EEVblog Electronics Community Forum

EEVblog => EEVblog Specific => Topic started by: EEVblog on August 24, 2012, 02:48:56 pm

Title: EEVblog #340 - USB 3.0 Eye Diagram Measurement
Post by: EEVblog on August 24, 2012, 02:48:56 pm
EEVblog #340 - USB 3.0 Eye Diagram Measurement (http://www.youtube.com/watch?v=o8DPlqWVmzk#ws)

Dave.
Title: Re: EEVblog #340 - USB 3.0 Eye Diagram Measurement
Post by: poptones on August 24, 2012, 03:18:20 pm
Tear it down! I want to see 20MB of 25ps memory...
Title: Re: EEVblog #340 - USB 3.0 Eye Diagram Measurement
Post by: amspire on August 24, 2012, 05:13:35 pm
Dave,

I had assumed that that scope was 40GS/s sampling scope, but it is a 40GS/s real time scope on all 4 channels?
 
Title: Re: EEVblog #340 - USB 3.0 Eye Diagram Measurement
Post by: AndyC_772 on August 24, 2012, 05:17:49 pm
Can't believe you chickened out of soldering the differential probe directly to the USB connector ;)

I was fortunate enough to get to play with a virtually identical scope earlier this week, though 'just' the 2.5 GHz version - and I agree, as a general purpose instrument it's nowhere near as useful as an 'ordinary' bench scope. Its inputs are fixed at 50 Ohm, so you can't even plug in a 10:1 passive probe to check if a clock is running. No variable display persistence either; it's infinite or nothing.

As a tool for a specific job I have to admire its capabilities, but I actually preferred using my 8 year old Tek TDS3034B. The first time a scope ever complained about needing to update its anti-virus software was a sad day indeed, IMHO.
Title: Re: EEVblog #340 - USB 3.0 Eye Diagram Measurement
Post by: BravoV on August 24, 2012, 05:42:37 pm
The first time a scope ever complained about needing to update its anti-virus software was a sad day indeed, IMHO.

Not a sad thing at all if you're an Iranian scientist or engineer working at the nuclear lab facility somewhere inside Iran ...  ;D  j/k
Title: Re: EEVblog #340 - USB 3.0 Eye Diagram Measurement
Post by: EEVblog on August 24, 2012, 05:44:37 pm
As a tool for a specific job I have to admire its capabilities, but I actually preferred using my 8 year old Tek TDS3034B.

For everyday use I'd rather use an old TDS210!

Dave.
Title: Re: EEVblog #340 - USB 3.0 Eye Diagram Measurement
Post by: EEVblog on August 24, 2012, 05:46:14 pm
BTW, the scope is now gone.
For those in Sydney who want to play with it, Agilent have a measurement seminar this Monday.
Or just ask'em, they have a couple of loaners!

Dave.
Title: Re: EEVblog #340 - USB 3.0 Eye Diagram Measurement
Post by: T4P on August 24, 2012, 09:45:45 pm
Looks like the 90000 was designed for WinXP but now it's Win7  ;D
Title: Re: EEVblog #340 - USB 3.0 Eye Diagram Measurement
Post by: AndyC_772 on August 24, 2012, 09:49:03 pm
The one I used was running XP. I'm sure Agilent's engineers really enjoyed porting the hardware drivers and getting the scope running properly under a new version of Windows... time very well spent pushing the boundaries of technology I'm sure you'd agree  :-\  :-[  ::)  :'(
Title: Re: EEVblog #340 - USB 3.0 Eye Diagram Measurement
Post by: EEVblog on August 24, 2012, 10:11:04 pm
Looks like the 90000 was designed for WinXP but now it's Win7  ;D

A sticker on the back says it runs XP, but is using a Win 7 license sticker.

Dave.
Title: Re: EEVblog #340 - USB 3.0 Eye Diagram Measurement
Post by: firewalker on August 24, 2012, 10:26:21 pm
A virus messing up your measurements would be fun!  :-X :-X :-X

Alexander.
Title: Re: EEVblog #340 - USB 3.0 Eye Diagram Measurement
Post by: ivan747 on August 24, 2012, 10:38:03 pm
A virus messing up your measurements would be fun!  :-X :-X :-X

Alexander.

Kinda like the old school viruses that just pissed you off and messed up with your DOS prompt  ::)
Title: Re: EEVblog #340 - USB 3.0 Eye Diagram Measurement
Post by: T4P on August 24, 2012, 10:58:41 pm
Dave,

I had assumed that that scope was 40GS/s sampling scope, but it is a 40GS/s real time scope on all 4 channels?

