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

[pascal_sweden]

Is it in the same range as the Siglent or Rigol  what? They both make  more than 1 model. Also, instead firing off questions for every little thing, is there some reason you can't look up most of this information yourself?

Of course I can look at the specs, but specs don't say a lot.
There is overall noise floor of the oscilloscope, and then there is also noise on the offset level.
Moreover some people might have both Hantek and Rigol or Siglent, and can contribute with their comparison and experience on both scopes.

[Macbeth]

Is it in the same range as the Siglent or Rigol  what? They both make  more than 1 model. Also, instead firing off questions for every little thing, is there some reason you can't look up most of this information yourself?

Of course I can look at the specs, but specs don't say a lot.
There is overall noise floor of the oscilloscope, and then there is also noise on the offset level.
Moreover some people might have both Hantek and Rigol or Siglent, and can contribute with their comparison and experience on both scopes.

You still aren't being exact enough for CustomEngineerer who clearly can't offer anything either way.

Of course in this thread we use common sense and know that you are referring to the many Hanteks that are all based on the same DSO board, compared with the popular low cost equivalents, Rigol DS or Siglent SDS equivalents in the low price range. All scopes that are crippled by software. Though hardware shows much room for improvement by all accounts of pioneers and modders over here. 

Unfortunately I don't have an answer to your question, but others might. I certainly don't trust China specification sheets (and joke "cal certs" where the red stamp is overprinted  ) at face value. But Mr CE is an arse on this one.

[tinhead]

Sorry tinhead, but... 

THE KING IS DEAD, LONG LIVE THE KING!...

Welcome, tatus1969! 

hrhr, no big deal

Reached my goal: Gaussian response, as much bandwidth as possible.

my goal was to get as much as possible "affordable" DSO with flat (and not gaussian) response, to replace my broken TDS754D, which i managed to get with my hardware hacks. Happy to see, that you reached your goal.

Btw, it seems they have some new engineers, they changed a lot of things in latest fw versions, equ sampling is implemented (that was funny to see), 9 digits frequency counter (at least on my old hw version MSO), some bugs (where is was begging for) has been finally fixed. Bit late, yeah, but at least my new DSO8202E have all the new things as well (except the 9 digit frequency counter) implemented.

How is that on other (latest hw) models?

[Macbeth]

Sorry tinhead, but... 

THE KING IS DEAD, LONG LIVE THE KING!...

Welcome, tatus1969! 

hrhr, no big deal

Reached my goal: Gaussian response, as much bandwidth as possible.

my goal was to get as much as possible "affordable" DSO with flat (and not gaussian) response, to replace my broken TDS754D, which i managed to get with my hardware hacks. Happy to see, that you reached your goal.

Btw, it seems they have some new engineers, they changed a lot of things in latest fw versions, equ sampling is implemented (that was funny to see), 9 digits frequency counter (at least on my old hw version MSO), some bugs (where is was begging for) has been finally fixed. Bit late, yeah, but at least my new DSO8202E have all the new things as well (except the 9 digit frequency counter) implemented.

How is that on other (latest hw) models?

Great to see you back tinhead 

I can't say my MSO5102D firmware has been updated by Hantek for a long time, it's still listed as "MSO5102D Firmware_V3.2.35(For Hardware V3.x.xx)". Let's face it the MSO part of it is nearly unusable

[aandrew]

I've got an MSO5074FG, which is similar enough to the other DSOs that I can recognize the input sections and have a pretty good idea what to change to upgrade the bandwidth.

My question is more on the software side; has anyone been able to drive any of the hardware and bring up basic scope functionality? I've noticed some UI issues and particularly some triggering issues with the software and I wouldn't mind trying to see what I could do to help.

[tatus1969]

with flat (and not gaussian) response

That is pobably a bit philosophical subject. Maximally flat response reproduces the slope best but adds significant overshoot. Gaussian response is slower but has no overshoot. Depends on what properties of your signal you are interested in. By the way, I think that your solution should actually be close to Gaussian, and the overshoot is probably mostly software related (sinc filter). Would be interesting for me if you could post a screenshot with equ sampling time.

Btw, it seems they have some new engineers, they changed a lot of things in latest fw versions, equ sampling is implemented (that was funny to see), 9 digits frequency counter (at least on my old hw version MSO), some bugs (where is was begging for) has been finally fixed. Bit late, yeah, but at least my new DSO8202E have all the new things as well (except the 9 digit frequency counter) implemented.

How is that on other (latest hw) models?

It is still crashing now and then, but not annoyingly often. And it seems to be related to specific (exotic?) settings.
I got a new firmware package from Hantek support, and can share it if you are interested. It is not on their website. And important to mention: they seem to have changed their crypto keys.

