My Car News

Motion Industries ordered the bearings for the Mazda rear end. Each came from a different place and was a different brand, but they were fine. I was the most surprised that Motion Industries could find the pinion seal because Mazda changed the spec for it later. The older design has a dust seal that fits into the dust cover on the yoke.I put it all together as directed by various websites (Google ring and pinion adjustment. Here’s one: www.richmondgear.com/01instructions.html). I am assuming now, that you have read that.Some logic: the differential is OEM Mazda RX7. It has over 140,000 miles on it, and it never had a problem. It never made noise, and it always worked. It shows no signs of damage or heavy wear. It has a spacer in the pinion bearings. The purpose of this spacer is to compensate for tool wear in the manufacturing process of the housing. Since, new bearings should be within hundredths of thousandths, the factory spacer should always work.However, instructions will tell you when to read the patterns on the ring teeth and use new spacers. I fell for this. I dug up some pinion spacers I had from a Ford 8.8 I overhauled a few years ago. Even back then, I used the factory spacers without measuring it, and it worked great (I had marked the positions of everything and put it back the way it was). It lasted another 100,000 miles with no trouble. But, because I wanted the Mazda to be just right, I planned on using new spacers.I bought a dial indicator with the magnetic stand (to measure the backlash) form Harbor Freight, and I searched for an inch-pound torque wrench to properly measure the pinion preload. The Haynes book for the RX7 says the preload should be around 12 inch-pounds, and I had to ask an engineer friend if that is 1 foot-pound; It is.Inch-pound torque wrenches are not cheap rare. Sears had one, but I wasn’t rushing to buy anything there. Amazon.com had one for 300 dollars. So I got the click-type one at Harbor Freight for 19.99. It measures down to 20 inch-pounds, but I figured it would be close enough as long as I used its lowest setting and made sure it didn’t click.I started by adding one pinion spacer and setting up the preload without the crush sleeve. This way, I could check the tooth patterns without having to use up my one shot at making the crush sleeve work.I put the carriage in with the new bearings and started measuring. The book says 3.5 to 4.3 thousandths for backlash, and I was having trouble getting it set up. It never looked right. Worse, it was making patterns that were inconsistent with any instructions I could find. Finally, I realized that the teeth patterns were indicating that the pinion was too far away. Yet, that didn’t make sense because I had added a spacer. It should have been too close.After hours of trying to figure this out, I looked closer at the dial indicator. I realized it only measures in the thousandths. I had assumed that it measures in the 10 thousandths because every single backlash spec I have ever seen was in 10 thousandths (3.5 to 4.3 thousandths = 35 to 43 10-thousandths). It took me a while comprehend this: The dial indicator, which was made for this specific use, does not display significant digits for this measurement. I cussed. Then, I reset the carriage to the right backlash, and the patterns correctly indicated that the pinion was too close. So I took the extra spacer out, reset the pinion with the factory spacer, reset the carriage to the right measurement, and it was perfect with the original spacer just like it should be.I reassembled the pinion with the used crush sleeve. I figured out how to reuse it and have it work just fine (there is no such thing as a new crush sleeve for this rear end, at least not that I could find). Mazda has decided that 85 RX7s are now too old to support. I placed a small spacer, which I had left over from that Ford 8.8 kit, with the sleeve, torqued down the yoke, and it all went together perfectly.Now the car is its fast old self again.Punch line:Know what you are measuring.