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

[rosasharn]

Could built-in waveforms like StairUp, StairDn, StairUD, PPulse, NPulse, Trapezia, UpRamp, Dn, Ramp, Cardiac, Earthquake and common mathematical waveforms be preloaded onto HDG2000B? 

[rosasharn]

Could built-in waveforms like StairUp, StairDn, StairUD, PPulse, NPulse, Trapezia, UpRamp, Dn, Ramp, Cardiac, Earthquake and common mathematical waveforms be preloaded onto HDG2000B?

Anyone has clue about it?

[rosasharn]

How to establish communication with the device using the USB port on the back of the HDG2000B?

[Starcruiser]

@fremen67,

I was looking up colorcoding for the chip inductors and depending on size they will either have no color coding, be a single color or have multiple dots which follow typical resistor coding nomenclature 1st 2nd, multiplier (zero's), see graphic

[Starcruiser]

So I finally received my 34401A 
Here is the modified version of the HDG calibration tool with:
- Excel x64 support (thank you smgvbest for the feedback)
- 34401A DMM support for DC values
- Rigol DS/MSO2000A support thanks to flatlander
- Excel sheet update improvement

If you don't want to use a device (DMM or DSO), just leave the adresss/alias empty.
The DSO detection type is done using the IDN string so it should be able to detect a DS/MSO1000Z and DS/MSO2000A (hopefully ).
Depending on the calibration value type, you should be prompted to connect the correct device to the HDG.

I also wanted to add a progress status bar but it's bed time here ...
Hope this will help!

Will this work with the DS5202P as a data source? I read that you wrote this for the MSO series but did not see anyting about the DSO mentioned, perhaps I missed it. I would use this with Win7 64bit OS as well.

[Starcruiser]

So I finally received my 34401A 
Here is the modified version of the HDG calibration tool with:
- Excel x64 support (thank you smgvbest for the feedback)
- 34401A DMM support for DC values
- Rigol DS/MSO2000A support thanks to flatlander
- Excel sheet update improvement

If you don't want to use a device (DMM or DSO), just leave the adresss/alias empty.
The DSO detection type is done using the IDN string so it should be able to detect a DS/MSO1000Z and DS/MSO2000A (hopefully ).
Depending on the calibration value type, you should be prompted to connect the correct device to the HDG.

I also wanted to add a progress status bar but it's bed time here ...
Hope this will help!

Do you know how this can work with the Hantek DSO? The HDG is loading with the IVI driver but the DSO5072p (upgrade to 5202p) is not recognized with the Hantek driver, I am looking for a compatible one but not had any luck yet.

I was able to use you spreadsheet tool to record and upload the config using the latest version you provided (thanks!) but wanted to automate. Any ideas on how we can do this?

The HDG2002B I have has;

HW: 1004.1
SW: 1.00.3.02.03(1504230)
Kernel: 3.2.35
FPGA: 30
Kybd: 3

Thanks in advance.

[fremen67]

Fremen67,
With your spreadsheet tool, do you know how this can work with the Hantek DSO? The HDG is loading with the IVI driver but the DSO5072p (upgrade to 5202p) is not recognized with the Hantek driver, I am looking for a compatible driver but not had any luck yet, any suggestions?

I was able to use you spreadsheet tool to record and upload the config using the latest version you provided (thanks!) but wanted to automate.

The HDG2002B I have has;

HW: 1004.1
SW: 1.00.3.02.03(1504230)
Kernel: 3.2.35
FPGA: 30
Kybd: 3

Thanks in advance.

