Advice on Overdrive installation

[Nick Webster]

Mention has been made above regarding notes on an A type overdrive fitted by Keith Rumsey. Although I think it likely that modern thinking would select a smaller and more modern unit, one word about the A type. I have run a non Jowett car fitted with the A type for many years. When I transferred this overdrive to another car I sought out the absolute correct unit to replace it in the original car. I was surprised to find it dropped a lot of oil and realised that the earlier A type design did not have a rubber O ring on the operating lever shaft. Easy to put one on there said I and went away for a long weekend where I was rewarded with an overdrive that would not disengage. The drop out relies on gravity on this design and I had made things a lille bit too tight. It occurs to me that it would be as well to investigate any type of overdrive that has Jowett fitting potential to ensure you have the best version of it.

Which brings me neatly to an associated point. Overdrives must not remain engaged while in reverse gear or the uni-directional clutch will literally explode and potentially split the box. In a standard installation this is prevented by switches that must be fitted to the gearbox (which may also be configured to prevent low gear application where torque is excessive) and these are very similar to the Jowett reverse light switch, but of course not every car has this. It might be relevant to this thread to discuss switching and solenoid wiring?

On a ligfhter note, some Healey 100 cars had a centrifugal arrangement so that the overdrive cut in and out as speed dictated. I am not aware that this was universally popular owing to its unpredicability. Which also reminds me that I once had an overdrive failure at 112 mph in a car only capable of 105 without it. In the seconds before I could figure out what was going on and dip the clutch it had much the same effect as if I had pulled the handbrake on.

Nick

[Keith Clements]

John Blankley had his Javs flywheel bolts shear in Slovakia on a Marathon when the overdrive disengaged when on full chat. I stopped and told John the bumper bolts were the same. He managed to fix the issue by the side of the road and arrived at end of leg in Merano with the bumpers in the back seat..

[PJGD]

The previous graphs that I posted in this string plotted road speed against engine speed for a Jupiter with wide ratio gearbox, a 22% overdrive, and the standard rear axle ratio. Hopefully people find that useful, but this time I have plotted brake horsepower against road speed for the same drivetrain combination. This provides a different perspective because now one can potentially optimize ones gear change shift points for performance if that is your objective. For this graph, I am using the official Jowett power curve for the Jupiter which I recall Charles Grandfield saying was corrected to SAE standards - in other words, it was reliably accurate in its day. Of course, your engine may be different, perhaps better due to modern lubricants, or worse due to worn components and non-optimum settings.

One common strategy is to run the engine up to the peak power speed and then change up to the next gear as you accelerate. Conversely, when climbing a gradient the strategy is to lug the engine down to around peak torque, which for the Jupiter is around 50 BHP [3000 rev/min], and then change down. So from this graph you can read across and see the road speed for peak torque in each gear, and the same for peak power. Also, maximum fuel efficiency usually coincides with peak torque, and so one should strive to cruise in as high a gear as possible and ideally close to 3000 rev/min.

As before, if there is an interest in what the graph would look like with alternate ratios, let me know and I can replot.

Now I have to yet again post another correction to my Tractive Effort graph last seen in post 275; the title was in error because it indicated that the RAR was with the Javelin's 4.875:1 instead of the correct Jupiter 4.56:1. The graph has not changed otherwise since I did use the 4.56 ratio; only the title was wrong.

[Keith Clements]

Fantastic Philip.
I am not sure how much theory performance drivers use or how much 'seat of the pants' feel and experience is used. Real life driving has to be more predictive by considering changes of curve and incline as well as the other driver when overtaking. It is amazing how quickly modern gearboxes change gear compared to the venerable Jowett so the driver does not need to consider the lag and temporary loss of power. The overdrive helps here providing a much quicker power shift.

[PJGD]

Looking at the schematics posted earlier of how the overdrive unit works it occurs to me that there could be a business opportunity for someone who has the resources and ability to design electric motors. The Laycock unit generally works quite well but the usual high mileage failure mode is loss of hydraulic pressure due to wear, causing internal leakage and resulting in clutch slip. So, the opportunity is to remove all the hydraulic and clutch band mechanism and replace it with an electric motor that is connected to the sun wheel; this would change it from a two mode overdrive gearbox into an infinitely variable ratio gearbox similar to that used in the Toyota Prius.

For those of you who still have the "Flash" app on your computer [I don't], there is a very interesting animation of the Prius power-split device that can be found here: http://eahart.com/prius/psd/ Note that it has to be watched with Powerpoint in presentation mode. A power-split device is just a single stage epicyclic [or planetary] geartrain that has two power inputs (or three in the case of the Prius); the engine drives into the planet carrier hence the planet gears, the annulus or ring gear takes the power out, and the electric motor controlling the sun gear speed is able to vary the ratio. This can be seen in the nomogram in the animation where you can move the sliders to change the relative speeds and thus the ratio between engine and output speed.

If you arrange for all three elements to be running at the same speed, then this is essentially direct drive as everything goes around as one monolith.

If you now stop the sun wheel from turning which is what the clutch in the Laycock unit does, then you get the overdrive on the annulus output:

Now watch what happens when you drive the motor and sun wheel in the opposite direction - you get an even greater overdrive:

Finally, if you overspeed the sun wheel, it gives you a lower ratio or under drive if you like:

Of course, what I am suggesting is not at all simple since you need batteries and power electronic controls to make it all work, but it is undeniably a neat concept that would enhance the drivability of most cars, and avoid the need for a multi-gear gearbox.

[Keith Clements]

Also look at the unit at front of the new Murray T50 road car engine. It is a starter, and energy recovery device providing a beltless and chainless engine
https://www.racetechmag.com/2020/08/how ... percar-2//

[PJGD]

Looking through my library recently, I found that I had a Workshop Manual for the 1956 Sunbeam Rapier. I looking through the manual I discovered that it had a section on the service and overhaul of the Laycock De-Normanville D-Type overdrive unit. I have to say that in comparison to the Jowett manual, I found this Sunbeam version very detailed and thorough. I scanned the relevant pages on the OD and have included them here. Although this describes the D-Type unit, the later units are very similar in the way they function.

I notice that at the top of page 12, they state that the speed increase is 32.2% which equals 0.765; in fact this is a misprint and it should be a ratio of 0.756:1. I see now that I made a similar mistake with the ratios in the graphs that I posted before for the Jupiter and Javelin, so there is a small error which I will correct in due course.

Laycock D-Type Overdrive Workshop Manual .pdf

(13.37 MiB) Downloaded 23 times

[Mike Allfrey]

Hi Keith and Others,

All very interesting!

Back in my misspent youth I used a Laycock overdrive in my Jupiter as a mid-ship bearing. The unit came out of a scrapped 1950s Humber Hawk. The trial installation did not work too well, but it got us to the Harrogate rally in 1967 or 1968. It came to Australia in the car and was in use for a few years.

Then, on my second rebuild, I decided to go back to standard. However, once back on the road I discovered considerable slack in the differential and went to our local rear axle 'specialist'. His advice was that there is a Morgan differential with the same ratio as the Jupiter's - but not many spares lying around. The suggestion was made that a Borg Warner differential (4.11:1 ratio) would do the job. Such a differential came out of a local Ford Falcon utility (pickup). This we undertook and now second and third gears are superb for local work and hill climbing. All good fun.

The overdrive went into a Bolwell, a Holden based GT car.

Stay well,

Mike A.