Hantek - Tekway - DSO hack - get 200MHz bw for free

[tinhead]

for those who not noticed yet, Hantek have published today new firmware 2.06.3_120224.0
It should fix the bad scroll mode crash, and yes it does. Now DSO is not crashing anymore,
or well at least i was not able to let crash my DSOs.

What not fixed - all new functions are not linked to help file, each time you push help button
in one of these function dso is crashing (so don't use help, lol)

What also not fixed - in equ mode when changing timebase from 4ns/div to 2ns/dov with menu
disable DSO crashing. Fix - don't use supid equ mode on 1GSs real time sampling scope? yeah, but still this is bug.

What new, go figure? Well, there is new point - save waveofrm. You could save
100 waveforms now as bin data (recorder, ref bin data, screenshot and csv of course still there and still working)
and of course reload them later. you can save to SD or USB (when yous DSO is BM/BMV model) or internal
NAND (which might be dangerous .. think about no space in NAND). But don't worry, this function will anyway
not work for you becase your DSO need a director named /param/sav/wav_sav ro allow saving in NAND.
Sure, when you create it manualy it will works for you. And no, don't use USB as location to save - it might work
but it is crashing the most time due the wrong kernel (yet, it is kernel dependant and need again BM/BMV models
which have newer kernel version)

New bugs? Of course ... In single window, sample memory sset to >=40k, both channels (ch1, ch2) enabled
and timebase from 400ns/div up to 2ns/div the position change (fine and coarse) control is not working.
Only scale is moving, but not the waveform. Luckily nothing crashing but still a bug.

Something else? Yep, it is sampling now like Rigol, with long memory (>=40kpoint) enabled it is sampling
now per channel maximal 500MSs and not 400MSs anymore. Sees to be stable.


These firmware are here:

http://www.hantek.com/english/down_list.asp?unid=30

use only the "B" versions, even if i speak about "BM/BMV model features" it does NOT means
you can use the BM/BMV model firmware!!!

Hantek/Tekway are trying to maintain the code as simple as possible, which means many features
are visible in all models, but only some available ! And of course the executable from BM7BMV models
is compiled againts kernel 2.6.30.4 and not 2.6.13 as on yoyur DSO today.

[paul]

I have found a problem if the trigger point is off of the screen, I found this when I was trying to check that my PIC programming was actually generating the two 100ms pulses I wanted it to at the  right time, which it didnt appear to be doing until I changed the way I captured the waveform, by using 1M memory depth and single sequence. Very strange.   

What I have found is :-
If the trigger point is adjusted / moved off of the visible screen area then the displayed timings are all screwed up.
You can show this easily by connecting a probe to the probe check terminals, set the timebase for 80uS per div and adjust the horizontal position so that the trigger point is 500uS (to the left), you will see the 1kHz square wave change from high to low in the middle of the display, just what you would expect from a 1kz square wave (a period of 1mS).
Now adjust the time base to 40uS per division and the crossover point is no longer in the centre of the screen, indicating that it is not at 500uS any more. If you adjust horizontal position so that it is at the centre again the trigger point is now 454.0uS, change the time base back to 80uS per division and the square wave is now offset to the right of the centre of the screen

Return the horizontal position to 500uS and press the `Single Seq` button, now change the sec/div to 40uS. The display will stay locked to the real 500uS point, you can even zoom in to 4us per division with no problems (or more with a higher memory depth), until you press the Single Seq or Run/Stop button and everything is screwed up again.

I have not seen this problem mentioned on the forum, is it only my scope doing this ?
It looks like a simple software bug, I upgraded to 120112.1 a couple of weeks ago (as recommended by Tinhead after my scope locked up, thanks again BTW)  and this has made no difference to the error.

Paul.

[tinhead]

I have not seen this problem mentioned on the forum, is it only my scope doing this ?

Paul,

i'm using these scopes since 2 years ... tested already tons of firmwares .. and never saw this bug !, Fantastic finding ...

The problem exists in every timebase from 40s/div to 800ns/div. Why? Well because the buffer is at the end.
Then from 400ns/div everything seem to be ok, but only as long you not exiting the buffer length again.

Still, this is bug, the scale should not move or not change in wrong way.

[paul]

Hi tinhead,
Thanks for confirming its a software bug, I was a bit worried it was only my scope.

As I am used to using older Tek scopes with 1K  memory I have gotten used to moving the trigger point and recapturing the waveform at the point I am interested in, which was exactly how I was using the Hantek.

