Hantek HDG2002B AWG: 5Mhz or 100MHz? Let's see!

[flatlander]

Quote from: flatlander on Today at 05:42:24 PM

Did some testing on my 'upgraded' Hantek HDG2002B and found that the amplitude of the generated signal (sine) quickly drops off past 5MHz. It seems that my unit's firmware is still limiting the bandwidth to 5Mhz even though it lets me increase the frequency to 100Mhz. I tried loading older FPGA firmware but that doesn't work at all (no output and all output related self tests fail). Possibly have to flash a 'matching' set of software and FPGA versions. I'd be interested to know if the amplitude issue also exists on earlier software/hardware versions.

Versions:
Software: 1.00.2 (140926.0)
Kernel: Linux 3.2.35
FPGA: 20
Keyboard: 3
PCB: 1004

>I think there is a filter in the output stage that you need to bypass in order to get full bandwidth.  There are details here somewhere...

I thought that fix was for reducing ringing on output signals > 4Vpp. I should have mentioned this in my post, but I was using 1Vpp.

[fremen67]

I am keeping a blog at http://hdg2002b.simplesite.com . So far I have consolidated all documentation I have: datasheets, hi-res scans & pics, scripts and upgrade procedures sourced and credited to the fine folks that have contributed to this forum. 

Seems I have been away for a long time, very nice blog !

The firmware partitions for 1.00.2 are not the same as 1.00.1. They added a recovery partition. I updated the nanddump script for the new geometry, and succeeded with dum pin mtd0 to mtd8.

Good job. Could you send me a link to your dump? I am still running firmware 1.00.1 ...

I have successfully JTAGed the s3c2416 and NAND with a cheapo 100ask.net USB adapter via openocd. The scripts for USB nandbackup and the openocd config for the AWG target is attached below, and posted on the above site. Rename the .txt file as .cfg.

I have the same adapter but could not manage to go faster with OpenOCD than H-Jtag with the parallel port. I also tried OpenOCD with a J-Link adapter but it was not faster. I know that Tinhead has the H-JTAG USB adapter and he can burn a K9F1G08 in less than 8 minutes...
The only device I use with OpenOCD which is faster than H-JTAG // is a Raspberry PI used as a Jtag adapter (see https://github.com/synthetos/PiOCD/wiki/Using-a-Raspberry-Pi-as-a-JTAG-Dongle ) but only from 20%.

[fremen67]

Got my HDG2002B today. System info below, FPGA version seems to have jumped a few versions.

Software: 1.00.2 (140926.0)
Kernel: Linux 3.2.35
FPGA: 20
Keyboard: 3
PCB: 1004

Did you dump your nand? It seems you have the latest here.
There are several 1.00.2 versions with different release dates. I am wondering wether Hantek is correctly following revisions or not ...

[fremen67]

TTSOURCE reads the unit and controls the AWG, BUT the LCD & button LEDs no longer update when I make a change in TTSOURCE! 
Anyone with the DM9000 mod have a similar issue?
Can anyone confirm this behavior with a unit running FW v1.00.2 ? Thx.

I have the same issue but I have FW 1.00.1 and I tried TTsource after adding the DM9000 mod ...

[fremen67]

Did some testing on my 'upgraded' Hantek HDG2002B and found that the amplitude of the generated signal (sine) quickly drops off past 5MHz. It seems that my unit's firmware is still limiting the bandwidth to 5Mhz even though it lets me increase the frequency to 100Mhz. I tried loading older FPGA firmware but that doesn't work at all (no output and all output related self tests fail). Possibly have to flash a 'matching' set of software and FPGA versions. I'd be interested to know if the amplitude issue also exists on earlier software/hardware versions.

Versions:

Software: 1.00.2 (140926.0)
Kernel: Linux 3.2.35
FPGA: 20
Keyboard: 3
PCB: 1004

I have a similar problem but it starts right after 30 Mhz: https://www.eevblog.com/forum/testgear/hantek-hdg2002b-awg-5mhz-or-100mhz-let's-see!/msg475956/#msg475956. It seems that only a few of us have this problem...

