DIY: Compression Test
 

This is a write up on how you can do a compression test with a standard compression tester. If you record the gauge while preforming the test you can also calculate you RPM and better analyze the results.

I had a compression test done at a Mazda dealer ~400 miles prior to this test and the results are compared at the end of the write-up.

NOTE: That my margin of “error” was as high as 20% from the Mazda results.


Tools Needed:
Jack
Jack Stand
Wheel Socket (21mm)
Spark Plug Socket 13/16" (21mm)
3/8" Extension
3/8" Swivel
3/8" Ratchet
Compression Tester
Schrader Valve Removal Tool or Small Flat Tip Screwdriver

Optional:
Video Recorder (60fps ideal)


Step 1: Warm the car to normal operating temperature.


Step 2: Prepare you compression tester while the car is warming up. Make sure you have the right size plug insert. Mine says it’s an M14 and the threads are about ½” long.

Attachment 227065

Then you can remove the Schrader valve. The Schrader valve is a threaded valve in the center of the plug, to remove it you just need to spin it counter clockwise. There is a removal tool that came with my Craftsman Compression Tester, or you can just use a flat tipped screwdriver to gently spin the center valve.


Attachment 227066


Step 3: Once the car is warmed up, turn it off and break the lugs on the driver’s side wheel. Pop the hood and remove the main engine cover to let some light in.


Step 4: Jack up the driver’s side and remove the wheel. (Don’t forget the jack stand!)

Step 5: Now unplug the e-shaft sensor. This will stop the engine from sending spark and fuel into the chamber. The plug is found on the bottom/drivers side of the crank pulley. Here is a shot with a view between the subframe and the oil pan. There is a small tab on the top of the grey connector, just push it in and slide off the connector.

Attachment 227063

Step 6. Now you need to remove a spark plug and screw in the compression tester. Only remove one plug at a time. This will prevent you from mixing up wires, and you need the other plug in place to maintain proper compression for the test. Start by disconnecting the plug wire, this just pulls off. Then use the 13/16(21mm) deep well or spark plug socket to remove the first plug. This is a good place to use that swivel and extension. Here is a picture through the view of the wheel well. All the plugs are labeled.

Attachment 227064

Step 7. Screw the compression tester in place. You want it snug, but not too tight. Most have a soft o-ring to maintain a seal with the housing.

Step 8. Setup your camera if you have one and start recording. If you don’t have a recorder, find a friend to do step #9 so you can watch the readings.

Step 9. Hop in and depress the clutch and accelerator and try to start the car for about 10 seconds(ie turn the engine over). Keep the accelerator to the floor while doing this. As the engine turns each of the three rotor faces will pass and give you a separate reading. If you are not recording pay close attention to each of pulse and note how they differ in amplitude. Be sure to write down the results the best you can.

Step 10. Stop recording and reinstall the plug you removed and attach the spark plug wire.

Step 11. Now you can repeat step 6-10 on the remaining three plugs. You are going to do this a total of four times, twice for each housing.

Step 13. Analyze the video. If you recorded the results you can use the video to get your RPM and a better idea of your results. Download VLC Media Player if you don’t have it: VideoLAN - VLC: Official site - Free multimedia solutions for all OS! Open the video in VLC and go to View>Advanced Controls. This will add four icons above the normal play buttons. We are interested in the one on the right that looks like a film strip with a red arrow. This will allow us to go frame by frame.

Here are my videos for reference;




Step 14. Open Excel next to the video or grab a big note pad. Using the frame-by-frame button you will skip ahead one frame each time you press it. For each frame record the reading on the compression tester dial. Keep in mind most cameras will have a slow shutter speed to compensate for low light in the garage, meaning the needle may be blurred. Do your best to record the highest point. You need to record the value at each frame to get the RPM.

Step 15. Find the RPM. After you log all readings over a period of a 2-5 seconds (I used about 120-150 readings) you should have enough of a sample set. Count how many times the rotor rotates during your sample set of recordings. For example, I had 11 rotations (1 face rotation of the rotor equals one rotation of the e-shaft) over 138 samples(frames). Since I recorded at 60 frames per second, my RPM was: 11/(138/60)= 4.78 rotations per second. RPS*60=RPM. So I was at ~287 RPM.

Step 16. Find your peak values per rotor face. For this I played back the video in slow motion. This gives a better idea of what the actual values are since the camera does not capture a frame exactly when the compression is at its peak. In playback, your eye can better interpolate the estimated value.

My readings were as follows:

Housing #2 Leading:
Face #1 – 72psi
Face #2 – 68psi
Face #3 – 68psi

Housing #2 Trailing:
Face #1 – 78psi
Face #2 – 75psi
Face #3 – 75psi

Step 17. Normalize your results. I use this calculator to adjust my readings per the RPM.
Foxed.ca - Rotary Compression Calculator

Here were my adjusted results:

Housing #2 Leading:
Face #1 – 68psi
Face #2 – 64psi
Face #3 – 64psi

Housing #2 Trailing:
Face #1 – 73psi
Face #2 – 70psi
Face #3 – 70psi

Step 18. Compare your results. Normal compression gauges will measure PSI, you need to convert this to KG/cm^2. You can use this calculator; Pressure Units Online Converter
IE:
Housing #2 Leading:
Face #1 – 68psi – 4.8 kg/cm^2
Face #2 – 64psi – 4.5 kg/cm^2
Face #3 – 64psi – 4.5 kg/cm^2

Housing #2 Trailing:
Face #1 – 73psi – 5.1 kg/cm^2
Face #2 – 70psi – 4.9 kg/cm^2
Face #3 – 70psi - 4.9 kg/cm^2

Mazda Specs at 250 RPM are:
~6.9 kg/cm^2 minimum
~8.4 kg/cm^2 standard

How accurate is the test?

To be honest, I don’t know. There are more variables than data points, but the logic of the test is reasonably sound. That said, there are differences between compression testers, sometimes big differences, and there could be differences in readings due to the technology of a Mazda tester versus a normal compression tester. There also is more of a chance for human error in reading the gauge.

Here are my results compared to the Mazda test:

Housing #2 Leading:
Face #1 – 68psi vs. 82psi = 17% Low
Face #2 – 64psi vs. 79psi = 19% Low
Face #3 – 64psi vs. 78psi = 18% Low

Housing #2 Trailing:
Face #1 – 73psi vs. 84psi = 13% Low
Face #2 – 70psi vs. 82psi = 15% Low
Face #3 – 70psi vs. 81psi = 14% Low

Given these results, I would assume that analog testers will read lower than the Mazda testers, but this type of test can give you a ball park of what to expect.

Over the 400 miles the engine went from hard starting when hot, to not starting when hot, so I can assume that the compression of the engine itself was also lower between the two tests.

I only did tests on housing #2 because I could not find my spark plug socket and I had to use a wrench.

Please feel free to comment and add any helpful info!

Cheers,
Lucky Kid (Roy)