specifics on compression testing, P0301

[lanxer57]

I did a compression test, the engine was cold, I tested for 10 "cranking noises" the results:
1;90, 2;110-145, 3;121, 4;130
added some oil:
1;100-145-123, 2;110-145, 3;110, 4;145-120
because I was not sure if this was correct (given the inconsistant results, and how far off they were from 180), I cranked it until it stopped on its own 10 times (usually 4-5 "noises"). this was with oil in the cylinders the results:
1;175, 2;189, 3;170-180, 4;180

How EXACTLY would I get the most accurate readings possible on a compression test? A thread posted in 2003 (that is 1 month older than I am) by canada-eh said it was best to "crank the engine over 10 times", what EXACTLY does that feel/sound like? The sounds that cranking an engine makes are divided into pulses, how many of these "pulses" counts as cranking the engine over 1 time?
I would like to do it correctly next time.

This is on a 1999 SC1 with the SOCH engine, nearly 170000 miles.
I can not drive it until it passes emissions, and it will not pass with a "service engine" light. I have gotten a code P0301 for a missfire in cylinder 1, as well as the "service" light with the wrench. The light usually comes on after a few seconds (although sometimes it does not come on until the engine is shut off and restarted), but my OBD2 scanner usually takes a very long time to detect the error code. I have tried changing the spark plugs/wires, ignition coil packs with the plate they attach to (could not figure out how to separate them), fuel injectors, ECTS (with coolant), idle air control, EGR valve, (and cleaned as much out of the holes as I could), and MAP sensor. I also sprayed carburator cleaner around the intake manifold, it did not run any different. Dad said it smells like it is running rich (my nose does not work) Any other ideas? (do not be afraid to tell me it needs an engine, I have a 2001 SOCH with the transmission sitting on a pallet in the garage)

[billr]

Did you hold the throttle wide-open while doing these crankings?

Run it for a bit, then take compression again; and be sure the battery is good enough that voltage doesn't drop below 11V while cranking. This time, don't do any of that "adding oil" routine. That is only confusing things. Yes, adding oil can help determine if the rings/bores are worn, but that is rarely the case with these engines because of their oil burning issues. And, if you add too much oil, compression will get falsely raised, possibly even to the point of locking the engine rom turning!

PS: Five compression strokes is enough to get full compression on a good cylinder...

[alordofchaos]

See post 2 in these threads

http://www.saturnfans.com/forums/sho...d.php?t=200832

http://www.saturnfans.com/forums/sho...d.php?t=236763

edit: I read your post again, missed this question

what EXACTLY does that feel/sound like? The sounds that cranking an engine makes are divided into pulses, how many of these "pulses" counts as cranking the engine over 1 time?

Hard to describe a sound with words, lol. Needle will jump on the compression gauge on a complete cycle. Another thing you could try, take a video with sound for what you think are 10 cycles. Post a link to it, and we can listen / count and tell you what we are hearing.

[2NDSOUT]

Your compression numbers, are WAY low...

Regardless if you do a wet or dry compression test... your numbers across all 4 cylinders should read at least 185 or better.

The standard PSI rating across all 3 cylinders for the majority of the S-Series run; was 185 on each cylinder.

As you may know- these engines were known for oil burning... and the only way to really do *Some* kind of fix for oil burning, is to drill small holes in the pistons to allow for some kind of oil draining.

I think you may be chasing your tail on this one; as much as I hate to say that... nothing that you are going to do outside of the engine, is going to bring those compression numbers to a decent working number.

[lanxer57]

Yes, I had the pedal all the way to the floor.
I do not know if the results for my compression test were accurate, because they are not very consistant, and I may have been doing it wrong.
I will record a video later today, and maybe edit numbers counting the "pulses" of noise that I am discribing.
If the engine is bad, I have another one that was not misfiring after 181,000 miles (at least, before I crashed the car I pulled it out of).

[lanxer57]

If 10 compression cycles is minimum, how would I count the number of compression cycles?

[billr]

Normal cranking (battery, starter, and wiring all OK) results in about two strokes per second. Crank for five seconds.