If you use the full memory there is no need to do this, most of the time
However the Hantek is a lot more responsive and faster with 4K memory selected, and after having a total lock up of the scope in 1M memory I am reluctant to use it again, or update the firmware until it has been out there for a while.

Paul.

[tinhead]

HW1007 UART over RJ45


as you maybe already recognized there are some optocouplers and other not populated parts around RJ45 jack.
As few ppl asked what they for i decided to publish some details .
I actualy was about to publish this when some day the full hw1007 schematics is ready .. anyway.

You can (as is out of the box) only use this when you not plan to use LAN in the future.
If you wish to use LAN later, i propose to wire rxd/txd directly to pins 4 and 5 (ithout any optos).

Pins 7 and 8 are today not used by the firmware (but maybe i oversaw something in code),
however they can be easily used e.g. as trigger out. They connected to GPIO pin,
so you can actually toggle that pin,anayway, sometimes is nice to have spare gpio pin.

I haven't defined any components values of these not populated parts, go figure yourself

Btw, J801 is the UART header (which you normaly using when you connect over UART)


EDIT: is there any reason why rating of this thread is that low?

[NCG]

Did anyone got the Voltcraft labeled version outside Germany? I managed to make order when there was stock left but now the order seems to be in somekind of limbo. Safe to assume I missed the train and the order is waiting for next batch of unknown sample rate?

[Lo_tse]

Just curious, has anyone tried hacking the 5102BM model? Supposedly, it is the same scope as DS5102B but with larger memory depth (2M one channel, 1M for 2 channes). Would the same hack work for this model?  Thanks!

[tinhead]

Just curious, has anyone tried hacking the 5102BM model? Supposedly, it is the same scope as DS5102B but with larger memory depth (2M one channel, 1M for 2 channes). Would the same hack work for this model?  Thanks!

yes it works on BM/BMV models too.

Do you plan to buy BM/BMV?

[Lo_tse]

Thanks! tinhead.  Yeah, I am interested in the BM model (for the larger memory) but they are a little hard to find.  The price quote that I got from a Chinese web-store (where else) was quite reasonable and the price  was only about 10% higher than the regular B version. 

I was also considering the Atten1102CML but there are hardly any user informations/feedbacks on the Internet (don't know why).  As for the Hantek, you have put out so much info here and I feel a bit more comfortable knowing that I probably can get some help from this forum if something goes wrong.

Regards,

[tinhead]

Thanks! tinhead.  Yeah, I am interested in the BM model (for the larger memory) but they are a little hard to find. 

the reason why they hard to find is typical for new chinese products. Why?
Well, in principle all these chinese stores are checking market for the manufacturer long time
before the production even started, sometimes even at development stage!
This helps out to find "the best price" and potential launch volume.
Typical it takes 1 to 6 months before something marked as "in stock" in these shops is really available.

Knowing this most non-chinese dealers are just waiting until the manufacturer is double confirming the availibility,
which is sometime mistakes because not every manufacturer will beg a dealer (or inform once again)

Anyway, in case of these BM/BMV models it was about 5 months until they got really available (since 2 weeks i think).
Hantek confirmed this too, so feel free to buy it from your fav. store.

The price quote that I got from a Chinese web-store (where else) was quite reasonable and the price  was only about
10% higher than the regular B version. 

well, you can save another 20USD when you buy BM or BMV model, make a backup of the firmware and send it to me.

[Lo_tse]

One of the rationale that I can think of is that many of those Chinese brands are unknown to the outside world.  More importantly, after sales services, which is extremely important in the Western world, are essentially non-existent since there are no dealers or distributors in the region to support the retailer.  As such, the dealers in the West are always skeptical (cannot blame them) about these products and they are generally no that interested.  Let alone new models that are coming on-line for a particular manufacturer.

Everytime when I want to buy electronic goods from China, I will ask myself the question - what if the thing broke after a few months?  Shipping the item back to China for service is impractical due to high shipping cost (usually buyer has to pay for that).  Thus, whenever I bought something, I pretty much assumed that I will have to find my own solution if the item broke.  Most of the Chinese sellers on the Internet realised that so they usually mark the price fairly low (they do not have much overhead anyway) - to compensate for the lack of after sale service.  If the items are legitimate and passed the QA/QC tests, they are usually OK.  But then there are so many of those unscrupulous sellers that sell factory seconds or fake copies.  These people really kill the reputation of the brand even before it was known. 

I do not know when I will buy the scope, probably in a few months time.   I'll let you know when I got it. 

