If my gut feel is right and that it is suppose to point at what part of the captured wave is being displayed, it would add so much and make the software so much more “user friendly.”

Thanks, Richard.  Thanks for your hard work – we are benefited by it.

Rick

[RichardK]

1. The start up seems to have a bit of delay.  A few times, I re-clicked and ended up starting multiple instances.

Fixed (I hope) in latest binary (see below)...

2. The interactive measurement is quite nice.  It however escapes me on how to catch the peak, min, or max – points that I cannot exactly put a cursor on.

It's harder to get it exactly when at lower resolutions (many waveforms in view), the original cursor measurement feature is probably better for finer measurements.

3. Wave form for Ch2 is weird.  I have Ch1 on a TTL square wave and Ch2 on the output part of the circuit which transform the TTL square wave transformed to Sine wave.  The two waves are same frequency and phase.  With the native software, the two wave forms are locked together as they should (since one is input and one is output of the same wave source).  With Open6022 however, the Ch2 wave keeps rolling left/right as if it is not related to the Ch1 wave.  I tried from low KHz to 4MHz using the UB1308S function generator.

Fixed in latest binary (see below)...

4. I was confused for a while as only one set of vertical control is visible only.  Until I found the drop down to change from Ch1 to Ch2, I was for a moment very confused by not able to deal with Ch2.  It was rather inconvenient to drop down to switch to control “the other channel” and back again.  I think keeping both Ch1 and Ch2 vertical control knobs visible would be nice.  Perhaps use the same check box: when wave for Ch? Is shown, the vertical controls are also shown.

The reason I did this was because I felt the original software interface was too busy, and I was sick of scrolling down to modify CH2, so I figured using the arrow key to select channels would be a little better than trying to cram them all in. I'll look into making it easier to interact with CH1 & CH2 without using the drop down list.

It does not seem fully functional yet since the T like pointer doesn’t seem to move consistently with my dragging.  If my gut feel is right and that it is suppose to point at what part of the captured wave is being displayed, it would add so much and make the software so much more “user friendly.”

I think you are using an older binary, because I optimized the wave context movement recently, it's smooth as butter now even in a debug build... Here is the very latest binary: http://jmp.sh/v/MWcYkfQZ5t1qRzXBWIZD

[Rick Law]

1. The start up seems to have a bit of delay.  A few times, I re-clicked and ended up starting multiple instances.

Fixed (I hope) in latest binary (see below)...

Fix confirmed!  The new binary starts right the way.

2. The interactive measurement is quite nice.  It however escapes me on how to catch the peak, min, or max – points that I cannot exactly put a cursor on.

It's harder to get it exactly when at lower resolutions (many waveforms in view), the original cursor measurement feature is probably better for finer measurements.

Yeah, I think the non-cursor measurement window needs to come back as well.  Otherwise, it would be hard to measure something that is not always there to put a cusor on - such as trying to catch the peak of an unknown/irregular pulse that is not on the current trace.

3. Wave form for Ch2 is weird.  I have Ch1 on a TTL square wave and Ch2 on the output part of the circuit which transform the TTL square wave transformed to Sine wave.  The two waves are same frequency and phase.  With the native software, the two wave forms are locked together as they should (since one is input and one is output of the same wave source).  With Open6022 however, the Ch2 wave keeps rolling left/right as if it is not related to the Ch1 wave.  I tried from low KHz to 4MHz using the UB1308S function generator.

Fixed in latest binary (see below)...

Fix confirmed!  I used the same wave and Ch2 is plotted exactly like Ch1.  I put Ch2 into a frequency that is a multiple of Ch1 and the wave appears exactly as expected.  Two cycles with Ch2 lined up with the single cycle on Ch1.  Ch2 is no longer rolling wild.

4. I was confused for a while as only one set of vertical control is visible only.  Until I found the drop down to change from Ch1 to Ch2, I was for a moment very confused by not able to deal with Ch2.  It was rather inconvenient to drop down to switch to control “the other channel” and back again.  I think keeping both Ch1 and Ch2 vertical control knobs visible would be nice.  Perhaps use the same check box: when wave for Ch? Is shown, the vertical controls are also shown.

The reason I did this was because I felt the original software interface was too busy, and I was sick of scrolling down to modify CH2, so I figured using the arrow key to select channels would be a little better than trying to cram them all in. I'll look into making it easier to interact with CH1 & CH2 without using the drop down list.

I understand - I agree with the "too busy look" and agree with your logic.  Perhaps make it shows only when a channel is enabled.  Or may be forget the "knobs" to change scale but just use drop down menus?

It does not seem fully functional yet since the T like pointer doesn’t seem to move consistently with my dragging.  If my gut feel is right and that it is suppose to point at what part of the captured wave is being displayed, it would add so much and make the software so much more “user friendly.”