Software 1.00.1 (140402.0)
FPGA 11
PCB 1002

[Cyber7]

Idpromnut wrote:

I received today a replacement board for my MSO and have exactly the same results. The amp of a 2V sine decrease above 30Mhz to 340 mV à 100MHz.
So it has to come from the HDG.
With my stock MSO board:
30 000 000.000 000 Mhz -> 1.98 V
30 000 000.000 001 Mhz -> 1.62 V !!!

Software "issue" of v1.00.1?

I have the same issue. Not quite the advertised range for the 2102: Amplitude Range ?20MHz:1mVpp—20Vpp; ?60MHz:1mVpp—15Vpp; ?80MHz:1mVpp—10Vpp;?90MHz:1mVpp—5Vpp; ?100MHz:1mVpp—2Vpp

Software: 1.00.2 (140819.0)
Kernel: Linux 3.2.35
FPGA: 12
Keyboard: 3
PCB: 1004

[flatlander]

Quote from: flatlander on Yesterday at 05:42:24 PM

Did some testing on my 'upgraded' Hantek HDG2002B and found that the amplitude of the generated signal (sine) quickly drops off past 5MHz. It seems that my unit's firmware is still limiting the bandwidth to 5Mhz even though it lets me increase the frequency to 100Mhz. I tried loading older FPGA firmware but that doesn't work at all (no output and all output related self tests fail). Possibly have to flash a 'matching' set of software and FPGA versions. I'd be interested to know if the amplitude issue also exists on earlier software/hardware versions.

Versions:

Software: 1.00.2 (140926.0)
Kernel: Linux 3.2.35
FPGA: 20
Keyboard: 3
PCB: 1004

I have a similar problem but it starts right after 30 Mhz: https://www.eevblog.com/forum/testgear/hantek-hdg2002b-awg-5mhz-or-100mhz-let's-see!/msg475956/#msg475956. It seems that only a few of us have this problem...

Software 1.00.1 (140402.0)
FPGA 11
PCB 1002

I temporarily renamed the htg.cal calibration file and rebooted just to see if that would make a difference. Without that file the amplitude didn't start decreasing until past 10Mhz as opposed to 5MHz with the calibration file. Need to look into manual calibration and to see if that makes a difference.

Could someone who's HDG2002B doesn't have this problem upload their htg.cal file. I'd be interested to see if this makes a difference on mine.

EDIT: OK, after a bit more experimenting, calibration seems related to this issue. I used the calibration function of the generator and found that it starts with a DC voltage calibration followed by a AC voltage (sine upto 100MHz) calibration. I stopped about halfway through as it's a very laborious process and only wanted to see if it made a difference (entered many more values than the 12 values reported in this thread without issue though). Looks like the calibration issue may have been fixed in my version of the firmware.

It definitely had an impact as the decreasing amplitude now starts around 25MHz versus the 5MHz earlier. I suspect Hantek started to only calibrate the generators up to 5Mhz.

I would still be keen to try a cal file from a unit that doesn't have the issue (or less severe, the calibration can't be too far off).

[Cyber7]

@Fremen67:

Good job. Could you send me a link to your dump? I am still running firmware 1.00.1 ..

I sent you a PM.

[fremen67]

OK, after a bit more experimenting, calibration seems related to this issue. I used the calibration function of the generator and found that it starts with a DC voltage calibration followed by a AC voltage (sine upto 100MHz) calibration. I stopped about halfway through as it's a very laborious process and only wanted to see if it made a difference (entered many more values than the 12 values reported in this thread without issue though). Looks like the calibration issue may have been fixed in my version of the firmware.

It definitely had an impact as the decreasing amplitude now starts around 25MHz versus the 5MHz earlier. I suspect Hantek started to only calibrate the generators up to 5Mhz.

I would still be keen to try a cal file from a unit that doesn't have the issue (or less severe, the calibration can't be too far off).

