changing the rigol DS1052E to DS1102E using USB , the dummy guide

[torch]

Ok, so what changes have been made? You have it, have you installed it yet?

My scope is running just fine with 2.04, and so I'm not anxious to upgrade at this point. Rigol has been known to introduce new bugs with new untested firmware versions (the infamous screen flicker of 2.05 is one example).

People are still getting scopes with 2.06 from the factory.  Nobody has claimed to have v3 except GM Davies a month or so ago, and he isn't talking. I am wondering if his scope escaped from the factory with an experimental firmware that turned out to be buggy and so Rigol cancelled it. I'm not ready to risk installing it on my working scope without more information.

[adrianblack]

People are still getting scopes with 2.06 from the factory.  Nobody has claimed to have v3 except GM Davies a month or so ago, and he isn't talking. I am wondering if his scope escaped from the factory with an experimental firmware that turned out to be buggy and so Rigol cancelled it. I'm not ready to risk installing it on my working scope without more information.

I wouldn't trust firmware from a file sharing site like that -- you're asking for a brick ... and if someone posts a picture of a scope running this new v3 firmware I may be less skeptical that it even exists at all.

[T4P]

New firmware for models DS1000E (D) (DSP) update_ Version 00.03.01.00.00 
? DS1000E(D)(DSP)update_ Version 00.03.01.00.00..rar ? rapidshare.ru
http://www.rapidshare.ru/2836652
http://www.rapidshare.ru/2836658

[adrianblack]

? DS1000E(D)(DSP)update_ Version 00.03.01.00.00..rar ? rapidshare.ru

LOL!!!!! Someone here must read Reddit ;-)

[amspire]

With the older DS1102CD model, when they updated the firmware from 02.xx.xx to 03.xx.xx, it was accompanied by a hardware change that made the newer firmware incompatible with the older models.

So there is a fair chance that the 00.03.01 firmware means that that scope has some different hardware in it. It could even be a short term hardware change as a result of a component shortage.

[Mikhail Ivanov]

Hi everyone,
I just received my scope and it has 00.03.01 hw 58.
I tried to downgrade it by changing header, but it doesnt work.
Has anyone successfully downgrade new firmware?

[ghostx7]

Why drop it patched firmware!!!!that you do not like about this firmware?

[adrianblack]

Hi everyone,
I just received my scope and it has 00.03.01 hw 58.
I tried to downgrade it by changing header, but it doesnt work.
Has anyone successfully downgrade new firmware?

Can you take a picture and upload it of the info screen with the extra info and versions?

[adrianblack]

Hi everyone,
I just received my scope and it has 00.03.01 hw 58.
I tried to downgrade it by changing header, but it doesnt work.
Has anyone successfully downgrade new firmware?

Can you post a photo of the info screen?

[Mikhail Ivanov]

Hi everyone,
I just received my scope and it has 00.03.01 hw 58.
I tried to downgrade it by changing header, but it doesnt work.
Has anyone successfully downgrade new firmware?

Can you post a photo of the info screen?

[ghostx7]

discuss the new firmware DS1000E(D)(DSP)update_ Version 00.03.01.00.00!!!

[Sfinx]

Mine DS1052D has the same new fw:

serial:            DS1ECXXXXXXXXX
software:       00.03.01.00.00
dsp:              00.01
fpga:             03.07
Hardware       58

The header of this new fw is similair to 2.06 (no DS1000 string at the beginning) :

0000000: 2966 b0d5 fde7 56fd 82a4 8184 8888 ff03  )f....V.........
0000010: 09f2 5823 5160 0080 ff04 0000 0010 00ae  ..X#Q`..........

Anybody tried to downgrade, hack and upgrade back to 03.00 ?

[ioclk]

I see that you have FPGA 03.07 / DSP 00.01 with HW 58, other HW 58 comes with FPGA 02.00 / DSP 00.00
It may be something new in the FPGA, not another board as HW but a major change and this can make a big difference for the firmware.
Did anyone figure what are the FPGA and DSP change?

[avp1]

I installed 3.01 firmware in my DS1052E ("upgraded" to 100MHz) using images posted above on this forum. So far it looks good. The only difference I see - change in noise pattern at 5mV range. Factory version was 2.05

[Sfinx]

How to downgrade from 3.01 for hack ? Seems like header is crypted or so.

[Sfinx]

Anybody has exact connection wiring between BF531 and FLASH chip ? Trying to read the flash through JTAG, but got only FF's for now. I want to dump the flash, find and replace the serial and write it back.