[tinhead]

By the way, I think that your solution should actually be close to Gaussian, and the overshoot is probably mostly software related (sinc filter). Would be interesting for me if you could post a screenshot with equ sampling time.

sure, will make some in 2-3 weeks, have first move to my new house

I got a new firmware package from Hantek support, and can share it if you are interested. It is not on their website. And important to mention: they seem to have changed their crypto keys.

are you sure that you got firmware for proper model?

EDIT: and yes, post that fw pls, as i do collect all versions on my 1drv

[zsoltyfm]

Firmware upgrade problem.
This is a non-hacked Hantek dso5202b.
The source of the update: http://www.hantek.com/down.aspx?url=http%3a%2f%2fwww.hantek.com%2fProduct%2fDSO5000%2fDSO5202B_Firmware2.zip
What should I do?

https://youtu.be/BgPsEDDNBOQ

[tatus1969]

are you sure that you got firmware for proper model?

They have firmware update packages for the 5000B on their website. But after I applied that I had a smaller revision number than before. The scope was still working though. The package that I sent you is from Hantek support.

[deathwarror]

I just bought a MSO7000BLG series used Oscilloscope for $400. 
They Changed the format of the front side bus trying to make it smaller.  The scope uses the Samsung S3C2416XH-40.  The Internal File Structure Appears to be the same, and the bandwidth hack apparently still works on this version. LAN is availible to be added.  Can't seem to Identify the memory chip.  It may be the 1GB NAND (K9F1G08U0E) that I'm assuming for the software.

Most of the internal boards except the main are marked with DSO5000E.


I'm planning on doing the Lan hack, 2GB memory hack, and seeing if I can add a Bluetooth UART to the debug connection.

With the 300MHz (aka dst1302b) hack turned on I measured the following data points from channel 1.
With 1V set @ 1KHz
800mv @ 80MHz
740-760mv @ 150MHz - 400MHz
1.1V @ 480Mhz
@ 490MHz Hardware Side Failed to read Properly
@ 500MHz Software Side Jumps by 12MHz and has long since been inaccurate (around 300MHz).
@545.9MHz it finally started reading 700-680mv jumping from accuracy and knee.

Take all measurements with a grain of salt since the 250KHz - 40GHz generator has been damaged before by other students.  I do have pictures of the higher end frequencies to prove it. Unfortunately, they are too big.

[TheSteve]

Does the scope have 50 ohm input impedance as an option? It is rather odd to see the amplitude go up @ 480 MHz unless you're seeing a bunch of reflections.

[tatus1969]

I just bought a MSO7000BLG series used Oscilloscope for $400. 
They Changed the format of the front side bus trying to make it smaller.  The scope uses the Samsung S3C2416XH-40.  The Internal File Structure Appears to be the same, and the bandwidth hack apparently still works on this version. LAN is availible to be added.  Can't seem to Identify the memory chip.  It may be the 1GB NAND (K9F1G08U0E) that I'm assuming for the software.

Most of the internal boards except the main are marked with DSO5000E.


I'm planning on doing the Lan hack, 2GB memory hack, and seeing if I can add a Bluetooth UART to the debug connection.

With the 300MHz (aka dst1302b) hack turned on I measured the following data points from channel 1.
With 1V set @ 1KHz
800mv @ 80MHz
740-760mv @ 150MHz - 400MHz
1.1V @ 480Mhz
@ 490MHz Hardware Side Failed to read Properly
@ 500MHz Software Side Jumps by 12MHz and has long since been inaccurate (around 300MHz).
@545.9MHz it finally started reading 700-680mv jumping from accuracy and knee.

Take all measurements with a grain of salt since the 250KHz - 40GHz generator has been damaged before by other students.  I do have pictures of the higher end frequencies to prove it. Unfortunately, they are too big.

Would be interesting if you could post pictures from the input stages. How have you terminated the transmission line at oscilloscope side, was that inline termination or does the scope have it integrated? It it was the first, then the frequency response does not surprise me. The DSO5000B's input stage had quite long traces adding inductance, and they had not compensated them. This caused excessive ringing, also when feeding 50R terminated signals.

[deathwarror]

I did direct line from waveform generator to scope, non terminated.  Would expect reflections.

The BNCs are most likely non terminated. Have identified the DM9000EP location and the on board DDR2 SRAM is 64mb from the serial.

Images are here: https://1drv.ms/f/s!AnmmyurkkYaJilHJ6x7BgxpD3wSC

Looked at i2c.log for hardware version number and it showed "102"

[tinhead]

ohhh MSO7000BLG, interesting to see that they not populated some sampling memory on both FPGAs, but there is no higher model but only with 32k (yet? who knows, they already have 3GHz SA - which was a shock for me as i never thought they will develop one, and now just found another one new gear from Hantek, http://hantek.com/ProductDetail_8230.html).

Wondering as well what for are these UC22/23 near trigger section. It looks like additional drivers, probably you can see all 4 channels on these resistors R192/193 and R258/259.