That makes sense. As we both started the calibration but could not go through the whole process, it could explain why we have this problem others don't.
I will have a look with the initial backup version of my calibration file.

Would you be able to dump your nand?

[flatlander]

>

Quote from: flatlander on Yesterday at 12:23:16 PM

OK, after a bit more experimenting, calibration seems related to this issue. I used the calibration function of the generator and found that it starts with a DC voltage calibration followed by a AC voltage (sine upto 100MHz) calibration. I stopped about halfway through as it's a very laborious process and only wanted to see if it made a difference (entered many more values than the 12 values reported in this thread without issue though). Looks like the calibration issue may have been fixed in my version of the firmware.

It definitely had an impact as the decreasing amplitude now starts around 25MHz versus the 5MHz earlier. I suspect Hantek started to only calibrate the generators up to 5Mhz.

I would still be keen to try a cal file from a unit that doesn't have the issue (or less severe, the calibration can't be too far off).

That makes sense. As we both started the calibration but could not go through the whole process, it could explain why we have this problem others don't.
I will have a look with the initial backup version of my calibration file.

Would you be able to dump your nand?

@ fremen67: nanddump uploaded to http://www.filedropper.com/hdg2002b10021409260

[fremen67]

@ fremen67: nanddump uploaded to http://www.filedropper.com/hdg2002b10021409260

Got it thanks!

[andrija]

I received my 20MHz version - was only 20 bucks extra, just in case I don't feel like hacking the software. I am planning to add LAN port and maybe the counter and have already ordered parts. The unit did not work out of the box though! Fan turns on but not the screen. I opened it up and fiddled with wiring and then it started working, only to stop when I put the case back on. Finally I left it without case in order to play with it.

This thing feels and looks like a toy. The quality of the plastic is way below my Rigol stuff and the metal chassis inside is like that Siglent, not quite as much rust but it's really nasty and dirty and full of smudges. I was expecting that, though, having seen pictures here. The user interface is, however, quite nice and intuitive, at least for basic waveforms. The frequency of the oscillator was 1Hz off at 10MHz and kept stable, if I was to trust my HP counter which isn't calibrated. It also stayed stable over time. Using external 10MHz input did not make any difference in the precision of the output waveform frequency at 1kHz. What I was impressed with is the quality of the sine wave - I measured THD of -85.3dB on channel 1 and 2dB worse on channel 2 with my Keithley 2015THD, with 10 harmonics. That is a spectacular result for something like this, isn't it?

I'd say this thing is a better option than any of those simple classic function generators. User interface is nice and performance seems good. It seems upgrading from 20 to 100MHz would not get me much in practice based on what people are saying about the drop in signal quality and limitations of the non-sine functions which in most cases are already reached in the 20Mhz "model". Not sure I'd bother to upgrade bandwidth but I will mess with LAN for sure.

[Cyber7]

@andrija: Keep in mind that unless you plan on tweaking the software, of writing PC software for the debug UDP port, there's isn't much to do with the LAN port at the moment. If you would like I can post you a DM9000A + the MDIX magnetics for a few $ as I have a couple of spare sets. It would save you some time from shipping halfway around the world. PM me if interested.

[andrija]

Oh so the LAN port does not actually do anything? I thought you could control instrument through it. After all there's network setup in the menu and everything. I already wrote some code to control Rigol ocilloscope and power supply without need for any drivers so I figured I'll sniff network packets on this toy as well. Ah well, I don't have any other function gen (other than sound card and the sine gen up to 20kHz in Keithley) and it controls well through GUI so it's ok.

Thanks very much for the offer, however I already ordered magnetics and chips from China 2 weeks ago (I read this thread, was in need of a gen, thought "awesome, I can hack" and ordered hacking parts before even making decision to buy the Hantek itself), just need some lan ports and capacitors and I'll get that from Mouser.

