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

[A Hellene]

Thank you Krater, I appreciate this!

By the way, this is my little contribution to your project:
A .CFG file for the ADSP-BF531 including almost all the processor's register and bit definitions (see the _BOOTROM_X entries!), built specifically for the Rigol firmware files. Please, rename the attached file from .txt to .cfg.

Speaking of which, have you seen my latest addition to the DS1000 schematics?


-George


EDIT: Wrong file! Updated now.

[dsokoll]

I just finished the upgrade. I have HW 58 and am running 00.02.04.SP1. So far it works perfect...tomorrow will be the big test. I am going to hit it with a 3000V ( properly attenuated ) pulse with a 150ps rise time... i will test it side by side with my Tek 6604 6gHz scope.

[dsokoll]

After completing the DS1052E upgrade, i tested the bandwidth with a pulse from a Barth pulse generator. The pulse was 3000V with a 300ns pulse width. I used Barth GLP-26, GLP-20 and a 142-NMFP-20B attenuators for a Vr of 1,986.42. The TEK TDS6604 registerd a rise time of 168ps while the Rigol came in at 3.320ns, i believe is just better than the 100mHz. With a measured Vmax of 1.46 that would put the measured voltage at 2900.18. Not bad for a $379.00 oscilloscope. Very happy with this product...

[saturation]

Thanks dsokoll!  Its good to compare the 1052e to your top notch gear to know the Rigol still lives up to its 100 MHz spec after the hack.

After completing the DS1052E upgrade, i tested the bandwidth with a pulse from a Barth pulse generator. The pulse was 3000V with a 300ns pulse width. I used Barth GLP-26, GLP-20 and a 142-NMFP-20B attenuators for a Vr of 1,986.42. The TEK TDS6604 registerd a rise time of 168ps while the Rigol came in at 3.320ns, i believe is just better than the 100mHz. With a measured Vmax of 1.46 that would put the measured voltage at 2900.18. Not bad for a $379.00 oscilloscope. Very happy with this product...

[pepo]

 
Thank you guys,
just wanted to confirm I made the downgrade of my brand new DS1052E from 2.06 to 2.02,
patched it according the description on page one and files found on page 61
and then upgraded it again to 2.04.
After calibrating it, it works as expected!

[alm]

Well done test. Rise time is consistent with a 100 MHz bandwidth. This is in contrast to some earlier tests that suggested it had much more than 100 MHz bandwidth. I wonder if this is due to differences in the scopes or test methodology.

[yanir]

Well done everyone, all the hard work has actually forced rigol to sell the 100Mhz version for 399!
http://www.saelig.com/product/PSPC017.htm

No need to work so hard on the hack .

[torch]

That's the exact same price they are selling the 50Mhz DS1152E for:
http://www.saelig.com/PSBEB100/PSPC016.htm

Looks like complete capitulation!

  I'm gonna miss this thread...

[Lightages]

There is no need to stop developing mods for the DS series. The 50MHz version is still available $50 cheaper, plus with the hard work that people have been doing we might have more or different functionality!

[p51dblue]

Hello, I just got a Rigol DSE1052 and I looked at the first page on this thread I got the soft version
00.02.05.02.00
= 58

I thought that I will be able to hack it but ...I am doing something wrong...cant do.
I got all the soft installed and try to do it. I am getting lost .
In my case do I need USB stick or not. If my version is 2.5.SP2 I will have to go to George stuff on Page 43.
There I do not know what to do, I do not get how to do it.
I was looking to find that Frank say was able to make .

Bottom line I am to dumb to figure out. Can any one help me out to upgrade mine?

Thank You
Adrian

[A Hellene]

After completing the DS1052E upgrade, i tested the bandwidth with a pulse from a Barth pulse generator.
[...]

Very nicely done, dsokoll! Thank you for the quite interesting results of the tests you've done!

Really, have you taken these readings by upgrading your DS1052E to the 100 MHz DS1102E model or to the 150 MHz DS1152E one (by entering "DS1152E" as the model type instead of "DS1102E" for the 100 MHz version)?

Though the pictures seem to be missing, there is a post by flolic, a fellow EEVBlogger, who broke the 200 MHz bandwidth barrier on his hacked '1052, by having also pushed the bias of the front-end differential amplifier LMH6552, by slightly changing the input and the gain feedback resistors from 365 to 270 ohms.


