3.54:1 Differential Overhaul

[8EL]

This has been a long process mostly due to things that happen in life, and then tooling up to aid the process being undertaken. The next one will be much faster due to the tools made here. The fixture mounting the banjo allowing horizontal to vertical rotation in both directions, has proven to be most useful in this process. One of the pinion nut adjusting wrenches had to be modified to clear the banjo mounting fasteners as shown. The design of this differential is unique to say the least. In my time I have overhauled hundreds of carrier type (Caddy/Olds/Pontiac/Ford 9"), and Salisbury types used on more modern GM vehicles. Never came close to what Henry put out here by the millions some 90 years ago.

Awkward to handle to say the least due to the 4' driveshaft that procedures I have read instruct to install early in the process. I think one guy bored a hole in his shop wall to allow clearance to mount the banjo in his mounting fixture. I made a stub shaft from an old drive shaft about 8" long to serve this purpose. Details of the procedures written indicate pinion bearing pre-load adjustments to be made after installation of the driveshaft. I substituted the stub shaft for the driveshaft, and used it throughout the rebuilding process. Once pinion and carrier bearing pre-loads had been established and final assembly was at hand, the banjo was removed from the jig, the drive shaft was installed on saw horses and the unit was simply assembled. Final rechecking of the pinion bearing pre-load indicated little if any change to the twenty inch pound setting developed with the stub shaft.

They say a picture is worth many words so without further delay, many picture have I, here are a few.

Please, thoughts and suggestions on this process undertaken are greatly appreciated. Much can be learned from another's perspective, all welcome here.....

[HD Rider]

Looks good.
I installed 3:54 gears in my 40. Dad did the A and regretted not regearing to 3:54.

[8EL]

Thank you, yes, I desired to bring the engine speed down, always seemed to be turning up too high, which it has, but it now seems to be a little under powered. At 45 MPH the throttle is barely open and the engine speed is noticeably slower than before. But slowing down for cornering, where I used to be able to leave it in 3rd gear and let her "cackle" a little, it now needs to be downshifted. Your 40 most likely has the little V8?

I overhauled this engine about 15,000 miles ago, completely stock except for upgrading to insert bearings. It burns no oil all season long, which amounts to 2,000-2,500 miles, and starts instantly and runs smoothly and is quiet. There are no markings on the cylinder head so I assume it to be the stock 4.2:1 compression ratio engine. The compression to me seems a bit low however at 50 psig, (number 4 is 45 psig) it does not increase with a little of oil in the hole. The valves when overhauled were lapped as the seats were not too wide, but if the compression is in fact low, they would have to be the culprit.

What do you guys realize for compression results on a engine at operating temperature, having a good battery fully charged, and the throttle wide open?

[8EL]

So I guess if we think about this, in a ideal application, the atmospheric pressure would be realized in our cylinder with the intake valve open, piston on BDC our gauge pressure (psig) on the compression tester would read 0 psig. Of course if the pressure were actually 0 the pressure in our bodies would blow us apart, that is why astronauts in zero pressure/gravity environments wear suits to maintain the atmospheric pressure applied to us here on earth.

Depending on elevation there is an absolute pressure (psia) association to that elevation. As elevation increases psia decreases, at sea level the psia is 14.7. So given this, whatever the compression ratio is, the final pressure read on our gauge which is calibrated to read psig at sea level, would not our final cranking compression look something like psia x C/R......

Applying this then my elevation is 663 feet above sea level, this would correspond to roughly 14.3 psia x 4.2 = 60.06 psia the gauge is calibrated to read 0 at sea level so then to compare the gauge pressure to absolute one would subtract 14.7 from the product = 45.36 psig.....which says my engine is fine.

So then the $64,000 question, what has been the realized performance characteristics of others having reduced their final drive ratio to 3.54:1?

Surely I am not the first one to have done this, anxiously waiting to hear from those who have......

