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

[morris6]

Back again..

On the 23rd of January, during a test after the MSO modification, I blew the s3c2440 of my scope. A few days ago I managed to get the broken chip off the board. Today I had a go at soldering a new chip in place. Miracle.. It is working!!

awsome!

Further testing now to find out if the LA board was damaged at the disaster of a month ago.

let's hope the best. In principle when you appy 3.3V to the LA pcb the fpga should start working, when you see
the proper ID ovr JTAG then at least this is working. The next step is lan ic, when current consumption is not more
than 100mA everything will be fine, then you can plug the LA PCB to DSO and run other tests.

In wort case the io communication between both FPGAs will not work (due the applied reversed voltage to io21 and 22),
then you will have to check which one of the FPGAs is broken (let's hope none, they much harder to solder than the SoC),
the best way is over jtag to test with boundary scan the status of FPGA pins/ports.

In best case nothing else broken, hope the best for you.

I managed to boundary scan this FPGA, the IO21 and IO22 pins are working, input and output. The FPGA of the main board I did scan as well but I am in the blind here. I don't know what to look for: the hw version 1007 is totally different from the published 1003.

There must be something wrong since the LA FPGA is not loading its program.
The LA_nCONFIG pin, number 6, appears shorted to GND, that doesn't look good. It is next to the GND pin at pinnumber 8. It was connected to GND of the main board during the "one row off event" while the GND of the LA board was at 5V level. I guess this FPGA pin is killed.
The main board, when the LA board is removed, gives high- and lowlevels on this pin during start up. Also a clock signal appears for a few seconds on LA_DONE_CLK. That seems to be OK.
To me it looks like I will have to solder a new FPGA on the LA Board.

Anyone with other suggestions or ideas?

[tinhead]

I managed to boundary scan this FPGA, the IO21 and IO22 pins are working, input and output. The FPGA of the main board I did scan as well but I am in the blind here. I don't know what to look for: the hw version 1007 is totally different from the published 1003.

there is something i can do for you, attached FPGA schematic from hw1005 (i think it's ok, not 100% sure) and hw1007 (verified)

There must be something wrong since the LA FPGA is not loading its program.

there is as well JTAG port on LA PCB, use it (the best would be to test while the LA PCB is NOZ mounted in DSO).

I've attached small design, load it over jtag, it should produce 25MHz clock on FPGA pin1 (the clock for LAN IC)

EDIT: attachement removed, please check the full schematic link on first thread page

[morris6]

Thanks very much, Tinhead. Once again you are very helpful.

I will load the design and try it. I was thinking about a something like this. I was also not sure about the nCE pin in the LA schematic, it is not connected. Now in the main FPGA schematic I see that there it is connected to GND, as expected.

[mazurov]

Just got my Hantek DSO5102B. Bought from goodlucksell on eBay for $388.88 shipped. Sw. version 2.06.3 (111025.0), hw.version 10070x555583e8. The scope came with the fan installed; first impression - fan is less noisy than Tek7104 but more noisy than Tek475a. Overall - nice liitle scope.

Running some baseline tests at the moment, will start hacking/taking pictures later today.

[tinhead]

Just got my Hantek DSO5102B. Bought from goodluckbuy on eBay for $388.88 shipped.

hmm

Sw. version 2.06.3 (111025.0), hw.version 10070x555583e8.

this is old stock, Hantek stopped 83E8 production somewhere in Nov 2011.
I would send it back or ask the seller to reduce the price.

Sure, you can replace the FPGA design with the 83E9, but hmm, i don't knwo, i would not pay full price for (not matter if really) "used" or old stock device.

The scope came with the fan installed; first impression - fan is less noisy than Tek7104 but more noisy than Tek475a. Overall - nice liitle scope.

well, replace the 7812 by 7808, turn fan in oposite direction - or replace the KA378R33 by dc/dc converter or Micrel HELDO.
The noise will be gone, or at least reduced enought to enjoy the scope.

[mazurov]

Thanks for the hint - I'll contact the seller and see if I can get some money back :-).

[mazurov]

Here's the risetime - BFR505 avalanche pulser, ~100ps risetime, ~10ns pulsewidth, 50 ohm in-line terminator. Second picture taken on 1G real-time analog scope w/50 ohm input. Third picture shows measurement setup. It would be interesting to measure the step right on the frontend to see how much crap is added by processing.

