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

[prpplague]

This are kernel and driver sources as well some tools (e.g. the fpga tool missing on Voltcraft DSOs).
For sure it is really nice to have these sources (regardless of GPL whatsoever thing), however don't forget
that the DSO code itself (FPGA , CPLD and ARM code) is not and will not be published as they
are Hantek / Tekway Intellectual Property.

Therefore the amount of "new DIY features" is reduced to only few things (like e.g. custom LAN)

You should also not forget that the firmware (ARM code) itself is checking the kernel version,
it need to be 2.6.13 or 2.6.30.4 - if not some features/the firmware will not work.

true on all points, however it is one step closer to a more functional system....

[prpplague]

This are kernel and driver sources as well some tools (e.g. the fpga tool missing on Voltcraft DSOs).
For sure it is really nice to have these sources (regardless of GPL whatsoever thing), however don't forget
that the DSO code itself (FPGA , CPLD and ARM code) is not and will not be published as they
are Hantek / Tekway Intellectual Property.

Therefore the amount of "new DIY features" is reduced to only few things (like e.g. custom LAN)

You should also not forget that the firmware (ARM code) itself is checking the kernel version,
it need to be 2.6.13 or 2.6.30.4 - if not some features/the firmware will not work.

already looking at the released sources shows the hantek folks have no clue and there major performance issues with their build. not only is there kernel a mess, but their root file system is a disaster.

[tinhead]

major performance issues

explain that point

[prpplague]

major performance issues

explain that point

i've counted seven spinlock issues that have already been fixed in newer kernels, as well as two atomic operations that are suspect. the root file system also has two processes that are zombies on startup and multiple issues with the init scripts. the usb host and usb slave devices drivers have _major_ improvements in new kernels that deal with errata about double buffering. the dma driver is missing work arounds for about a number of errata.

the whole kernel and root file system is a disaster. hantek should fire everyone involved in the development linux development of this device. in fact i plan to make to make it a case study on what _not_ to do for a commercial product.......

[tinhead]

i've counted seven spinlock issues that have already been fixed in newer kernels, as well as two atomic operations that are suspect. the root file system also has two processes that are zombies on startup and multiple issues with the init scripts. the usb host and usb slave devices drivers have _major_ improvements in new kernels that deal with errata about double buffering. the dma driver is missing work arounds for about a number of errata.

hmm, so nothing really (maybe except dma driver) important for an DSO.
Usb is anyway 1.1 - even with quad buffering and kernel 25.765 will be still exact that slow
(or even slower due blowed up kernel), zombies or init script are not an performance issues - nobody really
care if DSO need a half second longer to boot (as it have anyway to warm up).

I really doubt that newer overblown kernel on a box with reduced resources with what so ever file
system used only to boot will make an DSO faster ^^

the whole kernel and root file system is a disaster. hantek should fire everyone involved in the development linux development of this device. in fact i plan to make to make it a case study on what _not_ to do for a commercial product.......

maybe, but on the other side this firmware has been developed in 2008, released begin 2009 for testing and
end 2009 for production. At that time 2.6.13 was not that old (actually prefered version on many systems).
I do have here almost all firmware versions from the last years, and when i look inside it was already a chalange
to maintain and support all these versions.

The good things about new models (BM/BMV) - Hantek took the chance for a cut, they now based
on 2.6.30.4 and newer drivers (and usb slave is making me crazy on PC side ... double buffered=doubled issue)

[prpplague]

i've counted seven spinlock issues that have already been fixed in newer kernels, as well as two atomic operations that are suspect. the root file system also has two processes that are zombies on startup and multiple issues with the init scripts. the usb host and usb slave devices drivers have _major_ improvements in new kernels that deal with errata about double buffering. the dma driver is missing work arounds for about a number of errata.

hmm, so nothing really (maybe except dma driver) important for an DSO.
Usb is anyway 1.1 - even with quad buffering and kernel 25.765 will be still exact that slow
(or even slower due blowed up kernel), zombies or init script are not an performance issues - nobody really
care if DSO need a half second longer to boot (as it have anyway to warm up).

