Boxerworks.com

Garnet,
The shim plate also holds the shafts in place so they remain square. The desired nominal gap is 0.002" or 0.05mm. If measuring the shim gap without reassembling the cases, keeping the shafts straight ensures a better measurement. When we're talking about 0.002" clearance, that can easily be the error in moving a shaft off vertical. One post cited Matt Parkhouse not using a shim plate. I read every article Matt Parkhouse publishes and have a world of respect for him. I suspect Matt would agree that if you have access to a shim plate you should use it to improve the measurements.

I understand that you are proposing using solder to measure this gap by putting the transmission rear case back on. Having done the rebuild on my 1975 R90S transmission, I used a shim plate and I would use this method again because even when I heated the transmission case to 150 degrees F, I still had to carefully pry and do a lot of hammering with a dead blow to free the cases. As one article described working on these transmissions: The skills of both a boxer and a neurosurgeon are required. Here's my point: You CAREFULLY heat the case (I used an infrared thermometer), pry a tiny gap around the gasket surface and then hang the rear case in the air and bang CAREFULLY on the front case to separate the two. It is a wonderful moment when the cases finally separate and you've not destroyed anything. You then replace the very pricey bits within and start reassembling. I would not want to do the heating and "tapping" more than absolutely necessary and of course using solder requires you assemble and disassemble to get the measurement. And please be careful not to forget to account for the gasket thickness assuming you've scraped the gasket. I kept draping the case gasket in front of me so I wouldn't forget! And, a final thought: Thermal coefficient of expansion. Because each component in the assembly (bearings, solder, aluminum case) will all expand a different amount as you heat them up, I'd suggest trying to measure the solder quickly and as close to the same heat when you get the cases mated. To check if this is even a worry, I'd suggest checking to see how much the solder would expand over a room temp to 150 degree F differential. If its way less than 0.002", no worries. If its close, then consistent temps would help.

DaveM
1975 R90S
2000 R1100RT

Charlie your gaps are more tightly grouped than mine were. I had (without shims):
output - .7mm,
lay - .3mm and
input - 1.02mm

Even though there will be some axial play in the outer race of the bearing (I think that's the right word for the sideways wobble) the measurements have to be taken from the outer race don't they? That's where the shims sit. And there's a corresponding shoulder at the bottom of the bearing recess in the end cover that is the other measurement point.

I found that with two fingers pushing down on the outer race, one at say 12 o'clock and one at 6 o'clock you can feel when the race is sitting level. If you then take a reading at 3 o'clock you should be able to get repeatable results. You may need more than the standard issue number of hands to do this.

I also found that some of the new shims I bought had a very slight lip on them that made them appear thicker than they were. A quick rub on some fine wet and dry fixed that.

I've no doubt that you'll get it right and will soon be enjoying that warm glow of smugness as you're smoothly snicking through the gears as you wind along the river road.

Cheers
Ray

One thing that I confirmed this morning when remeasuring everything using the shim plate, is that the input shaft has 4-5 thou difference between the inner and outer race.
I have read that the inner race should be measured, but that is difficult on the lay shaft because the oil baffle is on the bearing. Difficult also on the output shaft because the shaft interferes with the depth gauge.
Also, the inner race is not what contacts the cover.
Using a shim plate will provide close to perfect alignment of the shafts as Dave says, so the outer race should provide the necessary measurements.