[tinhead]

best case - so after the hack - you can get something like this

https://www.eevblog.com/forum/testgear/hantek-tekway-dso-hack-get-200mhz-bw-for-free/msg127529/#msg127529

This is taken with internal attenuator disabled (=< 2V/DIV), i've used external 20dB. And yeah, don't wonder about the 1ns/DIV, special deep hacked version heh.

Btw, depends on firmware version the DSO might display wrong rise time, but yeah, you will see it then.

[mazurov]

Couple more pics - resistors in input stage and heatsinks on ADCs/FPGA. According to what I've read so far this is 200MHz input stage, is this correct?

[tinhead]

According to what I've read so far this is 200MHz input stage, is this correct?

well, yes and no. I've attached my current schematics (hw1005 and hw1007), as you can see since Jan 2012 Hantek is implementing
in every model the same combnation of 4 x 280R and 22R array (so far i know). I didn't saw real 200MHz model since a
long time, and Hantek/Tekway will no tell me the actuall values, so maybe real 200MHz are still different, but maybe not.

I personaly don't the current values, the circuit is overshooting up to +3dB at arround 150MHz, my combination with 4 x 270R and 33R array works
for me much better (deviations +- 0.6dB only). However my combination is the best for my probes (or 50R termination), with orignal probes it might
be that Hantek's solution works better (i never tested this, no reason as i use better passive or active probes).

I think i should post some frequency response measurments (with different J-FETs and some combination of resistors), some day

EDIT: attachement removed, please check the full schematic link on first thread page

[mazurov]

Got the hack done. Thank you, tinhead - this was a lot of fun!

[mazurov]

Risetime post-hack w/latest fw.

[mazurov]

if someone wish to test - here another one firmware for

 - Tekway DST1062B, DST1102B, DST1202B
 - Hantek DSO5062B, DSO5102B, DSO5202B
 - Voltcraft DSO3062C, DSO1062D
 - PCE DSOs: PCE-DSO5062B, PCE-DSO5102B, PCE-DSO5202B

http://www.mikrocontroller.net/attachment/170890/dst1kb_2.06.3_15202b_fact_130221.0_.zip

Is there a place to file bug reports?

[tinhead]

post it here, i will put on my list (an forward to Hantek/Tekway engineers)

[mazurov]

Here's the one - not sure if it's due to FRR measurement implementation, or the way the sampling implemented in this scope or me not quite understanding the instrument :-); likely the combination of all three.

The setup: two Hantek 160MHz probes (came with the scope) connected via a T to the output of the generator outputting 1.4V p-p sinewave. The scope is triggered from generator's sync out. Next screenshot shows ~1nS delay between channels as measured by FRR (output frequency 50MHz). The next screenshot shows output frequency (and consequently triggering rate ) of 5MHz, measured difference of 150pS is the best in series, could go as high as 800pS with no visible increase in trace separation.

The probes are pretty well matched for the price. The following two screenshots show the same probes moved to the analog instrument (Tektronix 7104, 7A26 plugin, chop mode, 7B10 timebase ). Swapping probes between channels of the Hantek makes no difference.

[snowl]

I might found another bug, not sure if it is considered normal...
SW: 2.06.3_130221
HW: 1007 /3Eb
When you use the tracking Cursors when the Acquisition is stopped and you change the Timebase or the V/Div Settings, the Cursors will still follow the "old" Curve instead of the now visible one.

[morris6]

There must be something wrong since the LA FPGA is not loading its program.

there is as well JTAG port on LA PCB, use it (the best would be to test while the LA PCB is NOT mounted in DSO).

I've attached small design, load it over jtag, it should produce 25MHz clock on FPGA pin1 (the clock for LAN IC)

I learned here that while pin to pin boundary scan is working the chip core is not guaranteed to be allright. The LA-FPGA was not loading the configuration file over JTAG. The shorted nCONFIG pin I found earlier made me decide to change the LA-FPGA.

Another soldering miracle was performed. With the new FPGA the start up log file shows the config file is loaded. However I had to overcome the pinnumber issue with the cable connector for the frontpanel connection of the LA. In the schematic the odd numbered pins are GND. The connector on the board is wired even pins to GND. Looks like the pin connector was on the bottom side during the board design stage. The silkscreen marking "1" is at the position of pin 2 now. When I sorted this out I made another adapter and now the Logic Analyzer is working, Finally!

The DM9000 chip most probably also not survived the accident because I can trace no network activity. The LnOE net is also routed to pin 1 of the DM9000 and this net was relatively negative 1.7 V. during the accident. Another soldering job; but I'll have to find a new DM9000 chip first.

