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

[smgvbest]

Just got my HDG2002B and did the bandwidth mod without any issue's (after I found my 3.3V USB-TTL adapter)
Verified everything worked as planned and it did now goes to 100MHz.

a few things I did notice before and and after the mod
Before opening the unit I ran s self text and everything passed except the CH1/2 Zone test which failed.
I searched this thread and saw one mention of if but no one replied so wondering if this is normal?

Next after opening the unit there were a few things I noticed
The metal shields are not present around the output section of my unit

it looks like a few relays were hand soldered and the melted the casing when soldering
See Image.   It looks like it's surface damage only but it bugs me on a new unit seeing this.
Also you can see the cans missing from the output section.



I Highlighted the RJ45, Magnetics and Crystal as well as hopefully labeled the Serial Port Correctly in this view

[alex.forencich]

Just got my HDG2002B and did the bandwidth mod without any issue's (after I found my 3.3V USB-TTL adapter)
Verified everything worked as planned and it did now goes to 100MHz.

a few things I did notice before and and after the mod
Before opening the unit I ran s self text and everything passed except the CH1/2 Zone test which failed.
I searched this thread and saw one mention of if but no one replied so wondering if this is normal?

Next after opening the unit there were a few things I noticed
The metal shields are not present around the output section of my unit

it looks like a few relays were hand soldered and the melted the casing when soldering
See Image.   It looks like it's surface damage only but it bugs me on a new unit seeing this.
Also you can see the cans missing from the output section.

The cans were missing on mine as well.  The relays also look hand soldered, but it seems they did a better job on the one I got.  They are bowed out on the top, but I think that is normal for these things. 

[andrija]

Those relays look like they were pulled from a 10 year old equipment used by a smoker. Not saying they aren't new but they sure don't look new to me. My own unit looks fine. Software, not so much.

[smgvbest]

Could someone run the self test on thier unit and see if it fails one of the zone tests on each channel please?

[Howardlong]

Could someone run the self test on thier unit and see if it fails one of the zone tests on each channel please?

Here are my results:

[smgvbest]

Here's mine
I wonder how the ebay seller will respond to this one.
It also after being powered down all night reverted to Chinese for the language when I powered up for this photo



And I just notices after posting that the date/time is off again

[Howardlong]

I wonder what a zoom unit is in this context?

Have you done a cal on both channels since doing any mods? Wonder if that has anything to do with it.

[smgvbest]

I checked it both before and after the mod and in both cases it failed.
I have not yet done a cal on it but it does need it and seems to work ok but like you I am not sure what a zoom unit is.
first thing that comes to mind is maybe the range limits controlled by those relays.
During that part of the self test I do hear the relays trip a total of 4 times,  2x for ch1 and 2x for ch2

would you mind checking again and listen for how many times you hear them trip?
I counted every time I head them trip after the FPGA Well test.

[fremen67]

I checked it both before and after the mod and in both cases it failed.
I have not yet done a cal on it but it does need it and seems to work ok but like you I am not sure what a zoom unit is.
first thing that comes to mind is maybe the range limits controlled by those relays.
During that part of the self test I do hear the relays trip a total of 4 times,  2x for ch1 and 2x for ch2

would you mind checking again and listen for how many times you hear them trip?
I counted every time I head them trip after the FPGA Well test.

There are 4 relays for each channel. You will find the schematics of channel 1 made by Cyber7 here:
https://www.eevblog.com/forum/testgear/hdg2002b-awg-firmware-reverse-engineering/msg538156/#msg538156

This is the same for channel 2, you just need to replace K4, K5, K6 & X5 with K1, K2, K3 & X4

K1(K4) is activating the x5 gain stage, K2(K5) is activating the x4 gain stage, K3(K6) is activating the channel output.
It's not clear what the X4 & X5 relays are doing.

I captured the data going to the shift registers (U24&U26) which are powering the relays when going from 0v to 10V with the DC function and when doing a selftest:
From 0 to 0.5V: no amplification
From 0.5 to 2V: K2(K5) is used: x4 amplification
From 2V to 10V: K1(K4) and K2(K5) are used: x20 amplification

During the selftest you should hear 5 clicks and not 4, each click beeing separated by 2 seconds :
1 - K5 ON
2 - K4 ON + K5 OFF
3 - K4 ON + K2 ON
4 - K4 ON + K1 ON + K2 OFF
5 - K4 OFF + K1 OFF

If I follow what is written on screen during the selftest, Zoom Unit 0 should be activated by K5 on channel 1 and K2 on channel 2, Zoom Unit 1 should be activated by K4 on channel 1 and K1 on channel 2

If K4 and K1 are dead (or not switched) on your HDG, you will hear only 4 clicks during the selftest (which is the case) and you should not be able to reach more than 2V on both output of your HDG. Could you check that?

