Since the conditions changed so much under the cowl, no carb could have provided a constant mixture. So I decided to test an idea that I had for a loooong time, but never had reason to test.
Instead of the in-fuse balso box, I wanted to use the static pressure tube approach, like in a pitot tube, but then only half of it, disregarding the dynamic pressure part for speed readings.
So I made a simple brass tube, bent it to shape, pinched one end shut, and made two opposite saw cuts in the closed end.
The other end was connected to the carb refernce pressure cover.
The pictures say all.
The staic pressure tube runs straight forward into the airflow, and about 12mm (1/2") out from the cowl