[T3] MPS calibration

[Jim Adney]

On 2 Jan 2011 at 13:03, Bryon Garvin wrote:

> Well my friend showed up today to take the inductance readings from my PS
> and then attempted to adjust his so that they match.  His '71 Fastback MT/FI
> has not been running all that well.  It died on a long trip and a pressure
> sensor from a late 60's model was installed at a shop and got him running
> again....but not great.  He limped home and has been trying to figure it all
> out ever since.  Because the "C" sensor would no longer run his car, and the
> one the shop put in did allow him to drive, he started by looking at the
> sensor.

The first thing to understand is that there's no certainty that the 
PS is/was the problem. My first thought would be to suspect the VR 
and check the system voltage with the engine running at medium RPM. 
It's quite possible that putting a different/wrong PS in there just 
partly compensated for the richness caused by low system voltage.

Of course it could also be any number of other things. It's way too 
easy to jump to conclusions about where the problem is.

> He came today and we pulled MY working sensor (we both have 71 Fastbacks
> MT/FI).  First thing we did was use an inductance meter to measure the
> values on both inner and outer pins at zero vacuum.  We then used a hand
> vacuum pump to take readings again at 4 inHg and 15 inHg.  We were following
> the instructions on this site:
> 
> http://members.rennlist.com/pbanders/manifold_pressure_sensor.htm#Adjustments

Personally, I don't think the inductances are the key. The PS acts as 
a transformer, so when I do my calibrations I work with them to try 
to make what's called their MUTUAL inductance equal to a known good 
PS. I suspect the inductances are somewhat important, but I doubt if 
they are the main thing.

> We then took HIS sensor and removed the screw.  With me holding the
> inductance meter probes on the pins, he turned the inner and outer screws
> and made the values match what my PS readings were.  When we were done, his
> values were exactly like mine.  It took some time as each screw affected the
> other and we had to go back and forth.  This brings me to something Jim said
> one time and I wanted to ask you, Jim, what you meant:
> 
> ".... the 3 adjustments will have to be done in some particular order to
> work.
> The problem is that the adjustments affect each other, so I need a
> proceedure that gets it right in a finite amount of time."
> 
> Now that I have tried to do this, why does it all have to happen in a finite
> amount of time?  I don't understand that.  Can you elaborate?

Well, because an infinite amount of time would take too long.   ;-) 

You need a process that converges on the right settings quickly. Did 
you find that it took long to get all 3 adjustments where you wanted 
them?

> We put the sensor in his car, and well....it didn't run the car.  It barely
> started and it sure didn't idle.  So he went home with the newly "tuned"
> sensor in his hand wondering why it all didn't work.  I'm no help!  I was
> just watching and helping take readings.  So, he's kind of back to square
> one.

I would have suggested that you then try his PS with your car. That 
wouldn't hurt your car, but it would have proven whether your 
adjustments got you what you were after.

> But I did want to post our procedure and results for anyone that was
> interested.  And Jim I'd love to hear your thoughts on what we did and why
> it didn't seem to work.
> 
> One other note of interest....I have 3 "C" sensors that are on my shelf.
>  Since Larry was here with his measuring equipment I figured I'd put them
> thru the tests and see what the values were.  Months ago, they all seemed to
> hold vacuum when I sucked on them with a new piece of hose.  One of them
> clearly was worse-off than I thought.  With a hand pump, we applied 15 inHg
> and it slowly dropped in a matter of 8-10 seconds back to zero.  But the
> other two held strong.

Slow leaks past the O-ring and the gasket that seal the diaphram are 
common and don't seem to cause a problem. Cracked diaphrams make for 
really obvious leaks.

> Out of the 2 that DID hold vacuum, one had inductance values nearly
> identical to the values of the working sensor in my car.  The 2nd one had
> values that were nearly HALF of what the unit in the car read.  So, what
> does this mean?  I have no idea :)

If you've run all of them and they all work with your engine, then 
that tells you that the inductance method isn't the whole story. I 
think the Rennlist article talks about long spring vs. short spring 
versions. Different versions may also have different windings in a 
way that doesn't effect the resistance or the operation but does 
effect the inductance. Maybe that is part of the explanation.

OTOH, I agree that a factor of 2 seems like a lot.

> Anyway, our findings are in this spreadsheet if anyone wants to take a look:
> 
> https://spreadsheets.google.com/ccc?key=0AmRobzkiJkYhdGx1aVF3TXpHV2xoaEVFS21oZlZpSnc&hl=en&authkey=CJucx48O

-- 
*******************************
Jim Adney, jadney at vwtype3.org
Madison, Wisconsin, USA
*******************************