General > General Technical Chat
The Rigol DS1052E
<< < (33/166) > >>
frederir:
Doing some research on DSO, and found this thread. From the photo we can say there is a least one other component overclocked, the FIFO.
Reference is IS61LPS25636A-200 (http://www.issi.com/pdf/61LPS25632TD.pdf), It is either 32 ou 36 bits wide and can run at a maximum of 200MHz. Now you have 1G sample per second, assuming 8bits, if you want to store it in a 32bits RAM you need to run at 250MHz. Then you are 50MHz over limit....
The other possibility is to reduce samples from 8bits to 6 bits and use the 36bits to store 6 words of 6bits at each clock, you can the store 1.2G samples per second but only with 6bits resolution.
Last solution you can store 4.5 samples of 8 bits each clock in 36bits but it would get you only 0.9G samples per second and the software would be a bit more complex.
My guess is they run the fifo in 32bits and overclock it at 250MHz. Can someone put a frequency counter on the pin 89 of the ISSI chip ?

F.
rossmoffett:
Can't wait until mine arrives so I can get a look inside.

I took another look at the 100 Mhz Rigol scopes we have at school and they are the 1102C model.  400 Ms/s!  Probably the same 100 MHz analog board with different clocks/firmware on the digital side, before they figured out they could reliably overclock.  Would it be useful to graph its frequency response curve?
bushing:

--- Quote from: TrentO on November 16, 2009, 03:20:29 pm ---
--- Quote from: dexters_lab on November 16, 2009, 10:20:04 am ---
they are very similar, but there's differences in the pcb for sure. The 1052 looks like it's using a Cosmo KAQY214S SSR ( http://www.cosmo-ic.com/object/products/KAQY214S.pdf) instead of the mechanical relay of the 1102. Could this limit the bandwidth?

other than the components ralated to the SSR everything else seems the same to me?

--- End quote ---

The board in my DS1052E is exactly like FLolic's-- down to the FT relays and silk-screened labels. The only difference is the minor 'patch' on the LCD driver IC on mine. I believe Dave has a slightly earlier rev board.  

I would suspect the 'shaved' IC's in the analog section-- there's has to be a reason why they insist on defacing the labels.

--- End quote ---

I just got my DS1052E in the mail (thanks to Brett for the tip!) -- is this where the 50MHz/100MHz investigation ended, or did anyone make further progress?
rossmoffett:

My scope is about like Dave's in build quality, looks very nice except for some similar mess that others have observed due to the rubbed off A/D converters.

Has anyone been probing around in there lately?  I have mine apart already and I'm looking at the Analog section.  While the 50 MHz and 100 MHz models have what seems to be an identical front-end, I have a theory.  Mind this comment from the teardown blog comments:


--- Quote ---Mastro Gippo October 13th, 2009 at 02:43

Mh.. The scratched 16 pin IC in the analog part should be an AD8370
http://www.analog.com/en/amplifiers-and-comparators/variable-gain-amplifiers/ad8370/products/product.html
but I don’t know why they are scratching it! There’s nothing to be ashamed of, it’s a 750MHz part…
--- End quote ---

Rigol makes some 300 MHz scopes too.  What if the analog front end on the 100 MHz scope is bandwidth limited also, and we've got filters on both models with different values?  What if our *real* bandwidth can be even higher than 100 MHz?

I've found something that I need to investigate with a better scope, will post details if it ends up being anything of significance.
rossmoffett:
I just threw caution to the wind and made the mod..  I'll do a write-up ASAP tonight, I've been up all night playing with it and I've got to go to a robot build meeting.  After you find out what I've done and why you can better decide whether it was a good idea or not.  I don't want to say what I did until I can thoroughly explain it.

Channel 1 (yellow) is unmodified, Channel two (blue) has the mod.  It seems to correct the curve that is shown in the other Rigol post (blog specific forum).  I don't know what happens around 100 MHz, honestly.  The function generator I'm using is a piece of trash and very unstable, 100MHz is where the road ends.  I was only able to get curves this good by using time averaging, you can tell that the VPP stability is crap, but since both channels are referencing it, the data should still be good for proportional reference.  Stability of the function generator around 80-90 MHz was too horrible to even post.

20MHz
30MHz
40MHz

V1 = .7(V2)
50MHz
60MHz
70MHz
100Mhz

EDIT:
My high frequency signals are fine, I shouldn't have taken measurements before closing the shielding back up again.  I just compared the scope with a Rigol DS1102C (100 MHz, 400 MS/s) and that scope has a hard time keeping up!  The modified channels signal is clean and consistently much stronger than the 50 MHz channel.
Navigation
Message Index
Next page
Previous page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod