Hacking the HDO1k/HDO4k Rigol 12 bit scope

[skander36]

there's a section in the code called "Change the data" which is completely commented out.

I have seen that increment loop few times when inspected the code but I didn't realized that I should decoment first
Active blindness ...
Now it works.

[trinacria]

I didn't realized that i should decoment first

The format of vendor.bin might change in the future, so it's good to check if it can recreate the original file first. Consider the whole file as example code, not as a black box ready to go.

[skander36]

I tested just for curiosity about the HW capability of the scope. I don't search for a deep mod as it is clear that without consistent modifications at main app, it cannot be used too much. at least for my two channel version. I did a backup of my original vendor.bin.

[trinacria]

is clear that without consistent modifications at main app, it cannot be used too much.

Agreed. There seems to be no value in doing this except to poke around.

[scopeman]

Try cursors.  One knob should do horizontal setting and the other vertical

Thank you, but believe me I have checked all contexts as I said. Both knobs have vertical and horizontal functions in different contexts. I have checked all functions many times.Switched from automatic to manual and back, etc.
The best method to check is to use integrated routine from Self Check->Key test. In attached pic you can see that this knob(1) don't take any value, while the others have values each other.
Fun part: Yesterday I have filled a request for assistance at Rigol, but soon after that the knob started to work. I didn't received yet an answer from Rigol.

Could you post the screen with HW/FW versions ?

My scope works like a charm encoders wise, they do no skip a bit even if i rotate them very fast, so in your case it could be a bad soldering or damaged device.

Question on the Flex Knobs: Can you assign the flex knobs so that Knob 1 is on Cursor 1 and Knob 2 is on Cursor 2?

Sam

[ebastler]

Ok, now i'm messing with the fan speed as suggested by markone.  My scope is running in a cool basement so what would be a practical pwm value for the fan in this case?

I understand that the DHO1000 (and 4000) has PWM-controlled fans, but always runs them at full speed anyway -- at least with the early firmware versions. Has this been improved in the latest firmware? Or are there instructions somewhere on tweaking the PWM settings? I could not find anything, be it from markone or other users. Thanks!

[skander36]

They are running at full speed.

[ebastler]

They are running at full speed.

Thanks. It's a bit disappointing that Rigol have not improved this so far. How annoying do yo ufind the fan noise?

Did anyone come across details about changing the PWM settings via some Android hack? (Is it even confirmed that the fans are PWM-driven? How do we know if they are always on at full speed?)

[tautech]

They are running at full speed.

Thanks. It's a bit disappointing that Rigol have not improved this so far. How annoying do yo ufind the fan noise?

Did anyone come across details about changing the PWM settings via some Android hack? (Is it even confirmed that the fans are PWM-driven? How do we know if they are always on at full speed?)

First hack Dave tried to do was quieten it.
Tried other fans and a bunch of stuff but only quantified noise by ear. 

Sorry that was the 800.

[NE666]

How annoying do yo ufind the fan noise?

That's a very subjective question, of course.  Personally, I don't find it to be a problem with mine (although to be fair, I've only run it ~ 1 hour so far, more to come this weekend if there is time).  However, we have to consider a) my age and b) my experience.  As you probably know, the high frequency response of human hearing tails off markedly early on in our lives.  I and others simply won't hear to any real degree some high frequency components of that noise that you and others might, regardless of whether it can be measured by electronic equipment.  And then I think life experience sets some expectations.  I started my career working in labs with bank upon bank of diagnostic instrumentation that ran 24x7, bathing rooms in constant fan noise.  I went on to spend nearly a decade working in data centres where fan noise, was and to this day remains, riotous.  So I'm likely biased - nothing from this scope compares to what I've been accustomed to and so I just don't notice/care about it, perhaps.

All that said, I still don't think you will find it so bad as to be unworkable for you.  It's louder that my old 2645B but much quieter than the Agilent I used to have.  YMMV.

Consider also that if it does bother you, your environment might play a part in ultimately how much you perceive the noise.  How you site it on your bench i.e. proximity to back wall, other objects, will have an effect on higher frequencies.  It might be a case of adding some deadening material to your wall behind your desk/bench to cut down reflected noise, for example.

Did anyone come across details about changing the PWM settings via some Android hack? (Is it even confirmed that the fans are PWM-driven? How do we know if they are always on at full speed?)

I can't speak to how control has been implemented (I have no intention of opening mine) but there is control of some kind.  There is a noticeable increase in fan speed during the boot process.  Whether this translates to full scale speed can't be known without measurement but subjectively, it doesn't sound like it to me.