[SAJ]

8EL, your calculations assume an isothermal (constant temperature) cylinder compression, whereas the rapid upwards piston stroke makes the event closer to an adiabatic (no heat loss) event. As an analogy, much as a bicycle pump heats up when you prevent the air escape with your finger.
So the temperature rises and the gauge pressure reading increases. I think to as much as 70-odd psig in a new motor without looking if up.
Lots of standard engines report about 50 psig WOT. Many say no difference with throttle closed or wide open, but my 6 volt engines jump up more than 5 psig when tested WOT with fully hot engine. They are overbored to 4 inch with 5.5 Snyder heads so not comparable to your readings. Closer to100 psig.
I notice 12 volt engines crank faster and give higher readings due to both higher adiabatic heat retention and less time for leakage through ring gaps, amongst other factors.
I find Your posts and experiments very interesting, nonetheless.
SAJ in NZ

[Rob Doe]

Sponsored Links (Register now to hide all advertisements)

8EL, thanks for posting this. I'm hoping some of the veterans chime in here. I'm a newbie to the Model A and have participated in only two rear end repairs. I'm in the midst of a 3.54 ring gear conversion on our coupe. I had to seek help as the two end bearings on the differential case were no longer a sound press fit. The journal on one side had to be knurled to restore the press fit. The other needed a bit of Locktite, but no knurling. Thus I took the parts to a professional for repair and assembly.

That said, I spent several hours in Tom Endy's articles plus I watched several differential videos to get a good grasp of backlash and pinion depth etc.

Here is a helpful youtube that does a good job of clarifying contact patch interpretation. After several days and hours of effort this is the straightest, clearest and shortest video I've found.

Measuring Ring Gear Contact patch on All Hypoid Differentials

If I recall correctly, the Model A differential is not a hypoid because the pinion shaft intersects at the center line of the ring gear??? A hypoid would intersect off center and that increases the tooth contact area increasing load capability....too much info.I think the graphic explanation of how the contact patch is interpreted is the same???

I don't believe it is common to shim the pinion on a Model A??? If true, that is one thing we don't have to adjust (pinion depth).

Your patch appears to be near the middle of the ring, heel to toe, on the drive side. What does your contact patch look like on the coast side???

As an amateur seeking to learn, I'd be interested in knowing if moving a couple thousandths of gasket material from the left side of the banjo to the right would allow a reduced backlash and move the contact patch lower on the ring tooth face. Maybe someone experienced will post suggesting an amount. What total thousandths gaskets, left and right, produced the contact patch shown???

Applying some rags and twisting at the axle ends may allow you to apply some resistance to the ring gear to get a better wiping of the patch???

The banjo with the fill plug offset to the left might allow you to take a reading of the ring gear backlash with a dial indicator??? I seem to recall Tom suggested .010 or less??? You might be able to read the patch via the fill plug???

Nice pictures and thread, I hope some of my research helps you.

[Rob Doe]

8EL, thanks for posting this. I'm hoping some of the veterans chime in here. I'm a newbie to the Model A and have participated in only two rear end repairs. I'm in the midst of a 3.54 ring gear conversion on our coupe. I had to seek help as the two end bearings on the differential case were no longer a sound press fit. The journal on one side had to be knurled to restore the press fit. The other needed a bit of Locktite, but no knurling. Thus I took the parts to a professional for repair and assembly.

That said, I spent several hours in Tom Endy's articles plus I watched several differential videos to get a good grasp of backlash and pinion depth etc.

Here is a helpful youtube that does a good job of clarifying contact patch interpretation. After several days and hours of effort this is the straightest, clearest and shortest video I've found.

Measuring Ring Gear Contact patch on All Hypoid Differentials

If I recall correctly, the Model A differential is not a hypoid because the pinion shaft intersects at the center line of the ring gear??? A hypoid would intersect off center and that increases the tooth contact area increasing load capability....too much info.I think the graphic explanation of how the contact patch is interpreted is the same???

I don't believe it is common to shim the pinion on a Model A??? If true, that is one thing we don't have to adjust (pinion depth).

Your patch appears to be near the middle of the ring, heel to toe, on the drive side. What does your contact patch look like on the coast side???

As an amateur seeking to learn, I'd be interested in knowing if moving a couple thousandths of gasket material from the left side of the banjo to the right would allow a reduced backlash and move the contact patch lower on the ring tooth face. Maybe someone experienced will post suggesting an amount. What total thousandths gaskets, left and right, produced the contact patch shown???

