-
Upgrading the Hantek DSO4072C and DS4104C oscilloscopes bandwidth up to 250MHz
Posted by Microcheap on 04 Jan, 2019 03:35 -
So, I got one of these cheap 4072C to play and then decided to see if it was possible to hack it from its original 70MHz to the 200MHz maximum bandwidth. I successfully managed to re-config it and found that it is rather simple, it is very similar to the other Hantek hacks with just a small different detail. Here is how I did it, if anyone is interested:
1 - You need to open the case to get access to the serial interface on the PCB, it is located beside the usb connector and the pinout is the same as the others, from the usb connector the pins are: unused, RX, TX , GND and 3v3 (not needed).
2 - Connect the cables to an USB-TTL converter and on your pc open the terminal emulator of your choice (I use Poderosa 4). Open a serial connection with the configuration 115200bps and 8-n-1.
3 - Turn on the oscilloscope and wait for the boot. Before change anything I did a full backup using the tool nandbackup_V2 provided by @fremen67 here (just extract the files to an usb flash drive, plug it to the scope and execute the script, i did ./nandbackup_1.00.2_OOB and ./nandbackup_1.00.2 just in case, it will take about 15min each).
4 - Now, to change the bandwidth you need edit just one line in the file i2c.log located in / (I would recommend you to make a copy of it to your flash drive as well). Open the file with the vi editor and change the last line: where you see [bw] 70 to [bw] 200 to get the 200MHz or [bw] 100 to 100MHz. Save and close.
5 - Now reboot the scope typing the command reboot [enter] (if you just turn off and on the scope, it won't work), and that's it
Opening the menu Utility -> System Info you should see the model changed to DSO4202C.
To be honest, I am not sure if there is really any difference among the models. I don't have a function generator or other equipment capable of test the full bandwidth, my hacked HDG2002B is useful up to just about 50/60 MHz so I couldn't see any practical change after the upgrade, if anyone could test it and confirm, I would be grateful.
Some pictures
-
Some other pictures
-
I understand that there is not much interest in these cheap Chinese equipment manufacturers, so I will keep the information in one thread only to not pollute the forum, just in case someone ever looks for it, they will be able to find it here.
So, I borrowed the newer model of the Hantek scope with 4 channels (DSO4104C) to try the same upgrade as above and indeed it works. The 2ch model goes up to 200MHz only but the new 4ch has a 250MHz version: http://www.hantek.com/en/ProductDetail_3_12167.html
The procedure is the same as described in the first post, the only difference is in the file to be changed. To get the full 250MHz bandwidth, edit the file system.inf located in /config/root and change the first line "Model=80M$DSO4084C" if you have the 80MHz or "Model=100M$DSO4104C" for the 100MHz to Model=250M$DSO4254C, save and reboot.
If someone is able to test and confirm it would be nice, as I don't have the equipment to test the effectiveness of this upgrade or the real performance of these oscilloscopes measuring maximum bandwidth.
-
Some more pictures
-
Just a small question, could you provide a closer photo from the wave gen output?
I'm wondering which IC is placed at U63 (and if possible the diode D1501).
I'm curious if simply adding a IC, diode and BNC would enable the wave gen on DSO4004B ... (the according GUI is already available in my case)
-
If it is anything like the AWG output on a handheld 2000 series then it is a LMH6702 opamp. Pinout looks correct. Probe the input pin to that opamp to see if there is any signal when you activate the AWG. If there is a signal then start pondering what parts to get.
-
Thanks for the hint. Nope, no signal around.
Perhaps, a DAC or some SW is also missing.
I'll leave it as it is for now and maybe get back to that topic later on ...
-
Sorry, I don't have the the scope with me anymore, it was a borrowed unit, but the output OpAmp U63 is a EL5166ISZ https://www.mouser.ie/ProductDetail/Renesas-Intersil/EL5166ISZ?qs=sGAEpiMZZMsko7UDAsUSISfAG4JYqL3li7G7U2Q8z3U%3d
D1501 is probably a clamping diode BAV99 (marking code A7). Check U62 too, it is a DAC902E, it is not a cheap iC, so I suppose it is not populated in your board as well. -
Thanks a lot for the detailed information!
Just order a few components, just to see (for fun) if I can get the signal gen to work.
-
Nice, please report back your results.
I was wondering if there is also the need to change some configuration in the software to enable the AWG. On the 2CH models, in the file i2c.log there is a line "[dds]" set to 1 that i suppose tells the scope software to activate the function generator option, but I didn't find anything like this in the 4104C, maybe just changing the "B" from the DSO4XX4B to a "C" in the file system.inf will do the trick, but it need to be tested. -
Sure, will be my pleasure to keep you and this thread updated!
And thank you again for all your helpful comments and information! -
I can confirm that the bandwidth hack really work.