For LAN hack it looks like on other models, LAN or rs232 with optocouplers and DM9xxx + crystal on mainboard.

Questions:

Can you make and share with us firmware dump?

What is the sampling rate when running 2 and 4 channels?

When you apply same signal to channels 1/2 and e.g. 1/3 how high is the skew between channels? (and is there function/menu in the firmware to change it?)

[tatus1969]

I did direct line from waveform generator to scope, non terminated.  Would expect reflections.

The BNCs are most likely non terminated. Have identified the DM9000EP location and the on board DDR2 SRAM is 64mb from the serial.

Images are here: https://1drv.ms/f/s!AnmmyurkkYaJilHJ6x7BgxpD3wSC

Looked at i2c.log for hardware version number and it showed "102"

From the pictures, the analog stage pretty much looks identical to the DSO5000 series. They probably have tweaked here and there, but they also made the trace from BNC to the range relay muuuch longer. I would be surprised if that amplifier is anywhere flat.

The relay at the BNC is to feed in the auto calibration signal, the one in the box is the range switch. I am confident that there is no 50 ohms option built in.

Without inserting an inline terminator, your initial measurement has not much use.

[MarkM]

Thanks for the hack!  I had it done in just a few minutes and also updated it the most recent stock firmware then with tinhead's modified firmware.  Unfortunately, this scope had issues before that. Most of the time it would read a flat DC signal(or square wave) up to 800mV too high depending on the time base and vertical settings.

It will self calibrate, but doesn't seem to help.  There's actually a new issue now. Seems to be reading the peaks a bit better(5.00v-5.20v), but look at the minimum, maximum and mean.  Any idea on this? 

[tatus1969]

Thanks for the hack!  I had it done in just a few minutes and also updated it the most recent stock firmware then with tinhead's modified firmware.  Unfortunately, this scope had issues before that. Most of the time it would read a flat DC signal(or square wave) up to 800mV too high depending on the time base and vertical settings.

It will self calibrate, but doesn't seem to help.  There's actually a new issue now. Seems to be reading the peaks a bit better(5.00v-5.20v), but look at the minimum, maximum and mean.  Any idea on this?

Had you disconnected your probes during calibration?

[MarkM]

Yes, probes are disconnected during calibration.

EDIT: I was wrong. It's not making it through the calibration, although it used to.  I can hear a relay click at step 24/36 then it resets without an error message.  Will let it warm up for much longer and try again.

[snik]

I can hear a relay click at step 24/36 then it resets without an error message.

If problem continues, please check the 595 in the Frontends. Had a similar problem after both 595 gone bad, solder new ones in and one was not soldered right at one or two  pin and than the calibrations resets at point 24/36 too. Maybe it can help

[MarkM]

I can hear a relay click at step 24/36 then it resets without an error message.

If problem continues, please check the 595 in the Frontends. Had a similar problem after both 595 gone bad, solder new ones in and one was not soldered right at one or two  pin and than the calibrations resets at point 24/36 too. Maybe it can help

Thanks!  It's worth a shot.  Just to make sure, you're referring to the 74HC595Ns, correct?

[snik]

Just to make sure, you're referring to the 74HC595Ns, correct?

In my scope both (74HC)595 TSSOP16  in the Frontends for Ch1 and 2 was completly faulty. Maybe one of yours also... Got a bunch of them as bargain on eBay, so the replace was very cheap.

[MarkM]

Just to make sure, you're referring to the 74HC595Ns, correct?

In my scope both (74HC)595 TSSOP16  in the Frontends for Ch1 and 2 was completly faulty. Maybe one of yours also... Got a bunch of them as bargain on eBay, so the replace was very cheap.

Thanks, I appreciate it!  I just ordered some.  I'll report back soon. 

[juppkk]

Hello,

I own a Hantek DSO5102P still 100 MHz S/W Version 3.40.0

My question is, how to change the background color.
I mean the black color behind the Signal, not the Menu-Background.

And secondary, please in detail, how my I trigger such a signal (Picture)
(Signal is from PT2272 a remote control decoder)
Small impulswidth is 34 µS, the signal package is 25 mS
I´ve tried a lot but no success.

Thank´s for help

[MarkM]

Just to make sure, you're referring to the 74HC595Ns, correct?

In my scope both (74HC)595 TSSOP16  in the Frontends for Ch1 and 2 was completly faulty. Maybe one of yours also... Got a bunch of them as bargain on eBay, so the replace was very cheap.

Thanks, I appreciate it!  I just ordered some.  I'll report back soon.

Replaced both and same issue.  Triple checked the soldering.  Ordering the Rigol 1054z, as I need a working scope within the next two weeks.  I appreciate the help, though.  I'll mess with it later.

[snik]

Sorry to hear that it was not the fault...
Please tell us what caused the Error when you find it