and here is the Nuvistor, flanked by a "miniature" vacuum tube on the left and a transistor from the same era on the right. RCA's Nuvistor was just a micro-miniaturized vacuum tube they felt could beat the transistor. (it did not.)

RCA published a really nice cutaway drawing with annotations.

my current theory is that the echo bus was included in the IBM 729 Model I because it could be used with older computers that were designed for use with the 727, and expected the echo bus to be there.

yeah looks like the 727 has a dual-purpose single read/write head. the 729 has two sets of heads, one for read, one for write. the 727 needs the echo bus because the read bus doesn't work during a write operation.

a clue in the earlier IBM 727 manual! there are pins listed as "echo bits" which are sent back to the computer from the tape unit when it is recording data being written. i think the 729 doesn't include those bits because it uses the read bus to echo the write bits. but for whatever the reason, the shield pin connections were left in place.

the pin list doesn't have those pins either, but it DOES document some "spare pins" -- they are shown as *extra data bits*. the 729 is a 7-track machine, so you have data bits 1, 2, 4, 8, A, B, and C. but this chart has additional bits X and Y!

ok, one IBM document shows the pins, but in a weird way. the schematics have the pins shown as blank boxes, but in the "tape signal terminator" section, they are listed as type Z pins (response line decoupling.)