Regards,

[tinhead]

there a good news for those of you who wished to have 1:2:5 instaed of 2:4:8 timebase ratio,
Hantek is (since last two fw versions) working on that. The timebase 1:2:5 is already partialy
implemented, not everything works smooth now so they have not yet enable it.

I hope they implement a firmware feature like "change ratio" to change timebase dynamically,
after 2 yrs of working with these DSO i would miss the 2:4:8

[Payti]

@tinhead this is checked from sw 111026.1 that I sent You?

Regards,

[tinhead]

@tinhead this is checked from sw 111026.1 that I sent You?

Regards,

no, from 2.06.3_120112.1 and 2.06.3_120224.0

[quarros]

Hello everyone! Can some of you help me to figure out why this scope displays in cursor mode twice as much as the actual reading? Probably just some silly mistake on my part, but I can't seem to figure it out. Thank You

[bilko]

Your vertical scale is 2V so the reading is correct
Are you looking at the 2.48V ? that is the trigger voltage

[quarros]

Your vertical scale is 2V so the reading is correct
Are you looking at the 2.48V ? that is the trigger voltage

Well the signal I'm feeding to channel two is 2.462V according to my multimeter. And the displayed reading is 5.04V. Realy I'm missing something basic.

[bilko]

Read up or google, RMS, peak to peak and average
Your meter is probably reading average

[quarros]

Read up or google, RMS, peak to peak and average
Your meter is probably reading average

As i said basic problem.... my mind is probably taking a tour around the moon because it ain't here.
Anyways thanks.

[alm]

Read up or google, RMS, peak to peak and average

The meter is probably indicating RMS. Whether this is an actual RMS value or an approximation by multiplying the average absolute value by sqrt(2) depends on the quality of the meter.

[quarros]

Read up or google, RMS, peak to peak and average

The meter is probably indicating RMS. Whether this is an actual RMS value or an approximation by multiplying the average absolute value by sqrt(2) depends on the quality of the meter.

Thank You for the clarification. It was clear form the moment he mentioned RMS, but somehow until that my mind froze up. Well that's what you get if you forget the basics when measuring PWM.

[tinhead]

IMPORTANT:

as you know Hantek/Tekway said long time ago "we will provide SDK". Well they didn't, even worse - they
removed almost all "3rd dev kit, LabView support" links.

However with some private effort, massive presure (yes, Hantek provided some information!) and hard work
of some users we have enough information to remote control these DSOs, get data, manipulate files and so
on (few things still not ready, if you wish to spend some time on them just do it - every information might be useful)

The major part or "our SDK" is ready, see here:

http://www.mikrocontroller.net/articles/Hantek_Protokoll

i'm working right now on #0x01_DSO_Einstellungen_lesen (means read DSO settings) and
#0x11_DSO_Einstellungen_schreiben (means write DSO settings). This is a 208 to 238 bytes block of data,
describing the fronpanel settings, state of DSO and so on.
During consolidation of collected information i recognized however an issue which might have large influence
on how good/bad the PC software is working for you.

In principle each firmware version might be unique, mean the length of these DSO setting blocks might differ,
also the order in which each of these settings is stored/read might differ. Looks complex? No it's not, Tekway
was smart enough as they designed these DSOs to provide a protocol.inf file which describes the order or
settings and size of settings.

However, at the time Hantek shareholder bought some Tekway shares someone "forgot to write down" what's
about the protocol.inf.

What happen then, well, everybody who bought DSO with firmwares 2.05.x to 2.06.2 and later updated to current
2.06.3 firmware will have complettly different protocol.inf file installed, this is because Hantek/Tekway never ever
again released firmware updates with proper protocol.inf.

As some of you guys did made backup of your firmware i got an small overview what you have, some of these dumps
didn't even have protocol.inf, some wrong version.

Anyway, i will create an firmware update which will only copy the current protocol.inf, it will not break your DSO is you
have already the proper version but it will help ppl who bought older model this.

[condor]

Tinhead, Thanks for the hard work. I wish you success!

[Payti]

@tinhead many thanks for Yours great job! Please don't stop ...

Regards,
Tom

[tinhead]

well, the User Written SDK is ready and published :
(actually only the SD part of SDK .. the "K" is something you or Hantek have to do).

http://www.mikrocontroller.net/articles/Hantek_Protokoll

There is already a nice tool developed on base of the informations from the SD(K), check this:

http://www.dreisiebner.at/dso-usb-tool/


I'm not using Matlab nor LabView, so don't ask me for compiled drivers - use your power and develop one.
It is not that complex, all you need to access these DSOs is docummented.