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

[firehopper]

Lol how did I miss that the update went smoothly. currently running self calibration. we shall see how it works/

its not for my dso5072p/dso5202p or I would have tried

i don't know how is your DSO5xxxP but i think it will work, check what i wrote:

That firmare is compatible with following models:

 - Hantek DSO5072P, DSO5102P, DSO5202P

[firehopper]

so far seems to be good. nice fast responses and seems to be fine. anything you want me to test?

Lol how did I miss that the update went smoothly. currently running self calibration. we shall see how it works/

its not for my dso5072p/dso5202p or I would have tried

i don't know how is your DSO5xxxP but i think it will work, check what i wrote:

That firmare is compatible with following models:

 - Hantek DSO5072P, DSO5102P, DSO5202P

[richardbb]

works good on my DSO5072p as well, sample rate info is useful

how about running a custom script on clicking the "sys status" button, the user could edit it to do what they want. probably not very useful to most people though

[tinhead]

how about running a custom script on clicking the "sys status" button, the user could edit it to do what they want.

that can be done very easy, there is place for 47 instructions in the affected firmware function (DoKeyOption5InMenuUtility1)

[tinhead]

Did you update anything to the logic analyser part except for the HELP button?
I only tried the DSO part so far, except I noticed while trying the HELP button that I still have spurious flashes of random data on all LA channels.

nothing what could have influence, i'm calling the LA in exact same way they doing. These random data, that seems to apears on my device especially when set to 200ns/DIV. As well of course when tri

LA is the most broken part as far as I am concerned.
   not all samples captured in LA memory can be zoomed to
   LA memory seems to be only 4k deep (should be 512k?)
   triggering in mixed mode leaves the LA data trailing the DSO data by one event (seems like it is only a misplaced refresh operation since if you force a display refresh by operating controls it shows the newest samples)

yeah, waiting since a year to get firmware which supports the full memory, shame on them. 49k was already working in one version, but unstable. Bad design? I don't know, i don't see issue why the data can be transfered from one FPGA to ARM but not from the other FPGA. So it must be something else. The other things are bugs which i'm trying to find/track/forward, i can't test everything so every feedback can help a lot

The support from Hantek has been horrible and unresponsive to the point that I have given up any hope they will fix the LA in the foreseeable future. Also beware of their support web site, they log your IP (that is OK i guess) and publish it in full for all posters (NOT OK). Pointed it out 5 months ago and they would look at blanking out parts of it to shield their customers... not a thing happens   

shall i really comment that

with the IP shown all the time was however nice to see that they posted problems with one user, and solved with another, yeah, how smart. Probably they tried to shwo that they doing something.

As a result I have moved on to a DS4014.  First priority on my home lab todo list at the moment is checking out the BW settings on the I2C bus of the DS4014 (looking for confirmation that BW limits are set as in DS2000).

oh well, this is what i told them so many times. They simply not fast enought with bug fixing, no idea why. Maybe it is hard to get good engineers, on thier chinese HR page you can see they have lot of open positions. Now they released new products and probably hoping to get some positive feedback about new hardware .. but i doubt that i will  be able (in case i receive one day the new 4ch hardware, i think/hope it is now on the way ...) to do so, how can be my feedback positive when the hardware, which exists since long time (talking about LA now, DSO itself is almost fixed if i don't think about digital filters or equ sampling), is still having such issues like the missing >4k memory depth in LA. This have to be there! People bought their MSOs with that feature, but it is not there, and that's really crap. One might wonder what have new 4ch model with "old" 2ch models, short answer - many things, as they continuesly improving. So when existing things not fixed, they taking them into new things as well and the discussion starts from begining. Best example - the FFT full span bug in curent MSO firmware. It is fixed in DSOs since long time, but it must be hard to write down what done on one side and to implement it as well on the other side :\

I think sometimes the engineer even don't know that x or y didn't work properly ...  the information flow is sometimes not given, no chance to achieve something without lot of presure. Sure, these products are not really expensive, but that can't be an excuse for bad work, never ever. On the other side this is the learning process, look at ATTEN - for years crap company (and i shall said look on Siglent because all that ATTEN DSOs was made by Siglent team). And Rigol wasn't better years ago as well. Rigol is now much better, Siglent at least good enough for LeCroy, whatever this means.