Have you checked the compression gauge in another engine?

[lanxer57]

"2 strokes per second" does that mean that the "pulsing" noises are called strokes?
5 seconds of that would add up to 10 strokes, but my starter seems to pull away on its own after 4-7 strokes, so I would need to turn the key more than once. Will that make a difference? I have not seen the pressure gauge visibly go down without manually releasing it, so I do not see why it would.

[billr]

Yes, the "pulsing" change in sound is from increased resistance to engine rotation when a piston is compressing air; a compression stroke. With all plugs out it is pretty distinct and easy to hear. I assume you do have all plugs out while testing compression.

The starter should be able to crank the engine continuously until either the battery runs down or the starter burns up; probably well past one minute. If your starting system can't do that, then any compression testing is somewhat suspect, since cranking speed matters. I would try to resolve the cranking issue first.

The point to trying the gauge in another engine is to check its calibration. Even if it doesn't leak down, it may be reading quite wrong. Or you may get a hint that your procedure is not quite right if another engine gives similar poor results.

You first reported a spread from 90-145 psi. If that were really 175-230 it would not be ideal, but adequate for the engine to run OK. I really don't think this is a "gauge calibration" problem, but it would be sad to condemn the engine and then later find it was the gauge that was bad.

Point is, it is rare to find these engines with poor compression on all cylinders (bad rings), so I am suspicious of your testing. Usually there will be bas compression on only one or two cylinders. A bad valve will kill one, a bad head gasket will kill two.

Are you sure the throttle is going wide open, and that there is no obstruction in the intake ducting/filter while cranking? Low compression on all cylinders is inevitable if the engine can't breathe freely. What altitude are you located at?

[lanxer57]

Here is the video of how I was counting "pulsing" noises. https://youtu.be/h_00p-YWgDs
Are these actually strokes?
If not, how many noises are in a stroke?

[billr]

It did not crank log enough for me to identify the sound of the compression strokes, but each of those gauge "ticks" was a stroke.

Not having continuous cranking is a problem, most of us have little experience with testing compression that way. I don't know if the usual compression limit (180) is even valid testing like that.

Also, I'm guessing you are at about 5000' elevation. That will lower compression, even on a good engine, but I don't know how much (guessing about 170 for the lower limit). Maybe somebody living at a similar altitude can advise what they normally get.

[Signmaster]

Not enough cranking time. Based on the video you're only getting a few full turns of the engine through the compression stroke.

With the car not able to start, it should crank as long as you hold the key to crank. Once you get that figured out, you can watch the gauge "ticks" up and estimate 10 or so full revolutions. I only see three or four bumps on your gauge as you crank, with the pause in between.

Do you have all the plugs out or just the one?

[lanxer57]

Elevation:5770 feet, or about 0.85 of sea level according to some motorcycle forum. Does that apply to any engine? Even fuel injected ones? If so, than the number for a SOCH at this altitude would be closer to 148-175 PSI.
The compression tester was in one hole, the others plugged with spark plugs.
I will try to get it to crank more next time. The battery DIED as I was finishing up, so that may be a clue...

[fetchitfido]

Start with a fully charged battery for each test, otherwise the results are invalid due to slow engine speed. Also need the throttle wide open, either a weight on the pedal or locking pliers/clamp at the TB.

Crank until the gauge stops increasing in value. Run your own start button to the starter if you can't see the gauge from the driver seat. Looks like you stopped cranking 3-5 times in the 8s video...it'll likely take 8-10s per cylinder to get a valid number.

If you add oil, drop 1 teaspoon in the cylinder, crank 1-2 cycles with nothing blocking the cylinders then retest. You just want oil on the cylinder walls to aid in ring sealing, you don't want the added oil on top of the cylinder artificially increasing compression numbers.

According to here 6000ft gives 80% of the normal readings...drops 160psi sea level to 128psi. 180psi as the service manual state becomes 144psi. I don't know how accurate the 80% at 6k feet is, but it's inline with other sources at higher and lower altitudes.