I think you are using an older binary, because I optimized the wave context movement recently, it's smooth as butter now even in a debug build... Here is the very latest binary: http://jmp.sh/v/MWcYkfQZ5t1qRzXBWIZD

Fix confirmed!  The dragging you provided is really a joy to use - whereas the stock software, using the stock software's dragging is really a drag.

Once you get "Normal" and "Single Shot" working, I can see using this to catch a whole wave form and then find it easily.

***
And now for something completely different.  I got an Abort testing Interactive Cursor. Probably a divide by zero or unexpected values in a calculation.
(see attached Access Violation error screen print.)
I am able to reproduce it with these steps:
- I had both channels connected.
- I had both on 100KHz, Ch1 Sine wave, Ch2 Square wave.
- Time base is 20uS/division
- (EDITED: adding info on the vertical:)
  Vertical is 1V/division.
  Ch1 (sine) max and min are 1.82V, 3.64V peak to peak according to stock version measurement window
  Ch2 (square) max 816mV, min -753mV, 1.57V peak to peak
- Interactive cursor measurement is turned on
  The cursor positions likely is important. I am not sure this what other settings will trigger it, but the settings below is a sure way of making it happen:
  Ch1 measurement point is point #2 is about 2 divisions right of point #1 and about 1 division below point #1.
  Ch2 measurement points are left as default.
- I changed Ch2 to 0Hz
- I changed Ch1 to 0Hz
- I changed Ch1 to 0.1Hz
 (So far so good...next step cause error)
- I changed timebase to 5 seconds
 The system pauses for a brief moment waiting for the trigger, then the program aborts.
***

Hey, Richard.  Thank you so much for your hard work.  I think I speak for the rest of us: "We appreciate it!"

Rick
(re-edited for typo and format error and typo)

[RichardK]

And now for something completely different.  I got an Abort testing Interactive Cursor. Probably a divide by zero or unexpected values in a calculation.
(see attached Access Violation error screen print.)
I am able to reproduce it with these steps:
- I had both channels connected.
- I had both on 100KHz, Ch1 Sine wave, Ch2 Square wave.
- Time base is 20uS/division
- (EDITED: adding info on the vertical:)
  Vertical is 1V/division.
  Ch1 (sine) max and min are 1.82V, 3.64V peak to peak according to stock version measurement window
  Ch2 (square) max 816mV, min -753mV, 1.57V peak to peak
- Interactive cursor measurement is turned on
  The cursor positions likely is important. I am not sure this what other settings will trigger it, but the settings below is a sure way of making it happen:
  Ch1 measurement point is point #2 is about 2 divisions right of point #1 and about 1 division below point #1.
  Ch2 measurement points are left as default.
- I changed Ch2 to 0Hz
- I changed Ch1 to 0Hz
- I changed Ch1 to 0.1Hz
 (So far so good...next step cause error)
- I changed timebase to 5 seconds
 The system pauses for a brief moment waiting for the trigger, then the program aborts.
***

Thanks for detailing this bug, I fixed it and will upload the latest binary later on... The problem was a little more complicated than a divide by zero, it was a small oversight in some calculations I do to scale mouse cursor positions from the scope image to data array indexes and vice-versa.

The mistake was when writing the scaling code, I had made the assumption that the Raw Read Length (the length of the waveform data array) would always be larger than the Raw Draw Length (the length of the raw data to draw to the screen)...

Well as it turns out (and I knew this but it didn't pop into my head when writing that part of the code) that when the Timebase is 2 seconds or higher, the Raw Draw Length is actually higher than the Raw Read Length and you can tell, because the waveform becomes shrunken inside the scope view.

The fix was to get the ratio between the Raw Draw Length and Raw Read Length and use this in the scaling calculations, otherwise they would be invalid from 2S/DIV upward!

[RichardK]

Here is the very latest binary: http://jmp.sh/v/UPtQLGTWm7zJrftGiad2

[Rick Law]

Here is the very latest binary: http://jmp.sh/v/lNrJZjN58d1SpM42tz8O

RichardK,

Sorry!  This fix doesn't work.  The cursor just jump around and I cannot grab/set it at any point  to do testing.  The cursor is not even on the trace.  See attached short video at about 38 seconds in and you will see what I mean.



Rick

[RichardK]

Here is the very latest binary: http://jmp.sh/v/lNrJZjN58d1SpM42tz8O

RichardK,

Sorry!  This fix doesn't work.  The cursor just jump around and I cannot grab/set it at any point  to do testing.  The cursor is not even on the trace.  See attached short video at about 38 seconds in and you will see what I mean.