[bigmessowires]

I've been lurking in this thread for a couple of days now - thanks for all the great information! I don't actually own a Hantek scope, but I'm considering getting one soon. Actually I was first looking at the Hantek MSO with the logic analyzer functions, before I read about the issues with it here. So I will probably stick with one of the DSO models.

I wrote a story on entry-level oscilloscopes, technology and some comparison of a few popular models (including the Hantek) for my blog, and included a shout-out to the thread here for more Hantek details. Check it out if you're inclined: http://www.bigmessowires.com/2013/09/23/entry-level-oscilloscopes/

[tinhead]

I wrote a story on entry-level oscilloscopes, technology and some comparison of a few popular models (including the Hantek) for my blog, and included a shout-out to the thread here for more Hantek details. Check it out if you're inclined: http://www.bigmessowires.com/2013/09/23/entry-level-oscilloscopes/

so far ok, maybe bit short about the (necessary) sample rate:
- for single shot more sample point is always better, just think how many are necessary to see
  the exact shape of e.g. 7th harmonic
- for repetitive signals the "4x bandwidth vs. sampling rule" apply only to DSOs
  with gaussian frequency response, but not with flat response; see Agilent 5988-8008EN

In case of these DSOs from this thread, they does have flat response, which is unique for that price class.
This is good from point of measurment accuracy from zero to near bandwidth limit, but it might be easy
misunderstood (ringing) when measuring signals with rise time significant faster than the DSO can measure
accurately (just watch Dave's measuring very fast edge signal in his Tekway review, regardless of the fact that
he have used broken DSO, the extra ringing was misunderstood).

When talking about necessary sample rate it make sense as well to check if DSO is doing interleaving (and if so, then
how and when), sometimes it is better to sample slower (without interleaving) , see Agilent 5989-5732EN

[tinhead]

I've been lurking in this thread for a couple of days now - thanks for all the great information! I don't actually own a Hantek scope, but I'm considering getting one soon. Actually I was first looking at the Hantek MSO with the logic analyzer functions, before I read about the issues with it here. So I will probably stick with one of the DSO models.

this is question of trust - do you trust manufacturer to be able to fix problems in acceptable time x? If so, buy MSO, if not buy DSO. The MSO itself is nothing else than DSO + additional PCB, so it can be still perfectly used as DSO until bugs fixed.

The annoiyng thing with these MSO is not the fact that only 4k is currently enabled (this is still 4k per channel sampled up to 500MSa/s) or that there is skew between DSO and LA trigger point on higher sample rate (one have to be anyway careful with that, a signal with very fast edge might produce some errors as well):



I don't like more things like these artifacts on LA, sometimes middle of measurment, sometimes when changi ng timebase, sometimes even when you stop acq, zoom in, move waveform and then boom - they appears on screen. That can ruin your work and be very frustrating:



Where other things like errors on 800ns/DIV are probably easy to fix, and easy to ignore if one know that they there:

[dr_p]

I'm back with an irritating problem on my handheld Hantek DSO-8060: I want to make screenshots (BMP, JPG or whatever) via USB, with the push of a button (in the PC app). If I want to document my work, it's annoying to keep plugging in a USB stick, going through the menus and options, wait, unplug USB stick, transfer to PC, save. And do it all over again for every screenshot, because if I screen capture 5-10 images one after another, and then transfer to PC, I don't remember what is what.

I installed the USB driver from Hantek's site, the DSO says PC connected, I have DSO-8060 in device manager.

The Hantek software(v1.04) is crap. I tried it's snapshot button, but it captures the data from the application, not the real screen capture that has many useful additional data (cursors, measurements, trigger status etc plus I like it more ). Even more heart stopping, the application screenshot includes the partially transparent "Save as..." window is on top of it. (as it was fading away in Win7) Genius, guys!!


I tried the USB tool from here: http://www.dreisiebner.at/dso-usb-tool/ but either my DSO-8060 is not supported or I don't understand how to make a screenshot. I tried ticking "handheld" boxes, connect, save, capture... it doesn't work.

I also tried the modified dso-usb-tool that tinhead posted in this thread, specifying it was a semi-working version for handhelds. Same results. I also tried a different USB driver, I don't remember where I got it - same result, nothing works.