[tinhead]

I learned here that while pin to pin boundary scan is working the chip core is not guaranteed to be allright. The LA-FPGA was not loading the configuration file over JTAG. The shorted nCONFIG pin I found earlier made me decide to change the LA-FPGA.

sure, if config pins are gone the scan will not show it. Additionally scan is as good as test matrix, but that's another story.

Another soldering miracle was performed. With the new FPGA the start up log file shows the config file is loaded.

that's good

However I had to overcome the pinnumber issue with the cable connector for the frontpanel connection of the LA. In the schematic the odd numbered pins are GND. The connector on the board is wired even pins to GND. Looks like the pin connector was on the bottom side during the board design stage. The silkscreen marking "1" is at the position of pin 2 now.

no comment on this

When I sorted this out I made another adapter and now the Logic Analyzer is working, Finally!

and that's really good news!

The DM9000 chip most probably also not survived the accident because I can trace no network activity. The LnOE net is also routed to pin 1 of the DM9000 and this net was relatively negative 1.7 V. during the accident. Another soldering job; but I'll have to find a new DM9000 chip first.

i will send you new chip

[morris6]

i will send you new chip

Well, what can I say, you really are very helpfull. I really appreciate this, great.

[mazurov]

Another bug:

If a USB flash drive is connected before power is turned on (for example, disk left connected after power was turned off the previous day) it won't get mounted - one needs to unplug and replug it. I probably even know where to look for a fix (had it myself while developing USB stack for some other platform) - ask me off-list if interested.

[Crenwick]

Hi,

I'm very happy with the MSO mod I made to my DSO (Thanks Tinhead  )

I was just thinking about from where (Tekway or Hantek) does I have to download the updates ?

The startup logo is Tekway (But I changed it back to my custom...)

I read that the last version of Hantek has a lot of bugs so I suppose it's best to wait 

Regards
Bernard

[crackintosh]

First of all - MANY MANY thanks to tinhead!!!
Great job! Great results!!! Great man!
Now I have 200Mhz DSO1202B (made from 1062B) with latest FW 2.01.1_DSO1202B(121229.0).
Till now - all is Ok.

Somebody have Hantek DSO1202B ?
Please read "disp" file from the root directory and place it here...
It can be read simply via USB using DSO-USB-Tool.
If need some help in reading it - just contact me via PM or:
Skype: er1asd
ICQ: 279972305

Thank you.

[tinhead]

Somebody have Hantek DSO1202B ?
Please read "disp" file from the root directory and place it here...

see attached ...


#############################################################

How to create own logo compatible with:

- Hantek Handheld DSO1062B/BV, DSO1102B/BV, DSO1202B/BV
- Hantek Handheld DSO1062S, DSO1102S, DSO1152S
- Tekway MSO MST1062B, MST1102B, MST1202B
- Hantek MSO MSO5062D, MSO5102D, MSO5202D
- Voltcraft MSO MSO5062D, MSO5102D, MSO5202D
- Hantek DSO DSO5062BV/BMV, DSO5102BV/BMV, DSO5202BV/BMV

(NOT FOR Hntek DSO5xxxB, Tekway DST1xxxB)

Simply download attached logo tools and read the readme.

You will in principle have to edit the (example) bitmap, save it, convert to 'ebm', merge (before splited) disp file and use it of course.
For Handheld you will need to find first the locatin of the bitmap, attached tools contains everything for benchtop models,
but yeah, to find the bitmap in disp file is 5min work.

[Mike Warren]

I just received a DST1102B but can't get TTScope to connect to it. The error message is:

---------------------------

TTScope
---------------------------
Not exsit configuration files to this machine!
---------------------------
OK   
---------------------------

Initially, I couldn't get the driver to work, but the version supplied in this thread solved that. The version of TTScope I have is 1.0.0.4, which is the same as the one on the Tekway site. I tried downloading the 1.0.0.5 version supplied in this thread several times, but the download is always corrupted. I also tried downloading Software.zip from the Hantek site, but that is corrupted as well. Not to mention the incredibly slow download rate of < 6kB/s.

Is there somewhere else I can download 1.0.0.5 from?

The driver seems to be fine as the DSO-BW-Changer tool works fine.

Edit: Oh, and I'm running Win7/64

[tinhead]

Tekway haven't updated the app since longer time, but the Hantek version of TTScope is working as well

check the "DSO5000B Series" section on

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