On 20 December 2011, Art Pearse in Ontario, Canada wrote:
"What is the difference in measured vacuum between the manifold tap and carb tap, at various throttle openings, or engine speed. Why not use the manifold tap"?

This was a tough discussion in the BBS, as it had a complex answer. A manifold port gives manifold vacuum, high vacuum at idle and lower vacuum with greater throttle opening. The carburetor port on the MGA is at bottom just in front of the throttle plate, therefore exposed to atmospheric pressure (no vacuum) when at idle. When the throttle plate starts to open it quickly passes the small drilled port, and the vacuum signal jumps immediately high, almost as high as manifold vacuum. This gives you a nice pop to spark advance with initial poke at the throttle.

With slightly more throttle, still relatively small throttle opening, the radially positioned port acts as a venturi port, generating vacuum due to the velocity of air passing the port. So the vacuum signal remains fairly high for moderate throttle opening and high velocity of intake air. This gives good spark advance for good fuel economy at cruising speed.

When you press the loud pedal to the floor for max power, vacuum at the throttle plate drops quite low, almost down to atmospheric pressure. Air velocity near the wall location of the vac port is immediately reduced some as the total mass air flow is distributed over the full area of the intake throat, and the (ported) vac signal drops dramatically. At low engine speed and wide open throttle air velocity in the throat is not so high, but intake air velocity increases with engine speed. With increasing engine speed the vac signal can rise slightly due to the venturi effect. Maximum air velocity is ultimately limited by engine displacement, engine speed, and restrictions to air flow in the intake and exhaust paths.

Even though the vac signal is rising a bit with increased engine speed, it is still relatively low (at full throttle). There is a threshold vac signal level required to start movement of the vac advance diaphragm on the distributor. Any time the throttle is wide open this ported vac signal is too low to move the diaphragm, so the vacuum advance goes away entirely with wide open throttle.

End effect of all this is, no spark advance it idle, immediate rather dramatic increase of spark advance with small throttle opening, and the vac signal is maintained to higher engine speed as long as the throttle is only moderately opened. Combination of vacuum advance and mechanical advance at cruising speed and partial throttle (2000-4000 rpm or 35-70 mph for instance) gives high total spark advance and good fuel economy.

The kicker comes with the kick, so to speak. When you floor it for max power the vac signal drops, and vac advance goes away. For full throttle operation the engine runs on mechanical spark advance only (and is why race engines often do not use vacuum advance). You can set maximum spark timing to about 32d BTDC at high engine speed when mechanical advance is maxed out, but you have to disconnect the vacuum line to do this.

At cruising speed with moderate throttle the vacuum advance piles on top of the mechanical advance, and total advance can be up around 45d BTDC. Punch it for max power and it's instantly back to 32 at speed (so you don't have pre-ignition or detonation with max power). Advantages of ported vacuum advance are quick throttle response and good fuel economy.

This is why you need to disconnect the vacuum line to check and set spark timing dynamically at any speed higher than dead slow idle. When you rev the engine up with no load and fairly small throttle setting, the vacuum signal is still there in a big way and does not represent the low vacuum signal you would have under load with wide open throttle. So do disconnect the vacuum line when setting timing with a strobe light.

For the "other" setup using manifold vacuum, the vacuum signal is high at idle, is reduced with increasing throttle, and goes very high on overrun (closed throttle at high engine speed). The vacuum advance function on the distributor has to be dramatically different when using manifold vacuum. The vac signal can be applied to opposite side of the vacuum module to create vacuum retard at idle.

When you hit the throttle for power, manifold vacuum drops, the vacuum retard goes away, and you get spark advance for quick acceleration. If you close the throttle at high engine speed you get very high manifold vacuum and maximum vacuum retard on overrun. This setup was used to cope with emissions issues in later model MGB.