Am I doing something wrong?

[tinhead]

I'm back with an irritating problem on my DSO-8060: I want to make screenshots (BMP, JPG or whatever) via USB, with the push of a button (in the PC app). If I want to document my work, it's annoying to keep plugging in a USB stick, going through the menus and options, wait, unplug USB stick, transfer to PC, save. And do it all over again for every screenshot, because if I screen capture 5-10 images one after another, and then transfer to PC, I don't remember what is what.

sorry, but my knowledge about DSO1060, DSO1200 ans DSO8060 is minimal, i know how they looks like inside/outside, i know that they complettly different product line and not comparable the DSO/MSO/Handhelds posted in this thread, but that's all.

The Hantek software(v1.04) is crap.

welcome in club, the PC software for DSO/MSO/Hndhelds mentioned in this thread is crap as well. It is doing something sometimes, but i'm sure the engineer who developed it never worked with DSO nor ever tested the software functionality.

I tried the USB tool from here: http://www.dreisiebner.at/dso-usb-tool/
...
I also tried the modified dso-usb-tool that tinhead posted in this thread
...
Am I doing something wrong?

no, they will not work, they only compatible to DSO/MSO/Handhelds based on this one specific platform, your Handheld is complettly different product line, it is using different hardware, software, firmware and protocol. So no need to try them.

[bigmessowires]

- for repetitive signals the "4x bandwidth vs. sampling rule" apply only to DSOs
  with gaussian frequency response, but not with flat response; see Agilent 5988-8008EN

In case of these DSOs from this thread, they does have flat response, which is unique for that price class.
This is good from point of measurment accuracy from zero to near bandwidth limit

I'll check out those Agilent tech notes, thanks! Regarding the frequency response, do you mean that with a flat response, you can capture the correct shape of the signal at speeds all the way up to the bandwidth limit? But that there will be extra ringing for signals above the bandwidth limit, as compared to a scope with a gaussian responose? I guess I should go read the tech note first before asking. 

this is question of trust - do you trust manufacturer to be able to fix problems in acceptable time x? If so, buy MSO, if not buy DSO. The MSO itself is nothing else than DSO + additional PCB, so it can be still perfectly used as DSO until bugs fixed.

I don't like more things like these artifacts on LA, sometimes middle of measurment, sometimes when changi ng timebase, sometimes even when you stop acq, zoom in, move waveform and then boom - they appears on screen. That can ruin your work and be very frustrating:

Wow, those artifacts look bad. In my case, I think I decided that even if the MSO features worked correctly, I'd rather have a real logic analyzer with complex triggering, a state listing, channel grouping and naming, etc. I have an old HP1631d now, which has pretty good logic analyzer functions, and I'd hate to lose some of the features it has. So I'm probably going to upgrade to a separate DSO and logic analyzer, instead of a single combo instrument.

[dr_p]

thank you tinhead

[bigmessowires]

I read the Agilent tech note Agilent 5988-8008EN about float vs gaussian scope response. Very interesting! But after reading it, I'm not sure if it makes me more likely to get the Hantek, or less. The tech note describes how a flat response scope will be more accurate at measuring signal rise times, and maybe have less timing wobble. But the flat response scope will have more overshoot and ringing than a gaussian response scope. For extremely fast signals the difference is pretty significant (technote figure 2), but even for signals below the scope bandwidth (technote figure 3) there is a difference. Personally I can't think if I've ever needed to measure rise time, so additional accuracy there doesn't seem very valuable, but checking for overshoot and ringing is more common.

By the way, figure 2 in that technote is almost exactly what Dave did in his video review, so it explains why he got the result he did. Cool!

[tinhead]

I think I decided that even if the MSO features worked correctly, I'd rather have a real logic analyzer with complex triggering, a state listing, channel grouping and naming, etc. I have an old HP1631d now, which has pretty good logic analyzer functions, and I'd hate to lose some of the features it has. So I'm probably going to upgrade to a separate DSO and logic analyzer, instead of a single combo instrument.

