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

[schlot]

there was just one animated cartoon video as a sample and nothing else when it worked.

Sure i will check on your tips.

Just a cartoon movie? I thought there would be tutorial videos. Anyway, thanks for the information.

Good luck!

[Fixup]

Hi again 

Today I stumbled on a problem with my Hantek DSO5072p that has the DSO5202p firmware and has the two capacitors removed at the AD converters.
When measuring on a differential DDR memory clock signal i noticed that the channels are not in phase even if i measure from the same pin.
It looks like one channel is filtering the signal little more and at the same a small phase shift can be seen.
The picture below is measured from clk_p and clk_n so they should be inverted without the phase shift. (Unfortunately the picture when measured from same point was not saved)


I opened up the DSO5072p to see the input stage.
The RX1_1, RX1_2, RX1_3, RX1_4 seem to be 499ohm on my device (marking 68A). I have read in the forum that the 200MHz model should have 280ohm 0,1% so I will give that a try for to correct the phase shift.
The following picture shows the input stage on my DSO5072p (Bought 2014)

The differential amplifier seems to have the numbers scratched of. It is probably still a LMH6552 as on the other boards.
Maybe the phase shift could be corrected from the trimmer capacitor on the input. Has anyone got any idea about the trimmer?

I will keep you updated how the resistor change at the differential amplifier affects the signals.
 

[Spirit532]

I just received a Hantek DSO5072P, and I was quite surprised to see that it came "hacked" from the factory.
It can go down to 2ns/div, the firmware says it's a DSO5072P, sw is 3.40.0(150521.0), hw is 1000xffff8108, serial number is >3000.
And I am absolutely certain it came right from the factory, since it was still sealed by the original branded scotch tape, plus the seller told me that they never touched it.
To say the least I'm pleasantly surprised, but I'm also quite confused as to why this happened.

[schlot]

I just received a Hantek DSO5072P, and I was quite surprised to see that it came "hacked" from the factory.
I can go down to 2ns/div, the firmware says it's a DSO5072P, sw is 3.40.0(150521.0), hw is 1000xffff8108, serial number is >3000.
And I am absolutely certain it came right from the factory, since it was still sealed by the original branded scotch tape, plus the seller told me that they never touched it.
To say the least I'm pleasantly surprised, but I'm also quite confused as to why this happened.

As I said, the same applies to my DSO1122S.

[Fixup]

Hi again,

In a erlier post I had some problems with skew between the channels when I measure the same signal on both channels. This is a very annoying problem since I would need to measure a 133MHz differential memory clock signal and USB signals that is also a differential signal. This problem has been discusses on page 27 on this topic. It seems like the the skew is varying also depending on scope settings. Has anyone got any information about this or is there any fix available for this?
In my opinion they could delay one signal in software so that the signals would show up right on the display.

When studying this problem I also modified the input stage little. I changed some resistors and changed the j-fet to mmbfj309LT1.

[schlot]

Has anyone encountered strange DMM readings on Hantek Handhelds?

I have a DSO1122s which is apparently hacked from factory (I have 2 ns div without hacking anything).

Now I try to use the DMM function to measure resistance. The readings were very odd, something like 48 Ohms on a short. I tried to measure resistors and got readings which are way to high or too low.

Continuity mode beeps all the time, but correctly shows "E Ohm" on open circuit and "0 Ohm" on shortcut.

Now the strange thing: When I switch to Amps measurement and afterwards to resistance or continuity, everything is OK. When I switch to voltage and then to resistance or continuity, readings are bad. Switching between V and A gives a "click" noise from the device (maybe a relay switch). Is it possible that the software erroneously reads a floating pin if modes are used in this order? Measurements are clearly correlated to the connected pins, you can see a reaction when shorting the pins during resistance measurement. But the readings are slow and severely off.

Any idea?

[commongrounder]

I have the same DSO1122S scope you have.  I have probably used the meter function on this scope only two or three times, as it isn't well protected (or nearly as accurate) compared to my Fluke meters.  I did give it a go this morning and all of the functions performed normally.  I do hear quite a lot of relay switching when changing meter functions, so I would suspect faulty relay contacts as the first thing.  Have you repeatedly changed functions to exercise the relays to see if that changes anything?  The next steps are probably going to be investigating cold/broken solder connections.  Also, what firmware version are you using?  Whatever you do, do NOT use the firmware posted on the Hantek web site.  It will brick your scope.  As far as I know they have not fixed this yet (someone feel free to update me on that).
Good luck!

[schlot]

I have the same DSO1122S scope you have.  I have probably used the meter function on this scope only two or three times, as it isn't well protected (or nearly as accurate) compared to my Fluke meters.  I did give it a go this morning and all of the functions performed normally.  I do hear quite a lot of relay switching when changing meter functions, so I would suspect faulty relay contacts as the first thing.  Have you repeatedly changed functions to exercise the relays to see if that changes anything?  The next steps are probably going to be investigating cold/broken solder connections.  Also, what firmware version are you using?  Whatever you do, do NOT use the firmware posted on the Hantek web site.  It will brick your scope.  As far as I know they have not fixed this yet (someone feel free to update me on that).
Good luck!

