Sniffing the Rigol's internal I2C bus

[Carrington]

At least the transistors seem to be the same. But the compensation stage is certainly different.
Note: I refer to the previous image.


@ Marmad.: You know if there are different 1.0 versions?
For example HW.v.: 1.0.1.0.0 / 1.0.0.0.0 / 1.0.1.0.1 ...

[EV]

Ok, I run the self calibration and new sweep. Maybe there is a small improvement. BW seems to be now about 310 MHz.

In the picture is sweep from 1 MHz to 350 MHz. Sweep time is 14 s, so one horisontal craticule division is 25 MHz. Cursor lines show the 3 dB levels.

No,  I run the calibration and then the sweep again. Wait a moment!

@EV: BTW, did you run the self-calibration after installing the 300MHz option (or before running the sweep)?

[EV]

The true 50 Ohm input is propably better than this feed through terminator.

Then it seems even more illogical that they didn't sell a 300MHz version DS2000 a year and a half ago - unless it was only because of the lack of 50 Ohm input.

[Carrington]

The true 50 Ohm input is propably better than this feed through terminator.

I've always thought the opposite.

[Carrington]

Inverted -> after calibrated.



Where is the difference in BW? 

[marmad]

Ok, I run the self calibration and new sweep. Maybe there is a small improvement. BW seems to be now about 310 MHz.

It looks more like 330MHz 315MHz - although it would be a little easier to read if the full graticule was showing. Of course, there is always some small percentage of +/- error in display, etc.

[marmad]

Inverted -> after calibrated.

The images don't match because EV had statistics ON when doing first sweep - so graticule was reduced to 700x360 - instead of the normal 700x400. One of the images needs to be adjusted for compensation.

[Carrington]

I've resized it better.

[darrylp]

The LMH6518 has bandwidth limits of 20, 100, 200, 350  and 650 750 and unlimited ie 900MHz.

So the 350MHz should be used for the 2302 model I reckon.

--
Darryl

[EV]

Here is one picture more!

[Rigby]

The gain after calibration is very, very small.  Seems like the gain is within measurement noise.  Am I wrong?  I guess I was looking for a huge improvement.

[marmad]

The gain after calibration is very, very small.  Seems like the gain is within measurement noise.  Am I wrong?  I guess I was looking for a huge improvement.

It's HW v.1, IMO, to be expected. I imagine HW v.2 and A models are better.

[EV]

No, the sweep is from 1 MHz to 350 MHz. I checked my text and picture titles there reads 350 not 300.

EV,

I'm not sure that I understand your measurement.
You said that the signal generator is producing a sine wave sweep between 1MHz and 300MHz over a time span of 14 seconds (24.93MHz per second assuming a linear frequency sweep).

The DS2000 display is showing that it's horizontal time base is set to 1 second per division and we see a waveform covering all 14 horizontal divisions of the display.
We also see that the trigger point is off the left side of the display by 1.5 seconds (display shows trigger delay of 8.5 seconds minus 7 divisions of 1 second each).
Wouldn't that mean that the actual sweep time from the generator is at least 15.5 seconds? 
If so, then the sweep would be 349/15.5 = 22.52MHz per second and the -3db point would be about 342MHz since there is an extra 1.5 seconds of sweep off the left side of the display (assuming the the sweep also ends at exactly the last graticule line of the display).

Or are you manually triggering the scope and then centering the entire sweep on the display?

[JDubU]

I see now.  I was confused by the offset trigger point.

[Carrington]

Photo montage 350 MHz BW 

[mightyzen]

I just had to do some testing on the bandwidth of my v2 ds2072. It still runs the original firmware unlocked to 200MHz, simply because I am not ready to lose the 50 ohm termination. May be upgrading it later though.

I have looked at the output of a 125MHz crystal oscillator via:

1. 3300A probe with the ground lead (no bw limit)
2. 3300A probe with spring clip (no bw limit)
3. Coax 20MHz limit
4. Coax 100MHz limit
5. Coax no limit
6. Coax on a crusty hantek dso5072p hacked to 200MHz

The different probing/coax testing was inspired by Mike's excellent video series on the nano lcd reversing. Just shows again that at these frequencies the way you are probing makes a lot of difference.

[m-joy]

Is it True that there was an 200Mhz trial on the ds2072?Because i can not Find that on my ds2072A.if this option has been removed, maybe there can not be keys for activating higher bandwidth on A-models...?

[JDubU]

Is it True that there was an 200Mhz trial on the ds2072?Because i can not Find that on my ds2072A.if this option has been removed, maybe there can not be keys for activating higher bandwidth on A-models...?

There was no 200MHz trial on a new DS2072.

[marshallh]

[gaijin]

I have a 2072 non A 1.0 hardware FW v.02.01.00.03 upgraded to a 2302 with DSHH.
Was wondering if anyone else tried accessing the CAN trigger/decode by SCPI

Querying the current trigger with :TRIGger:MODE? works for all but CAN.
Trying to set the trigger to CAN with :TRIGger:MODE CAN pops a message on the scope "Input is invalid"
:BUS1:MODE? will return CAN and :BUS1:MODE CAN will select CAN.
But I couldn't access anything under :BUS1:CAN like :OFFSet

[JDubU]

I have a 2072 non A 1.0 hardware FW v.02.01.00.03 upgraded to a 2302 with DSHH.
Was wondering if anyone else tried accessing the CAN trigger/decode by SCPI

Querying the current trigger with :TRIGger:MODE? works for all but CAN.
Trying to set the trigger to CAN with :TRIGger:MODE CAN pops a message on the scope "Input is invalid"
:BUS1:MODE? will return CAN and :BUS1:MODE CAN will select CAN.
But I couldn't access anything under :BUS1:CAN like :OFFSet

There is no mention of CAN specific commands in the DS2000A programming guide:
http://www.rigol.com/prodserv/DS2000A/document/?act=view&itemid=749

[AndersAnd]

What's this? Looks like snowy mountains.

[m-joy]

isnt it currently more efficient to buy an Rigol DS1074Z instead of ds2072A ? Because for the DS1074Z the serials seems to work because there is no A version...?
And with options you have at least 100Mhz and stuff compared to the 70Mhz ds2072A and save money.

Greets

[Marty_MC]

Finally I got a DS2072 HW version 1.0 with and version 00.01.01 from a small dealer in south Germany.
I had to drive 100kms one way but it was worth.
 

[Rigby]

isnt it currently more efficient to buy an Rigol DS1074Z instead of ds2072A ? Because for the DS1074Z the serials seems to work because there is no A version...?
And with options you have at least 100Mhz and stuff compared to the 70Mhz ds2072A and save money.

Greets

You have more reading to do.  As soon as someone gets a dump of the firmware on the device of a DS2000A scope, it is very likely that the keygen will be modified to accommodate the newer model.

Just take the time and read this entire thread.  It will take a while, yes.  You will learn many things if you don't rush it.