HP1631d is real logic analyzer, that integrated Hantek LA is definitely for those who don't have room for additional gear or wish to have something fast but maybe not that complex (16ch x 4 state levels is the most complex for it can be setup) logic analyzer. What i would like to see is serial protocol triggering and decoder, for LA and/or analog channels. However, for known reason - just look what happened to Rigol, Tekway/Hantek wasn't really very responsive to such "licensed addon" proposal.

[bigmessowires]

I'm thinking about getting one of the newer Hantek -P models without the long memory, and I'm hoping to confirm a few things:

- Is there any known difference between the DSO5072P and DSO5102P, or will they be exactly the same once they are bandwidth hacked? Maybe the front-end is different, or the probes that are sold with them?

- It seems like some -P models advertise 24K sample memory, but others say 40K. Will they all be 40K after the firmware hacking, or is it an actual hardware difference?

- Do the new -P models require replacing any resistors in the front end after hacking? Or is that only for the -B and -BM models?



I'd also like to do some before and after measurements for the bandwidth hack. I don't have a function generator or other lab equipment, though. Is it worthwhile connecting a scope probe to a 10 MHz or 32 MHz "can oscillator" and measuring that signal? Something like this: http://www.jameco.com/webapp/wcs/stores/servlet/Product_10001_10001_102779_-1 Or is there any other simple way of building a useful test source, that would have a demonstrably different measured signal before and after the hack?

[neggles]

I'm thinking about getting one of the newer Hantek -P models without the long memory, and I'm hoping to confirm a few things:

Having read most of this thread recently I can probably help a bit here.

- Is there any known difference between the DSO5072P and DSO5102P, or will they be exactly the same once they are bandwidth hacked? Maybe the front-end is different, or the probes that are sold with them?

No hardware differences, but the 70/100MHz ones come with 150?MHz probes. the 200 comes with 200MHz probes. Spend $50 on some Texas 250MHz probes and all is well though (or $20 on some cheap chinese 200MHz ones) - I wouldn't trust the frequency rating of the stock probes anyway.

- It seems like some -P models advertise 24K sample memory, but others say 40K. Will they all be 40K after the firmware hacking, or is it an actual hardware difference?

Early firmware versions only offered 24Ksamples, later firmwares give you 40Ksamples. The update to give 24K models 40K samples can be found here: https://www.eevblog.com/forum/testgear/hantek-tekway-dso-hack-get-200mhz-bw-for-free/msg271366/?topicseen#msg271366

- Do the new -P models require replacing any resistors in the front end after hacking? Or is that only for the -B and -BM models?

Not as far as tinhead seems to know, I recall it being mentioned in an earlier post - and the first page thread specifically mentions that it's for SN under 15000, which excludes the P models.
More info RE: resistors (but with no mention of P models): https://www.eevblog.com/forum/testgear/hantek-tekway-dso-hack-get-200mhz-bw-for-free/msg212054/?topicseen#msg212054

I hope I've been of help

[bigmessowires]

Thanks Neggles! According to eBay listings, it looks like the 70 MHz DSO5072P actually comes with only 80 Mhz probes (the PP-90), although the official Hantek website doesn't specify either way. The PP-150 probes are also 100 MHz rated by Hantek, not 150, despite their name: http://www.hantek.com/en/ProductDetail_15_73.html

Not a big deal if you're planning on replacing the probes anyway, or already have some.

[neggles]

Thanks Neggles! According to eBay listings, it looks like the 70 MHz DSO5072P actually comes with only 80 Mhz probes (the PP-90), although the official Hantek website doesn't specify either way. The PP-150 probes are also 100 MHz rated by Hantek, not 150, despite their name: http://www.hantek.com/en/ProductDetail_15_73.html

Not a big deal if you're planning on replacing the probes anyway, or already have some.

You're welcome Huh, I didn't manage to find a listing showing the probe bandwidth, but, like you said, given the cheapness of reasonably-good probes it's not a huge issue.

I've actually got a DSO5072P on its way from an AliExpress seller (here, if you're interested: http://bit.ly/1fDdk5c ) and I've ordered a couple of Uni-T 200MHz-rated probes to go along with it from a local seller down here in 'straya. It should be here within a couple of days, if i'm lucky, so I'll gladly check out anything you'd like me to once I get it

oh, a thing I found out that you might like to know: if you're viewing a thread in the forum, and you search for something with the search box while on that thread page, it'll search just the thread. Trés useful.

