6 speed gearbox for the Elan

[elanfan1]

Well I am one of the 12 with my car currently in the bodyshop I am hoping that I may see my car with the 6 speeder fitted later this year.

I spoke to Neil this morning - the test car has been out and about already and Neil says 'it's fantastic'. Obviously some wider testing will be carried out over the next few weeks. The tests will also work on the speedo calibration and cross checked via a GPS and if necessary a new speedo gear will be sourced.

I don't know how well Ken and Neil are known outside their specialist area (those who know them will already know) but they are life long enthusiasts Ken was one of the first Lotus dealers and a personal friend of Chapman. Their preparation of anything they do is meticullous (you only have to see the cars they have prepared and the prices they command {hoping here for a Lottery win!} to know that they do not accept second best.

I can appreciate the doubters on here with the many false dawns and promises that have been made about various replacement boxes/conversions etc however you can be assured that when they set their mind to sorting something it will be done correctly. I know the price will probably be considered steep but as far as I am aware at the moment it includes fitting and the MX5 box will have been overhauled/checked before installation.

Speculation cannot be stopped I'm aware but there are a few on here posting updates from personal knowledge which is great and as far as I can see accurate - it won't be too long now for a definitive appraisal of the box.

[GrUmPyBoDgEr]

There is surprisingly little tooling required, with the sandwich plate/spacer the most important piece. This is CNC cut from steel plate and the design effort is what has taken the time. As I said the existing Mazda gearbox casing is modified, so no tooling for this, but some skilled alloy welding and machining is required. The propshaft is a custom made item, but these again are not difficult to fabricate at reasonable cost. Not sure about G/box mount, but existing speedo drive is to be retained. Even the existing gear lever can be retained with some modification. With regard to the Spyder Elan chassis, I can't comment, but it was my understanding that they usually gave better access - is this wrong?

As to ratios, I don't know details, but trust their combined judgement.

Jeremy
PS if no-one answers the phone in the Myers' workshop, it is probably because they are working on a car, so have a clear message ready to leave before you ring! It may be my car he's working on so don't disturb Neil too much....

The Spyder chassis certainly provides much improved engine bay access but the 4 square tubes at each corner of the g/box transmission tunnel areas make things tighter than with the standard Lotus folded sheet steel chassis.
The "standard" Spyder chassis consists of a 1" x1" square tubing frame with 1" round tubing diagonals. The outside dimensions match those of the "Lotus" chassis but those tubes rob it of internal space.
For the Zetec/MT75 frames they replaced the rearmost set of diagonals with ovalised tubing (dimensionally the same as their rear wishbone tubing) to make some more space for the fatter MT75 box.
Also at the entrance to the Tunnel the piece of sheet metal that is welded in there to give torsional stiffness is either heavily relieved (or maybe removed completely?), again to provide the req'd space for the MT75.
My "normal" Spyder chassis has non of these changes, maybe newer ones have them as "standard" because they don't affect the integrety of the chassis & would avoid chassis build "variants" in production at Spyder.
I'm afraid that to do those changes to an existing "round diagonal tube" chassis would mean major surgury

Cheers
John

[RotoFlexible]

I know the price will probably be considered steep but as far as I am aware at the moment it includes fitting and the MX5 box will have been overhauled/checked before installation.

I hope their business plan includes some means of shipping the kit to customers, rather than requiring customers to come to them. The roads between the US and UK are in poor repair...

Seriously, if this is for real (and it appears to be) and the price isn't too astronomical, there will probably be enough demand just through this site to put together a bulk shipment to North America. It certainly sounds like what I am looking for.

[garyeanderson]

In order to get the ratios where they need to be for the Elan, Plus2 or other Kent block powered lightweight car you will need to overdrive all of the ratios by 15 to 20 percent. this is a fairly common way of turning a bog standard box into something with some sporting ratios. This is how Ford Came up with the Ultra close box for the Lotus Cortina (2.51, 1.64, 1.21). The process is pretty simple, just chuck the first motion shaft in a lathe and remove the gear, look for the suitable replacement with a few extra teeth and weld back to the original first motion shaft. The same is done on the second motion shaft (countershaft) except that the new first gear on the countershaft has a few less teeth.


The ratios on the bottom row are a guess, I just multiplied .80 by the current gear ratio.
Miata 6-spd 3.760 2.269 1.645 1.257 1.000 0.843
gear x .80 = 3.008 1.815 1.316 1.006 0.80 0.674

[RotoFlexible]

The ratios on the bottom row are a guess, I just multiplied .80 by the current gear ratio.
Miata 6-spd 3.760 2.269 1.645 1.257 1.000 0.843
gear x .80 = 3.008 1.815 1.316 1.006 0.80 0.674

Is the Miata box all-indirect? If 5th is direct, it would remain at 1.000.

