New Hantek DSO2X1X models?

[DavidAlfa]

AFAIK, libanolis is their library for graphics, I/O, etc .

[DavidAlfa]

I'm about to order new encoders.
After thinking it twice, I'll avoid the mechanical steps in the small encoders, adjusting certain values require a lot of turning, dealing with a ton of clicking will make the experience even worse!

So the idea is keep it just like it is now, but with more steps per turn to make it less painful, and decent quality parts.

The PEC12R-4220 (7mm shaft) series are a direct replacement (As we already know, thanks to pu6k1n and vistorik ).
If not being able to source them, the PEC12R-4225 (12mm shaft) will almost fit, the small caps have a depth of 12mm while the large ones have 11.2mm, it would only require trimming a bit off the shaft end.

- Pushbutton + stepped rotation (Sec/div knob):      PEC12R-4220F-S0024 / PEC12R-4225F-S0024
- Pushbutton  +  smooth rotation (Small knobs):      PEC12R-4020F-S0024 / PEC12R-4025F-S0024
- No pushbutton + stepped rotation (V/div knobs):   PEC12R-4220F-N0024 / PEC12R-4225F-N0024

Also the PEC12R-3x series will also work, it's the same P/N, just replace the 4xxx with 3xxx.
The difference lies in including a bushing.

The shaft offset is 2mm larger in the PEC12R-4222 and PEC12R-4230, so the knob sit 2mm away from the base, though still possible with additional modifications to the shaft.
The difference can be seen here:

[alias095]

I am replace FPGA with GW2AR-LV18EQ144PC8...
After the replacement, the buttons became work, the calibration passes but the Auto mode does not work, + incorrect data in 1GSa / s and 500,
Has GW2AR-LV18EQ144PC8 been used correctly or should GW2AR-LV18EQ144 be used without P
or such GW2AR-LV18LQ144

[DavidAlfa]

I think the SDRAM will be too slow?
But I have no idea about the exact FPGA model..

[vistorik]

I am replace FPGA with GW2AR-LV18EQ144PC8...
After the replacement, the buttons became work, the calibration passes but the Auto mode does not work, + incorrect data in 1GSa / s and 500,
Has GW2AR-LV18EQ144PC8 been used correctly or should GW2AR-LV18EQ144 be used without P
or such GW2AR-LV18LQ144

alias095, the best is to ask Hantek (Han Zhangqiang, Email: zhangqiang@hantek.com) to sell you a native FPGA LEMONTREE chip of the corresponding series (numbers under the name of the chip) with an explanation of the reason. Maybe they will sell it to you. They once sold me separately a new DSO2D10 motherboard.

[Cihan]

Is it worth buying DSO2D10 in 2023 or should I be looking forward for some other 2 channel model under 300 bucks? 4 channel is preferable if I could find one below 300 USD. By the way there are many oscilloscopes with 2Gpts memory depth today but I have no idea when they'll be under 400 USD.

[BillyO]

Is it worth buying DSO2D10 in 2023 or should I be looking forward for some other 2 channel model under 300 bucks? 4 channel is preferable if I could find one below 300 USD. By the way there are many oscilloscopes with 2Gpts memory depth today but I have no idea when they'll be under 400 USD.

The Siglent SDS1102CML+ runs about $320.  It may have slightly fewer "features", but it actually works and is properly supported by a reputable company.  Probably easily "upgraded" to 150MHz too.

I wouldn't bother with Hantek bench scopes.  They are just junk.

[dirtmover]

Is it worth buying DSO2D10 in 2023 or should I be looking forward for some other 2 channel model under 300 bucks? 4 channel is preferable if I could find one below 300 USD. By the way there are many oscilloscopes with 2Gpts memory depth today but I have no idea when they'll be under 400 USD.

I purchased and subsequently hacked a 2C10 -> 2D15 just over a year ago. It's has many shortcomings but if you don't get too hung up on it's flaws and focus on what it can do, it still offers some decent functionality for the ~$200 you'll pay for it. Now, if you expect it to meet all it's published specs you're going to be disappointed. Unless you can score something 2nd hand, I don't believe you're going to get anything better below $300.

[dirtmover]

So the idea is keep it just like it is now, but with more steps per turn to make it less painful, and decent quality parts.

I wonder if the software can keep up with more steps per turn or is it just going to start skipping steps?

[Aldo22]

I purchased and subsequently hacked a 2C10 -> 2D15 just over a year ago. It's has many shortcomings but if you don't get too hung up on it's flaws and focus on what it can do, it still offers some decent functionality for the ~$200 you'll pay for it.

Prices are low at the moment.
You can get the DSO2D15 (150MHz + AWG) for under $200 shipped.
https://www.aliexpress.us/item/3256803584138756.html
But i can still resist! 

[aztecboy]

I'm with Dirtmover. I also don't use all the features but have fixed many a radio with it now.

[DavidAlfa]

I finally ordered the new encoders. They had become completely unusable.
Look at the waveforms V0 makes when turning faster than 1/2 rotation a second!
Pure noise, missing pulses... the contacts are clearly floating.

So add +$25 to the scope during the first year.
Plus another vertical line appeared in the display so soon another $30.

[DavidAlfa]

The leg said no more, it broke a year ago and got it more or less fixed by melting it.
But the plastic had the strength of old pizza, broke in no time.
Made a 3d print, nothing fancy, the idea is to insert screws from the inside of the scope into the part.
I doubt I will be the only one with this problem so I'm attaching the STL.
Will be updating the design as needed.

[aztecboy]

Very cool David!

[tautech]

The leg said no more, it broke a year ago and got it more or less fixed by melting it.
But the plastic had the strength of old pizza, broke in no time.
Made a 3d print, nothing fancy, the idea is to insert screws from the inside of the scope into the part.
I doubt I will be the only one with this problem so I'm attaching the STL.
Will be updating the design as needed.