I really doubt that newer overblown kernel on a box with reduced resources with what so ever file
system used only to boot will make an DSO faster ^^

the whole kernel and root file system is a disaster. hantek should fire everyone involved in the development linux development of this device. in fact i plan to make to make it a case study on what _not_ to do for a commercial product.......

maybe, but on the other side this firmware has been developed in 2008, released begin 2009 for testing and
end 2009 for production. At that time 2.6.13 was not that old (actually prefered version on many systems).
I do have here almost all firmware versions from the last years, and when i look inside it was already a chalange
to maintain and support all these versions.

The good things about new models (BM/BMV) - Hantek took the chance for a cut, they now based
on 2.6.30.4 and newer drivers (and usb slave is making me crazy on PC side ... double buffered=doubled issue)

thats just the start. if your position is that "it works why complain?", then i don't accept that.  i've have made a wide range of commercial products using linux. have i created crappy stuff that made it to market? yes. have i learned from my mistakes? yes. the first principle in moving forward is the simple statement and acceptance of "there is a problem".

even if you want to argue that is "works", do we really want people to use the das_oszi as an example for future development? i sure as hell don't.

[rf-loop]

Thanks to the hard work of the folks at gpl-violations.org and a coordinated effort against Hantek's distributors, all open source license code for the Hantek/Tekway DSO has been released:

http://ftp.gpl-devices.org/pub/vendors/Voltcraft/VOLTCRAFT_dso3000series.zip


we will be dissecting and making it available from a github repo within the next few days.

thanks gpl-violtations.org!!

gpl-violations.org   wau...  maybe manufacturers are shaking with fear... no..  in real world they just laugh.

Such declaration and the papers can  use just as toilet paper in China.

There is no force of law in any real case, and certainly not in China. China have the Chinese laws and regulations of China and China's laws are in force - and only them. The only way is not to buy them outside of China. And, it is unlikely anyone would want to practice.

 I want to see that these GNU dream ideologists go and try with some western company, example with Agilent etc..
 Who equipment manufacturer they really open firmware.
What manufacturer can do better. They can make more difficult to look anything inside equipment. End - user need nothing other but these things what are for end user implemented in equipment. Nothing other need be open for end user - nothing. Even case can close so that end user can not go inside. Without breaking and damaging. If there is FW update need, it can do so that nothing is available for end user. After this, what for there need be open source codes or anything other open than user manual. 
 
I hope Chinese manufacturers learn this and protect they products in future better. Becouse there is starting fourth world. There are soon so that company in China design and then manufacturing is outside China in some undeveloped countries where prices and work is really cheap.

It is totally different case if we make product what is open example son that "everyone" can do aplications for this product.
Oscilloscope, signal generator, power supply, spectrum analyzer etc. They are stand alone test equipmenst what have they functions and specifications and closed box. How you can show there is something inside what just you or other have right to get source codes. This you can try in Chines law room.

If there is free linux just use it and hide all source codes as well as Possible or what feel for Relevant protect product.

In reality, there is no legal way to force the manufacturer to comply with any of the paper which has no legal force, over the manufacturer.

Suitable for a try. The manufacturer can just smile and push these ideologists papers to the trash. I do not know any cases where someone could have been in China through judicial or other means to force a manufacturer to publish product source code.

Totally different case is if manufacturer look that it may be useful to publishe sources and then they want do it.
No one need give any source codes from ready made end user product where is user interface and  not need touch embedded FW. 

Anyone but the manufacturer does not need to get the software. The interface is a device intended for the use of buttons and controls as well as a screen for the device or the device coming out of the data which is intended for the end user.

These are NOT as called "open project" even if manufacturer have been littlebit lazy, loosy and... let some gates and windows open. 