Hello Starcruiser,
The Excel calibration tool is only able to connect to instruments that support SCPI commands (via VISA driver).
As far as I know,  the DSO5072p does not support scpi commands.
I have a DSO5062D and I can't use it for the calibration of my HDG2002B. It would require to write a driver on the PC side which would be able communicate with the specific protocol of Hantek DSOs.
You will find informations on that protocol here: http://elinux.org/Das_Oszi_Protocol
It seems that someone got a calibration tool from Hantek that works with Hantek DSO ... https://www.eevblog.com/forum/testgear/hantek-hdg2002b-awg-5mhz-or-100mhz-let's-see!/msg577081/#msg577081 but I don't know how to get it

The SW version of your HDG is a new one. Would you be able to dump your HDG and post somewhere the files so that I could create a update file? You will find an "upgrade" package in my onedrive that will allow you to backup your nand  https://onedrive.live.com/redir?resid=C666E0DC0AF0C9E2!114&authkey=!AMtRrjm2d33uI8Q&ithint=folder%2c (in the backup folder).

Thanks in advance!

[smgvbest]

@Starcruiser and @Fremen67
I'd love to see that dump also,  maybe they fixed that zoom selftest finally 

[robert_]

Im having some really bad effects after the upgrade, namely at any frequency beyond 5Mhz, the amplitude goes nuts.
Ch1: looks right below 1V,above that it jumps to 22V
Ch2: about 6x the value set, above 1V (till 20V) fixed 22V
Adding offset also causes total f..up.

With sweep, the amplitude stays stable IF the frequency is set to max. 5Mhz before turning on sweep.

Any idea what causes that, and how to fix it? Some sort of countermeasure?
At the moment, i can just downgrade it again to avoid breaking everything

Just remember that i tried to do the DC calibration, and it decided to reboot after a few values. Maybe related... Didnt check the amplitude flatness before that, so i dont know if it ever worked.

PCB 1004.1
SW 1.00.3
FPGA 30
Keyboard 3

[robert_]

Ok, looks like the failed calibration is to blame.
Did a quick and dirty cal of Ch2 again, and now it looks much better. I dont know why it crashed on the first attempt.
Amplitude is surprisingly stable even up to 100Mhz, maybe -2db. Flatness is so-so but probably caused by my QaD Cal. I could do much better with some more time and motivation (and using my LeCroy 9354 in my main lab instead of the old 9304 here).

[smgvbest]

Ok, looks like the failed calibration is to blame.
Did a quick and dirty cal of Ch2 again, and now it looks much better. I dont know why it crashed on the first attempt.
Amplitude is surprisingly stable even up to 100Mhz, maybe -2db. Flatness is so-so but probably caused by my QaD Cal. I could do much better with some more time and motivation (and using my LeCroy 9354 in my main lab instead of the old 9304 here).

@robert
Have you run a selftest on it?  I'd be curious if you get the same zoom failure I get as you have a slightly newer version of the firmware.
Could you run the selftest and let me know??

[robert_]

Yes, zoom unit 1 still fails on both channels.

[smgvbest]

Thank you,
I wonder what they fixed in your release.   if anything

[fremen67]

Im having some really bad effects after the upgrade, namely at any frequency beyond 5Mhz, the amplitude goes nuts.
Ch1: looks right below 1V,above that it jumps to 22V
Ch2: about 6x the value set, above 1V (till 20V) fixed 22V
Adding offset also causes total f..up.

With sweep, the amplitude stays stable IF the frequency is set to max. 5Mhz before turning on sweep.

Any idea what causes that, and how to fix it? Some sort of countermeasure?
At the moment, i can just downgrade it again to avoid breaking everything

Just remember that i tried to do the DC calibration, and it decided to reboot after a few values. Maybe related... Didnt check the amplitude flatness before that, so i dont know if it ever worked.

PCB 1004.1
SW 1.00.3
FPGA 30
Keyboard 3

Hi Robert,
You seem to have the latest HW (PCB 1004.1). I think I missed something... What sw version did you have before the upgrade : 1.00.x(yymmdd.0) ? What version exactly do you have now after the upgrade 1.00.3(yymmdd.0)?... and where did you get it from?

If you encounter a problem, performing a complete calibration after an upgrade is the first thing you have to try.

