So, driving a small signal relay ought to be easy...
An NFet as a low side driver and a diode across the coil and one is good to go!
Unless, of course, the small, 3 V, SMD PCB mountable, small signal relay's coil is polarized!
As far as I can recall, none of the relays I've every used before had a polarized coil UNLESS they had an internal diode, which this one doesn't.
Unfortunately, I missed the small (+) and (-) signs on the relay's pinout diagram, and Murphy's Law being what it is, I laid out the wiring to the coil "backwards".
I couldn't figure out why the relay wasn't switching!
I swapped out the NFet to no avail.
I tested another relay and it switched fine on 3 V.
I went and studied the data sheet and then I noticed the polarity markings, and sure enough, when powered correctly, the little relay switches just fine.
I assembled five PCB's for this project, and I had to destructively remove the five relays and do a little surgery on the PCB's this afternoon, to swap the coil drive signals for the new relays.
The small tan rectangular relay shows the coil pins insulated from the PCB with electrical tape, and jumper wires crossed over and connected to the coil's diode.
And yes, I broke down and used a SFE breakout board for the LIS3DH accelerometer as I didn't wish to attempt soldering it.
They now work fine.
The Relay is a NEC EE2-3NUX, (non-latching, just a simple DPDT relay).
I'm a little bit smarter today, as I learned about polarized relay coils, even when there is no internal diode.
Some lessons are more painfully learned than others.
Edit: Relay Pinout is for the EA and EB series, my is the EE series, but the concept is the same.