[Fungus]

How annoying do yo ufind the fan noise?

That's a very subjective question, of course.  Personally, I don't find it to be a problem with mine (although to be fair, I've only run it ~ 1 hour so far, more to come this weekend if there is time).

The DHO800 fan is noisy, for sure, but for some reason I don't find it too annoying. Not sure why.

Also noteworthy is the complete lack of any discussion on fan replacement in the DHO800/DHO900 threads. Go figure.

[ebastler]

Thanks all for the feedback, especially NE666. The fan characteristics should be quite different between the 800 and 1000 series: A much straighter air path on the 1000, so I would expect less high-frequency noise from the airflow itself. But then there are two fans, and they must have put those in for a reason, so the airflow volume must be on the high side.

Agree, it is subjective after all, so I will have to listen for myself. My DHO1074 is scheduled to arrive today. Maybe I can justify a quick incoming inspection before I put it away again until Christmas. 

[NE666]

The DHO800 fan is noisy, for sure, but for some reason I don't find it too annoying. Not sure why.

It's your age 

[skander36]

They are running at full speed.

Thanks. It's a bit disappointing that Rigol have not improved this so far. How annoying do yo ufind the fan noise?

Did anyone come across details about changing the PWM settings via some Android hack? (Is it even confirmed that the fans are PWM-driven? How do we know if they are always on at full speed?)

It's somewhat annoying but I have another scope much more annoying (Micsig TO3004)  You can work with it ...
In this thread I remember Markone has find a setting for lowering the speed. You can try the following command:   echo 100> /sys/devices/platform/pwm_fan/hwmon/hwmon5/pwm1  where echo can have values from 50...255. It will lower the fan but not at 0 speed. at 50 it is barelly audible but some of the chips temp increase to 72 degree C (from 64) (it has a hw monitor in board test section).

[ebastler]

In this thread I remember Markone has find a setting for lowering the speed. You can try the following command:   echo 100> /sys/devices/platform/pwm_fan/hwmon/hwmon5/pwm1  where echo can have values from 50...255. It will lower the fan but not at 0 speed. at 50 it is barelly audible but some of the chips temp increase to 72 degree C (from 64) (it has a hw monitor in board test section).

Aha, thank you! I had seen a mention of @markone's fan exploration ealier in this thread, but could not find the actual posts when I skimmed the thread. The command you quoted gave me something to search for; these seem to be the relevant discussions in the thread:

https://www.eevblog.com/forum/testgear/hacking-the-hdo1khdo4k-rigol-12-bit-scope/msg4576183/?topicseen#msg4576183
https://www.eevblog.com/forum/testgear/hacking-the-hdo1khdo4k-rigol-12-bit-scope/msg4618909/#msg4618909

[skander36]

Yes. As I said after i used it for a period, I did not find the noise level as annoying as to try to find a way to lowering it. In a home lab where is perfectly quiet maybe can be perceived as annoying but on a work lab where there are other tools running (AC, vents, sources) it is not obvious.

[ebastler]

I was looking through the older posts and found this discussion on the contents of the .GEL file. (Which is just a compressed archive and contains common Android software for the DHO1000 & 4000, but model-specific FPGA configurations.)

The readme.txt file in the FPGA folder has this content: [...]
I ran it through a translator:

Code: [Select]

FPGA bit file description
[...]
HDO2000/1000-FPGA-A100T
bit1 SPU_H12S2_SelfTest.bit Complete chip and board level self-test; including DDR calibration parameter readback.
bit2 SPU_H12S2.bit Regular oscilloscope program, no Bode map function
bit3 SPU_H12S2_bode.bit Bode function version, no high resolution
[...]


Has anything ever transpired about the "HDO2000" model which is mentioned in the list of FPGA configurations?

I was wondering whether there might be a planned intermediate model based on the same PCB, which Rigol never shipped but which is supported in the firmware. In a perfect world, this would use a single ADC but enable higher bandwidth (400 MHz using up to two channels?) and the 50 Ohm inputs -- such that a DHO1000 could be made to believe it's a DHO2000 without causing side effects. 

Not sure how this "make believe it's a DHO2000" would work though. When people tried to run their DHO1000 as a 4000 (with heavy side effects), I assume a vendor.bin from an actual 4000-series scope was swapped in? Or did decrypting, modifying and re-encrypting the vendor.bin actually work for someone? From the recent discussion here I could not figure out whether a model change was successfully made this way.

