Oscilloscope for CD eye pattern

[abhi1981]

I'm planning to buy an oscilloscope. Will the Siglent SDS1052DL(50Mhz, 500MSa/s) oscilloscope be able to display a CD eye pattern, like an analog oscilloscope?
Or do I need 1GSa/s?
Please advise.

[tooki]

500MSa/sec is more than enough for that.

[wasedadoc]

An audio CD carries 2 channels of 16 bits at 44,100 samples per second.  That is a net data rate of less than 1.5 Mbyte/s.  There is extensive error correction which means the gross rate read from the CD is significantly higher but nowhere near the rate that would trouble even a 50 MHz bandwidth scope to show the eye pattern.

[tooki]

It’s eight-to-fourteen modulation, so 44100s^-1x 2ch x 16bits x (14/8) = 2.4696Mbps.

But yeah, there’s a screenshot elsewhere on the forums of a CD eye diagram on a digital scope, and it was running at just 100MSa/sec.

[paschulke2]

The raw data rate of a compact disc is approx. 4.3Mbit/s (audio data + data for error correction + subcode + ...). With the help of the eight to forteen modulation this datarate falls in a frequency band of 20kHz to 1.5MHz on the disc (Heemskerk and Shouhamer Immink (1982) Compact Disc: system aspects and modulation, Philips tech. Rev. 40, 157-164). Therefore neither sample rate nor bandwidth of any half-decent scope will be a problem.

But: In case of a digital scope you need a scope with a high number of waveforms per second and good intensity grading . Typically an eye pattern is done at 500ns/div (5µs for the complete screen). For a good eye pattern you should have more than 5000 waveforms per second. A bad digital scope - even with high bandwidth and sample rate - will be worse than almost any old 20MHz analog scope.

[abhi1981]

Thanks everybody for the valuable explanations.
Actually, this video confuses me the most -
A 100MHz 1GSa/s DSO is not able to display the eye pattern. 
Maybe because it's a cheap DSO.

I had bought an OWON VDS1022(25MHz 100MSa/s) but the eye pattern was not smooth -
Luckily I was able to return it.

In this video - ,
the RIGOL DS1054Z (50MHz 1GSa/s) is displaying an eye diagram perfectly. But it is quite expensive.
The Siglent SDS1052DL is within my budget. But don't know whether it is good or not. Because once I buy it, I will not be able to return or exchange it.

[paschulke2]

[...] the RIGOL DS1054Z (50MHz 1GSa/s) is displaying an eye diagram perfectly. But it is quite expensive.
The Siglent SDS1052DL is within my budget. But don't know whether it is good or not.

The Rigol DS1054Z has intensity grading and 12Mpts memory. The Siglent SDS1052DL has 30kpts memory and no intensity grading (no guarantee for this - I only had a short look at the data sheets). I'd therefore say that the RS DS1054Z is good for the job and the Siglent SDS1052DL isn't. Again: It's not the samples per second or the bandwidth that matters, but the intensity grading, the amount of memory (which is needed for the intensity grading) and the number of waveforms per second.

[wolfy007]

The Owon VDS-1022 is a fairly cheap PC oscilloscope with 100Ms/s, and looking at that video, a low refresh rate (low WFPS rate) probably due to the fact it needs to communicate to a PC to display the waveform. As paschulke2 said earlier, Im pretty sure the waveform update rate will impact eye diagram quality not just sample rate.

The Rigol DS1054Z apparently has a max update rate of 30000WFPS and the SDS1000DL family isnt particularly fast, but I suspect significantly faster waveform update rate than the Owon PC oscilloscope but I couldnt find any figures for it on the net. I think I still have an SDS1102a at home, might power it up and see if I can gauge its update rate.

[tooki]

I can confirm that the Rigol DS1054Z is suitable for CD player eye diagrams: I have used it for exactly this purpose.