Checking the datasheet reveals that it does 40GS across all 4 channels
80GS on the X models ( half a million ... ) for 2 channels and 40GS all 4 channels for the X models

BEASTS. Well considering the agilent DSOX3054A already costs about 12k for 500MHz ( of course there's the 1GHz upgrade Dave is getting )
It's about fair it costs 140k ... well 10x more sample rate 10x more speed and about 250x more memory (if you get the 1GB upgrade which i guess costs a bomb)
or just 10x more memory (assuming both have no upgrades)
And then there's the PCI-E x4 upgrade for the memory ...
Title: Re: EEVblog #340 - USB 3.0 Eye Diagram Measurement
Post by: AndyC_772 on August 24, 2012, 11:53:38 pm
A virus messing up your measurements would be fun!  :-X :-X :-X
Isn't that pretty much what Stuxnet did?
Title: Re: EEVblog #340 - USB 3.0 Eye Diagram Measurement
Post by: firewalker on August 25, 2012, 12:04:53 am
Isn't that pretty much what Stuxnet did?

Yep. StuxNet was inserted to normal Desktop computer though. I don't believe you can run binary files from the usb port of the Agilent.

But with Windows you never know. You might be able to download a virus through the probe.  :P :P :P

Alexander.
Title: Re: EEVblog #340 - USB 3.0 Eye Diagram Measurement
Post by: vl400 on August 25, 2012, 12:42:04 am
We run anti virus software on all our agilent spec ans and VNAs at work that run windows, nothing like AVG running a scheduled scan in the middle of trying take a measurement  ::) They all run a full windows environment though, can minimize the measurement software back to a normal desktop. Perfectly capable of being infected or BSOD. Not sure if these high end scopes are the same like that?
Title: Re: EEVblog #340 - USB 3.0 Eye Diagram Measurement
Post by: AndyC_772 on August 25, 2012, 12:52:50 am
If it's based on a PC running Windows, I can't see any reason why it shouldn't be as vulnerable as any. Probably more so, since it's unlikely to be kept up to date with all the latest patches.
Title: Re: EEVblog #340 - USB 3.0 Eye Diagram Measurement
Post by: mikeselectricstuff on August 25, 2012, 01:21:27 am
If it's based on a PC running Windows, I can't see any reason why it shouldn't be as vulnerable as any. Probably more so, since it's unlikely to be kept up to date with all the latest patches.
..although there is much less reason to connect it to a network than a PC. And it probably won't be running a web browser.
And if it's been done properly it will be running a Windows Embedded flavour with all the system and application on a read-only hard disk partition. If.....
Title: Re: EEVblog #340 - USB 3.0 Eye Diagram Measurement
Post by: ee851 on August 25, 2012, 01:47:15 am
Very cool machine!   I hope I get a chance to play with one too.
Looks as if it has about 359 different ways to measure & display clock jittter, hahaha....

I wonder how much Agilent pays Apple Computer in royalties.    Surely Apple Computer patented the touchscreen technology that makes this machine so easy to operate ?

Does this machine connect to ethernet too ? to download software upgrades & to upload measured data to the www ?
Title: Re: EEVblog #340 - USB 3.0 Eye Diagram Measurement
Post by: AndyC_772 on August 25, 2012, 01:59:51 am
And if it's been done properly it will be running a Windows Embedded flavour with all the system and application on a read-only hard disk partition. If.....
Nope, just XP. The one we rented actually popped up with the 'found new hardware' wizard on power-up, which was slightly unnerving!
Title: Re: EEVblog #340 - USB 3.0 Eye Diagram Measurement
Post by: tnt on August 25, 2012, 02:14:18 am
I think the one that had 'overshoot' might actually be the TX of the HDD box, just using pre-emphasis to make sure the signal looks nice at the RX end.
Title: Re: EEVblog #340 - USB 3.0 Eye Diagram Measurement
Post by: SeanB on August 25, 2012, 02:36:54 am
Scary are bank ATM's some running WinXP, and a heck of a lot running Win2k, mostly the older ones. The oldest are running a flavour of Unix, but those are not common these days.  I have often seen a crashed bank ATM showing a BSOD, or worse a dragged off to the side message ( with the now visible mouse cursor) where it wants to restart.