[fremen67]

Yes, zoom unit 1 still fails on both channels.

Failed also on mine before a complete calibration. If you have some time, try at least on one channel...

[robert_]

Still fails even after full cal on both channels. Did also fail right out of the box. Probably doesnt mean much, as everything works so far.
As for the newer firmware, it seems to be somewhat more stable, expect for the one time it rebooted at my first cal attempt, i didnt witness any crash yet.

Oh, and now for some bad news: The whole thing is seriously dangerous.
The power supply is constructed with totally inadequate isolation between primary and secondary, and the secondary side is NOT grounded. So in case the isolation fails, the outputs will be live.

Why? Have a look at the transformer, trough the slits at the top of the bobbin. Primary and secondary windiwngs are made with normal magnet wire (approved TIW looks different), with no margins on both sides of the bobbin and only one layer of thin tape in between. As the secondary is not grounded, it would have to fulfill reinforced isolation, ie 3kV proof voltage, and (depending on material group, PD and installation category) at least around 5mm of creepage distance, and 3 layers of tape. What it has is pretty much ZERO creepage. From my experience, crap like this break down with <2kV, often less than 1kV when hot. Spikes of that magnitude are not uncommon, especially on 230V mains.
Also, the isolation is crossed by a 4.7nF "Y1" capacitor that looks too small for its ratings. Looks more like a Y2 from the ceramic thickness.
Apart from that questionable part, this capacitor results in the non-grounded outputs having significant voltage (and quite some peak current when shorted) to ground, which can kill whatever circuit it is connected to (and you can even feel it).
At least, the secondary common needs to be solidly grounded to improve safety for now, but Hantek really has something to explain as they stuck a CE mark on a non-compliant device.

[fremen67]

As for the newer firmware, it seems to be somewhat more stable, expect for the one time it rebooted at my first cal attempt, i didnt witness any crash yet.

Could you give more details one your firmware version? Did your device came with a new version or did you upgrade it yourself?

Oh, and now for some bad news: The whole thing is seriously dangerous.
The power supply is constructed with totally inadequate isolation between primary and secondary, and the secondary side is NOT grounded. So in case the isolation fails, the outputs will be live.

Why? Have a look at the transformer, trough the slits at the top of the bobbin. Primary and secondary windiwngs are made with normal magnet wire (approved TIW looks different), with no margins on both sides of the bobbin and only one layer of thin tape in between. As the secondary is not grounded, it would have to fulfill reinforced isolation, ie 3kV proof voltage, and (depending on material group, PD and installation category) at least around 5mm of creepage distance, and 3 layers of tape. What it has is pretty much ZERO creepage. From my experience, crap like this break down with <2kV, often less than 1kV when hot. Spikes of that magnitude are not uncommon, especially on 230V mains.
Also, the isolation is crossed by a 4.7nF "Y1" capacitor that looks too small for its ratings. Looks more like a Y2 from the ceramic thickness.
Apart from that questionable part, this capacitor results in the non-grounded outputs having significant voltage (and quite some peak current when shorted) to ground, which can kill whatever circuit it is connected to (and you can even feel it).
At least, the secondary common needs to be solidly grounded to improve safety for now, but Hantek really has something to explain as they stuck a CE mark on a non-compliant device.

Any modification to suggest?

[robert_]

My HDG came with the SW 1.00.3 installed.

Well, as the PSU is crap, the only way to sort of inprove the situation is to solidly connect the secondary common (BNC shield etc) to the earth conductor. With hopefully intact grounding in the IEC cable and bulding installation, the risk of dangerous failiure is greatly reduced. Now it would blow the fuse when the transformer breaks down, and ther will also be no more common mode voltage on the outputs.

However, the transformer is still not adequately insulated (to comply with IEC60950 or 60010, both of which have pretty identical requirements), as it still would require basic isolation which needs a creepage distance well above zero.

