Possible stupid question on a situation I've never encountered, could use help!
01-13-05, 02:05 PM
#26
Some notes about compression:
In general compression ratio doesnt seem to affect static compression numbers measured in a rotary at cranking speeds...I build all sorts of engines and test them all, and with roughly equivalent parts being used in the build, they all give about the same numbers.
Mazda spec for new is around 125, and dying is 85 and below. You start to see flooding issues at 95 and below, the lower the compression, the more flooding you'll encounter. I consider above 100 to be pretty healthy. It is possible to use rough housings in a rebuild with new seals, and make relatively poor compression (say, 90-95psi broken in) due to the housing wear, yet still have a LONG time before the seals break, since they are new and close to original height. There is a big difference between an ORIGINAL engine making 90psi on old, worn, lowered hight seals that could break at any time, and a REBUILT engine making 90psi on new, taller seals that are stronger and won't break.
Compression will be greatedr when cold than when warm. Rotary engines lose compression when warm, which is also why they tend to flood more when warm then cold. It's normal to lose 5-10psi when a rotary warms up. You check when warm to get the "worst case scenario" numbers.
You can check compression using T or L plugholes. Using L plugholes shows a few psi more compression since it's later in the "stroke". The mazda testers are designed for use with the T holes, I've had better results using piston testers in L plugholes, personally.
Also note that the presence of any oil/atf/mmo/whatever in the chambers will throw off compression readings a lot (adding 5-40psi). This is why atf will start a low compression engine...it adds compression instantly until it burns off. So, if you have an engine that the oil rings are weak on, and oil runs down and puddles up in the chambers, and then you run the compression check, you might see really awesome numbers, but it's because of the oil that's there. ON a related note, running too much or not enough premix or OMP injection can affect compression numbers a little, by the same token.
Cranking speed has some effect on numbers, too. Cranking at 200rpm will give numbers 10% lower than cranking at 250rpm.
Forgetting to open the throttle during a test will throw the results off, since there is less available air to the engine. Since a compression test is by definition a measure of the amount of air that an engine moves, restricting the air intake by letting throttle plates close will throw those results off. A cranking engine doesn't require a lot of air, though, so just having the plates cracked open is enough to satisfy this requirement.
Any questions, class?
In general compression ratio doesnt seem to affect static compression numbers measured in a rotary at cranking speeds...I build all sorts of engines and test them all, and with roughly equivalent parts being used in the build, they all give about the same numbers.
Mazda spec for new is around 125, and dying is 85 and below. You start to see flooding issues at 95 and below, the lower the compression, the more flooding you'll encounter. I consider above 100 to be pretty healthy. It is possible to use rough housings in a rebuild with new seals, and make relatively poor compression (say, 90-95psi broken in) due to the housing wear, yet still have a LONG time before the seals break, since they are new and close to original height. There is a big difference between an ORIGINAL engine making 90psi on old, worn, lowered hight seals that could break at any time, and a REBUILT engine making 90psi on new, taller seals that are stronger and won't break.
Compression will be greatedr when cold than when warm. Rotary engines lose compression when warm, which is also why they tend to flood more when warm then cold. It's normal to lose 5-10psi when a rotary warms up. You check when warm to get the "worst case scenario" numbers.
You can check compression using T or L plugholes. Using L plugholes shows a few psi more compression since it's later in the "stroke". The mazda testers are designed for use with the T holes, I've had better results using piston testers in L plugholes, personally.
Also note that the presence of any oil/atf/mmo/whatever in the chambers will throw off compression readings a lot (adding 5-40psi). This is why atf will start a low compression engine...it adds compression instantly until it burns off. So, if you have an engine that the oil rings are weak on, and oil runs down and puddles up in the chambers, and then you run the compression check, you might see really awesome numbers, but it's because of the oil that's there. ON a related note, running too much or not enough premix or OMP injection can affect compression numbers a little, by the same token.
Cranking speed has some effect on numbers, too. Cranking at 200rpm will give numbers 10% lower than cranking at 250rpm.
Forgetting to open the throttle during a test will throw the results off, since there is less available air to the engine. Since a compression test is by definition a measure of the amount of air that an engine moves, restricting the air intake by letting throttle plates close will throw those results off. A cranking engine doesn't require a lot of air, though, so just having the plates cracked open is enough to satisfy this requirement.
Any questions, class?
01-13-05, 04:53 PM
#28
Lives on the Forum
Originally Posted by RotaryResurrection
You can check compression using T or L plugholes. Using L plugholes shows a few psi more compression since it's later in the "stroke". The mazda testers are designed for use with the T holes, I've had better results using piston testers in L plugholes, personally.
Mazda wrote this up in an SAE paper a few years back.
Taking compression readings on the L holes will result in 5 - 10 psi lower readings.
Now, most "getto" compression testers might not be sensitive enough to show this kinda deviation, but they certainly do show up on a Mazda digital compression tester - down about 0.5 bar or kg/cm^2.
Mazda did all kinds of testing on trailing hole sizing, and there's a reason why it's so tiny - to keep compression loss minimize when the apex seal sweeps past the trailing hole first.
I try to adhere to the Mazda standard procedure when doing compression tests, and my compression testers (I got a Snap-On one too) don't show that much difference between L and T either.
But, if my eyes are working really well, I can detect a slight drop - maybe a needle width or two - in compression.
Man, I dunno what kinda engines you're working on, by the trailing spark plugs are a LOT easier to get to for myself!
-Ted
01-13-05, 05:16 PM
#29
I just did a hot compression test on an engine I just built, running on the engine test stand. First I pulled the front trailing plug. Inserted the mazda digital checker, and got 6.5, 6.3, 6.2 (150k used housings, couple of hours of idle breakin only).
Then I removed that and inserted a piston tester, with the valve installed and closed (for the highest reading in the chamber, equivalent to the above 6.5 face). That got me just a HAIR below 90psi on the guage, which I don't believe to be all that accurate since it doesnt sit directly at zero when idle anyway.
Then I repeated both tests for the lower hole. Got 6.3, 6.2, 6.1 for that, and the piston tester showed exactly the same thing, just a hair below 90.
I coulda SWORN that I've done this before and got higher numbers in leading holes. Maybe I got it backwards, it must have been a while back that I did this with a piston tester. Anyway, looks like you're right about that, so I stand corrected.
Then I removed that and inserted a piston tester, with the valve installed and closed (for the highest reading in the chamber, equivalent to the above 6.5 face). That got me just a HAIR below 90psi on the guage, which I don't believe to be all that accurate since it doesnt sit directly at zero when idle anyway.
Then I repeated both tests for the lower hole. Got 6.3, 6.2, 6.1 for that, and the piston tester showed exactly the same thing, just a hair below 90.
I coulda SWORN that I've done this before and got higher numbers in leading holes. Maybe I got it backwards, it must have been a while back that I did this with a piston tester. Anyway, looks like you're right about that, so I stand corrected.
01-13-05, 05:38 PM
#31
Lives on the Forum
Originally Posted by RotaryResurrection
Oh...and about the plugs. You're saying you actually PREFER to work around that dumbass knock sensor and those big wiring harness plugs found on FD/s5 t2 front rotorhousing T plugholes?
I yank all the crap out!
-Ted
Thread
Thread Starter
Forum
Replies
Last Post
msilvia
3rd Generation Specific (1993-2002)
15
09-11-15 12:13 PM