Could you also tell us what firmware/hardware version you have?

Edit: I forgot to mention I have PCB 1002 with FPGA 11 and software upgraded from 1.00.1(140402.0) to 1.00.2(140926.0)

[smgvbest]

Both CH1 and CH2 I hear the relays kicking is around 100mV, 2V and 5V

PCB       1004.1
SW        1.00.3(141225.0)
KERNEL 3.2.35
FPGA     30
KEY       3

and I got up to 20Vpp out

[fremen67]

Both CH1 and CH2 I hear the relays kicking is around 100mV, 2V and 5V
and I got up to 20Vpp out

I suppose you are not talking of the DC function. Is it the values with the Sine function?

This is strange that your selftest is not working but your HDG seems OK... Did you already try to see if some specific function is not working?

In fact I just realized that the ranges are not the same depending on the function your are working with   
I just captured data to the shift register with the DC function from 0 to 10V and and I only saw 2 ranges. I didn't pay attention up to now but I can hear also 3 clicks on the sine function range: around 100mv, 1.3V and 1.4V. It's very strange it is so close for the 2nd and the 3rd on my HDG.
I will open my HDG again and will try capture with other functions. Maybe I will figure out what X4 & X5 are used for...

PCB       1004.1
SW        1.00.3(141225.0)
KERNEL 3.2.35
FPGA     30
KEY       3

That's at least the good news with your HDG. You seems to be the first person here to have this new version
I wonder what Hantek solved with this last release.
The .1 in the PCB version is surprising... let's hope that PCB 1004 and 1004.1 are compatible so that other people will be able to use the new FPGA 30.
From what I saw, FPGA versions are linked to PCB versions (which seems quite logical). As I have a PCB 1002 for example, I have to stick with FPGA 11 even if I could upgrade to Software 1.00.2.
Would you be able to backup your HDG and PM me a link to it? Don't post it directly here unless you want to share your serial 
I suppose that now they won't change the partitions on the nandflash anymore so that you can use the 1.00.2 backup scripts posted here: https://www.eevblog.com/forum/testgear/hantek-hdg2002b-awg-5mhz-or-100mhz-let's-see!/msg540248/#msg540248

To make it easier, I also created today 2 scripts that use the update process of the HDG to backup the nand flash partitions directly to a sd or a usb stick. You can now backup the HDG without opening it (plus the calibration file and the system.inf file are also saved just in case). The backup will be in the "backup_hdg" folder, either on your SD or USB stick, depending on the script you choose (The backups are not compressed like they were in the older script in order to speed up the backup process).
To use the new backup scripts, just unzip the one you want to use and put it on your backup media, go in the Utility menu, select Update, select your disk, select the backup script... and launch the update. It will last for 3 to 5 minutes, depending on your media speed. Don't be afraid by the "Updating..." message. It does not write anything on the HDG. After the backup is finished, the HDG will reboot automatically.

Are some of you interested in real upgrade scripts? I upgraded from 1.00.1 to 1.00.2 by flashing the nand with an external tool because of the nand layout change but I think this could also be done via the upgrade process. Upgrades from newer versions will be much easier to do.

[smgvbest]

I ran your script and PM you the ftp information to grab it

[fremen67]

Thanks to Sandra's backup I was able to upgrade my HDG from 1.00.2 to 1.00.3 (I still have an old PCB 1002)

This version (1.00.3) seems to be a big improvement in term of stability. I played with it for 1 hour and I had no freeze nor reboot.
The calibration file from the previous version is not compatible. I had to run the calibration process again (I just run the DC part as it takes a lot of time).

The way the 3 ranges are used also changed and now, like Sandra, I can hear the relays clicking around 100mV, 2V and 5V for a Sine function at 1kHz (it also changes, depending on the frequency).
The .1 from PCB 1004.1 seems to be related to a modification on the output relays of channel 1: just after upgrading the software to 1.00.3, I replaced the FPGA config file (FPGA v30 now) with the v11 corresponding to my PCB 1002 but kept the PCB version unchanged in the system.inf file (1004.1). I had only channel 2 which was working. No click for channel 1. With "PCB 1002" in system.inf, channel 1 was back on duty!
With "PCB 1004" in system.inf , my 2 channels are still working and I have an extra menu in Utility which is "Word Generator". This "Word Generator" Menu is not displayed when I put "PCB 1002" in system.inf. As there seems to be an extra chip on the PCB1004, I suppose it is related to it..

Regarding the Self Test, apart from the 4 first steps which are succesfull, the 6 last steps are failing... but I did not see anything wrong during the time I played with the HDG.
I will perform a full calibration on both channels and will try again later on.

