Eye diagram demonstration

[nicedwarf]

Hi,

I have a 2 analog channel 100 MHz scope and would like to generate an eye diagram for demonstration purposes.  This is a standard MDO style scope (it doesn't have any waveform db, or application software running on it to help).

What would input would you recommend I use to generate the signal?  I don't mind spending up to ~$500 for a demo board that I could configure.

Thanks!

[ataradov]

What would input would you recommend I use to generate the signal?

Any source of square signal? Including the embedded test signal in the scope. You can make it less square by purposefully mis-compensating the probe or adding small capacitor.

[Someone]

You could use the internal square wave with the trigger set to alternate or both edges and some storage (if available on your scope) or a faster signal and adjust the intensity, but the classic eye diagrams use a temperature map or other coloured display of the "hits":
www.tek.com/dl/55W-17445-0.pdf
Which is only available on some digital scopes. The characteristics people look for on eye diagrams only show up with a more complex waveform, you could replay some simple patterns from an arbitrary waveform generator or build a small digital circuit with the needed patterns.

[w2aew]

The simplest thing would be to use a LFSR (linear feedback shift register) based pseudo random binary sequence generator (PRBS). This is done with a simple shift register with XOR feedback from a couple of taps in the shift register back to the input. Trigger the scope on the clock that is running the shift register, and the output of the shift register will be displayed as an eye diagram.

This can be built in hardware using logic chips such as a serial input shift register, a few XOR gates, and a oscillator. It can also be done with a micro platform such as an Arduino, and a bit of code to generate the clock and data signals on a pair of output pins.

Google PRBS generator, or LFSR circuit, etc. to get the details of some of the possible circuit configurations.

[daqq]

A common place where such a signal is used is the USB bus. Though with a 100MHz BW you might not see all of it - USB Full Speed would probably be the limit.

[tggzzz]

What would input would you recommend I use to generate the signal?  I don't mind spending up to ~$500 for a demo board that I could configure.

Oooh! I'll design you one for that price,paid in advance

All you need is a something with a regular clock where the outputs don't change every clock cycle. Trigger the scope from the clock (use either a second channel or the external trig input), and look at the waveform on a channel.

Thus a binary counter is sufficient, although you may have to twiddle the holdoff period to avoid the scope's trigger locking onto the divider output. A non-binary counter, e.g. a divide by 10, is slightly easier, and a pseudo-random number generator (an LFSR) would be best.

[mrpackethead]

The audio out of a PC.. :-)

[nicedwarf]

Thanks everyone for your responses! 

ataradov for the win!  I used the square wave with a small capacitor, attached is a pic.  Thanks!

[David Hess]

I was going to say linear feedback shift register but that has been covered.

For a more interesting display, use a counter to vary the theshold voltage of the clock and create pattern dependent jitter.

[Someone]

I was going to say linear feedback shift register but that has been covered.

For a more interesting display, use a counter to vary the theshold voltage of the clock and create pattern dependent jitter.

Now thats an analog scope.....

[NiHaoMike]

The audio out of a PC.. :-)

S/PDIF actually is a good signal to make practice eye diagrams with. It's high frequency (a few MHz) but still low enough to view with a budget scope.

[David Hess]

I was going to say linear feedback shift register but that has been covered.

For a more interesting display, use a counter to vary the theshold voltage of the clock and create pattern dependent jitter.

Now thats an analog scope.....

It is very analog but still sampled.  The measurement shows jitter in a TTL logic based delay from a TTL counter chain modulating the supply voltage.  It was made using a Tektronix S-4 sampling head (*) and 7S11/7T11 plug-ins in a 7834 analog storage oscilloscope operating in variable persistence mode.  I processed the photograph to make it more suitable for printing which gave it a unique look.

The S-4 was overkill for bandwidth and an S-1 (1 GHz) or S-2 (4 GHz) would have worked just as well or better but at the time the S-4 was all I had.  My best DSO at 300 MHz could just barely see the jitter although I suppose my 7904 at 500ps/div should have shown it but probably not as well.

(*) In the immortal words of Jim Williams, "I'm sorry, but 14GHz is the fastest 'scope in my house."

[Someone]

I was going to say linear feedback shift register but that has been covered.

For a more interesting display, use a counter to vary the theshold voltage of the clock and create pattern dependent jitter.

Now thats an analog scope.....

It is very analog but still sampled.  The measurement shows jitter in a TTL logic based delay from a TTL counter chain modulating the supply voltage.  It was made using a Tektronix S-4 sampling head (*) and 7S11/7T11 plug-ins in a 7834 analog storage oscilloscope operating in variable persistence mode.  I processed the photograph to make it more suitable for printing which gave it a unique look.

The S-4 was overkill for bandwidth and an S-1 (1 GHz) or S-2 (4 GHz) would have worked just as well or better but at the time the S-4 was all I had.  My best DSO at 300 MHz could just barely see the jitter although I suppose my 7904 at 500ps/div should have shown it but probably not as well.

(*) In the immortal words of Jim Williams, "I'm sorry, but 14GHz is the fastest 'scope in my house."

Mostly joking about the op wanting help with their analog scope and then returning with a photo of a Tek DPO/MSO.

[ataradov]

Mostly joking about the op wanting help with their analog scope and then returning with a photo of a Tek DPO/MSO.

OP clearly mentioned MDO. Where do you get analog from this? Are there analog MDO scopes?

[David Hess]

Mostly joking about the op wanting help with their analog scope and then returning with a photo of a Tek DPO/MSO.

OP clearly mentioned MDO. Where do you get analog from this? Are there analog MDO scopes?

There sure are or were.  The old Tektronix 7000 series can produce a simultaneous mixed domain display and I have used mine for a real time and sampled display at the same time but so can the even older Tektronix 556:

Test equipment aficionados may wish to consider how this picture was taken. Hint: Double exposure techniques were not used. This photograph is a real time, simultaneous display of frequency and time domain information.

http://readingjimwilliams.blogspot.com/2011/09/app-note-35-part-1.html