Rick

Opps sorry about that, I forgot to update a piece of code... Should be fixed now: http://jmp.sh/v/UPtQLGTWm7zJrftGiad2 NOTE: Same zip filename as before, make sure you remove the old one!

Also in that new binary the waveform dragging (dragging the scope, not the waveform context area above) and the scrolling with left and right arrow keys is now scaled to TimeDIV so the waveforms will move the same regardless of Timebase.

[Fraser]

@RichardK,

I just wanted to send you a big THANK YOU for your work on this software. You are very generous with you time and skills. My 6022BE has not been used since discovering the poor performance of the supplied software. I will have to dig it out in preparation for its first exposure to decent coding 

[RichardK]

@RichardK,

I just wanted to send you a big THANK YOU for your work on this software. You are very generous with you time and skills. My 6022BE has not been used since discovering the poor performance of the supplied software. I will have to dig it out in preparation for its first exposure to decent coding 

You're welcome, I'm just glad other people are finding my work useful to them somehow, even if it's purely a spare time project and I am grateful for all the various feedback and suggestions

[Rick Law]

Here is the very latest binary: http://jmp.sh/v/lNrJZjN58d1SpM42tz8O

RichardK,

Sorry!  This fix doesn't work.  The cursor just jump around and I cannot grab/set it at any point  to do testing.  The cursor is not even on the trace.  See attached short video at about 38 seconds in and you will see what I mean.



Rick

Opps sorry about that, I forgot to update a piece of code... Should be fixed now: http://jmp.sh/v/UPtQLGTWm7zJrftGiad2 NOTE: Same zip filename as before, make sure you remove the old one!

Also in that new binary the waveform dragging (dragging the scope, not the waveform context area above) and the scrolling with left and right arrow keys is now scaled to TimeDIV so the waveforms will move the same regardless of Timebase.

Fix Confirmed!  That problem is history.

Thanks for the hard work.

Rick

[frenky]

@RichardK:
I just ordered the 6022BE because of your improved software.
Thank you for all the time & energy that you put into this software.

[callito]

@RichardK

Wow, god dam this is awesome! Just a few questions.
you hacked the SW for 1x to 10V,... so if I use the 10x probe it goes to 100V up??

and is it possible to hack this SW, so it goes up to 35V at 1x??? and 350V at 10x probe?

that would be unbelievable!! I normally want to use this scope for 3-16V at 10-100Hz, so it would be great if I can use 1x probe up to 35V instead of the factory-set 10x probe at 50V .

By the way,.. I did't order this 6022BE yet,.. just saw this thread today and since you do something with the SW, I wanna have this scope!


Thank you very, very much for your hard work!!!

Greetings, and sorry for my bad english.

[womai]

But even with the new software the scope has a lot of issues (see the first few pages of this thread...) - as they say, you can put lipstick on a pig but it will still be a pig :=)

The +/-5V is a limitation in the hardware, so software won't be able to get around this restriction.

I would definitely NOT recommend this scope in its current state.

[RichardK]

But even with the new software the scope has a lot of issues (see the first few pages of this thread...) - as they say, you can put lipstick on a pig but it will still be a pig :=)

The +/-5V is a limitation in the hardware, so software won't be able to get around this restriction.

I would definitely NOT recommend this scope in its current state.

The +/-5V is a limitation in 1x mode, it's not like you can't measure any signal over +/-5V period, in fact the limitation is not hardware, as you can see from the front end, after the 900K & 100K resistor, the input signal is 1/10th lower, so it's NOT being clipped by those protection diodes.

The signal is being clipped either in the firmware, or the acquire DLL, hopefully the latter as that will be easier to hack.

[RichardK]

@RichardK

Wow, god dam this is awesome! Just a few questions.
you hacked the SW for 1x to 10V,... so if I use the 10x probe it goes to 100V up??

and is it possible to hack this SW, so it goes up to 35V at 1x??? and 350V at 10x probe?

that would be unbelievable!! I normally want to use this scope for 3-16V at 10-100Hz, so it would be great if I can use 1x probe up to 35V instead of the factory-set 10x probe at 50V .

By the way,.. I did't order this 6022BE yet,.. just saw this thread today and since you do something with the SW, I wanna have this scope!


Thank you very, very much for your hard work!!!

Greetings, and sorry for my bad english.

It's a software hack of 10V/DIV @ 10x but with software attenuation, so you lose some resolution. It's just a placeholder for now as I plan on trying to hack either the driver or the DLL to get access to the real raw data, not the tweaked data provided by the SDK.

[callito]

It's a software hack of 10V/DIV @ 10x

Ahhhh.. ok I understand.... i thought it would be @ 1x ... but still very cool and usable!

If you are realy gonna hack the driver or dll's.. wow,... but i think it is very hard to do so...
...it would be dubble the price, if they will find out.