Uniguely In this special Hantek case it have been useful tah there have been possible to touch FW and maybe in this case it is better if source have been open. 
But even without open sources (source codes) there have be posibility to do something thanks for tinhead, who have done and do  lot of work to make and develop this product better than Hantek alone can do with they EE's who have  maybe lack of experience for do good job and finished product.

So, use this gpl poems in this case as toilet paper.

[Gall]

Uh. Finally.

That's much better than completely reverse engineering everything.

[bjoernx]

@rf-loop: WTF!? 
IMHO you seem to have little to no idea, what you're talking about... it's just ridiculous, not worth to go in any detail. 

[alm]

There is no force of law in any real case, and certainly not in China. China have the Chinese laws and regulations of China and China's laws are in force - and only them. The only way is not to buy them outside of China. And, it is unlikely anyone would want to practice.

Agreed. Unless that Chinese company wants to have international sales, which Owon apparently does, since they struck a deal with Conrad. At that point the distributors will require them to comply with local laws, which apparently happened in this case. Few western companies will be willing to take the risk of being found responsible, so they'll request compliance or drop the product. Not exactly a good deal for said Chinese company, is it?

I want to see that these GNU dream ideologists go and try with some western company, example with Agilent etc..

Plenty of cases against western companies, did you care to take even ten seconds to do research before starting off on a rant?

In reality, there is no legal way to force the manufacturer to comply with any of the paper which has no legal force, over the manufacturer.

This comment is based on the large number of lawsuits being dismissed, and the fact that Conrad/Owon ignored this request and refused to release any source code? It's extremely rare for these cases to even make it to the courtroom, presumably because defendant's council does not consider the piece of paper worthless.

[jap]

Hi,

I intend to buy a Hantek DSO, but the new versions have problems:

"many of you with 555583e9 version have issues (and these can be replicated by it seems only on your hadrware)."

Do you have more info
- if this is still an issue with new DSO-s, and
- if it is hardware or firmware related?

[rf-loop]

Conrad/Owon ignored this request and refused to release any source code? It's extremely rare for these cases to even make it to the courtroom, presumably because defendant's council does not consider the piece of paper worthless.

Why Owon need release any source code. How Owon is related to this case?
Hantek, no need release any source codes. 

_________
Edit:
Start Of Add:
It need tightly understand what is difference between derived work and  (x-made) propietary software.
Of course hantek may give this embedded linux source code. This can easy find and download many places.
dso.exe afaik  is NOT derived work and it is out from GPL.
There are also other important things inside oscilloscope what are out from GPL. Just these most important things if example think modifications. Hnatek do not need give any source codes about these and even not binaries.

Some manufactures may use unique key and hide nearly all.. even kernel and also this is legal. And agen, there is no any legal force for get sources or even binaries. If GPL "black side" peoples (there is good side and black side related to this and I think one Linus Thorvalds belongs to this good side) are happy if they get embedded linux kernel source code - oh well --- laughing. Be happy.
If I make equipment where is embedded linux and then there is my propietary software for this equipment... I need not give source codes or binaries or anything. I can give link to you where you find linux kernel sources and thats all. Just becouse my propietary software is my own and it is not any way under GPL  If you do not believe me ask Linus. He have tell it extremely clear.
End Of Add
------------------


Threaten, blackmail and pressure special in sensitive markets can lead to the disclosure of a minor but real legitimate reason for not selling because Hantek oscilloscope and it is a product. Hantek is provided at least at a level no access to the operating system and the user should not have any access to the internal Hantek adopted to modify the software or parts of any kind.
 Besides, if they wants, they can close the whole system and even to conceal the system and change it so that there is no access to it or it is very difficult, and in addition to binary may choose to conveniently change the way that it is no longer a fact of which one could prove it to be that of the GPL hence the concept .

 I definitely side with manufacturers. I encourage that do not publish the source code or any parts of programs in any situation where they have embedded software and a user does not have as product feature to access to neither the need or opportunity to install their own or third-party software or plug-ins.

