Elan Diffs

[gordonlund]

Week before Donington I just finished swapping the dif in Rhubarb from the std 3.77 to a 3.54 out of an Escort Mexico. I looked at numerous options of 5 speed boxes but still wanted to keep a std layout and retain the sweet change of the old bullit 3 railer. I was looking for a slight up gearing for our French tours, mainly on N roads interspaced with D and unclassified roads. White roads to Ordinance Survey buffs. My brief excursion to Donington was pleasant enough now with a handbreak that works and not so much jiggling from underneath, courtesy of new handbrake tree and extra springs. New donuts, to me, added for extra confidence that we get back under own steam, touch wood.
Oh, by the way the car runs like a dream.
See pics. This car car is for touring not concours

Gordon Lund (Rhubarb's ready for his hols and has just passed his MOT)

[steveww]

How much difference does this make to rpm at cruising speed?

[mcclelland]

Hi Gordon, great to see you at the show, had a good time. Didnt know you had fitted the new diff, but I would like to know with the new diff is your speedo now reading better or worse. Best regards, George...

[garyeanderson]

the math works out to about 6 percent.

Gary

[msd1107]

Gordon,

The other question is, what happened to the odometer reading?

There are alternative speedometer drive gears you can fit. I believe they were color coded so you didn't have to count teeth.

Take your car out to a measured mile (or even better a 2, 5, or 10 mile stretch). Take your odometer reading between points. If your odometer was accurate before, it should read 9.39 miles over a 10 mile stretch now.

If it reads between 9.65 and 9.12 miles, putting in the the speedometer drive gears for the 3.55 differential should be acceptable.

There is a firm that makes a correction box for any mileage mismatch. Sorry I don't have the link right to hand, but I probably could retrieve it if necessary.

After the odometer is to your satisfaction, check the speedometer reading by going at a fixed rate of speed through a measured distance and see if there is unacceptable (to you) inaccuracy. Most speedo shops can handle the recallibration for you, since most speedometers have a standard input rev/mile.

There are also other techniques to establish the odometer calibration.

Of course, you can mentally recalibrate the speedometer and interpolate on the fly. Keeps the brain cells active. Or use the tach as a speedometer.

David
1968 36/7988

[oldokie]

I have a dumb question, slightly off track, does anyone which alloy the diff case is, can it be welded? I had a slight problem with a 'donut' and when the dust settled I had a hole in the bottom of my boot, 1/2 a spare, a very ratty car cover, jumper cables that were now 60 feet long, and an ear broken off the top of the diff case!!

[crannyr]

We raced weld repaired diffs for several years without a problem related to the repair itself.

[gordonlund]

I did some calcs a while ago and drew a graph to compare the ratios. Here they are for you techi types. Afraid they are a bit rough, I never got round to inputting them to the computer. Must get round to doing that sometime.

Gordon

[types26/36]

I have a dumb question, slightly off track, does anyone which alloy the diff case is, can it be welded? !

Yes, with a good welder.

[gordonlund]

Cast Aluminium can be welded with the right kit and some one who knows how to use it. Argon arc TIG (Tungsten inert gas) is really the only process for such a high stressed component. Properly preped, ie V notched at the correct angle and welding done in stages with the appropriate pre heating in an electric heat treatment oven. Unless you don't want it to break in service, have it welded by an expert. Don't use a back street welder.

Other course of action though is to buy a decent second hand one, they are out there but be careful of rubbish and scoundrels who want silly money for them.

Gordon

[oldokie]

Thanks for the welding advice, it will go to an expert as I don't have a TIG machine, and won't try it if I did! Thanks again!!

[msd1107]

The general question was what is the true speedometer and odometer reading with variations in diff ratio, tire size, etc.

A previous post gave a cost free way to determine this.

But we can bring technology to bear on this problem, if we are will to part with some hard earned cash.

For UK readers go to:
http://www.race-technology.com/ap22_4_620.html

For US readers, go to:
http://www.race-technology.com/ap22_10_1529.html

These are basic units, but there are much more capable units, depending on the depth of your pocket book.

Seems like a club or neighborhood group could buy one and share.

David
1968 36/7988

[elansprint71]

I fitted a 3.55 diff last year to my Sprint, although I had the correct speedo drive gear (black), I did not bother to fit it. Comparing the speedo reading with my top of the range GPS Sat Nav I find that the speedo is now showing the true mph right across the scale from 10mph to 90mph, and presumably beyond.
I guess that by doing this swap I have just removed the over-reading factor which was built in, as in most cars?

Cheers,
Pete.

[gordonlund]

For the techis again, the original fitment of tyre to the +2 was the old 165 x 13 which had an aspect ratio of 100%. I just happened to have one in the garage on an old rim and measured the od, 24" then I measured a modern 165 x 13 80% aspect ratio, 23". that 1 inch reduction in dia is 4.1% the difference between a 3.77 and 3.54 diff is 6.1%. I reckon that that albeit for a smidgen which I am sure my bouncy needles wouldn't know the differance answers your question. The 22 1/2 " in my calcs is measured on the car with the flat bit at the bottom.

Gordon (Rhubarb & Custard)

[msd1107]

With respect to possible tire diameters for 165-13 tires.

A 165/80-13 (modern day tire) is nominally 23.4" diameter, and 887 rev/mile. (The rev/mile figure is a nominal figure by the tire manufacturer and is based on a nominal rim width, inflation pressure, car weight, and tread depth. A variance of any of these factors will change the actual rev/mile on a specific vehicle).

If the tire was a 83% aspect ratio, it would be 23.8" diameter, and 873 rev/mile. It may have been that 40 years ago you could get a 83% aspect ratio, but the tire probably would have been marked 6.5-13 or 650-13. Or perhaps an old bias ply tire had a larger static unloaded diameter. It is not too probable that there was a 100% aspect ratio tire. That would be 26.0" diameter and 799 rev/mile.

In the post that had the graph, a figure of 906 rev/mile was used, which is about that for a 155-13 tire.

David
1968 36/7988