-George

[A Hellene]

Adrian,

See this post and this one, also.


-George

[dsokoll]

Very nicely done, dsokoll! Thank you for the quite interesting results of the tests you've done!

Really, have you taken these readings by upgrading your DS1052E to the 100 MHz DS1102E model or to the 150 MHz DS1152E one (by entering "DS1152E" as the model type instead of "DS1102E" for the 100 MHz version)?

George,

I apologize for not enclosing pictures of the test setup. This should answer your question.

Regards,
Doug

[A Hellene]

Ah, thank you Doug.

It keeps bugging me, though, the "what if" question. I mean, having access to such powerful test equipment, like you do, I would certainly try to find out whether instructing the DS1052E to become a DS1152E device would make any substantial difference from being a <EDIT>DS1052E DS1102E</EDIT> model or not.


Thank you, again,
-George

[dsokoll]

George,

Forgot to send these....This is the 6kV 300ns charge line and the attenuation that was used.

Doug

[dsokoll]

George,

My understanding is the timebase is still limited to only 2ns. Maybe is would affect the overall bandwidth and allow a faster rise time, which would still be good. I can't answer that though.

Doug

[A Hellene]

Forgot to send these....This is the 6kV 300ns charge line and the attenuation that was used.

Ah, this is pure electronics geeks porn!
Thank you for sharing!


On the other matter, yes the timebase will remain the same. But, somehow, I am a little bit optimistic since the device does not have a -3dB knee at 100 MHz, as expected by a 100 MHz model. Actually, the attenuation of the DS1102E seems to be only 0.29 dB (= -20*log(2900/3000)) at 100 MHz, which means that the device is, probably, more than capable of handling higher bandwidths when the software filtering of the top frequency will be pushed even further, to the 150 MHz point.


Anyway, thank you for your contribution,
-George

[dsokoll]

The above test data was from a single pulse from a 300ns charge line. I was just measuring the rise time from this very fast pulse generator to see if the conversion worked... I think the original rise time was around 6.8ns. and now its anywhere between 3.3ns and 3.6ns when sending a burst of pulses at 10Hz. No noise issues. I actually wish i had bought the 1102 with the logic analyzer though....

[dsokoll]

I am a little bit optimistic since the device does not have a -3dB knee at 100 MHz, as expected by a 100 MHz model. Actually, the attenuation of the DS1102E seems to be only 0.29 dB (= -20*log(2900/3000)) at 100 MHz

I will run a more detailed series of tests and post the results in the near future.

[Mechatrommer]

I am a little bit optimistic since the device does not have a -3dB knee at 100 MHz, as expected by a 100 MHz model.

this is tested with a 300ns pulse, i would say/certified this as 3MHz signal attenuation test. to properly profile the signal attenuation, a clean sine signal should be used AFAIK. i already PM dsokoll for this, it seems there will be light on this.

[Mechatrommer]

I actually wish i had bought the 1102 with the logic analyzer though....

ISTR there is not so good review on Rigol LA module. so be careful on what you wish. even DS1052E+LA will cost twice. so DS1052E still bang for buck you can get. YMMV though.

[alm]

I am a little bit optimistic since the device does not have a -3dB knee at 100 MHz, as expected by a 100 MHz model.

this is tested with a 300ns pulse, i would say/certified this as 3MHz signal attenuation test. to properly profile the signal attenuation, a clean sine signal should be used AFAIK. i already PM dsokoll for this, it seems there will be light on this.

Try looking at the frequency spectrum of a square wave. It has infinite odd harmonics. If you pass the signal through a low-pass filter, the rise time will be increased, which is what this test is based on. For a Gaussian one-pole filter, this depends on the -3dB point of the filter. The large pulse width is optimal, since it makes it easy to determine the 0% and 100% points and calculate the 10% and 90% points. This is an issue with the semi-dirac pulse used by some hobbyists.

[vl400]

Standard unmodifed DS1052E fed from an Agilent E8257D freq synth..

[Mechatrommer]

thanks for the pic vl400. any tabulated data? ok i saw it, bottom left, thanks alot. all the frequencies are using 1Vpp sine signal right? sorry but qurious.

[vl400]

Correct, 1v p-p signal for all. The time base changed, otherwise the scope settings remained unchanged. The pics were just for my reference point prior to doing the hack (still not done).