OB99W

HCTs do indeed have some accuracy issues; Hall-effect devices even drift with temperature changes. However, for the application being discussed (assuming the basic goal is to determine that a charging current is flowing), it seems that accuracy is secondary to spotting a trend. Although it should be possible to obtain accuracy within a few percent, it would appear that even 10% would easily be sufficient for the purpose. Some further info: http://www.multimic.com/clump_e/index.html Click on the links at the left of the page under "CT SENSOR FOR DC LINE CURRENT" for typical specs.

You're welcome! Please do let us know what the final outcome is at the shop.

You seem to be discussing the "true" current transformer (CT). They have turns ratios as you described, dividing the current by a certain factor. These devices only work with AC, and are indeed prone to outputting high voltage if the secondary is open-circuited. However, DC current measurements are not done with a true CT, but rather a "pseudo-CT" that encorporates Hall-effect sensors (known as an HCT). These don't have the same transforming effect, and there isn't a high voltage developed if the "secondary" connections are left open.

You mentioned 141k miles, but didn't say when the t-belt had been last changed (along with tensioner, pullies?). It's certainly possible that the timing has slipped due to problems in that area. When cam (and crank) sensors start to go bad, they can throw codes intermittently, so the code not being consistent doesn't mean the part is okay. It's possible the ECU is "limping", depending only on crank sensor data for timing. For some insight into how the cam sensor and related parts function, see: http://endwrench.com/images/pdfs/Fuel.pdf http://endwrench.com/images/pdfs/DirectIgnition.pdf (includes warning to disconnect injectors, not ignition wires, when doing cylinder-balance testing; you may want to do that to determine which cylinder(s) the problem lies with.) By the way, there are errors in the "DirectIgnition.pdf" info. In the "Ignition Coil Testing" section, the terminal numbers listed are reversed between the primary and secondary; however, the diagram is labeled correctly. 2.2 info including timing: http://endwrench.com/images/pdfs/2.2SingleOverWin01.pdf http://endwrench.com/images/pdfs/2.2Liter.pdf http://endwrench.com/images/pdfs/TBeltEWWin05.pdf

Easy engineering fix: Have an alarm condition mute the sound system. That will get noticed.

It certainly could be piston slap, since the noise abates as the engine warms. However, you may find that engine oil initial viscosity, and the change in it as the oil "ages" (including any dilution with fuel) have an effect. Has the oil (and filter) been changed recently? Do you always use the same brand/type of oil and filter?

I don't see "anyone" having posted that the elevation at Bushkill, PA isn't what either you ("340 feet") or Nipper ("390 feet") said. The point is, whether at Newark, Bushkill, or the highest Pocono peak, the engine vacuum readings shouldn't be as low as 16 inches Hg if all is right.

Thanks for the reference, Nipper. No, this particular "anyone" didn't use "Google Earth". By the way, the Poconos certainly have peaks that are about 1,000 feet higher than 700, if that's of any consequence in the discussion of diagnosing a certain Subaru with possible exhaust restriction. :-p

Congratulations!

I didn't find a ready reference for Bushkill, so the effect on vacuum I mentioned was worst case, based on an entry in the Columbia Encyclopedia: "Pocono Mountains, range of the Appalachian system, c.2,000 ft (610 m) high, NE Pa." The point is, even at that, the vacuum yarikoptic measured was too low, as I wrote before. We seem to have been diverted from the diagnosis by a valid (but in this case not very impactful) point made by frag; I suggest that the vacuum testing be run again, using the criteria previously mentioned. The reading should be close to 20 inches Hg, and remain in that vicinity when the engine speed is raised to and held at 2000-2500 RPM for 15 seconds or so.

Good find. The 2001 Legacy manual is actually an H6 supplement, but much of it can be applied to the 4. If you want to know what's in each manual, use the Russsian-to-English translation of web pages available via AltaVista Babel Fish (http://babelfish.altavista.com/). Here's how it comes out for the 2001 Legacy: http://babelfish.altavista.com/babelfish/trurl_pagecontent?lp=ru_en&url=http%3A%2F%2Fwww.subaru-car.ru/left/doc/art/leg1/ It's not a perfect translation, but a lot better than my ability to read Russian.

With the "winks", I'll assume you're not being serious. If you are, then subtract about 2 inches Hg from what's "normal" at sea level; you should still be seeing 18+ inches, not 16.

In very rough terms: The tip projects beyond the portion of the pole piece that's fully round by about 1/8"; it's the same width as the diameter of the rest of the pole piece, and about 1/16" in thickness from flat to flat. As Nipper succinctly put it, "Yes".