Edit: No failures anymore during the selftest after a calibration of both channels 

[smgvbest]

Did you have all tests pass again in the self test?
I am curious if this version just behaves differently or if I have a bad unit.

Got my first reply from the seller on this.   
They suggested to not power cycle the unit so fast and wait at least 30 seconds before turning back on.
Replied.  It had been off for at least 8 hrs then was on for at least 5 minutes before running the self test. 

FWIW, no where in my message to them did I mention the unit was ever power cycled.  OYE, its going to be one of these

[smgvbest]

@fremen67
So with the 1.03 software you are seeing failures now in the self test also?
if so that could be they broke the self test and did not fully regression test the unit out.  from what I've been reading that would not surprise me of Hantek.   

I've not run into any real issues with it other than lack of a decent manual in english

[fremen67]

@fremen67
So with the 1.03 software you are seeing failures now in the self test also?
if so that could be they broke the self test and did not fully regression test the unit out.  from what I've been reading that would not surprise me of Hantek.   

Yes I did but they just disappeared after the calibration process. Did you tried it on your HDG?

I've not run into any real issues with it other than lack of a decent manual in english

At least that's a good thing you did not notice any special issue 
Did you download this manual? https://www.eevblog.com/forum/testgear/hantek-hdg2002b-awg-5mhz-or-100mhz-let's-see!/msg454279/#msg454279

[fremen67]

@fremen67
So with the 1.03 software you are seeing failures now in the self test also?
if so that could be they broke the self test and did not fully regression test the unit out.  from what I've been reading that would not surprise me of Hantek.   

Yes I did but they just disappeared after the calibration process. Did you tried it on your HDG?

I just powered on my HDG today and run a selftest: lasts 6 steps are failing again 

[smgvbest]

The seller say sure they will be happy to exchange or refund my HDG2002B but I have to pay return shipping to china.
Since your seeing failures now with the 1.03 version you loaded I'm wondering if I do have a problem or it's a bug in the self test.

I was trying to start a cal but when I follow the instructions on the manual they do not match well at all
I checked if anyone had posted a better description of the calibration procedure but do not see one.

Would anyone be willing to post the process and maybe some reference as to what to expect.
As I do understand the first 9 tests are DC values and I can use my meter to read those out.   I did start that and set the meter to run an average and I let it run until it gave a steady value (about 500 samples, I am using a DM3058E) and it seemed I had to input the value, press value then next to go to the next one which didn't agree with the manual.

I do have a backup thanks to fremen67 script for the current calibration.

At this point I am not so sure that the self test is not just broken in this version

@fremen67 what would be your thought on downgrading to the 1.02 version and testing under it?
next would be do you have a copy of it and how to install it?  Only if you think its worth a try though

[fremen67]

I was trying to start a cal but when I follow the instructions on the manual they do not match well at all
I checked if anyone had posted a better description of the calibration procedure but do not see one.

Would anyone be willing to post the process and maybe some reference as to what to expect.
As I do understand the first 9 tests are DC values and I can use my meter to read those out. I did start that and set the meter to run an average and I let it run until it gave a steady value (about 500 samples, I am using a DM3058E) and it seemed I had to input the value, press value then next to go to the next one which didn't agree with the manual.

Yes that's almost it. When you are in the calibration menu:
1 - Select CH1 or CH2 with F1
2 - Press F2 to start
3 - Press F2 again (Next) to start with the 1 st value
4 - Measure the value, enter it and push F3 (Value) to validate the value
5 - repeat step 3&4 till the last value (90 values  )
6 - Press F4 save and you are done

With SW 1.00.3 the calibration output values seems to have changed. It seems that the HDG outputs now only the maximum values for each range.

The 12 first values are DC values so you just need to use your DMM (I used my Metrix MX54C and waited  some seconds for each value so that it stopped changing. Your DM3058E seems more than enough accurate  for that  )
The DC values seem to be now:
- 2 times the min&max Voltage with x1 gain
- 2 times the min&max Voltage with x4 gain  (only the x4 amplification stage)
- 2 times the min&max Voltage with x20 gain (the 2 amplification stages in serial,  x4 & x5)
For example I had 0,5781 -0,5805 0,5782 -0,5806 2,2476 -2,2664 2,2478 -2,2671 11,145 -11,259 11,145 -11,262 V

For the last 78 values you will need a scope as the signal is a sine wave at different frequencies with the maximum output for the 3 ranges (I connected directly the probe with an adapter on the HDG output).
The 26 frequencies are 0,5 1 2 5 6 10 12 18 21 25 26 30 35 40 46 50 51 55 56 60 62 68 72 78 85 90 Mhz
You have to enter the Vpp value for each frequency (so in my case 27x values starting at 1.16V, 27x values starting at 4.5V and 27 values starting at 22.5V). I used for that the average acquisition mode of the scope, waiting some seconds between each value so that it stopped changing).

