What's the reason why you rarely if ever see a centrifugal blower BETWEEN the throttle plate and the manifold?

On positive displacement roots type blowers, the only way is to have the throttle plate before the blower, as they by definition pump a fixed amount of air per cycle, if it this were not limited by limiting the upstream pressure, it would no doubt blow the intake pipes.

But, can centrifugal blowers operate under a vacuum?

Reason I ask that if a centrifugal blower intake is open to atmospheric pressure, it will always pump an amount of air proportional to RPM, regardless of load, and under part throttle, most of this is released through a bleed off valve. This does however cause a huge inefficiency as there is no need to compress and release all that air under part throttle. If the intake side had a throttle plate, then only the air used would be compressed, which would give you a big gain in efficiency and mileage.

I'd guess an upstream throttle could result in decreased throttle response. A solution to that might be dual throttle plates, one at the standard manifold location, and another upstream one. This should be pretty easy to accomplish with a DBW upstream throttle.

The real question I have is is there any reason a centrifugal blower would have issues operating under vacuum with an upstream throttle?

 

I don't see a big efficiency increase here by doing this, most if not all of the inefficiency comes from parasitic losses from a belt driven accessory aka the blower. The blowers step up ratio relative to rpm is very long compared to a roots also, which means it cant force/overcome itself as an obstruction in your proposed config like a roots which builds boost right off idle. Driving under vac is really only done in steady state cruising, which can be sizable part if you like to take long steady state roads trips for hours on end sure, but as soon as you apply throttle it quickly exceeds vac. Putting the blower in-between the manifold and TB would seem to cause more inefficiency by reduced throttle response and obstruction before sizable boost is made. Your talking about overcomplicating a set up here with no clear upside. That's my initial thoughts on this, right or wrong.