Also one other recommendation. Do not use embedded linux. Or if use, modify it very hard so that it is not anymore "linux".
Do commercial mature products.

Who, by the way, is responsible about product warranty if some other than manufacturer itself have done any change without manufacturer accepted just this change.


nm eot

[marmad]

I finally received a message back from Conrad that they can't give me an expected delivery date for the Voltcraft DSO-3062C  which basically equates to 'never' in retail-speak.  And in the meantime, they 'hold on' to my money?  I let them know that it's fine - as long as they pay me interest on my money in the meantime  which of course means a refund and no Hantek. Oh well, on the bright side, Agilent is offering free Wavegen and DMM options on their scopes until the end of August.

As an aside: this is just one of the reasons Europe is still in the dark ages (compared to the States) when it comes to online buying. When you order something from a real shop here, the most they usually take is a small down-payment; the fact that Internet sites can take ALL of your money for an item - and then not deliver for long periods of time - is simply ridiculous and unacceptable.

Edit: BTW, for those who don't know, most established US online sites don't charge you anything UNTIL they ship your product - which is the way it should be.

[pullin-gs]

@rf-loop: WTF!? 
IMHO you seem to have little to no idea, what you're talking about... it's just ridiculous, not worth to go in any detail. 

Actually he does.

[alm]

Conrad/Owon ignored this request and refused to release any source code? It's extremely rare for these cases to even make it to the courtroom, presumably because defendant's council does not consider the piece of paper worthless.

Why Owon need release any source code. How Owon is related to this case?
Hantek, no need release any source codes. 

Indeed, I confused Hantek and Owon. I saw your name and assumed it was about Owon.

I guess it's an interesting ethics question. Is something like copyright infringement wrong as long as you don't get caught? Does this also apply to Hantek/Owon/Atten/Rigol IP? There are significant differences in ethics between different cultures. Things sometimes gets interesting when said cultures meet, for example trade with each other. I'm sure plenty of western companies that tried to outsource work to Chinese companies have figured this out.

[bjoernx]

@rf-loop: WTF!? 
IMHO you seem to have little to no idea, what you're talking about... it's just ridiculous, not worth to go in any detail. 

Actually he does.

Well... not in this case.

There is one point here, that almost everybody is completely missing, and that is, that hacking a product adds value to it. Any device around, that has been successfully hacked, proves to drive demand for that product, makes it gain longer life-time/usability and higher second-hand reselling price-points. Look at the DBox2 in germany or the WRT45GL. These two hacks showed even another aspect of this - the Dbox2 especially - that is: Hacking leads to innovation! Streaming A/V via ethernet, which has bevome ubiqitous now for a few years, was driven and first put to a working design by customized linux firmwares on the DBox2 - and that was possible with these boxes +10 years ago!

Now, if Hantek were smart, they'd gone open-source with their scopes from day one. Actually they were bound to do so by the GPL, at least to some minimal extent. They did not. Ignoring the license, it is THEM who broke the law and crapped on other people's intellectual property. And now again them hackers are blamed, who are allegedly doing illegal stuff by simply hacking stuff- that they are legally allowed to do in this case!? WTF!! rf-loop's argument simply turns around this obvious question: who did wrong and who is to blame here.

But at least, if Hantek were smart, they would open this product from now on, create a small community to work with, with the best achievable quality of design meeting their customer's needs in mind, nothing else - no ideology here whatsoever! This would not mean, that they'd have to open all of their sources. They could keep their own binary blobs (doing fancy dsp stuff, possibly using other IP-blocks for example) closed, and this is widely accepted by the large majority of users.
But because product management will probably be lead by MBA style thinking, this will not happen and they will miss a great opportunity to make them a name in the industry and the market by making their devices unique. An opportunity to lower their r&d costs, making  bigger chunks of money by driving demand and a respectable name for themselves, that is. And who wouldn't want that actually... but I'm afraid it is their MBA ideology (there you have it!) that hinders them from even realizing opportunities like this, one that is open to them to take on right now in this somewhat difficult situation they brought themselves into.