[Sfinx]

Ok, seems I replying to myself I've managed to read the flash through jtag. So the model and serial was found in flash dump as ascii strings at offset 0x1ff440. Simpy changing them and flashing back do not work - seems like firmware checks something else beside these strings too. It will be good if somebody can tell here how to patch the flash dump as writing reading the whole flash using jtag takes 2 hours per one operation

[Sfinx]

Forgot to say: I will be appreciated for ds1102d flash dump too.

[torch]

You probably need to talk to Drieg. Apparently the Blackfin processor is not compatible with certain JTAG interfaces. A special (read: expensive) one is required. Some FLASH signals routed differently and you have to understand how to generate those signals.

I seem to recall some discussion about how the flash was subdivided as well -- something along the lines of each successive firmware upgrade is actually stored in the opposite half of the flash memory space from the half currently in use. I don't remember if that was ever definitively resolved though.

Make a mistake and you could brick your scope.

[Sfinx]

The BF is fully JTAG complaint until you do not use their internal debug protocol that triggered my special EMU pin. The ds1052 flash routing is clear and was investigated by tester in 10 min. The flash can be read with the cheap LPT ByteBlaster cable (can be assembled for $1-$2 in 3 minutes). I've already have read it, and can change it and flash back again and again - so the brick do not afraid me. I will post all the info here just after the hack will succeed. Will try to contact Drieg about details.

[KTP]

I just read through some of this thread (haven't been up to date on the Rigol hacking since I did mine in 2010).

I recently bought an Agilent MSOX3024a 200mhz scope and while playing with it I also went back to looking at my modded Rigol DS1052E.  I have software version 00.02.02 SP2 (I pretty much did not change anything since late 2010).

So...two questions:

1)  Should I bother upgrading to 00.02.04?

2)  Why is my rise time so good on this scope?  I am measuring about 2.2ns rise time with a 750ps risetime, 20ns pulse from a tinylogic buffer circuit (4V into 50 ohm externally terminated at the Rigol channel 1 input).  From what I read in this thread, people were seeing around 3.2ns.  Just lucky hardware draw, something special about 00.02.02 or a flaw in my setup?

Here is a pic of the pulse on the Rigol (reused from my other post about a clock generator)

[KTP]

I modified the pulser circuit a bit by adding 50 ohm series resistance at the source.  This made a marked improvement in the pulse ringing and is quite a beautiful fast 4.5V 750ps pulse for almost no work and zero money (the tinylogic parts were free samples  ).

**Edit:  It is roughly a 2.5V pulse now because of the addition of the 50ohm source series resistance.  Rise time 600ps on Tek7104 1Ghz scope.

The interesting thing, and a reason why I am commenting in this thread, is that the Rigol really does seem to have a bandwidth of 140mhz to maybe even 180mhz if I am not doing anything wrong.

In the first pic I measured the pulse with my 200mhz Agilent MSOX3024A and found a rise time of 1.59ns (limited by the BW of the Agilent, as this pulse is less than 750ps on a 1Ghz scope).

In the second pic, I measured the pulse on the Rigol modified DS1052E firmware 00.02.02 SP2 and found a rise time of 2.512ns (normal mode aquisition, real time sampling, sinx/x)

Lastly, in the third pic, I used the same Rigol scope only this time with 16 averages, equal time sampling, sinx/x and found a rise time of 1.928ns!  This calculates using .35/rt to a BW of over 180mhz.  And the pulse looks great, almost as good as the Agilent (sadly for 1/15 the cost)

Maybe I will keep firmware 00.02.02 after all 

[torch]

In pages 50 and 51 of this thread, I played around and eventually achieved a consistent rise time of 1.84ns (+/- 0.04ns) by telling the Rigol it was a DS1152-EDU (I have 2.04SP1 and HW58).

As a DS1102E the rise time was consistently 1.96ns (+/-.04ns), which is remarkably close to your 1.92ns screenshot. So I would say that either you are doing it right, or we made the same mistakes ;-)

I think there is no question that the 50mHz Rigol is batting way beyond it's class. A remarkable little machine at a phenomenal price/performance ratio. After the appropriate hack, of course!

[saturation]

... with so many pages on this thread new readers should remember the sampling rate of 1Gs/s is only using a single channel and lowest record length, engage the second channel or the next higher memory length on a single channel and the sampling rate halves.   So, always keep the aliasing in mind.

[torch]

Good point. My results were the absolute best case scenario.

Still, the results are pretty impressive for a $300 50MHz scope...