This is an often hidden portion of the SP-D trimer, as it can be hidden under the CRD domain which can flop around and lie in different positions during processing.  It seems that this peak will be something around 5 nm (top figure) depending upon how the CRD domains have obscured it.  It might not have any subpeaks (the mode for this peak is 1 and it seems reasonable for this relatively tightly wound helical neck portion of the trimers (bottom figure).

It will be interesting what adding four more measurements to this dataset will confirm or refute. It is nice that the small N of four (composed of 50 to 100 separate grayscale plots of trimers (from a tracing of the hexamer) after image and signal processing, show some consensus.