[tinhead]

right, Hantek designed then to hve 24k but then decided to allow the use of 40k. For all who bought before that they provided me a upgrade firmware, which was linked above.

Regards resistors, these P models they does have changed frontend. The first part is still the same, and probably it can be still bit optimized, but the part which was responsible for bad frequency response after a hack of 60/100MHz models didn't exist anymore. It has been replaced (THS4504 instead of LMH6552), which seems to run smooth. But i wrote about that already, and when you look on the first page of this thread, you will find a section:

what inside Tekway/Hantek

and in that section all the schematics i've made are posted, as well the schematic of the P model frontend (i don't have P model, but based on pictures i've made schematics of what i can see there).

https://www.eevblog.com/forum/testgear/hantek-tekway-dso-hack-get-200mhz-bw-for-free/msg264085/#msg264085

So long answer short : no, afaik you don't need any changes on P models.

[firehopper]

Thanks Neggles! According to eBay listings, it looks like the 70 MHz DSO5072P actually comes with only 80 Mhz probes (the PP-90), although the official Hantek website doesn't specify either way. The PP-150 probes are also 100 MHz rated by Hantek, not 150, despite their name: http://www.hantek.com/en/ProductDetail_15_73.html

Not a big deal if you're planning on replacing the probes anyway, or already have some.

You're welcome Huh, I didn't manage to find a listing showing the probe bandwidth, but, like you said, given the cheapness of reasonably-good probes it's not a huge issue.

I've actually got a DSO5072P on its way from an AliExpress seller (here, if you're interested: http://bit.ly/1fDdk5c ) and I've ordered a couple of Uni-T 200MHz-rated probes to go along with it from a local seller down here in 'straya. It should be here within a couple of days, if i'm lucky, so I'll gladly check out anything you'd like me to once I get it

oh, a thing I found out that you might like to know: if you're viewing a thread in the forum, and you search for something with the search box while on that thread page, it'll search just the thread. Trés useful.

'
I got one of those 5072p's off ebay, $280 including the fastest shipping from china I've ever seen.. 2.5 days from when it was shipped till it got to me.. I was like WOW

but anyway, mine came with those 80 mhz probes as you said.

[tinhead]

you now, these P90 probes, when you read careful the specs page you will see they all the same (except bit bigger compensation range on 60MHz probes).
And actually whn you check, the PP150 probe (so specified as 100MHz) is working up to DSO full bandwidth, so it is more than 200MHz probe.

However, they do NOT have RF compensation, so the waveform looks like:



where on proper RF compensated 500MHz probes (Testec HF512) it looks like:



or on this one, as well RF compensated 250MHz probe (Texas TX250/II) it look like:

[bigmessowires]

I'd appreciate any opinions on how useful the deep memory is in real life. I think I've narrowed my choices to:

DSO5072P $270 direct from China
DSO5062B $320 direct from China, or $350 from USA supplier

Is it worth an extra $50-$80 to get 1M memory with the -B model instead of 40K with the -P model? From what I've read, using the 1M memory makes the scope UI perform very slowly, and limits the timescale to slower speeds, so maybe it's not very useful anyway?

Most of my work is microcontroller and digital stuff, looking for signal integrity and general troubleshooting at speeds under 16 MHz.

[firehopper]

also the B version may still have the spot for the logic analyzer inside. The P does not.. I might have tried to find a B version if I would have known that and had the extra money.

[neggles]

Quick Q: does anyone happen to know the output voltage and current specifications of the 5000P series' internal PSU, & is there any spare space inside the case? (5000B info would be appreciated too as it's probably the same/similar)

I'll find out for myself on Monday, but I just thought I'd ask anyway (considering possibilities for building a small battery pack into it, or making an external battery pack for portable operation)

Thanks!

[tinhead]

Quick Q: does anyone happen to know the output voltage and current specifications of the 5000P series' internal PSU, & is there any spare space inside the case? (5000B info would be appreciated too as it's probably the same/similar)

for P models not, but for B/BMV/MSO models, see the first page, there are links to schematics. I would say best reference is hw1007 schematic (is the newest i've made).