Oh. I didn't know that firmware thing, so thanks for the hint about the corrupted Hantek link.

edit:
I just checked: The DMM once again did not work after startup: Continuity beeps all the time without cables connected and resistance shows random values. I restarted the DSO and again switched to DMM mode. Then everything seems ok. Voltage readings sometimes are a bit strange as it seems to "charge" as soon as nothing is connected. Open circuit, the voltage meter reaches about 400 mV which I think is a bit too much for background noise.

I hear relay switching noise whenever I change from a DMM mode with a voltage source to a mode with high impedance.

No clicking between V <> A.
No clicking between Diode <> Continuity <> Resistance <> Capacitance.
Clicking between V <> Diode, V <> Continuity, V <> Resistance, V <> Capacitance, A <> Diode, A <> Continuity, A <> Resistance, A <> Capacitance.

The SW version is 2.01.1 (150108.0).
HW Version is 1004 0x57029100

[commongrounder]

I checked my unit and the relay clicking follows the same pattern as you have determined.  My open circuit DC voltage fluctuates between about 15mV and 300mV, depending on what mode I am switching from.  Going between Resistance and Volts DC it settles the closest to zero. 
You seem to have the most recent firmware on your unit.  It is later than what I have on my unit - 2.01.1 (140508.0). 

[schlot]

Thanks. It seems to be a common behavior of the DMM unit. At least the DSO part works as I expect it.

Is the number in brackets of the SW version the build date (YY-DD-MM)?

[commongrounder]

Thanks. It seems to be a common behavior of the DMM unit. At least the DSO part works as I expect it.

Is the number in brackets of the SW version the build date (YY-DD-MM)?

Yes.  That is my understanding.

edit:

I forgot to mention there is a Hantek forum where you could raise your scope's meter mode problem:
http://www.hantek.org/asken/
It can be a little hit or miss, as there can sometimes be a language barrier, but it wouldn't hurt to try.

[schlot]

Thanks for that hint. Somewhere I have already seen that board. There is somebody who complained about a defective resistance measurement function and hantek answered they will repair the unit.

My thought was like "C'mon, at eevblog there are people who analyze and modify the input stage and know about each and every part in the DSO. They know the device better than the manufacturer does" :-)

[commongrounder]

The real expert on Hantek stuff (and the other brands made by them) is the OP of this thread tinhead.  But he hasn't posted since January of this year, so I am not sure what his status is.  He has done a lot of hardware analysis, as you can probably tell by browsing this thread.

[Fixup]

Hi,

I made a modification to the analog filter on the input stage on my Hantek DSO5072p.
Now the varactors D01_2 and D02_2 get 10V when the filter is off. This should give a capacitance less than 2.7pF. The cutoff frequency will thereby also be higher when the capacitance is lower.

This could be useful when measuring higher frequencies.
I also increased the amplifiers filter frequency to 350MHz (Made a file /tst in root containing [filter] 35)
Frequenciy settings over 350MHz was having distorted shapes and therefore I chose to leave it at 350MHz.

Please comment what you think about the modification?

[tupac]

after nearly a year of absence i'm back. I have still to check/reset my emails and other data to be able to login here again as "tinhead", so for now just with temp account. I bet there are lot of PMs in "tinhead"s inbox, to make a clean cut i will simply purge them all, if you have still something important please re-send PM (in a week or so, i need to "reboot" myself first).

tinhead

[RavenSpy]

after nearly a year of absence i'm back. I have still to check/reset my emails and other data to be able to login here again as "tinhead", so for now just with temp account. I bet there are lot of PMs in "tinhead"s inbox, to make a clean cut i will simply purge them all, if you have still something important please re-send PM (in a week or so, i need to "reboot" myself first).

tinhead

Welcome back!

[Fixup]

Hi,

I documented the changes that I've made to my DSO5072p. Hopefully they will be to interest for someone. I am trying to modify the input stage so that it would be more stable to a higher cutout frequency than the original one has. t would also be nice to get the noise in the oscilloscope lower.

Yes, It would probably be cheaper and easier to buy a better oscilloscope. But this is little like a challenge and it is good also for learning purpose. 

Please give some comment about the changes or point out if you see some other tweaks that you would recommend or not.

Thx Tinhead for the information found on the forum!!!


 
 

[Macbeth]

after nearly a year of absence i'm back. I have still to check/reset my emails and other data to be able to login here again as "tinhead", so for now just with temp account. I bet there are lot of PMs in "tinhead"s inbox, to make a clean cut i will simply purge them all, if you have still something important please re-send PM (in a week or so, i need to "reboot" myself first).

tinhead

Yeah! Welcome back tinhead!

