> If you are having your turbo overhauled anyways,
> consider having flats milled on the shaft at it's
> mid-section. Machine the center casting to accept an
> inductive prox sensor that senses the flats as the
> pass the shaft. Now you've got a rugged, reliable
> method of measuring shaft speed. This is the method
> Allied Signal used when we requested instrumented
> turbos for test purposes.
>
> Roger Enns.
There are many ways to skin the "turbo tach" cat.
I've read about opposing black/white paint on the shaft end, painted nut
faces, shaft magnets being picked up with inductors, etc. They all
have
their drawbacks.
OPTICAL
-------
Optical solutions usually have dirt/dust/grit problems, and involve an
illuminator and detector (LED and photosensor) in the intake path. This
interrupts airflow, can be ingested, or can move in fast airflows. The
best solution I've seen used drilled holes in the turbocharger housing
surface to mount the optics/components rigidly, to one side, out of the
intake airflow.
MAGNETIC
--------
Magnetic solutions usually involve magnetizing the shaft or
embedding/attaching a magnet. Any shaft modifications require precise
balancing, since this thing will spin FAST. Then we have to get a
sensor in there to detect the magnet. Close, but out of the airflow,
otherwise we end up with movement and ingesting problems again.
Somewhere between the compressor and turbine sections seems reasonable.
Roger's idea of a flat on the shaft should work, but again, that shaft
would have to be precisely balanced.
*****
I'm a fan of the optical method.
1) Drill a hole in the shaft end, as if to put in safety wire. The hole
needs to be big enough to let light through easily.
2) Drill opposing holes in the compressor intake housing, 180 degrees
apart, big enough to see through. More importantly, big enough for a
photosensor to see a LED/laser diode through.
3) MAKE SURE the shaft hole lines up with the housing holes.
As the shaft spins, you will get TWO pulses through for each rotation.
>From there, it's a simple counting and timing game. Simple circuits.
I like it because there is nothing in the airflow, and the components
are securely mounted on the exterior of the intake housing.
Now, go ahead and shoot this one up guys. :)
1) Can an LED/photosensor work 3-5 inches apart?
2) Are there any convenient housing hole locations that will line up
with the end-of-the-shaft hole?
3) Will dirt/dust/grit be a problem?
I'll do the circuits if someone drills up turbo charger for testing. :)
Paul - If you would, please 3-D this for everyone. Your pictures are
worth a thousand words.
--
Vince Orton
OK Thanks Vince :)
--
Paul Lamar ...No rotor no motor.
We're talking about a device for R&D characterization, not so much one
which would be required in a permanent sense, right?
Here's a tach that goes to 100K RPM..
http://www.valuetesters.com/Extech-461700-Mini-Photo-Tachometer-PocketTachT.php
I have seen similar devices with a fiberoptic extension on the nose to
allow remote sensing of RPM. Paint half of the compressor blades
black, and the other half white. Put a bung on the compressor side, and
mount the end of the extension in the bung and call it good. You might
have to develop a pressure capable "window." I lense of tempered
glass. If the window gets dirty in 20 or 30 hours of operation, pull it
apart and clean it.
If someone decides this is a useful instrument for a fully debugged
turbo system (to warn of some impending failure), then design something
special.. I suspect it won't be needed however, as once the system is
characterized and optimized, turbo RPM deviations are probably a symptom
of something else going wrong - something measurable through pressure
ratios or similar.
Matt-
The Rotary Engine NewsLetter. Powered by Linux.
ACRE NL web site.
http://www.rotaryeng.net
Copyright 1998-2006 All world wide rights reserved.