[tinhead]

Alm,

in general default answer to ethics question and China is -> in Chinese culture, it's generally okay to steal ideas
and copy others. But this is not the exactly truth, chinese manufacturers trying to produce similar things as other does,
trying to learn, there is nothing worng with that. Sometimes (oh well, depends on what product this might be 99% of the products)
they just don't know exactly what to do and wat such specific product need to fulfil/provide/guarantee.

A good example:
One of my customers in China - they actually producing rigips walls - have started (with some help from ppl whom
i know in EU) to copy "good known German" manufacturer.
In principle nothing wrong with that, so far i know no copyright issues or whatsoever problems - except quality.
The probelm was -> the customer was not aware that there are many kinds of Aluminium! Not a joke.
They know it now and fixed production, but it was somehow funny to hear somethign like that.

So this i exactly the problem - they think they can do something but they can't. From a culture point
of view something like "fail" didn't exists .. so next turn of self made problems start here.
This kind of thinking is producing a lot of bad copys, bad products. Sure, everybody started someday,
everybody spend money for nothing (well, learn process). But the chinese market is groving that
fast that milions of tons of crap is produced every day (and bilions of good product as well).
So what ppl see are exactly these milions tons of crap (and well, there will be always someone trying to sell
such things to reduce the lose), and therefore everything what "made in china" is bad (where still 50-70% of
things we use daily are "made in china" and have superior quality).


Now with Hantek is complettly story: the DSO market is a bilion of Dolar market, there are many competitors.
Without ASICs you just don't have so many options how a good product can be made, just look on
almost all (except Owon and HanTekway) other manufacturers - everything is like a clone of itself - the same DSP,
the same ADC (sometimes different), "same" FPGA, display controlelr and so on ... even firmwares like
from same mother.

So you can imagine why Rigol stopped to focus on low-range market, Owon is using CryptoMem to secure
their design and why HanTekway (actually only management) was worry about potential clones due the fact
of "anything open source" (they actually still pissed of because i draw and published the schematics)

Now, after some presure has been made, they spend some time and understood the GPL thing - they know now
that there is nothing wrong with that and no risk for their products. You might ask where i know that?
Well, they even published GPL sources of a product which is not yet on the market! (but it will be in few weeks) -
there was no need to do this, a 1-3 months after product has been launched would be still perfect from GPL point of view.

[bilko]

But because product management will probably be lead by MBA style thinking, this will not happen and they will miss a great opportunity to make them a name in the industry and the market by making their devices unique. An opportunity to lower their r&d costs, making  bigger chunks of money by driving demand and a respectable name for themselves, that is. And who wouldn't want that actually... but I'm afraid it is their MBA ideology (there you have it!) that hinders them from even realizing opportunities like this, one that is open to them to take on right now in this somewhat difficult situation they brought themselves into.

Well said, I would like to add that Hantekway el al should concentrate their efforts on hardware design and manufacture, which they seem to be reasonably good at. They can easily keep that proprietary and let the hackers loose on their product to fix and enhance the firmware which is where their skills are seriously lacking

[tinhead]

Now, if Hantek were smart,

they are, this is why they refused a proposal of complette open source DSO.

From a marketing point of view disaster in China - it is already (well not for Chinese company to sue another Chinese company
but for other) a disater to sue someone because of copyright problems - but with open hardware/source there is nothing you
can do anymore - you lost.  Day later someone from the street will sell a "bad clone" for 5USD cheaper - and gues what
a typical western buyer will do ? exactly - buy the cheaper product ...

