Trottle Position Sensor Testing

[svx_commuter]

Here is a way to test the Throttle Position Sensor (TPS).
Get a piece of ¼" plywood approximately 2 feet square.
( I had one left over from the Honey Do medicine cabinet .) Using the TPS for a template, cut out a hole in the lower left corner. Mount the TPS from the bottom with a couple of screws to hold it in place. Insert a 7/32 allen wrench into the TPS and mount a long arm to the allen wrench. ( http://www.subaru-svx.net/photos/user.php?svx_commuter ) Using the arm as a compass, draw about a 22" radius on the board. Mark off 1/8" increments on the radius line, number the marks and then start a chart. This works well. For every 1/8" of movement at 22" radius, the resistance changed about 10 ohms. The TPS is only moving about .003" for each 1/8" movement at 22". (I can’t do that when I try to hold it in my hand and check it!)

The ECU in my car had given me the TPS trouble light code some time ago. I took it out and cleaned it and the trouble code has not come back. However, I have a 2-3 shift flare, so I thought I would try a new TPS.
Now having completed this test, the TPS was indeed bad.


During this testing on the "BOARD" I duplicating a small range of displacement, similar to the light throttle driving I was doing when the slip occurred. This resulted in some very erratic readings. The motion was duplicated by moving the long arm on the TPS back and forth between positions 32 and 40 several times for an example. Sometimes the readings were okay, the resistance dropped as it should but other times the resistance went way up instead of going down. It isn’t supposed to do that! This was very surprising. I really thought the old TPS was not going to be a problem. Check out the TPS_CHART in the locker.

To remove the TPS, remove the SVX plastic cover over the throttle body. Pull it straight up in the back (the side near the fire wall) about 2" inches (don’t pull it more than that!) and the plastic "fake screw" expansion pins will pop out. Hold the front of the cover down while pulling up because there are tabs under the center aluminum cover that can break. (One of mine was broken when I got the car.)

Use some "white out" to match mark the TPS to the throttle body right on the top of the TPS. (Everybody has white out and that stuff has lasted a long time on my TPS.)
The TPS will be back in the same place when the match marks are lined up. Remove two vacuum hoses and the large hose that goes over the TPS. These will get in your way otherwise.

Remove the TPS wire harness connector. This is a lot easier to get off than the O2 connector. Just push in the release strip and tug. It comes off easy.

Remove the screws holding the TPS and drop them into the area under the intake manifold "No mans land". Have a long magnetic ended pointer available to retrieve the screws or extra screws or just don’t drop them.

Connect the ohm meter leads with alligator clips to the TPS. This keeps them from falling off the TPS when it goes back in. These clips should have insulation jackets on them so they don’t short against each other. Connect to pins 2 and 3. There are 3 pins, 2 is in the middle and 3 is on the bottom when the TPS is installed. Put it back in place and take your time because the alligator leads have to go under the intake manifold and they can fall off.

Measure the resistance with the throttle closed. This is the reference resistance for the new TPS. For the TPS was mounted on my car, it read about 4630 ohms with the throttle closed.

Well this is the method I like. After burning my hands on the intake manifold while trying to attach leads to the pins and then trying to hold leads on the pins and then given up, this appears to be the easiest way.

The tranny slip from 2nd to third has not shown itself in 5 days.

[Aredubjay]

John, this is fan-freakin-tastic documentation. Thank you for taking the time to figure this out. This (as you may know) was the diagnosis for my 2-3 flare before the last Reading meet (which eddycat poopooed). Unfortunately, the tranny rebuilders took too long to diagnose it and burnt up my high clutch. I think you've proven, that, if caught in time, you can save yourself the cost of a new tranny with a 120.00 (give or take) part. Thanks again for your efforts in documenting this procedure. If you don't object, I'll cut and paste it into the "how-to" locker along with a link to your picts.

[svx_commuter]

Randy,
Thank you for the thank you.
I am more than happy to contribute. Heck I can't possible make a web site like this or keep it running but without it where we post stuff to help each other?
It would be my honor to be in the "how-to". So YES please copy
I paid $150 for the TPS. That was with 25% off from Liberty.
I hope the price has gone up since you checked last.

[Aredubjay]

Originally posted by svx_commuter
Randy,
Thank you for the thank you.
I am more than happy to contribute. Heck I can't possible make a web site like this or keep it running but without it where we post stuff to help each other?
It would be my honor to be in the "how-to". So YES please copy
I paid $150 for the TPS. That was with 25% off from Liberty.
I hope the price has gone up since you checked last.

I got mine for 130.00 from my CDA Subaru fiasco. I call it a fiasco because I needed it 'RIGHT AWAY' and paid 20.00 for OVERNIGHT shipping. I realize that, sometimes they have to order the part in and that takes a while, but, I expect OVERNIGHT shipping to at least mean I get it the day after they get it. Turned out my URGENT order took 11 days to get to me (OVERNIGHT shippint to them meant second day air), and, not only did they charge me sales tax on the order (something you're not supposed to do on mail order), they charged me sales tax on the 20.00 "OVERNIGHT" shipping charge -- that was my first and last order from them. I see that Subaruparts.com has it for 153.46, plus shipping, so, all in all, you got a good deal.

I'll try to cut and paste your instructions into an HTML document and put it into the "how-to" this afternoon. Thanks again for your work and your willingness to share.

