DG4000 - a firmware investigation

[Gandalf_Sr]

I tried Ted's open, no change, save CAL technique (on HFLAT only) but it does not seem to have worked; at some point, the values that self-populated for the steps took a jump from 16's to -2's.  I get significant drop in output level (Vpp) when going up to 100 MHz and more at 200 MHz.  I guess I'll have to do a real CAL 

[ted572]

I tried Ted's open, no change, save CAL technique (on HFLAT only) but it does not seem to have worked; at some point, the values that self-populated for the steps took a jump from 16's to -2's.  I get significant drop in output level (Vpp) when going up to 100 MHz and more at 200 MHz.  I guess I'll have to do a real CAL 

Gandalf:

Please read all of the below so that you know what to expect:

For a DG4202 the CW RF Output should be flat up to 200MHZ within +/- 0.8dB after the Calibration recovery has been completed.  Note that the Sine Sweep Mode is still limited to a maximum frequency of 160.000,000 MHz for a flat RF Output.  Any Stop frequency above this at all, will screw up the frequency response even below 160MHz.  So if want to go beyond 160.000,00MHz do it in Sine CW mode only.

Still think your DG4202 has a problem:

Check that your DG4202 is Set for (under Utility, Settings) 'Power, Default', Not 'Power, Last'.  Then I suggest going back to FW 00.01.06.  Verify that you still have a DG4202 model installed per the DG's 'Utility', 'System', 'Information' menu.  If not use the appropriate CEN file to change it to one.  Perform the Calibration Recovery routine for Amp-AC*, LFLAT*, and HFLAT.  Then go back up to to FW 00.01.08.    How is it now?

Note: *   Cal for Amp-AC and LFLAT Cal. may not be required, but I do it to prevent other potential glitches.  I had a few people tell me that doing this fixed some issues for them, but I don't have any conclusive evidence.  It's easy to do, but you can always go back and do these two later (if you chose to) after you have determined your unit is now working Ok in manual CW mode to 200MHz.

The above is what I have successfully done and has worked for others, but if there is something unusual with your unit (i.e.  FW/hardware) I'm NOT responsible.  I think you will be OK, but I live far, far away, so don't come looking for me.

If you have questions, but no bitching please (Ha Ha), feel free to contact me via a PM.

       Good luck, Ted572

[TooOldForThis]

One thing I found is that when using a DG4062 (FW rev 1.07) to measure ~1 second  periods,  the value displayed on the LCD and the value sent over the USB interface don't match.  The value that comes back from ":counter:measure?" is always ~13.5nS higher than the value on the display. 
For example,  LCD value =  1.000,000,001,1  while the USB value = 1.000000014E+00 
This problem doesn't happen at higher frequencies.

I reported this to a Rigol app eng.  The last I heard from him was "I have reproduced the issue. We are investigating. I'll let you know what I find."    That was six months ago.

[ted572]

One thing I found is that when using a DG4062 (FW rev 1.07) to measure ~1 second  periods,  the value displayed on the LCD and the value sent over the USB interface don't match.  The value that comes back from ":counter:measure?" is always ~13.5nS higher than the value on the display. 
For example,  LCD value =  1.000,000,001,1  while the USB value = 1.000000014E+00 
This problem doesn't happen at higher frequencies.

I reported this to a Rigol app eng.  The last I heard from him was "I have reproduced the issue. We are investigating. I'll let you know what I find."    That was six months ago.

Install FW 00.01.08 and then try this again.  There were some bugs in .07 that were fixed in .08.  Even the installation of FW .07 was a little different than normal.  In fact several reported that they didn't think the installation completed properly.  Although the DG reported that FW .07 was installed (via. Utility, System, Information).  I don't know that the issue you found was fixed, but you certainly do want to go to .08.

[TooOldForThis]

I had updated to 1.08 in July.  It didn't help.

Here are the results using rev 1.08 firmware:
Setup:  Measuring the period of a GPS pulse-per-second output. The DG4062 is running on an external 10MHz Rubidium time source.  Gate Time = 10sec. HF suppression on.

Front panel period measurement = 1.0000000011   (dang, off by 1.1nS  )

At the same time, the SCPI output for  “:counter:measure?” is
9.999989975E-01, 1.000000014E+00, 1.000000047E+01, 1.000029678E-01, 8.999970466E-01
(Frequency, Period, Duty Cycle, Positive Pulse Width, Negative Pulse Width)

The problems are:
    The period value displayed over SCPI is always 13-14nS higher than the value on the display.
    The period and frequency values in the SCPI data are not reciprocals of each other.
 
On the plus side, the Positive Pulse Width and Negative Pulse Width values add up nicely to the Period value in the SCPI data.
When test equipment gives you conflicting data for the same measurement, which data, if any, should you believe?

[ted572]

The NEW 2014 Rigol DG4000 Performance Verification Guide is attached:

[ted572]

The new Rigol 2014 DG4000 Calibration Guide is attached:

[Gandalf_Sr]

A friend has a DG4162 that's sw vn 00.01.04, he contacted Rigol who've sent him firmware vn 00.01.09.00.01 along with a new vn 00.06 bootloader which he's been told to apply before the firware.  He contacted me because the instructions from Rigol on how to apply the update don't seem to work.  Any ideas?

I wonder why there's a new bootloader? It's possible that they are trying to close down the loopholes.  I'm thinking that he should try the firmware update without the bootloader first.