Applying some rags and twisting at the axle ends may allow you to apply some resistance to the ring gear to get a better wiping of the patch???

The banjo with the fill plug offset to the left might allow you to take a reading of the ring gear backlash with a dial indicator??? I seem to recall Tom suggested .010 or less??? You might be able to read the patch via the fill plug???

Nice pictures and thread, I hope some of my research helps you.

[Rob Doe]

8EL, thanks for posting this. I'm hoping some of the veterans chime in here. I'm a newbie to the Model A and have participated in only two rear end repairs. I'm in the midst of a 3.54 ring gear conversion on our coupe. I had to seek help as the two end bearings on the differential case were no longer a sound press fit. The journal on one side had to be knurled to restore the press fit. The other needed a bit of Locktite, but no knurling. Thus I took the parts to a professional for repair and assembly.

That said, I spent several hours in Tom Endy's articles plus I watched several differential videos to get a good grasp of backlash and pinion depth etc.

Here is a helpful youtube that does a good job of clarifying contact patch interpretation. After several days and hours of effort this is the straightest, clearest and shortest video I've found.

Measuring Ring Gear Contact patch on All Hypoid Differentials

If I recall correctly, the Model A differential is not a hypoid because the pinion shaft intersects at the center line of the ring gear??? A hypoid would intersect off center and that increases the tooth contact area increasing load capability....too much info.I think the graphic explanation of how the contact patch is interpreted is the same???

I don't believe it is common to shim the pinion on a Model A??? If true, that is one thing we don't have to adjust (pinion depth).

Your patch appears to be near the middle of the ring, heel to toe, on the drive side. What does your contact patch look like on the coast side???

As an amateur seeking to learn, I'd be interested in knowing if moving a couple thousandths of gasket material from the left side of the banjo to the right would allow a reduced backlash and move the contact patch lower on the ring tooth face. Maybe someone experienced will post suggesting an amount. What total thousandths gaskets, left and right, produced the contact patch shown???

Applying some rags and twisting at the axle ends may allow you to apply some resistance to the ring gear to get a better wiping of the patch???

The banjo with the fill plug offset to the left might allow you to take a reading of the ring gear backlash with a dial indicator??? I seem to recall Tom suggested .010 or less??? You might be able to read the patch via the fill plug???

Nice pictures and thread, I hope some of my research helps you.

[8EL]

Originally Posted by SAJ

8EL, your calculations assume an isothermal (constant temperature) cylinder compression, whereas the rapid upwards piston stroke makes the event closer to an adiabatic (no heat loss) event. As an analogy, much as a bicycle pump heats up when you prevent the air escape with your finger.
So the temperature rises and the gauge pressure reading increases. I think to as much as 70-odd psig in a new motor without looking if up.
Lots of standard engines report about 50 psig WOT. Many say no difference with throttle closed or wide open, but my 6 volt engines jump up more than 5 psig when tested WOT with fully hot engine. They are overbored to 4 inch with 5.5 Snyder heads so not comparable to your readings. Closer to100 psig.
I notice 12 volt engines crank faster and give higher readings due to both higher adiabatic heat retention and less time for leakage through ring gaps, amongst other factors.
I find Your posts and experiments very interesting, nonetheless.
SAJ in NZ

It is true that the heat of compression is valid, increasing temperature of the air in an enclosed cylinder will cause molecules to be further apart in the fixed volume of the cylinder taking up more space, would have to increase pressure. Valve timing was not considered here either which will along with throttle opening affect the engines volumetric efficiency. Cranking speed as you mention would also be a factor. Cylinder pressures are also very dynamic and very greatly as engine speed and operating conditions such as ambient temperature humidity and barometric pressures change.

As a standard we try to remove as many variables (or at least control them), This said compression test are taken with a engine at operating temperature, battery fully charged, throttle wide open.

What I was trying to determine here, knowing that the air fuel mixture was correct (evident by spark plug appearance), and spark timing was properly set, if the compression was what it should be then it could be assumed that the engine was most likely producing the power it was intended too.

From this then would it be reasonable to assume the initial compression test yield more or less an average compression of 50 psig on these engines with their original 4.2:1 heads?