Im going to build my own PSU though, since i want to keep the outputs floating to avoid ground loops. This however requires the digital isolators to be fitted, because otherwise the output will be grounded (not really safely, just noisy) as soon as the USB cable is connected.
Looks like the board was intended to have 2 ADUM2400/3400 series isolators fitted. Pinout matches, and it also seems to be the x400 variant with 4 channels in the same direction.
Also the processor/USB side has spaces for 2 1117 linear regulators to produce 3.3 and 1.8V which are not fitted, as these voltages are supplied by the analog/FPGA side by the 0R resistors. Removing the resistors and installing 2 regulators should resolve that, the 4pin connector for supplying 5V on the processor side is already fitted. 1.8V seems to need the LT1117-ADJ because the feedback divider is already there, 3.3V has no divider and ADJ pin grounded so should get the LT1117-3.3.

[fremen67]

My HDG came with the SW 1.00.3 installed.

Would you be able to dump your HDG and post somewhere the files so that I could create a update file? You will find an "upgrade" package in my onedrive that will allow you to backup your nand  https://onedrive.live.com/redir?resid=C666E0DC0AF0C9E2!114&authkey=!AMtRrjm2d33uI8Q&ithint=folder%2c (in the backup folder).

Well, as the PSU is crap, the only way to sort of inprove the situation is to solidly connect the secondary common (BNC shield etc) to the earth conductor. With hopefully intact grounding in the IEC cable and bulding installation, the risk of dangerous failiure is greatly reduced. Now it would blow the fuse when the transformer breaks down, and ther will also be no more common mode voltage on the outputs.

However, the transformer is still not adequately insulated (to comply with IEC60950 or 60010, both of which have pretty identical requirements), as it still would require basic isolation which needs a creepage distance well above zero.

Im going to build my own PSU though, since i want to keep the outputs floating to avoid ground loops. This however requires the digital isolators to be fitted, because otherwise the output will be grounded (not really safely, just noisy) as soon as the USB cable is connected.
Looks like the board was intended to have 2 ADUM2400/3400 series isolators fitted. Pinout matches, and it also seems to be the x400 variant with 4 channels in the same direction.
Also the processor/USB side has spaces for 2 1117 linear regulators to produce 3.3 and 1.8V which are not fitted, as these voltages are supplied by the analog/FPGA side by the 0R resistors. Removing the resistors and installing 2 regulators should resolve that, the 4pin connector for supplying 5V on the processor side is already fitted. 1.8V seems to need the LT1117-ADJ because the feedback divider is already there, 3.3V has no divider and ADJ pin grounded so should get the LT1117-3.3.

This is interresting, any pictures for details?

[rodlaird]

Sorrt dumb question for Fremen67...

I downloaded your zip file with the various backup and upgrade scripts - loaded afg_HDG2002B_to_HDG2102C.upk on a FAT32 formatted USB and....

...nothing. 

I get an error -129 when I try to perform the update (or -4 if I retry).

newb error on my part obviously... But advice appreciated...

[fremen67]

Sorrt dumb question for Fremen67...

I downloaded your zip file with the various backup and upgrade scripts - loaded afg_HDG2002B_to_HDG2102C.upk on a FAT32 formatted USB and....

...nothing. 

I get an error -129 when I try to perform the update (or -4 if I retry).

newb error on my part obviously... But advice appreciated...

I suppose that the script can't find system.inf... What firmware version do you have?
This script should work on a HDB2002B with firmware >= 1.00.2 where system.inf is located in /dso/root. It won't work with firmware 1.00.1 where system.inf is in /etc.
I am not at home at the moment so I can't check further 

[pascal_sweden]

I haven't been following this thread at all, and just read the initial posting.
Here are my questions to make a long story short, and jump in the subject area

Is the latest HW version of HDG2000B (5MHz) still hackable to HDG2102B (100MHz)?

Where can I buy HDG2000B in Europe? Which distributor?
How much does HDG2000B and HDG2102B cost respectively? What is the price difference?