I usually turn about 4,000 on the highway. With Gary's speculative ratios, I would be turning 3,200 in 5th (should be no problem) and 2,700 in 6th (questionable, even with 1700cc).

[garyeanderson]

I dunno, I am just guessing. What ever is direct would still be direct, my mistake but you get the idea.

[msd1107]

Gary,

If I am not mistaken, the rally gear set (2.51 1st) came before the Cortina gear set (2.97 1st).

The rally set had 17:32 1st, 22:27 2nd, 26:24 3rd, and 21:28 main transfer.
The Cortina set had 17:32 1st, 22:28 2nd, 26:23 3rd, and 19:30 main transfer.

So it wasn't just a new transfer pair.

If they had just changed the main transfer pair, the box would have had 17:32 1st, 22:27 2nd, 26:24 3rd and 19:30 main transfer for 2.972, 1.938, 1.457, and 1.000 ratios. I doubt that anyone would have accepted the 3rd-4th drop.

Incidentally, 40+ years ago, I did roughly the same thing with my motorcycle gearbox, changing 4 (different from what Ford did) gears to go from a wide ratio gearbox to a close ratio gearbox with properly staged ratios.

As far as I am concerned, it is easier machining a gear set than designing the ratios for the gear set, but then I have a warped mind, a math degree, a program that selects the optimum gear pairs, and no mechanical skills.

David
1968 36/7988

[msd1107]

Elanfan1,

It is good to hear that Vincent et al are looking at proper speedometer calibration.

There are a total of at least 4 speedometer drive gears needed to accommodate the usual range of Elan tire sizes, differential ratios, and speedometer turns/mile.

I have furnished Vincent with a link to a spreadsheet that should allow his team to determine the necessary gears.

Let's hope for the best.

David
1968 36/7988

[garyeanderson]

My generalizing has got me in trouble again.

Gary

Gary,

If I am not mistaken, the rally gear set (2.51 1st) came before the Cortina gear set (2.97 1st).

The rally set had 17:32 1st, 22:27 2nd, 26:24 3rd, and 21:28 main transfer.
The Cortina set had 17:32 1st, 22:28 2nd, 26:23 3rd, and 19:30 main transfer.

So it wasn't just a new transfer pair.

If they had just changed the main transfer pair, the box would have had 17:32 1st, 22:27 2nd, 26:24 3rd and 19:30 main transfer for 2.972, 1.938, 1.457, and 1.000 ratios. I doubt that anyone would have accepted the 3rd-4th drop.

Incidentally, 40+ years ago, I did roughly the same thing with my motorcycle gearbox, changing 4 (different from what Ford did) gears to go from a wide ratio gearbox to a close ratio gearbox with properly staged ratios.

As far as I am concerned, it is easier machining a gear set than designing the ratios for the gear set, but then I have a warped mind, a math degree, a program that selects the optimum gear pairs, and no mechanical skills.

David
1968 36/7988

[msd1107]

Jeremy,

It is good to hear that Vincent, Ken and Neil are working hard on the conversion. From history, I know they have a wealth of experience and are very skilled.

I would have hoped that they would have approached the project from a systems engineering point. Namely, the requirements are:

Drive car into garage.
Disconnect gear stick, speedo cable and reverse light and misc.
Remove engine and transmission.
Remove old transmission, bolt up new transmisson.
Replace engine and transmission in car.
Reconnect back up light, speedo cable, gear stick, and misc.
Drive away.

Any thing less is not a completely engineeered project. Since our cars are 40 years old, the builder cannot expect to see anything other than the standard output splines and input splines for any particular target car.

In particular, a new driveshaft may not be compatible with various users' cars.
A new clutch may not be useable or compatible with a users' clutch/flywheel.
A new bellhousing may not be compatible with or the user may not wish to replace their current bellhousing.

It seems that the current state of the project is a demonstration of feasibility. The fact that at the current time it needs a custom drive shaft, clutch, adapter plate, and bellhousing means the engineers have some detailed engineering to do before the project is complete.

All these extra pieces just drive up the cost, especially the shipping cost for potential customers outside the UK.

Let's hope for a satisfactory conclusion to this project.

David
1968 36/7988

[RotoFlexible]

In order to get the ratios where they need to be for the Elan, Plus2 or other Kent block powered lightweight car you will need to overdrive all of the ratios by 15 to 20 percent. this is a fairly common way of turning a bog standard box into something with some sporting ratios. [...] The process is pretty simple, just chuck the first motion shaft in a lathe and remove the gear, look for the suitable replacement with a few extra teeth and weld back to the original first motion shaft. The same is done on the second motion shaft (countershaft) except that the new first gear on the countershaft has a few less teeth.

Hmmm, if this process were applied to the MX-5 Miata 5-speed, with a factor of .93 (and assuming 4th is direct), you'd have:

Original.. 3.1360 1.8880 1.3300 1.0000 0.8140
x .93..... 2.9165 1.7558 1.2369 1.0000 0.7570

With the exception of the 16% shorter 1st, those ratios are nearly identical to the Lotus CR, with a (for me) ideal 0.75 5th.