@fremen67 what would be your thought on downgrading to the 1.02 version and testing under it?
next would be do you have a copy of it and how to install it?  Only if you think its worth a try though

If you still have self test issues after the calibration process, it could be worth trying with sw 1.00.2.
You will just need to keep your FPGA 30 configuration file because of your PCB 1004.1.

If you want to do it, just tell me and I will send you a custom 1.00.2 file plus another one to go back to 1.00.3 after the test.
There are several ways of doing the downgrade, depending on the tools you have. The easiest way for you would be that I provide you with an .upk "upgrade" file that your HDG could directly use (2 in fact, one to downgrade to 1.00.2 + one to upgrade back to 1.00.3).
The second way would be to do it from u-boot (via the serial console and a PC connected on the USB port of the HDG).
You can also do it via JTAG with OpenOCD (very slow) or H-Jtag (much better but you need Windows XP or Win7 32bits when using a Wiggler JTAG device).

Anyway I will try to package an upgrade file so that others having 1.00.2  can also upgrade to 1.00.3.

I wonder when Hantek will start posting official upgrade files on their site ...

[smgvbest]

Thank you very much for clarifying the process
It's probably not good to do while tired so I'll work on it tomorrow and let you know the results

[smgvbest]

Definition of PITA = Calibrating the HDG2002B to 100Mhz   
Definition of Sucks = Realizing afterward you did the whole thing with 20Mhz BWL on   

Oh FWIW, at 100Mhz @ 1Vpp with BWL on it read 100Mhz @1.01Vpp.   Now it's 3-4Vpp without BWL on   

[Howardlong]

Definition of PITA = Calibrating the HDG2002B to 100Mhz   
Definition of Sucks = Realizing afterward you did the whole thing with 20Mhz BWL on   

Oh FWIW, at 100Mhz @ 1Vpp with BWL on it read 100Mhz @1.01Vpp.   Now it's 3-4Vpp without BWL on   

I feel your pain, I failed to backup mine after a full calibration, then it decided to rebuild itself a few days later :-(

There is apparently an autometed way of doing it but I think you need a Hantek scope to do it.

[smgvbest]

I Reran the Calibration and after a total of 127 Measurements total, 12 on the DM3058E in the 200mV, 5V and 20V ranges and the rest on the MSO1104Z I made it through it all 

The good news it it's allot more accurate now
The bad news CH1 Zoom 1 still fails


More good news is I found where i can change the default key to Print Screen instead   
but you can't print the Hardware/Software Pages 

@fremen67
If you have the firmware stuff so I can downgrade to 1.02 and give that a try I would appreciate it.
But also to upgrade it back to 1.03 after.

I am really hoping its a bug in the self test in 1.03. and this should prove it

[fremen67]

I Reran the Calibration and after a total of 127 Measurements total, 12 on the DM3058E in the 200mV, 5V and 20V ranges and the rest on the MSO1104Z I made it through it all 

Hey you cheated, you were supposed to enter 180 values! That's not fair, 127 are too easy 

@fremen67
If you have the firmware stuff so I can downgrade to 1.02 and give that a try I would appreciate it.
But also to upgrade it back to 1.03 after.

I am really hoping its a bug in the self test in 1.03. and this should prove it

No problem. I will try to create an upgrade package so that you don't need to connect to the serial nor to the JTAG port. I just have to figure out the best way to enter recover mode without manually stopping the boot process.

[fremen67]

There is apparently an autometed way of doing it but I think you need a Hantek scope to do it.

Yes I PM Scope Sam who posted the info https://www.eevblog.com/forum/testgear/hantek-hdg2002b-awg-5mhz-or-100mhz-let's-see!/msg577081/#msg577081 but had no feed back...
I did some tests yesterday with the HDG and my DS1054z using Agilent Connection Expert. With the right driver (IVI) you can control the HDG calibration process with SCPI commands. You can also easily control and read values from the DS1054z. That means that it should not be that difficult to write a script that would automate the calibration process. At least a part of it for me as I don't have a SCPI compliant DMM.
One advantage of calibrating the HDG via SCPI, apart of the PITA part  , would be that you may be able to jump directly to one of the 90 calibration steps (for SW 1.00.3). That means you could calibrate separatly the DC part and one of the 3 amplification stages.

The documentation is quite lite but there are at least some information on those calibration commands:

CALibration:SETup <step>
CALibration?
CALibration:VALue <value>
CALibration:STORe

There is a 1 offset in the steps between the manual calibration in front of the HDG and the calibration via SCPI. This you can also see on the video posted by Scope Sam
CALibration:SETup 2 outputs the first DC calibration value and CALibration:SETup 91 outputs the last AC calibration one .