I've not seen a S-Series engine (SOHC or DOHC) happy about running at less than 150psi on all 4 and have not seen a running engine at all under 130psi (AllData says 100psi is the min...lmao). Also shouldn't have more than 10psi difference between cylinders, if all 4 are low the engine is just worn out...if 3 are high and 1 is low then that cylinder has a major issue (both usually end up with a replaced/rebuilt engine).

[fdryer]

Originally Posted by lanxer57

Elevation:5770 feet, or about 0.85 of sea level according to some motorcycle forum. Does that apply to any engine? Even fuel injected ones? If so, than the number for a SOCH at this altitude would be closer to 148-175 PSI.
The compression tester was in one hole, the others plugged with spark plugs.
I will try to get it to crank more next time. The battery DIED as I was finishing up, so that may be a clue...

Save your battery from grief - remove all spark plugs when performing compression tests. This removes useless compression straining the battery while allowing the starter to draw a little less power as the starter runs faster. The engine turns over quicker for better compression results. And remove either the fuel pump fuse or pump relay to stop injectors from adding fuel to each cylinder during compression tests.

The altitude may skew compression values downward since the higher one goes, less dense air occurs.

[lanxer57]

After having a better idea what I was doing, (at least I thought) I tested again.
I used CYLINDER;COMPRESSION/NUMBER OF STROKES format. The results were:
1;160/6,2;145/6,3;155/6,4;147/6
Added 1tsp oil to cylinders with syringe:
1;150/5,2;160/6,3;164/5,4;154/5

Cylinder 1 is the one with the misfire, can oil in there DECREASE compression?
These results look minimal at best, is this what is causing it to misfire?

[billr]

Are you still doing "interrupted" cranking; or have you fixed that and now doing continuous stroke for compression? No, I don't expect oil in the cylinder to ever lower compression.

Back to your actual misfire issue:

What do the plugs and wires look like? What gap are you using?

Have you checked fuel pressure with a gauge?

Have you tried swapping the ignition coils, tried swapping #1 injector with one from another cylinder?

Are you using live-data to check MAP, TP, and ECT sensors?

Post a picture of how the plug wires are routed.

What is the electrical system voltage when the engine is running at a fast idle (~2000 rpm)?

[lanxer57]

After removing all other spark plugs, and putting the battery on a trickle charger, it seems to be able to crank continuously most times. I released the pressure whenever it stopped before the needle did.
The plug in cylinder 1 is black like coal, from excessive fuel if my memory is working, the rest are a lighter color, likely from oil that I squirted in there during testing. (there was lots of smoke from burning it off afterwards) The gap should be no different than when it left the NGK factory, I need to find the feeler gauge.
I have not heard of a gauge for fuel pressure, but I will ask for one.
I have not figured yet figured out how to separate the coils from the plate on the back, but have tried swapping them both with the back plate, and the towers all seem to have power going through them, I also swapped the entire fuel injector rail with one from a DOCH in a junkyard, and it acted exactly the same. I could try swapping them individually.
The live data from the sensors mentioned seems to be working with the cheap OBD2 scanner.
It is dark outside now, so pictures would not work until maybe tomorrow.
I have not seen it idle faster than 1200, it usually drops to about 800 in a few minutes, I also need to borrow a multimeter.

[TomM96]

Good that you did a compression test (at altitude).
This demonstrated that:
1.) compression is adequate to run (which it does)
2.) There is no deficit in the misfiring cylinder

Moving on:
You observed carbon on spark plug of misfiring cylinder. That plug
is likely defective Now; it may have been bad previously, causing
the misfire. THEREFORE
> Replace black plug with a new clean one of similar design. Do not
use the thin-wire type.
Check if engine runs, and if code set.

Re assess situation.
It is still possible that there is an injector issue, but a fouled plug will
foul/mask any information from testing.

To remove coils from IgCtrlModule, pull out/up OFF the plate/circuit board,
maybe one end before the other (lift one end up, it rocks up, releasing other end).
Check spark across (grounded) plug or coil paks.

[TomM96]

Best compression test is with WARM engine.