EDIT: After re-reading trinacria's post (just below the post linked above) it seems that a change to just the model name in vendor.bin did have some effects. But the "fake DHO4000" worked to an even lesser extent than one would expect, without 50 Ohm inputs and triggering even in those two channels which have an ADC. So there might be more software "switches" that need to be set, or subtle differences on the PCBs?

[ebastler]

Alright, so the Auklet.apk (Android app which runs the scope) inside the .GEL file also mentions the HDO2104 and HDO2204 models. Which probably means that these model designations can be used in vendor.bin and will configure the scope app in some more or less meaningful way.

No HDO2404 model is listed, so there's little hope of support for 400 MHz bandwidth on two channels max. (Still worth a shot to try the BW2T4 upgrade on it in a second step?) But maybe 50 Ohm termination, or more memory, or I²S decoding will be available?

My DHO1074 is currently back in its box, waiting for Christmas eve. But unless someone beats me to it, I will definitely try a HDO2204-flavored vendor.bin once I get to play with it!

[gburdzin]

So I just received my DHO1104 and am attempting to use the rgtool1000 code to upgrade it.  Of course the code is setup to generate licenses for the BW7T20 and RLU options, but since my model is already enabled for 100MHz, the BW upgrade license is reported as invalid (RLU works fine though).

I have tried updating the code to use the BW100T20 option and then also modified the device id to be HDO1104 instead of HDO1072, but none of the generated licenses are accepted when using these permutations.  The licenses created using the BW100T20 option don't seem to do anything at all (scope doesn't even give an error message), while the ones for BW7T20 at least give a message about "invalid license".

Any thoughts on how to debug this further?  I'm not sure who found the list of options, but is it possible to double check what the correct option for going from 100->200MHz BW would be?

Thanks very much...

[ebastler]

I have tried updating the code to use the BW100T20 option and then also modified the device id to be HDO1104 instead of HDO1072, but none of the generated licenses are accepted when using these permutations.  The licenses created using the BW100T20 option don't seem to do anything at all (scope doesn't even give an error message), while the ones for BW7T20 at least give a message about "invalid license".

Isn't the option called BW10T20? If you typed the extra zero in the script too, that might explain the problem.

[gburdzin]

The original code has "BW100T20", so that's what I tried.  I just tried "BW10T20" and that has no affect either.

[mm1]

So I just received my DHO1104 and am attempting to use the rgtool1000 code to upgrade it.  Of course the code is setup to generate licenses for the BW7T20 and RLU options, but since my model is already enabled for 100MHz, the BW upgrade license is reported as invalid (RLU works fine though).

I have tried updating the code to use the BW100T20 option and then also modified the device id to be HDO1104 instead of HDO1072, but none of the generated licenses are accepted when using these permutations.  The licenses created using the BW100T20 option don't seem to do anything at all (scope doesn't even give an error message), while the ones for BW7T20 at least give a message about "invalid license".

Any thoughts on how to debug this further?  I'm not sure who found the list of options, but is it possible to double check what the correct option for going from 100->200MHz BW would be?

Thanks very much...

[gburdzin]

Thanks.  The changes that you made to the code is exactly what I had tried already unsuccessfully.

Today I tried again after leaving the scope off for the evening yesterday, and it worked now.  So I suspect there is some kind of silent lockout feature to prevent generation attempts on the license (which makes total sense).

[mm1]

Thanks.  The changes that you made to the code is exactly what I had tried already unsuccessfully.

Today I tried again after leaving the scope off for the evening yesterday, and it worked now.  So I suspect there is some kind of silent lockout feature to prevent generation attempts on the license (which makes total sense).

I don't think so. The rigol1000 file didn't work for me either.

[gburdzin]

Thanks.  The changes that you made to the code is exactly what I had tried already unsuccessfully.

Today I tried again after leaving the scope off for the evening yesterday, and it worked now.  So I suspect there is some kind of silent lockout feature to prevent generation attempts on the license (which makes total sense).

I don't think so. The rigol1000 file didn't work for me either.

After I ran the rigol1000 file, I modified and ran the code many times trying to figure out the correct option code/device id combination.  I know for sure that I finally tried HDO1104/BW10T20 (after ebastler had suggested it below) and it didn't work, so it was definitely ignored by the scope at that point.  It only worked the next day when I tried again after checking what you had done in your version of the code.

Anyway, it is done and working now, so all is good.  I only mention my experience in case someone else encounters similar behaviour and wonders why the upgrade doesn't work.

Thanks