Today I hacked my Hantek DSO4104B and it has now -3 dB point at around 280 MHz. You can even observe signals beyond 300 MHz.
The upper limit is not exact as I used an old LSV-350 VHF Sweep generator. Still, it is very obvious that the input station of the oscilloscope permit signals higher than 250 MHz.
Microcheap gave us good information at the beginning post so I can only add minor remarks:
You can find information on how to open the oscilloscope at https://youtu.be/fxgJdKeOmlo
You have to pull the black power ON button cap from the outside before you open the oscilloscope. Use some cloth between pliers and plastic cap to prevent scratches.
Info on serial connector is at: https://youtu.be/fxgJdKeOmlo?t=1921
RX goes to TX of the USB-TTL converter
TX goes to RX of the USB-TTL converter
Ground goes to converters ground.
For the terminal emulator, you can also use Putty as it is available on all platforms.
The DSO4104B is almost perfect budget oscilloscope for basic or should I say classic tasks and it is nice that it can be even a little better now
-
I'm curious what that SATA cable goes to on the main board there? Does the unit have a disk drive?
-
No, I don't think that SATA cable is for the disk.
They use SATA cable to connect the main board with I/O board at the back of the case that has network and USB sockets. At the back can be also mounted HDMI socket.
It is interesting to see a memory card slot located at the edge of the main board. -
I can confirm that the bandwidth hack really work.
That's good to hear, thx for the confirmation zoran -
Hi all,
I finally received the missing ICs required to get the function generator to life.
Two items I notices, hoping someone could help me with ...
I noticed that some resistor arrays (RA1601-RA1604) close to the DAC (U62) are missing. Does anybody know the used resistance?
I also ordered a BNC-connector by guessing, but noticed it does not fit. Does anybody has part number for those?
-
these not populated Relais for 50R input, they asking to get populated ^^ Wondering only if you guys change the model number, the 50R will get enabled in firmware?
-
I noticed that some resistor arrays (RA1601-RA1604) close to the DAC (U62) are missing. Does anybody know the used resistance?
I found some pictures I took when I disassembled my scope, I hope it will be helpful.
-
Hi just schort update.
It took a while to get the parts and the time to put them together.
But finally, after adding the missing parts:
- DAC902E (U62)
- 4x 20Rx4 (RA1601-RA1604)
- EL5166ISZ (U63)
- BAV99 (D1501)
- and a somewhat fitting BNC-connector,
and the function generator appears to work.
Only the bandwidth seems to be low.
Sinus looses amplitude @ f > 1Mhz (as I remember) ... Will look into that later on.
-
The problem may be calibration. There is a file called dds_calbration.dat (sic) in the folder /config/root that I suppose contains calibration data for the function generator. The calibration is made at the factory, and I think the devices without DDS option won't have this file. I don't know if it is possible to calibrate de DDS later but you can try to copy the file from another device, obviously it is not ideal, as this calibration data is unique for each device but it may improve things a bit.
-
But finally, after adding the missing parts:
- DAC902E (U62)
Are you shure it's the correct one?
When looking at the (not really sharp) pics all 14! data lines bit 1 thru 14 seem connected.
Compared to my shortly arrived 6074, where bits 13 and 14 are not!
So maybe DAC904 is used at DSO40nn... -
Are you shure it's the correct one?
Yes, the DSO4004C series uses DAC902E and, as far as I know, the 6074BC used it as well, unless Hantek has changed it in a new PCB revision.
For the new Hantek2000 series devices, Hantek is using now a Chinese DAC (made by 3Peak) but, it is pin compatible with the DAC902E. -
I've just checked on my scope. On the left resistor network is only the clock signal active.
-
I recently bought a Hantek DSO4084C and did the upgrade to DSO4254C.
The Info screen now shows the upgraded model but it seems not really work on 250MHz.
When I set 20ns timebase I see the signal with high time resolution (amazing!). When I switch to 10ns to take a closer look the waveform shows this with higher resolution and suddenly jumps to a filtered one (frustrating!). So obviously the software does not fully recognize the Model cheating.
I'm using:
Software 1.1.2(190701.0)
Hardware 501 001.001.000.000.000.000.000
FPGA a005
Does anybody know how to fix this behaviour? Did I miss something in hacking? Should I use an older firmware (is a downgrade possible?)?
Rainer
-
Do you maybe have the AutoRange enabled?
Check that AutoScale -> AutoRange is OFF.
I did the same as you did and converted a DSO4084B to a DSO4254C. Everything went fine so far. The signal stays stable even at 200 MHz. I use the same specs You do.
I wrote a little hacking summary, that I will attach as pdf.
Edit: Uploaded updated hacking manual with added EEPROM hack. You can always see the current document on my github .