After 12 months doing bird you certainly need to stretch your wings before returning to the fold 

[TheManual]

Sorry,
Can anyone advice the  .wk file structure (it is the output file of
tektronix software "wavestar for oscilloscopes").
Is it possible to get the  arrays of points from this file?

[luizhdinizf]

anyone knows a way to comunicate this instrument with c, python or another interface with a ubuntu 14.04 computer over usb?

[hobbes_dmS]

Hi,

I've got an Hantek MSO5062D
Software Version 3.2.35 (141129.1)
Hardware Version 1020x55778344
Serial Number T 1G/012 023xxx (so higher than 15000)

In this version the PINs for the serial interface are completely missing, but I just plugged some in.

I can connect with my USBtoSerial adapter and see the starting of Linux. But after booting the console cannot be reached by pressing "Space and Ctrl-C" (and enter)
I've googled and read a lot of pages of this topic but I did not find any solution to this problem (nor anybody else who has this issue).

Any hints? what did I miss?
If it's any help, I use linux with putty to connect.

Oh, and no matter which backup.up file I put on the USB Stick to get a backup, the update-menu will always tell me that the software Version I'm trying to install is too old and that there's a newer one already installed...

Any help is appreciated.
Thanks!
Wolf

[Fixup]

Hi.

I see that you have 4 wires connected. 3 wires should be enough.
-Don't connect that+3.3V to the USBtoserial adapter.
-Check that RXD line is connected to the right pin.
-Try to use your adapter alone and see that the characters are echoed. This can be done by connecting txd and rxd lines together on the usbtoserial adapter. The character should ehco when you type a letter.
-Check that flow control and parity is set to None.
-Check in device manager that the comport has 115200 speedsetting. This is normaly 9600.

For the backup I used files from "Tools_P_models.zip" to my P type osciloscope. Check that the backup is for your D model. I don't know if there is some difference.
In the "Backup_P_models" folder there is a text file describing the commands. Possibly you could run the backup commands from commandpromt if you get it to work.

Hope that these advises help  

[hobbes_dmS]

Hi.

I see that you have 4 wires connected. 3 wires should be enough.
-Don't connect that+3.3V to the USBtoserial adapter.

it's not connected (sorry I forgot to write that)

-Check that RXD line is connected to the right pin.

yes, rdx to tdx and vice versa. otherwise the terminal window stays empty

-Try to use your adapter alone and see that the characters are echoed. This can be done by connecting txd and rxd lines together on the usbtoserial adapter. The character should ehco when you type a letter.

hmm... then I just see the character I type (once, as in a 'normal' terminal

-Check that flow control and parity is set to None.

was not, changed it, no difference

-Check in device manager that the comport has 115200 speedsetting. This is normaly 9600.

set in putty - do I need to set it somewhere else?

For the backup I used files from "Tools_P_models.zip" to my P type osciloscope. Check that the backup is for your D model. I don't know if there is some difference.
In the "Backup_P_models" folder there is a text file describing the commands. Possibly you could run the backup commands from commandpromt if you get it to work.

Thanks, I'll try that.

And the next step is probably to try on a Windows machine...

edit: btw, when i type Ctrl-C, the terminal writes ^C - is that normal? or does it escape anything it should not

[Fixup]

Hi.

I made a script for the oscilloscope so that the tst file (filter variable) is copied from the USB stick to the root.
This is good if you wan't to use different filter settings, then you just need to modify the tst file on the USB memory.
The file will be copied if it is different or missing from the oscilloscopes memory.
The file is copied when USB memory is mounted. In that way the tst file will stay on the oscilloscope even if you power it off.
The tst file is removed from the memory if the tst file is removed from the USB stick.

Please try it on your own risk And also let us know if you see some issues with it or if it can be improved.
 

[timofonic]

This is amazing.

I'm sure that with enough funding and reviewing, some people involved in this forum thread can be able to be part in a project to make a really nice "FOSSHW" oscilloscope.

It's a shame the best reasonable and not so expensive oscilloscope is the Rigol 1054Z, yet they fail to fix certain bugs even if tons of hardcore geeks pointed at them. They are getting so much interesting feedback and free debugging, hey aren't able or don't want to contribute back in form of fixing them.

These cheap ass companies don't deserve this free labour.

Fortunately, their messes inspired many people to become better skilled at making an oscilloscope then the companies themselves. But a project like this isn't easy at all and would require a massive teamwork plus funding, something not so common in electronics.

I understand developing even a basic oscilloscope isn't easy, but I wonder why a somewhat resourced company isn't able to deliver an okay product.

What's the magic smoke behind an oscilloscope? Isn't it a very glorified voltmeter after all? I understand it need very precise readings and filterings that even an high end DMM lack, but I'm unsure about the hardware cost and difficulty to design one by using modern hardware (you can get a nice ARM SBC that is quite powerful for a tiny fraction of a cheap Rigol, for example). Is the cost related to hardware design and software development? being a niche market, money eating corporations with too much fame?

Sent from my ZTE Blade A450 using Tapatalk