MIMO Cable signal integrity/pin-out

[cochenob]

I'm looking to use a MIMO cable to lock two USRPs which will be packaged in separate underwater housings.  A sketch is below.  I'm anticipating having to snip up the MIMO cable so that I can interface it in/out of the housing bulkhead connectors.

Not knowing much about the exact nature of the signals that are flying over the MIMO cable to begin with...should I be concerned about signal integrity?  The MIMO cable is about 1m long.  Will lengths greater than this cause any issues from a timing perspective?  What kind of tolerance do the timing/synchronization signals have to amplitude/noise?  What kind of cabling/wire gage actually makes up the MIMO cable?

The only information I've ever seen regarding in MIMO cable signals is that instead of a PPS signal...the MIMO cable shares "time encoded signals" between USRPs.  That doesn't tell me much.  What are these signals?  What kind of volgate levels/data rates do they require?  Does the mere mention of snipping up the MIMO cable to begin with raise any immediate red flags?

Appreciate any thoughts or additional information you may have.

---

Brandon

[ErikL]

The MIMO cable is for convenience only. Supplying 2 Ethernet drops and a shared PPS and 10 MHz ref will accomplish the same result. Also length of the MIMO cable is accounted for in software for better time synchronization.

[cochenob]

Thanks Erik-

This is sort of what we thought. 

By the way...the documentation on both the Ettus and NI sites just say that "timing messages" are shared over the MIMO cable instead of the PPS.  What exactly are these "timing messages"?

[cochenob]

Actually...one additional question.  In the external 10 MHz ref/PPS method....do those two need to be phase locked?  It seems like the PPS is just used to latch a common time to each radio when you initially fire them up.  It's not obvious that it would have to be locked to the 10 MHz ref as well.

---

Brandon