Now, if Hantek were smart,  they'd gone open-source with their scopes from day one.

but there is something they can do! 
they (actually Tekway does) do have older product, 1GSs DSO with only 2.5k memory
and 5.6" display (searc for Tekway DST1102). It runs Linux as well, and a Samsung SoC ...
http://www.microsofttranslator.com/bv.aspx?from=zh-CHS&to=en&a=http%3A%2F%2Fwww.tekwayins.com%2Fproduct.asp%3FArticleID%3D7


This is in my opinion something they can release as open source kit - for small money.
Such product - if copied, cloned - would be not dangerous for "main products line".
Such product could be used by the community to improve the "main product firmware",
it could be used for educational things, it would tie endusers/hackers to HanTekway products.
For 150USD i would buy such kit immediatelly - and i'm sure when in volume produced there
would be still room for HanTekway to made some money.

If something like would works, then there could be a room to think about releasing
the main product line as open source - if not there would be no lose

There are so many ppl buying these cheap crap pseudo-handheld DSO-quad crap things ...
a proper 1GSs scope with only 2.5k sample memory is light years better - and for same amount ofmoney not to beat!

Maybe we should start a voting - to see how many ppl would buy such small kit ?
Who knows, maybe HanTekway wish to make ppl happy and tie them ?

[carloscuev]

Maybe we should start a voting - to see how many ppl would buy such small kit ?
Who knows, maybe HanTekway wish to make ppl happy and tie them ?

Me in, where do I sign?

[Gall]

Me too.

[tnt]

A couple of points:

 - AFAIK The request was never made to the manufacturer itself, but to the re-seller. Since the licence is a 'distribution' license, you can just ask the guy who provided the hw to you and he will in turn need to ask his provider and so on. And since the distributors are in the EU, they will want to comply and will make pressure to the manufacturer to get the source (i.e. "provide us the source so we can comply with GPL or we stop distributing your product" in short).

 - I actually made a couple of GPL request myself. One at Agilent for the 3000-X and another to vodafone. And in both cases I got the sources I asked for. Took time each time ( > 1 month ) but it worked out fine.

[Mark]

What is the best firmware for the DSO-1202B handheld? 
I have firmware version 2.01.1 (111212.0)

How do I get measurements for channel 2 to display? 

Also I found a bug, when viewing a low freqwave, such as the 1khz prob compensation output, trigger on the rising edge and scroll the waveform to view the falling edge in the middle of the screen (I want to do this to measure jitter in the pulse width).  Start with something like 400us/div.  Trigger marker shows approx 500us from rising edge trigger to falling edge in middle of screen. 
Now change the timebase to 20us/div and note how you have to scroll the waveform again to get the falling edge in the middle of the screen again, and now the time between the trigger and the falling edge is displayed as 350us. 
At 8us/div, the time from rising edge trigger to falling edge is displayed as 290us.
At 4us/div, the time from rising edge trigger to falling edge is displayed as 270us.
At 2us/div, the time from rising edge trigger to falling edge is displayed as 260us.
At 800ns/div, the time from rising edge trigger to falling edge is displayed as 255us.  etc etc.  Should be 500us obviously. 

[carloscuev]

Hello, Im running HW 0x555583e8 SW 120224.0 and have been using my scope a lot recently, being no master I found an annoying bug, When having just captured a single seq and pressed any menu button, the graph gets compressed in order to get space for the menu bar, this seems Ok, but if you want to measure time with the cursor, you end up measing an horizontally (time) compressed  graph, aka, you measure more time than the actual real time.

I couldn't made myself clear heres a video:

[tinhead]

What is the best firmware for the DSO-1202B handheld? 
I have firmware version 2.01.1 (111212.0)

How do I get measurements for channel 2 to display?

Mark

There is newer firmware since day or two available on Hantek website, it does not change that much (if any)
but it is GPL compliant^^

Regards the measurments, sure click Meas button, chose with blue cursor keys the item you wish to change, click F5
to modify and select with F1 the source of measurments.

There are 4 measruments visible in same time, however you can scroll with these blue cursor keys between 4 pages - so you can setup on page 1 and two channel1 and pages 3 and 4 channel 2 - then in principle just scrol up and down between
pages.