This is a real problem! The great scopes keep getting better, but also further away from our wallets. An alternate approach is to download captures from less-capable scopes, then do the analysis on a PC. If the scope supports LXI (as nearly all recent o'scopes do), then you can use any of several LXI remote front-end tools to fully automate the process. That is to say, you don't need your o'scope to run all the fancy software internally: That's what PCs are best at doing. Much of the profit for high-end scopes is burdened on the software "features". Get the minimum o'scope hardware that has the bandwidth and trigger capabilities you need with enough trace depth, then do the rest on a PC. You'll save a ton of money and gain a ton of flexibility. The downside is that you'll need a PC in order to make full use of your scope. But given that a mid-range laptop will have a much better screen and a more powerful processor than even high-end o'scopes, this isn't much of a negative. And it is a huge positive when doing complex signal analysis. Some may recommend USB o'scopes, but I still prefer to play with the knobs for doing quick measurements. And the hardware capabilities are nearly identical for the money. Another concern with super-smart o'scopes is that many of them run common operating systems, such as Windows. And they stop getting updates surprisingly soon after you buy them, meaning they could become security risks on your network. Having minimal software in the scope makes sense, so why pay extra for more? Pay for the o'scope hardware you need, and let a laptop do the rest.
As someone learning a lot about data transfer recently I think I need a scope or logic analizer but I'm not sure if it's worth half a year or more of saving money.
There are some great lower cost options out there for both scopes and logic analyzers, but if you need to do jitter testing or RTE type measurements you'll have to consider higher cost options.
So true. A jitter analysis instrument is like a calibration standard, in that it needs to be waaaay more accurate than your DUT. And, that level of precision and accuracy ain't cheap. :-)
I have the same model scope and use it for eye analysis of serial data. The part where I lack confidence is in the probing of high-speed signals. The probe is the N2752A 6Ghz , but I don't know how much the probe is disturbing the signal. I only have one of these probes at the moment so I cannot double up the probe loading to visualize.
Hey, i am wondering if I can measure the eye diagram of clock signal itself? Or it doesn’t make sense at all since the clock signal is the base for triggering.
Yes! That's a fairly common measurement because clock jitter tends to propagate throughout the system. What you can do is measure the jitter from the trigger point. So, the trigger point is consistent but the edges around it will allow you to measure jitter.
Generally you want to use gear that has that ability built in, but you could also set one up using persistence and/or mask testing. Just move your trigger to the left of the screen and crank up the persistence
You can have a set-up like this: CH1→ Bitstream Signal CH2 → Clock Signal (External Waveform Generator) 1.) Find waveform generator that is able to: (a) output a sufficiently high frequency for the application's bit rate (b) phase shift its waveform. (Look on eBay if not on hand.) 2.) Plug bitstream signal into CH1 of your oscilloscope. (Ideally, using differential probes. Differential probes have better CMRR than using two oscilloscope probes, so I'd use them.) 3.) Plug waveform generator into CH2 of your oscilloscope. (Waveform generator is your clock source.) 4.) On the oscilloscope, trigger on the rising edge or falling edge of clock signal. 5.) Turn on your bitstream. 6.) Adjust the phase of the clock signal so that you can align it with your bitstream signal appropriately. 7.) Turn on persistence.
They buy a Rigol MSO5000 series , bottom of the line 4ch, for far far less, then you hack it up to the highest model and you'll have your eye diagram for in the order of $2,000
I knew what an eye diagram looked like, and I've seen them on scopes, but now I know what they mean and what they are used for, thank you!
also used on ABC TV in Australia - th-cam.com/video/sL0FhqGtV4U/w-d-xo.html
Old problem: understanding eye diagrams
*watches video*
*looks up price for the scope*
New problem: not enough money on bank account
This is a real problem! The great scopes keep getting better, but also further away from our wallets.
An alternate approach is to download captures from less-capable scopes, then do the analysis on a PC. If the scope supports LXI (as nearly all recent o'scopes do), then you can use any of several LXI remote front-end tools to fully automate the process.