Good luck, so, I'm "staying tuned"


Greetings

 

[mrphil]

I did notify today that the printing option is not available,
Am I doing something wrong or you just haven't reach this point yet?

[RichardK]

I did notify today that the printing option is not available,
Am I doing something wrong or you just haven't reach this point yet?

Printing is not implemented yet, but will be soon... Typically, if you notice anything in the interface that doesn't appear to do anything, it's not implemented yet. Exceptions being of course, things that worked in previous builds, as this may be a bug!

[Prana]

@RichardK
Hi, is it possible to use your software with Hantek 6052BE (50 Mhz version of this scope) ??

[RichardK]

@RichardK
Hi, is it possible to use your software with Hantek 6052BE (50 Mhz version of this scope) ??

It does not, however that model might work with OpenHantek for reasons stated below...

That model uses vastly different hardware and thus a vastly different software architecture than the 6022BE, in fact I have looked at all the SDKs for all USB scopes & Logic Analyzers and most of them use the same architecture as the 6052BE, and the 6022BE software architecture is unique, I have not encountered it's design in any of the other SDKs.

[Mark_O]

I did notify today that the printing option is not available,
Am I doing something wrong or you just haven't reach this point yet?

Printing is not implemented yet, but will be soon... Typically, if you notice anything in the interface that doesn't appear to do anything, it's not implemented yet. Exceptions being of course, things that worked in previous builds, as this may be a bug!

Not that anyone has been complaining, but there's been a fair amount of "this or that doesn't seem to be working?".  It's good to keep in mind that RK is an outsider (doesn't work for Hantek, and is uncompensated), with no internals info other than what he's managed to figure out for himself, working part-time, for a couple weeks.  And even with all those constraints, he's managed to out-do the full-time Hantek programmer(s) with full access to all the internals, and other resources, for the past ?? years! 

Let's let him proceed at his own pace, without any pressure.  The capabilities and limitations of previous beta releases have been documented, and I suspect he'll continue to document improvements as he announces new versions.

Richard,

one thing that might help minimize confusion is if in your ReadMe you had a simple list of the major Features, broken into Implemented and Not-Implemented Yet lists.  Then just move one line between the 2 lists as their status changes.  That's about as simple a project-management & status tracking technique as I can think of.

You could even have a 3rd list (Not-Planned), if folks start requesting things that aren't possible, or you don't plan on doing.  That would forestall repeated requests (and explanations), since it's fairly obvious that folks aren't going back through and reading the thread postings.

[Oh, and thanks for all your efforts.  You're doing a great job!]

[RichardK]

one thing that might help minimize confusion is if in your ReadMe you had a simple list of the major Features, broken into Implemented and Not-Implemented Yet lists.  Then just move one line between the 2 lists as their status changes.  That's about as simple a project-management & status tracking technique as I can think of.

You could even have a 3rd list (Not-Planned), if folks start requesting things that aren't possible, or you don't plan on doing.  That would forestall repeated requests (and explanations), since it's fairly obvious that folks aren't going back through and reading the thread postings.

I was thinking about doing that for the next binary, since I was not originally going to periodically release binaries for a functionally incomplete project (I assumed nobody would want to use a partially functional program).

I will release the next binary soon, I know it's been a while between releases it's just that I have a methodology of "Get it working first, optimize it later" and I have to do more testing due to a switch from Borland's TTimer to true multi-threading for Capture & Drawing, and as you would expect with multi-threading you have to track down race conditions and I think I nailed them all, but I'm just being extra cautious.

The latest binary will have better performance (due to the multiple threads) and almost fully functional triggering, as well as initial support for FFT (which is already partially implemented in the last binary, but who's feature is hidden in the GUI)

[Mark_O]

I was not originally going to periodically release binaries for a functionally incomplete project (I assumed nobody would want to use a partially functional program).

Yeah, but what you overlooked is that's what everyone with a 6022BE has always been doing, so far.    Whether they wanted to or not.  You're raising the bar.

[Merlyn]

Richard, I've seen in the past people developing software for big companies' products which despite their big names released products with mediocre software to say the least and then offering fully functional software against small charge and/or donations.
Or maybe you can contact and make some deal with Hantek against some of their higher quality product/s or something so they can give you the software source as they will profit from it as well?

[Matchless]

Hi I joined this forum after reading this great thread. I have also bought the Hantek 6022be for all the reasons already mentioned.  The software RichardK is busy with just made it more promising. I still need to understand how to use this DSO properly, as this is all new to me.

I will try to post the noise shown on both my channels when connected to the internal square wave. It interesting to see that channel 2 is nearly clean compared to channel 1.
Both probes were set to X10.
Any comments please?