CrackHeadMel

iTrader: (2)

I just got the Garrett gear head email this morning mentioning this product.

Anyone else hurd of it prior to this? Im on the fence about it, i think it would be great to use to fine tune a turbo setup, however for just myself, i feel i would only use it once every time i build a new system, which my wallet hopes isnt much

product link is here

http://www.turbobygarrett.com/turbob...d_sensor.html#

-Jacob

philiptompkins

iTrader: (4)

that look pretty neat, how much does it cost?

Evil Aviator

Given my aviation experience and education, I am amazed that automotive tuners (other than mega-buck race teams) do not use an N1 gauge.

Yes, you would probably just use it for designing new systems, although monitoring N1 after tuning isn't such a bad idea unless the probe restricts airflow.

I wonder what kind of machining is required for installation?

It looks like it will be priced somewhere around $400.

CrackHeadMel

iTrader: (2)

Interesting.

I read on a Honda forum (was searching for more info on this product) that this unit has a 0-5v output, and some where mentioning using this for shaft speed driven boost control, which to me does sound rather precise

renns

For aircraft applications, turbocharger shaft speed must be known as part of the certification testing. In years gone by I've seen Garrett/AlliedSignal send out modified aircraft turbos for this purpose. To measure speed a flat is machined on the shaft, and the rotating assembly is then rebalanced. The center casting is machined to accept a 2-wire inductive sensor that is mounted facing the rotating shaft, triggered by the flat on the shaft.

From looking at the pics on that link posted, this product appears to be using the same measurement technique. If so, complete turbo disassembly, machining, and re-balancing will be required.

philiptompkins

iTrader: (4)

oh damn, I thought it wold just be a senor looking at the nut on the compressor wheel that you would just drill into the inlet.

Turblown

iTrader: (12)

Its requires removal of the compressor housing and machining.

Evil Aviator

Yeah, I was wondering if the machining was maybe on the nut, or some other part that was easier to deal with than the shaft. I think they would sell more units if they used an optical sensor that reads the painted compressor blade.

philiptompkins

iTrader: (4)

yeah, i was actually pictureing one that would read the corners, or flat spots, on the nut.

renns

Machining the nut doesn't really save any effort, as that still results in a rotating assembly that needs re-balancing. Once that's done you've got a sensor stuck right across the compressor inlet to detect the flats on the nut. That'd be a major flow disruption in a critical location. I still think this product is based on the requirement of a machined shaft. A telephone call would probably end the speculation, but what fun is that?

An optical sensor would be a good alternative for the diy'er, but setup and alignment would be critical. That's tough to do with a commercial product though, as the inlet piping varies so much between installations.

NoDOHC

I have measured turine speed with an optical sensor on the hex nut attaching the compressor wheel. This does work, as the light is only reflected back to the sensor when one of the nut faces is aligned orthogonally to the laser beam. This gives you 6 pulses per revolution.

The adjustment is very difficult and the signal is not perfectly reliable, as any humidity in the intake air tends to fall out of solution in the low pressure region at the turbo inlet and block the light.

My guess is a VR sensor watching a flat spot on the shaft.

fastrotaries

The best aviation school PERIOD! GO EAGLES!!!

I agree. an N1 gauge could prove to be beneficial.

auricomXL

we use speed sensors on the Garrett Tutbos at work, but i think ours are the older model (they look a bit different). anyway, you can get turbo RPM using a cheep microphone and a laptop using free software. count the number of enducer blades, record the turbo in real time. run the sound through a real time fourier transform to get it in the frequency domain, filter out the stuff you know isn't the turbo (low frequencies) and then take the big spike you see, divide by the number of enducer blades you counted earlier, and then take the reciporical of that, there you have cycles per second, then convert to rpm or whatever unit you wish. you could proabbly write some code to take care of all that for you and just display a single rpm value.

Also worth mentioning: the Holset tubos we have at work do measure rpm with a VR pointed at a groove in the shaft, the borg warners point at the enducer and pick up the blades.