A valid point, but yarikoptic is in Newark, NJ, a port city whose elevation is slightly above sea level.

Yes, that link is in the post that I referred to as well. Umm, there are a few variables. One is how many cylinders the engine has; 4-bangers "pulse" more than 6/8/12 cylinders at the same RPM. Another is where the vacuum is taken off, and whether the hose (or the gauge itself) has any restrictions; a large-bore vacuum connection to the gauge will reflect the piston strokes more readily. Some gauges have built-in restrictions to slightly dampen needle response, so you just don't see the rapid pulses even though they may exist. Yes, typically near 21 inches of mercury; minimum should be 20, although I've occasionally seen as high as 22.

I'm not sure where you connected the vacuum gauge or what exact procedure you used, but 16 inches of vacuum is low, and the zero reading might indicate a problem (depending on how that was arrived at). A more-usual reading is 20 inches or above, with some needle vibration at idle which diminishes as RPM is increased. See my previous post: http://www.ultimatesubaru.org/forum/showpost.php?p=626858

Yes, that certainly makes more sense than there being a burned valve. Thanks for getting back with the resolution to the problem, and good luck with the new car.

There is a 5EAT, but it's not original to a '97.

You're welcome; I hope that lead gets things resolved. As Cougar said, the control is solid state, and I've never seen an internal schematic for it. However, I can tell you that both sides of all bulbs that are dimable connect only to the control; neither side of the controlled bulbs are schematically shown connected directly to battery positive. This is going to be conjecture; I suspect that the fuse for the Illumi Con Unit isn't there to prevent failure of it from overload, but to prevent the harness from overheating if the control shorts in a severe enough manner. Apparently the control was able to initially sustain the connection you made, then there was a semiconductor failure that drew enough current to blow the fuse. See http://en.wikipedia.org/wiki/Scientific_method .

Nipper, ask Jamie to look up part number 83023 FA000; that's the "Illumi Con Unit" (illumination control unit, sometimes referred to as a "rheostat" in a throw-back to older times ). Simple to get to, if the lower dash panel can be pulled; may or may not be your (only) problem. Anyone needing to control illumination of aftermarket gauges might want to take a look at these: http://forums.nasioc.com/forums/showthread.php?&threadid=338562 http://www.scoobymods.com/forums/showthread.php?s=&threadid=3115

I don't seem to have the correct wiring diagram either, but I can provide some typical info that may be helpful. 1) In A/C mode, the radiator fans are controlled via relays by the ECU; if the ECU detects certain A/C problems, it won't turn the fans on. 2) The refrigerant pressure switch isn't the only thing that can cause the compressor relay coil circuit to not be completed. Have you checked the evaporator thermo switch, or wiring to it? It usually only comes into play when the evaporator temp gets out of range, but it can fail and cause the A/C relay coil circuit to be open even if the evap temp is normal.

You might want to electrically disconnect the fuel injectors as well. Gas can wash oil off the cylinder walls, making compression testing inaccurate.

Just in case anyone thinks my post above (#6) was some kind of joke, let me try again. Fuse number 15 in the box above the driver's side kick panel is a prime suspect. That fuse has a 10 amp rating, and is first in the lowest row. It's marked on the cover as "METER - SRS LAMP", but that might be misleading; it handles power for several things, two of them being the alternator field and the dash warning lights. I'd suggest checking that fuse before spending hours troubleshooting this problem. If it's not that, you won't heve wasted much time. Naturally, if it's blown, the next thing to determine is what caused that.

I'd start by checking at the under-dash fusebox, specifically fuse number 15. While if that fuse is blown it will cause more than what you mentioned to be out, you may not have noticed the other things as of yet.

Exactly. In fact, if charging was 14 volts at idle, it would probably be excessive at higher speeds, vastly shortening the battery life. Before cars had alternators (okay, I'm showing my age now ), DC generators were used and nobody expected any significant charging at idle, even with a very low load. When alternators were introduced, the claim was that they would keep the battery fully charged even under idle conditions. That turned out to not be completely true; with heavy loads, idling sometimes isn't sufficient. In fact, for vehicles that typically idle a lot (taxis, police, etc., especially in cities with heavy traffic), they have "heavy duty" charging systems that output more at low engine RPM to handle the problem. However, most of us don't sit for long periods with the engine at 600-700 RPM (500 is indeed too low) and all the accessories on, but probably run at least an average of 1500 much of the time, and often 2500 (yeah, I know some of you can claim yet higher numbers ); that's easily enough to keep the battery well-charged if the alternator is doing the job.