The oldest ATM I saw had a massive burn on the screen, which said " This machine temporarily out of order/hierdie mashien tydelik buite werking" which somebody had placed a sheet next to the machine which said " temporarily working". There was a first generation Post Office ATM that would deliver notes one at a time, you remove it, and it whirred for 30 seconds before presenting the next note. Took it a minute to print the receipt as well, using a built in hammer printer, then it would eventually give the card back. 15 minutes per transaction on average.
Title: Re: EEVblog #340 - USB 3.0 Eye Diagram Measurement
Post by: StubbornGreek on August 25, 2012, 07:39:11 am
EDIT: Almost forgot to comment...the hell with the manual, I really enjoyed this upload. Thanks Dave.

We run anti virus software on all our agilent spec ans and VNAs at work that run windows, nothing like AVG running a scheduled scan in the middle of trying take a measurement  ::) They all run a full windows environment though, can minimize the measurement software back to a normal desktop. Perfectly capable of being infected or BSOD. Not sure if these high end scopes are the same like that?

Consider using Microsoft Security Essentials - I've tried most others (including nod32, AVG, Avast, Comodo - I could list them all but there are dozens) and its the least intrusive package out there, MS uses the same definition database as Norton, McAfee, etc, and not to mention that MS does have an intrinsic understanding of their operating system. I haven't looked back - I highly recommend.
Title: Re: EEVblog #340 - USB 3.0 Eye Diagram Measurement
Post by: Zad on August 25, 2012, 09:16:34 am
Why would Apple own the touch screen technology? Just like MP3 players and GUIs, touch screens were going decades before Apple decided to use them.

Title: Re: EEVblog #340 - USB 3.0 Eye Diagram Measurement
Post by: StubbornGreek on August 25, 2012, 09:30:01 am
Why would Apple own the touch screen technology? Just like MP3 players and GUIs, touch screens were going decades before Apple decided to use them.

They claim capacitive touch (multi-finger touch) is their 'patented' technology.
Title: Re: EEVblog #340 - USB 3.0 Eye Diagram Measurement
Post by: T4P on August 25, 2012, 12:29:13 pm
HTC had capacitive touch screens way before Apple.

Dickheads at Apple
Title: Re: EEVblog #340 - USB 3.0 Eye Diagram Measurement
Post by: Hypernova on August 25, 2012, 03:10:34 pm
Got two LeCroy's at work, one running W7 and another XP, the universe would probably implode should I hook up my own rigol and use the PC interface with the touch screen.

Imagine that, controlling a rigol from a LeCroy...

On the subject of ATM's I saw one rebooting at a bank and it ran OS/2! God knows how old they are. This is at Three Kings shopping center at Auckland.
Title: Re: EEVblog #340 - USB 3.0 Eye Diagram Measurement
Post by: SeanB on August 25, 2012, 07:42:19 pm
Those are OK security wise, OS/2 has very good built in paranoid security, and banks used it a lot. My local bank has ditched XP aside from in the Diebold ATM network and has gone to RHEL, so that they could keep all the old computers and get an upgrade without a hassle. All run a terminal client, and connect to a VM on the branch server, so they are a lot more secure, and have paranoid access control ( verisign football, password and swipe card needed to log on).
Title: Re: EEVblog #340 - USB 3.0 Eye Diagram Measurement
Post by: StubbornGreek on August 26, 2012, 12:36:36 am
HTC had capacitive touch screens way before Apple.

Dickheads at Apple

I believe the slogan has become: "First to patent..." Unfortunate.
Title: Re: EEVblog #340 - USB 3.0 Eye Diagram Measurement
Post by: calin on August 26, 2012, 01:42:10 pm
Gotta say .. cool tool :) ... 140 K ... Just for the sake of it here is another extreme toy - http://www.home.agilent.com/agilent/product.jspx?nid=-33202.1010838.00&lc=eng&cc=US (http://www.home.agilent.com/agilent/product.jspx?nid=-33202.1010838.00&lc=eng&cc=US) - the price made me choke. The darn thing is more expensive than my house !!!

BTW .. OS2 running on an ATM is nothing out of ordinary. I work for a very big credit card company (can't say which one but is one of the big ones) - if you believe they have old mainframes which are programmed in assembler. You heard and read right I said ASSEMBLER, not a mistake :) . And don't believe that because of that their system are fast or reliable, a cluster with few nodes will blast that mainframe into dust in terms of speed ... is just that they a so afraid to touch them that nobody is allowed to even clean the dust off the damn thing. Just an old monstrous piece of junk that gets afraid any CTO only when you show it to him :) . Decision .. do not touch it !!! build anything that you can around it and hide it from anyone's eyes. And if you need an CPU replaced ... about 4 millions $ a pop (yes not a mistake again).