FYI, see the sticky atop the TE board ?
https://www.eevblog.com/forum/testgear/replacement-knobs-feet-and-fittings-for-test-equipment/

Suggest you make a post there. 

[DavidAlfa]

I'll do when it's fully assembled and tested working!
New encoders didnt arrive yet, but just for curiosity, I tried tearing V0 apart, bending the contacts so they make more force and assemblying back.
Instant improvement! But I guess it won't last long.
On close inspection they look identical to the genuine Bourns shown online... might they be original after all?
I really hope their products aren't such crap, started failing after only a few months.

Edit: I decided to also get rid of the LCD, bought a new AT070TN94 at Ali.

[Z_E_V_S]

Hi everyone.

I had the misfortune to acquire DSO2c10.
Recently, I detected an error when switching horizontal resolution from 50ms to 100ms at a memory depth of 400k or more.
The oscilloscope does not display the signal correctly.
I wrote about this in Hantek, but there is no answer.
https://www.eediscuss.com/forum.php?mod=viewthread&tid=17937&extra=page%3D1
Maybe this is only an error in my oscilloscope, I ask you to check this mode on your oscilloscope.
Thank you in advance.

P.S.   I apologize for my English, I use a translator.

[DavidAlfa]

Indeed, there's a new firmware 2023.03.27, though they uploaded it 2 months  later! (2023.07.28) 
We already had it, they've been shipping new devices with the updated firmware all this time, so we extracted it from a working device.
Anyways thanks for telling!

Yeah the trigger doesn't work at very low speeds.
About the signal levels, have you calibrated the scope and later pressed "Default setup" ?
Show some pictures! If the signal is very noisy, with random spikes, try updating the FPGA to FW3205.

The statistics issue has been there since day one, it's a shame.

Check the FAQ in my signature!

[Z_E_V_S]

About the signal levels, have you calibrated the scope and later pressed "Default setup" ?
Show some pictures! If the signal is very noisy, with random spikes, try updating the FPGA to FW3205.

The problem is not due to noise, the signal is clean.
The largest error when switching from 50ms to 100ms at a memory depth of 400k or more.
To check, you need to give a signal with a frequency of about 3 hz, set the trigger to normal,memory depth of 400k or more, and scan 50ms per division.
In this case, the oscilloscope shows the correct signal.
If you switch to 100ms per division, the oscilloscope will show a very stretched signal.
If you press the "single" button in this mode, the oscilloscope will show a normal signal after capturing.
Please check this on your oscilloscope.
On the Hantek website, I described everything in detail.
My oscilloscope is calibrated.
The error occurs at a horizontal resolution of 100ms or more.


Thanks.

[Z_E_V_S]

Activated the generator using your program,DavidAlfa,  for which thank you very much.
I continue to investigate the oscilloscope.
Found that the oscilloscope in Roll mode moves the image in jerks.
This is a consequence of poor program optimization.
Linux is not a real-time operating system.
I think that Hantek engineers devoted very little time to synchronizing processes and threads, which led to such consequences.
It is a pity that there is no source code.
This can all be fixed.

[DavidAlfa]

Yes, it'll be worse as you use it and discover more "features". We've been like this for more than 2 years.
Nothing to do, Hantek laughts at their customers.
It seems like the bottleneck is in the communication with FPGA, it uses 10MHz SPI to transfer the data to the CPU, that's a pretty small bandwidth (1.25MB/s).

Edit: Inspected it in Ghidra again, it seems it does at 50MHz.

Edit2: But it seems the Linux devicetree will limit this to 1MHz!

I made some tests here, feel free to try.

[groomit]

Hi there, I'm new on this forum and overwhelmed by the amount of information shared.
I have the Hantek DSO standing around on my desk for quite a while now. It still has the factory firmware installed. After digging through all of the information on this thread as well as on the hacking thread, I'm still a bit confused regarding the backup steps necessary before tinkering around with firmware updates.
Can anyone please explain to me the difference between Flash-, EEPROM- and Backup-Builder? What contents are saved on each backup? I assume it is not sufficient to read the ROM with a CH341A and a clip - correct?

I don't mean to bother you with my request. It is just that I'd like to better understand what and what for I'm doing each backup.

Thanks in advance!

[DavidAlfa]

Did you read the FAQ?
Backup builder does a copy of the important data like model, sn, generator calibration, current software, FPGA firmware and also the eeprom.

Backup builder files are enough recover but it's always good to make a 1:1 backup with  dsoflash just in case.

[groomit]

Did you read the FAQ?
Backup builder does a copy of the important data like model, sn, generator calibration, current software, FPGA firmware and  larz eeprom.

Backup builder files are enough recover but it's always good to .ake a 1:1 backup with  dsoflash just in case.

Hi David,

Thanks for your fast reply. Yes, I went through your extensive FAQ section - Thank you for your work! But still it was confusing what backup does what.
So now I understood that Backup Builder already saves a lot of data:
- Generator Calibration Data
- Model and SN
- Software and FPGA-Firmware
- EEPROM

As you stated, that should be enough to recover. What more/else does dsoflash do? And do I understand it correct that no seperate EEPROM backup is necessary after running backup builder?
I think most of the confusion I'm running into is the lack of knowledge about what types of storage are present and what is stored on each storage.

[vistorik]

What more/else does dsoflash do? And do I understand it correct that no seperate EEPROM backup is necessary after running backup builder?

David's dsoflash creates a complete memory dump, including the spare area of 64 bytes for every page of 2048 bytes in one binary file of 132Mb, like the programmer CH341a with an option to read the spare area does.
If you run David's backup builder you don't need to run seperate EEPROM backup.