[Cyber7]

Rigol Ultravision scope support VXI out of the box; Agilent IO picks it up quite nicely and you can control it very well via SCPI commands tcip/visa. Check RUU here: https://www.eevblog.com/forum/projects/software-tips-and-tricks-for-rigol-ds200040006000-ultravision-dsos/

There's a network setup menu...and that is pretty much everything  . While the HDG2002x supports a number of SCPI commands over USB, Hantek has been dragging their feet on supporting VXI on their their equipment.

[andrija]

That's too bad. I am aware of the mermad's utility and was wanting to write my own so that I can control all the instruments and set up a testing chain, e.g. test transient response of a DIY power supply or something like that which would involve setting power supply to do a pulse and the scope to trigger on it. I managed to communicate to 2702A and the DP832 and was hoping this Hantek was a cheap alternative to Rigol 4602 since I don't need uber-MHz and that thing is a cool $1k all in. But I guess you get what you pay for .

Oh, and good job on making a website to keep all info in one place. Forums S/N ratio usually gets so bad that you can't figure out anything. Appreciated .

[idpromnut]

That's too bad. I am aware of the mermad's utility and was wanting to write my own so that I can control all the instruments and set up a testing chain, e.g. test transient response of a DIY power supply or something like that which would involve setting power supply to do a pulse and the scope to trigger on it. I managed to communicate to 2702A and the DP832 and was hoping this Hantek was a cheap alternative to Rigol 4602 since I don't need uber-MHz and that thing is a cool $1k all in. But I guess you get what you pay for .

We shall see, but hoping you will be able to get more than you paid for, if we can get an open-source firmware up and running for this device.

[alex.forencich]

That's too bad. I am aware of the mermad's utility and was wanting to write my own so that I can control all the instruments and set up a testing chain, e.g. test transient response of a DIY power supply or something like that which would involve setting power supply to do a pulse and the scope to trigger on it. I managed to communicate to 2702A and the DP832 and was hoping this Hantek was a cheap alternative to Rigol 4602 since I don't need uber-MHz and that thing is a cool $1k all in. But I guess you get what you pay for .

Might not be a bad idea to try to piggyback on an existing project instead of starting from scratch.  There are some open source cross platform instrument control suites available that might be worth looking at.  For example, Python IVI, which supports the DP832 out of the box.  I would add support for the HDG2000, but mine is in pieces right now so I can work on the FPGA and frontend. 

[fremen67]

The firmware partitions for 1.00.2 are not the same as 1.00.1. They added a recovery partition. I updated the nanddump script for the new geometry, and succeeded with dum pin mtd0 to mtd8.

I upgraded today my HDG from FW 1.00.1 to 1.00.2. thanks to Cyber7 and Flatlander backups. The process is not straightforward:
- The new 1.00.2 u-boot is supposed to allow ubi partition restoring via usb or Tftp ... but it's not working a the moment.
- Restoring ubi partitions directly from nand dumps via JTAG (I used H-Flasher //) is very hazardous due to the bad blocks layout beeing specific to each HDG nand. Luckily enough, the mtd8 restore partition backup from Cyber7 had no bad blocks and the corresponding set of blocks on my nand neither so I could easily restore it. My rootfs partitions have bad blocks so I could not restore it directly from a nandbackup from another HDG...Luckily enough, the recovery process of the FW 1.00.2 seems to work.
The problem I have now is that the relays are not powered any more. All the menus seem to be functionnal but I don't have any output on CH1 nor CH2...no clicks... no signal.
My PCB is rev 1002, the new HDG have now PCB1004. Maybe this could explain the problem. Any idea?

I also realized that backup scripts were excluding badblocks from the dumps and I could not use Fatlander recovery partition backup which has 2 badblocks. I modified the scripts to include them as we need the dumps and the partitions to have exactly the same size.

[andrija]

My PCB version is also 1002 and I just received it on Monday. Software 1.00.1, FPGA 11 and keyboard 1, linux 3.2.35. Mine is the 2022B model though. Looks like it came from an old stock then.