[8EL]

Originally Posted by Rob Doe

8EL, thanks for posting this. I'm hoping some of the veterans chime in here. I'm a newbie to the Model A and have participated in only two rear end repairs. I'm in the midst of a 3.54 ring gear conversion on our coupe. I had to seek help as the two end bearings on the differential case were no longer a sound press fit. The journal on one side had to be knurled to restore the press fit. The other needed a bit of Locktite, but no knurling. Thus I took the parts to a professional for repair and assembly.

That said, I spent several hours in Tom Endy's articles plus I watched several differential videos to get a good grasp of backlash and pinion depth etc.

Here is a helpful youtube that does a good job of clarifying contact patch interpretation. After several days and hours of effort this is the straightest, clearest and shortest video I've found.

Measuring Ring Gear Contact patch on All Hypoid Differentials

If I recall correctly, the Model A differential is not a hypoid because the pinion shaft intersects at the center line of the ring gear??? A hypoid would intersect off center and that increases the tooth contact area increasing load capability....too much info.I think the graphic explanation of how the contact patch is interpreted is the same???

I don't believe it is common to shim the pinion on a Model A??? If true, that is one thing we don't have to adjust (pinion depth).

Your patch appears to be near the middle of the ring, heel to toe, on the drive side. What does your contact patch look like on the coast side???

As an amateur seeking to learn, I'd be interested in knowing if moving a couple thousandths of gasket material from the left side of the banjo to the right would allow a reduced backlash and move the contact patch lower on the ring tooth face. Maybe someone experienced will post suggesting an amount. What total thousandths gaskets, left and right, produced the contact patch shown???

Applying some rags and twisting at the axle ends may allow you to apply some resistance to the ring gear to get a better wiping of the patch???

The banjo with the fill plug offset to the left might allow you to take a reading of the ring gear backlash with a dial indicator??? I seem to recall Tom suggested .010 or less??? You might be able to read the patch via the fill plug???

Nice pictures and thread, I hope some of my research helps you.

It is not uncommon from what I have learned to have too much clearance on the carrier bearings inner race, I had one on mine that I knurled and applied 271 (red) high strength loctite.

Correct, these units are not hypoid, the pinion runs on the driveshaft center line. More than one mechanic has found this out as the carrier can be installed on either side of the housing. If the mistake was made mounting it on the passengers side, their Ford now had 3 speeds in reverse, and 1 forward! The drive side of the tooth is most important as its contact with the driven gear. I have attached a pattern of the coast side of the ring.

The pinion bearing cup could be shimmed out to change ring gear contact, but I think for the most part the cup is pressed up against it's shoulder. Carrier bearing cups can be shimmed to obtain proper backlash if this could not be done by gasket thickness selection. To get what you see here, I had .005" on the left side and .017" on the right for gasket thickness. Keep in mind that no two units will be the same, but .020" is from what I read the total normally encountered of the two sides cumulative.

I put the marking compound on a little heavy, the idea you had with the rag might help also, but try putting it on a little thinner than I did and see what you get. I would not be too terribly concerned with the backlash, I would be concerned to see that adequate axle shaft end play exists especially if installing new axles and/or side gears. There are a lot of substandard replacement parts on the market so look into where they are coming from and what others may have used before buying.

Be sure and let me know about the drive ability of your machine when you get it back together...It sounds like you are heading in the right direction....

[Rob Doe]

8EL, thanks to your post, other newbies like myself will have a good chance to catch some details of rear end repairs here. Here's an additional bit of info that I'd like to share specifically for those who read through this thread at a later time.

Be sure to examine your axle housing flanges for true. Our coupe had a shallow bend in a rear radius rod. When the rear end was disassembled for a cracked axle along the key way, the axle housing on the same side had a slightly bent flange in one area for several degrees. I was told this happens when the car hits something like a curb with a major force. I could just barely see that the rear wheels were out of alignment all along, but didn't catch on until removing, working on the rear end.

Brent Terry provided advice regarding how I might be able to straighten our bent radius rod. I will try his suggestion as I have a tig welder and the skill to use it. He suggested that I cut perpendicular slits in the stretched side of the rod and apply splints to straighten it, then weld it up. (paraphrased)

[8EL]

I hope this will be helpful to others Rob, Be sure and let me know what the performance pf your machine is when you get it together.......