[Orange]

A friend has a DG4162 that's sw vn 00.01.04, he contacted Rigol who've sent him firmware vn 00.01.09.00.01 along with a new vn 00.06 bootloader which he's been told to apply before the firware.  He contacted me because the instructions from Rigol on how to apply the update don't seem to work.  Any ideas?

I wonder why there's a new bootloader? It's possible that they are trying to close down the loopholes.  I'm thinking that he should try the firmware update without the bootloader first.

Most likely the 1.09 firmware can only be loaded with that new boot loader. The new boot loader prevents you then from loading an older version of the firmware.
Try to get firmware 1.08. This can be loaded without to upgrade the boot loader.
BTW, the patch that is available makes a 200MHz generator. Since your friend's model is already a 160MHz generator loading 1.09 firmware is only a small improvement. It's a different story if you own a 60 or 100MHz model.

I forgot to mention that you need a small size USB stick, 1 or 2 Gbyte. Larger sizes give errors
 

[Rory]

What bugs does the 1.09.00.01 firmware fix?

[rhost]

I forgot to mention that you need a small size USB stick, 1 or 2 Gbyte. Larger sizes give errors

 I used a 32GB Sandisk Cruzer on mine. I have 3 other brands of various sizes and they all didn't work. I think the brand makes more of a difference than the capacity.

[MiataMuc]

Hi all,

just a short question: anyone bought a DG4062 in the last few days and knows what firmware will be delivered?

Thanks,

Flo

[rhost]

Hi all,

just a short question: anyone bought a DG4062 in the last few days and knows what firmware will be delivered?

Thanks,

Flo

I bought one within the last month or so and the firmware version was

Software 00.01.08
FPGA 00.01.09
Hardware 01.03
Keyboard 05.01

[dave3533]

Hi all,

just a short question: anyone bought a DG4062 in the last few days and knows what firmware will be delivered?

Thanks,

Flo

I purchased a DG4062 from TEquipment on 10/01/2014 and the system info was:
Firmware: 1.08
FPGA: 1.09
Hardware: 1.03
Keyboard: 5.01

The calibration paperwork from Rigol states it was calibrated in April of 2014.

[pedjasns]

Just got a new DG4062

SW: 00.01.09
FPGA: 00.01.09
HW: 01.03
Keyboard: 06.01

Can it be hacked now or not? 

[MiataMuc]

Seems not to be the case with the methods we know.
Maybe someone who knows the internals is willing to help? I can offer any data which is needed; I do have an Olimex JTAG adapter and will provide any data which is requiered..

Florian

[PA0PBZ]

Just got my DG4062, unfortunately SW: 00.01.09. Does anyone have the 00.01.09 update GEL file?
Maybe all they changed is the key, so I'd like to have a look at the update.

[MiataMuc]

I could just ask them for the software, but they want to know the serial, and might find out that they delivered my DG4062 with  the latest firmware. Anyone with an older firmware might be able to help you..

Florian

[PA0PBZ]

I could just ask them for the software, but they want to know the serial, and might find out that they delivered my DG4062 with  the latest firmware.

Yes, I'm facing the same problem...

[Sparky]

Just got my DG4062, unfortunately SW: 00.01.09. Does anyone have the 00.01.09 update GEL file?
Maybe all they changed is the key, so I'd like to have a look at the update.

I could just ask them for the software, but they want to know the serial, and might find out that they delivered my DG4062 with  the latest firmware. Anyone with an older firmware might be able to help you..

Hope this helps! Piranha(DSP)Update_00.01.09.zip

[PA0PBZ]

Hope this helps!

Thanks Sparky! Now for some serious disassembling 

Edit: Looks like they encrypted the update GEL file now... Well, the answer must be in the boot loader 

[TooOldForThis]

Hope this helps! Piranha(DSP)Update_00.01.09.zip

I applied this update a week ago.  I found that you have to rename the GEL files to DG4000Update.GEL before the boot loader will recognize them on your USB stick. 
I don't know what 1.09 fixes, nothing that I could see.  The counter function is still gives different values on the LCD vs SCPI output for slow frequencies (~1Hz)  and the frequency and period values for a single measurement are not reciprocals of each other. Just kind'a close.

[EV]

I tried this "Piranha(DSP)Update_00.01.09.zip" update without luck.

When Mod and Utility buttons ware lit I inserted the USB stick with DG4000Update_Bootloader.gel file in it. Something was read for a while and Store button was also lit. Then nothing happened for many minutes and I powered off the generator.

I have now 00.01.08.00.02 FW.

[MiataMuc]

 PA0PBZ:

I'd really like to help, but I do not have any experience in Diassembling :-(

Florian

[PA0PBZ]

Actually I made some progress. The GEL file is not really encrypted, they just added 00-FF repeatedly starting at address 0080, so that was a quick fix.
Comparing with 00.01.08 I found that the key is still the same, they deleted the (nonexisting) DG4202 model and they sortof halfway deleted the .GEN file support 
I'm not sure how to go from here, I could try to downgrade to version 00.01.08 by lying about the version maybe, or try to find out if there is still support for .GEN files hiding. I'm trying to find a blackfin plugin for my IDA but having a hard time to find a Visual C 6.0 to compile it. Anyone have a spare blackfin plugin for IDA 6.1?