And... as I said the unit wouldn't turn on when I received it. It turned off again and there's no voltage going to the board (but yes to the fan and the unused 5V header). There is also faint clicking once a second that appears to be oscillator trying to turn on but failing; my voltage detector lights up in sync with that sound every second at the big transformer. No idea what could be wrong and I'm no fan of debugging a mains switching power supply. Sigh. Might have to buy a third party supply and try to attach it in the box. Any idea how much power is needed on 17.5V and 5V rails?

[fremen67]

The problem I have now is that the relays are not powered any more. All the menus seem to be functionnal but I don't have any output on CH1 nor CH2...no clicks... no signal.
My PCB is rev 1002, the new HDG have now PCB1004.

I downgraded the FPGA version to v11 which was the last one working on PCB 1002 with FW 1.00.1 and the HDG is back on duty. So FW 1.00.2 + FPGA rev.11 is working, at least I have now signals on both channels.
There is still something wrong as the selftest is failing when testing ch1/ch2 output and zoom.

The next step would be to identify the modifications between PCB rev 1002 and 1004 and make them on my board (if possible) so that I could have a stock new 1.00.2 firmware.

[alex.forencich]

The next step would be to identify the modifications between PCB rev 1002 and 1004 and make them on my board (if possible) so that I could have a stock new 1.00.2 firmware.

Yes, this would be very good to know.  Also, I have no idea if this is possible or not, it would be good to know HOW the revisions are differentiated by the firmware.  Could be tied off pins or a resistor value, or it could be written in an EEPROM somewhere.  Might not be possible to know completely, though.  Did the board revision reported by the unit change when you upgraded the FPGA?

[fremen67]

Did the board revision reported by the unit change when you upgraded the FPGA?

The PCB version reported by the unit changed. It was reported as 1004 after the upgrade. I changed it manually back to 1002 in the system.inf file but it did not do any good and was not changed back to 1004 after a reboot.

I would say that firmware versions are just supposed to  match the PCB versions and that no check is done in the firmware.
I was running a stock FW 1.00.2 on my PCB 1002 without any warning and it is obviously made for PCB 1004.
I suppose that it is just written in system.inf, like the model and serial number.

[fremen67]

Did some testing on my 'upgraded' Hantek HDG2002B and found that the amplitude of the generated signal (sine) quickly drops off past 5MHz. It seems that my unit's firmware is still limiting the bandwidth to 5Mhz even though it lets me increase the frequency to 100Mhz. I tried loading older FPGA firmware but that doesn't work at all (no output and all output related self tests fail). Possibly have to flash a 'matching' set of software and FPGA versions. I'd be interested to know if the amplitude issue also exists on earlier software/hardware versions.

Versions:

Software: 1.00.2 (140926.0)
Kernel: Linux 3.2.35
FPGA: 20
Keyboard: 3
PCB: 1004

I run a complete calibration on ch1 after the FW 1.00.2 upgrade and the drops off issues I had after 30Mhz are gone. It seems that is was just related to the calibration file.
So you just need to take som time and also run a full calibration on your channels  (91 values to measure and enter  )
You were right, the FPGA version must match FW and PCB versions. So with your FW and PCB versions, you need at least a FPGA v12

[alex.forencich]

Did the board revision reported by the unit change when you upgraded the FPGA?

No the PCB version did not change. It was reported as 1004 after the upgrade. I changed it manually back to 1002 in the system.inf file but it did not do any good and was not changed back to 1004 after a reboot.

I would say that firmware versions are just supposed to  match the PCB versions and that no check is done in the firmware.
I was running a stock FW 1.00.2 on my PCB 1002 without any warning and it is obviously made for PCB 1004.
I suppose that it is just written in system.inf, like the model and serial number.

Ah, I forgot about system.inf.  That's a rather annoying way to do board revisions.  Anyway, I presume most of the changes will be on the FPGA side of the board.  Do you have the ability to do JTAG boundary scans?  It would be good to compare 1002 vs 1004.  The UCF file for the 1004 revision is here: https://github.com/alexforencich/hdg2000/blob/master/fpga/fpga.ucf .  Can you possibly check this against your board and see what matches?  Don't bother with the DDR2 part, this should be fixed due to the embedded memory controller.