[msd1107]

Andrew,

Capital idea! Now, to apply some geeky analysis. For 155-13 tires, 3.55 diff, and 6500 rpm:

MX-5 5-speed
Gb ratio % diff speed.....diff......err......FoM=62.7
3.136....66.1....038.4...25.4...00.5
1.888....42.0....063.7...26.7...00.0
1.330....33.0....090.4...29.8...07.3
1.000....22.9....120.3...27.5...03.4
0.814..............147.8

Without the fifth gear ratio in the Figure of Merit calculations, the FoM is 93.0. The best FoM is 97.0 with a 5th of .79 instead of .814.

Applying a constant factor to all gear ratios, the FoM for 4 speeds increases to 99.3 with a factor of .99 for ratios of 3.105, 1.869, 1.317, 1.0, and .806. For 5-speeds, the FoM is 95.5.

The FoM for all 5 speeds increases to 96.4 with a constant factor of .98 for ratios of 3.073, 1.850, 1.303, 1.000, and .798.

If we accept the original FoM of 62.7, the constant factor can decrease to .9694 for ratios of 3.040, 1.830, 1.289, 1.000, .789.

For a constant factor of .93 we get

MX-5 5-speed with .93 constant factor
Gb ratio % diff speed.....diff......err......FoM=-376.0
2.916....66.1....041.2...27.3...06.7
1.756....42.0....068.5...28.7...00.8
1.237....23.7....097.2...23.0...95.8
1.000....32.1....120.3...38.6...39.5
0.757..............158.9

Here, you will notice the oddball gear is 4th, with it being close to 3rd and a long way from 5th.

The Figure of Merit for only the 1-4 gears is -18.9

For comparison, here is the Lotus gearbox

Lotus CR 2.51 1st, 4th direct,
Gb ratio % diff speed...diff....err..FoM=99.8
2.510....53.4...047.9..25.6..00.0...51.9 at 7,000
1.636....33.0...073.5..24.2..00.0
1.231....23.1...097.7..22.6..00.0
1.000............120.3

The gear ratios that are the closest between the tweeked 5-speed and the Lotus 4-speed are 3rd and 4th. Even if the 1st and 2nd look reasonably similar, mathematically they are not, and it requires an analysis tool to show this.

Incidentally, 45+ years ago, I used a sheet of paper, handwritten data, and a circular slide rule to design gear ratios around these basic design guidelines. Lots of trial and error, and the results were not nearly as good as using the software design tool.

I have offered the tool to Vincent free of charge just in case they want to analyse gear ratio possibilities. Just think of the advertising bling they would get by being able to advertise "optimum computer generated gear ratios"

David
1968 36/7988

[msd1107]

This does not directly address the 6-speed under development.

Last night, I received an email from John Esposito of Quantum Mechanics announcing the long awaited availability of a direct drop in T9 for the Elan.

This uses the Alan Voight tail piece and centre section. There are various options that include an aluminum gearbox case and the BGH E7 gearset.

This is probably as close to an ideal add-in gearbox as we are likely to see. Light weight, close ratios. The price is not that of a beater second hand T9, but if someone wants a 5-speed, find a way to order this.

I am including John's announcement email as a word document, and the specification document as an attachment in case he has not contacted anybody directly. Also included is a spreadsheet with an abbreviated technical analysis of the various gearset options

There is a downloadable spreadsheet at viewtopic.php?p=82911 at the bottom that allows for any ratio set to be properly evaluated. This includes help text to assist the user. As usual, the spreadsheet is ZIPped and renamed to a .xls extension. After downloading, rename it back to .ZIP and extract the spreadsheet

David
1968 36/7988

[bill308]

David,

The spreadsheet you posted above:

20040227xGearRatios.xls [55.5 KiB]

does not work.

The computed cells I looked at only have fixed values, not formulas and some of the values make no sense. Please take a look at it and post a corrected version.

Thanks,
Bill

[msd1107]

Hi Bill,

You are correct, there is no computation performed in the spreadsheet. 20040227x is just an extract of the figures from several gearboxes (the Lotus CR, Semi-close, and several T9 ratio sets) so that the data can be viewed together for comparison purposes. The computation is done by the original spreadsheet, and the relevant figures extracted. I did the same thing in a previous post in this 6-speed thread for various 6-speed gearsets.

The spreadsheet at viewtopic.php?p=82911 at the bottom will allow you to input any set of ratios and get all the gory, geeky figures calculated, together with help text to explain everything. The only thing not computed is mph/1000 rpm that is done in later versions of the spreadsheet. That computation is also performed in one of the sheets in elan-f15/updated-spreadsheets-t18445.html at the bottom.

Hope this helps. Any other questions, email me through the link in my profile page.

David
1968 36/7988