I just used a non rotary specific compression tester on my friends 93 r1 with 108,000 miles. I performed the test following the correct procedures. When testing, I held down the release valve so the gauge would not hold a peak reading and instead enable me to view the "bounce" of the needle as the rotor turns. I have heard this works ok. Does anyone know for sure if the results are reasonably accurate (as long as the gauge reads correct) using this method?
I feel certain the compression is low and the gauge is reading anywhere from 58-62psi when I hold the valve open. If I leave the valve closed it works its way up to a peak of approximately 75psi. Why the difference? I also find it strange I get the same readings whether I hold the gas to the floor (to open the butterflies) or not. Both rotors give almost the exact same readings.
So would you trust these readings?

 

13B rotary should have compression closer to 100 PSI or more on a healthy stock engine, but the readings you get depend on several things including the type of compression tester. The FWM has charts that you can use to adjust your readings for cranking speed, air temperature, and altitude during the test.

You say that you held down the 'release valve', but this doesn't sound like the correct procedure. Screw-in style compression testers have two valves and to use these to test a rotary you must remove the lower shrader valve. It's the one screwed in to the threaded end that goes in to the spark plug hole. You use a valve stem tool to remove/re-install this valve.

You should pull the fuel pump relay or fuse to prevent excess fuel from being dumped in to the engine during the testing. As you crank a fully warmed-up engine without any spark plugs installed and with the throttle held wide open, you'll see the needle pulse as each of the three faces come up on its compression phase. You'll have to watch closely for their peak readings and note the similarity between each of the three faces; then check the other rotor. Ideally all readings on all faces will be the same, but expect some fluctuation. Bigger variations or lower peak readings normally indicate worn seals, etc.

 

Thanks for the help. I didn't remove the lower valve where it screws in to the plug hole. I'll try to remove it and see if the readings change.
I have to wonder though if it will indeed make a difference. I assume the valve is there in order to keep the air in the tube which is attached to the gauge, thus maintaining the peak reading. If I am holding the pressure release valve open will the air not pulse to the gauge as it should? Certainly the lower valve would not restrict airflow to the gauge, would it? It is designed to allow the air to pass through it. If anything it makes more sense to me removing the release valve.
Am I totally off here?
All info is appreciated so everyone please share their thoughts!

 

Your initial procedure is correct. If you remove the valve stem you will not be able to get the max compression reading.
In general, you do the test once holding the pressure release. This will show the compression on each rotor face, the needle should bounce back to around the same reading each time. And do it again without holding the pressure release, this will give you the maxmum compression reading. The results from the two tests should be relatively close. A bounce in the readings (in the first test) means you've got a bad apex seal.
If it's low compression overall, it could be a bad engine/coolant seal or warped housing.
The acceptable compression should be 100 psi at 250 rpm and no more than 21 psi swing from each rotor face.
Just for reference my last compression test was 135 for the rear and 130 for the front, with <5 psi differential.

 

Remove the valve stem to get the correct reading. I've tried both ways and removing the valve stem show the correct reading. Also make sure that the engine is warm, or the readings will be off.