Cheap Scope ?

Oops,...just forgot to say required BW need to be at least 4 GHz

We're look for "something" to see how good the TMDS, with a low budget pocket and we hit this $10K PicoScope 9200 USB

Has anyone ever heard of or used this? I'm open for other choices with budget... let say below ~$20K. The BW need to be at least 4 GHz

Renting is almost out of choice since we will need it for infinite...

A used, working Tek 11801 with a 20 GHz TDR/sampling head is going for around $2K. But it's big.

Thank you John,

Sorry for the delay, I've been busy around on other stuffs...

This 11801 looks like the most feasible solution

Is that true we can only look at one diff pair on one head? If so we may need two heads, one for the clock (to trigger) and one for the data...and for the clock I think we can use the slower speed

A dual-channel head can sample two signals or one differential signal.

The SD24 TDR head can generate diff and single-ended steps, too. The level is -0.25 volts into a 50 ohm load, per channel.

SD24's are relatively expensive, because of the TDR capability. The SD22 is a low-noise dual-channel head, 12 GHz, no TDR, lots around and cheap.

The calibrator output of the 11801 is also a damned fine picosecond step generator.

You can't trigger an 11801 through a sampling head. It has a separate, single-ended trigger input on the mainframe.

We recently got a LeCroy 4-channel 7 GHz scope. It can acquire long records and display stuff like eye diagrams with no trigger at all. It synthesizes clock-recovery PLLs and stuff from the data in waveform memory, all in software. But that cost $50K.

Wow, those LeCroy are fancy for sure, and big buck too :-)

For the 18101, sounds like we need to recover the clock to single end first before feeding it to the trigger on the mainframe. In the old day I remember to treat one signal of the dif pair as single ended, and trigger on it with out big trouble :-)

and yes, we also need it to evaluate the eye diagram