[SVX10]

You are one intelligent man. I look forward to using your documentation to test my TPS. GREAT JOB!!!
Mat-1

[svx_commuter]

Hi Mat-1,

Thank you very much for the compliment. What you see in this post is what I seem to do best, find mechanical problems and fix them. Everybody has something their good at and also stuff their not so good at. I got alot of stuff I'am not so good at, so I guess it all balances out. Heck, just ask my wife and she will tell you.

[Aredubjay]

Originally posted by svx_commuter
Hi Mat-1,

Thank you very much for the compliment. What you see in this post is what I seem to do best, find mechanical problems and fix them. Everybody has something their good at and also stuff their not so good at. I got alot of stuff I'am not so good at, so I guess it all balances out. Heck, just ask my wife and she will tell you.

I may need to do some testing myself. The 2-3 flare that the TPS replacement was supposed to fix, seems to be rearing its ugly head again. Or, heading its ugly rear, one or the other. Damn! I can't seem to get rid of these tranny gremlins. Grrrrrr.

[svx_commuter]

Originally posted by Aredubjay

The 2-3 flare that the TPS replacement was supposed to fix, seems to be rearing its ugly head again.

That is not good news You put the new TPS in last spring. How many miles ago was that? I was hoping that this was a once in 100k mile replacement. I figure I helped my TPS end it's useful life my stomping on the accelerator so much recently when I was trying to get the rear wheels to kick-in. That is when the 2-3 flare started.

Mine hasn't come back after 7 days and I even tried to make it happen by driving at the constant speed of 27.67321 mph, on level ground and shifting between 2nd & 3rd.

[immortal_suby]

I cleaned the TPS and have been messing with its position a little now. I always thought that my wifes svx jumped off the line a little quicker than mine. I rotated the TPS back slightly (clockwise about 1/8th inch) and now my car jumps off the line like hers always did. It also changed a couple other things though - it unlocks the torque converter at a higher speed when coasting - it used to hold it down to at least 40 or so but now it will unlock around 45 with the foot off the gas. It also is a little more reluctant to go into power mode - something I had always noticed in her car as well. Mine used to go into power mode normally without really trying but now requires a very quick stomp of the throttle. It also is more prone to the 2-3 flare up at light throttle. I'm gonna do some more experimenting with the position but I wonder if there is a spec in the service manual about the resistance that it should be putting out at closed and wide open throttle. I noticed that mine was around 5000 and stayed near there for a small amount of turning with a small up and down variation until it then steadily came down to around 0 turned fully (not a digital ohmeter) I didn't try the wood mounting stuff yet, that was just cranking it in my hand with a screwdriver.

So is there any recommendations in the service manual on how to test it and what it should be at?

Also - is there a way to open it up and clean it more thoroughly instead of just spraying electrical cleaner in there?

[svx_commuter]

From the 92 service manual:
Section 2-7, K: trouble Code (31) - Throttle Sensor, Page 77
1) turn ignition switch to "OFF".
2) Disconnect throttle sensor connector.
3) Measure resistance between throttle sensor connector terminals.
Terminal/Specified resistance
No.1-No.3/ 5 Kohms
{My old one is 5,120 ohms}
4) Measure resistance betwen terminals while slowly opening throttle valve from the "close" position.
Terminal/Specified resistance
No. 2 - No. 3/10 - 12 Kohm [Fully close]
/3 - 5 Kohms [Fully open]
Ensure resistance increases in response to throttle valve opening.

Well this is what the service manual says. I think this came up in another thread about the resistance value. 10 to 12 ohms and not Kohms would make more sense for fully closed.

I haven't found any place in the service manual that talks about how to set it and I think they all have that white match mark on them. Hmmmmmm?

Please note that terminals 1 & 2 will read from low to high as it is opened and terminals 2 & 3 will read from high to low as it is opened. Depending on the pair of terminals used they will read opposite each other.

The digital ohm meter is only way to test this thing. Your looking for very small changes in resistance as it moves.

[immortal_suby]

I am going to have to get a digital ohmeter and test mine as well as my wife's.
I did some more experimenting and found that if I turn it slightly counterclockwise from its original position power mode engages very easily, it holds the torque converter locked right down to 40mph, but it also raises idle rpms up to around 1000 and does not start easily - it will start but it will idle erratically for a few seconds unless I start it with my foot lightly on the throttle in which case it starts fine.
It has not revved up on the 2-3 shift yet since moving it counterclockwise. Hmmmm....

Thanks for the specs from the service manual.

[1994SubaruSVX]

and time availability. i will just buy a new one and have it installed. nice how-to for those that are mechanically inclined. as for me.....

[svx_commuter]

Originally posted by 1994SubaruSVX
and time availability.

Time is a man made concept.

[1994SubaruSVX]

deserved that. ok, i dont want to make the time.....you got me.

[Beav]

It's a good test for the sensor but be sure to go as slow as possible when swinging the arm. Chances are that you won't be able to see errant spikes or drops on your meter. Those midrange flaws are what condemn most sensors that we see in the shop. Normally if we suspect the sensor we'll isolate it and use a Fluke Scopemeter (aboout $2000) to watch for such errors. But the all-time favorite, no mistakes about it solution is to exchange a known good sensor. Even the boys at the dealer have to resort to it occasionally. It's just difficult for independants and owners to exercise that option....

Beav

p.s. a lot of times we find ground flaws to be the biggest culprits in electronics. Make sure to check all grounds and clean them if suspect to keep from wasting time barking up the wrong tree...