How does the HDG2000 series perform?
Let's say I want to use it in the 25MHz region. Does it perform as good as the Rigol SDG1025 in the 25MHz range?
Let's say I want to use it in the 100MHz region. Does it perform as good as the Rigol DG4102 in the 100MHz range?

Is there a formal document with all the details for performing the hack? I don't like to scroll through 35 pages of postings I am looking for a summary that comprises the details on how to do the "upgrade", what the pros are, what the cons are, which risks there are, impact of future FW upgrades on performed "upgrade", etc.

Your help will be highly appreciated. I jumped straight from the initial entry to the last entry, and want to get up to speed on this "upgrade"

[fremen67]

Is the latest HW version of HDG2000B (5MHz) still hackable to HDG2102B (100MHz)?

The last known version is 1.00.3. It is still hackable.

Where can I buy HDG2000B in Europe? Which distributor?

I got mine there: http://www.rigoloscilloscope.fr/wholesale/Hantek-HDG2002B-2-Channels-16Bits-250MSa-s-Function-Arbitrary-Waveform-Generator-5MHz-695.html
They also have Websites for Uk, Germany... but it comes from china.

How much does HDG2000B and HDG2102B cost respectively? What is the price difference?

Have a look there: http://www.aliexpress.com/store/group/Function-Arb-Waveform-Generator/803942_257042015.html

How does the HDG2000 series perform?
Let's say I want to use it in the 25MHz region. Does it perform as good as the Rigol SDG1025 in the 25MHz range?
Let's say I want to use it in the 100MHz region. Does it perform as good as the Rigol DG4102 in the 100MHz range?

Good hardware potential but low software quality. Hopefully a opensource version will be available in the next weeks/months.
alex.forencich worked on the FPGA software and I am currently working on a GUI version.
I don't have other AWG to compare with.

Is there a formal document with all the details for performing the hack? I don't like to scroll through 35 pages of postings I am looking for a summary that comprises the details on how to do the "upgrade", what the pros are, what the cons are, which risks there are, impact of future FW upgrades on performed "upgrade", etc.

Hack is just done by changing the name of the device in a file.
You can also do it using the "upgrade" functionnality of the HDG with a.upk file. You can download it on my onedrive:
https://onedrive.live.com/?id=C666E0DC0AF0C9E2!124&cid=C666E0DC0AF0C9E2&group=0&authkey=!AMtRrjm2d33uI8QThere is a readme.
pros: 100Mhz limit
cons: you have to perform a complete calibration (I wrote a tool working in Excel to automate the process, available in the previous posts)
no known risk as long as you perform a full backup of your device (scripts on onedrive too).

[Howardlong]

How does the HDG2000 series perform?
Let's say I want to use it in the 25MHz region. Does it perform as good as the Rigol SDG1025 in the 25MHz range?
Let's say I want to use it in the 100MHz region. Does it perform as good as the Rigol DG4102 in the 100MHz range?

I can't (yet!) directly compare it to an SDG1025.

I can however compare it to the dual 25MHz signal generators integrated into the DS/MSO1000Z-S scopes, because despite the inevitably more fiddly interface on the scopes, the Rigol is far better just from a software reliability aspect.

This AWG had the prospects of being a really useful piece of test gear, but regretfully out of the box it's little more than a toy I'm afraid. This is almost all down to the quality of the software being really very bad. There isn't much redeeming to say about it other than the hardware itself, which isn't too bad.

The proof is in the pudding, and regrettably this unit doesn't get used that much as the software just is not trustworthy enough: I have to double check what it's outputting before applying it to a DUT for fear it might do some damage!

If you're a tinkerer, with time on your hands, you might just about suffer its limitations, but a serious piece of TE it's not I'm afraid. Looks like it's the result of the software dept going out for a boozy lunch and never coming back.

[smgvbest]

Looks like it's the result of the software dept going out for a boozy lunch and never coming back.

 ,  so true