Rbe over temperature?

[Grant Richter]

I am wondering if transistor Rbe is temperature dependent?

There seems to be less Rbe compensation required at lower temperatures. The
oscillators I calibrated during the summer seem to sharp at C7 now, although
the C2-C3 range still looks OK. Calibrated at 80 degF tested at 65 degF. I
am only seeing a small apparent slope error.

The transistor gain goes down with increasing temperature, so the tempco
increases the gain to the drive stage as temperature goes up. As I recall
that silicon has a theoretical -3300 ppm/C slope change. The tempco has a
+3500 ppm/C indicating a possible overcompensation of +200 ppm/C for
INCREASING temperature. So the high end should go sharp as temperature
increase.

The fact that they go sharp with DECREASING temperature I can only explain
if the Rbe is decreasing with temperature?

Any thoughts?