That is to say, you don't need your o'scope to run all the fancy software internally: That's what PCs are best at doing. Much of the profit for high-end scopes is burdened on the software "features". Get the minimum o'scope hardware that has the bandwidth and trigger capabilities you need with enough trace depth, then do the rest on a PC.
You'll save a ton of money and gain a ton of flexibility. The downside is that you'll need a PC in order to make full use of your scope. But given that a mid-range laptop will have a much better screen and a more powerful processor than even high-end o'scopes, this isn't much of a negative. And it is a huge positive when doing complex signal analysis.
Some may recommend USB o'scopes, but I still prefer to play with the knobs for doing quick measurements. And the hardware capabilities are nearly identical for the money.
Another concern with super-smart o'scopes is that many of them run common operating systems, such as Windows. And they stop getting updates surprisingly soon after you buy them, meaning they could become security risks on your network. Having minimal software in the scope makes sense, so why pay extra for more? Pay for the o'scope hardware you need, and let a laptop do the rest.
Awesome tutorial!
How do I calculate the bitrate of my Signal if I don't have your measurement function? Is it just two times the frequency of my Signal?
The inventor of the Eye Patten has passed away 4/25/22 RIP Dennis
Nice video..
How can we do the same with a DPO 4054B scope. Thanks
Nice! Patiently waiting for part 2.
th-cam.com/video/vMR1r0nDK7s/w-d-xo.html
As someone learning a lot about data transfer recently I think I need a scope or logic analizer but I'm not sure if it's worth half a year or more of saving money.
There are some great lower cost options out there for both scopes and logic analyzers, but if you need to do jitter testing or RTE type measurements you'll have to consider higher cost options.
Those digitizer "scopes" do both for a fraction of the cost of mixed mode 'scopes. There's about a dozen OEMs building these boxes.
So true. A jitter analysis instrument is like a calibration standard, in that it needs to be waaaay more accurate than your DUT. And, that level of precision and accuracy ain't cheap. :-)
I have the same model scope and use it for eye analysis of serial data. The part where I lack confidence is in the probing of high-speed signals. The probe is the N2752A 6Ghz , but I don't know how much the probe is disturbing the signal. I only have one of these probes at the moment so I cannot double up the probe loading to visualize.
I can understand the basics of eye diagram.Thanks
What is used to analyze the frequency and diagram in the oscilloscopy?
Thank you !
Дякую !
Hey, i am wondering if I can measure the eye diagram of clock signal itself? Or it doesn’t make sense at all since the clock signal is the base for triggering.
Yes! That's a fairly common measurement because clock jitter tends to propagate throughout the system. What you can do is measure the jitter from the trigger point. So, the trigger point is consistent but the edges around it will allow you to measure jitter.
Keysight Labs so, can I measure an eye diagram of clock signal itself?
Nice explanation!
Why didnt u show the physical connection??
How do people that don't have a $22,000 scope do an eye diagram?
Generally you want to use gear that has that ability built in, but you could also set one up using persistence and/or mask testing. Just move your trigger to the left of the screen and crank up the persistence
You can have a set-up like this:
CH1→ Bitstream Signal
CH2 → Clock Signal (External Waveform Generator)
1.) Find waveform generator that is able to: (a) output a sufficiently high frequency for the application's bit rate (b) phase shift its waveform. (Look on eBay if not on hand.)
2.) Plug bitstream signal into CH1 of your oscilloscope. (Ideally, using differential probes. Differential probes have better CMRR than using two oscilloscope probes, so I'd use them.)
3.) Plug waveform generator into CH2 of your oscilloscope. (Waveform generator is your clock source.)
4.) On the oscilloscope, trigger on the rising edge or falling edge of clock signal.
5.) Turn on your bitstream.
6.) Adjust the phase of the clock signal so that you can align it with your bitstream signal appropriately.
7.) Turn on persistence.
They buy a Rigol MSO5000 series , bottom of the line 4ch, for far far less, then you hack it up to the highest model and you'll have your eye diagram for in the order of $2,000
Nice explanation. You can remove the name university because most universities are at-least 2 to 10 years behind current technology in engineering.
Great video but there's usually a lot more swearing and cursing involved.
Usually there is more frustration and no Erin East.
This is just sad...