High revving 3.2: what is needed?

[911hillclimber]

Good detail as ever Chris, thank you for posting it.
As you can see from the damaged parts, the rod bolts are in one piece. The rod itself was so bent that it was 'wrapped' round a case web and I could partly open the cases, but not split them at the time. Took a lot of messing about to undo the road end to free the cases.

Bob W could not give a reason for the failure, the other 5 cylinders were fine for #6 was totalled, as were 5 rockers and ALL the valves bent.
Shards of aluminium were throughout the engine and oil system, this must have been the piston skirt.

Not an episode I wish to repeat.

I think I'm resigned to simply leave the car alone, and if a rebuild is indeed needed this year (I am expecting a bearing failure) then I will fit 964 cams to add a small amount of sparkle.
Should I do this I'll send you a rod bolt!

While I have some viewers:
Carbs.
The PMO's on the car are 50mm and Bruce Anderson's book calculates a 40 or 46mm size for the 3.2, 50mm for 3.5 litre and twin plugs and bigger capacities.
Being a hill climb car WOT is the order of the day, but also WOT from quite low rpm, sometimes as low as 3000 rpm.

The car seems to cope well with this but I wonder if more can come from the engine with smaller carbs?
Bob coaxed 247 bhp from this engine.

[fetuhoe]

I really don't think having carbs that are 'too large' is much of an issue.

It may influence initial throttle response slightly as the column of air you need to accelerate has more inertia but this should only affect initial response at low engine speeds and not a car being driven mostly at WOT.

As long as the venturis will allow sufficient airflow and the AFR is OK I wouldn't worry.

I would think that you need a venturi of around 38mm for a 3.2 running at 6000 to 6500rpm and I always disagree with our American colleagues who think you multiply the venturi diameter by 1.1 to arrive at the barrel size of the carb.

I believe you need to multiply by 1.25 for an optimum result and this gives a carb with a 48mm barrel size.

The closer the ratio of venturi to barrel size the nearer you are to having a simple tube which reduces the venturi effect and hence the manner and the point at which fuel will be drawn from the main jet and when the two sizes are close it will effect bottom end behaviour.

All of the formulae used to calculate carb sizes are very approximate so a 50mm diameter would be well within an acceptable tolerance band.

[911hillclimber]

Thank you again.
Bob seemed ok at the time, he had far more issues with the Aldon Automotive ignition and threw it out for conventional ignition and got the results.
Throttle response is very sharp, helped by that excellent Fenn Lane Motorsport flyweight flywheel!
Venturies are 38mm.

Terry Davison ran this engine in his Nomad sports racer on 46mm Webers that were snaffled off the engine the day before I bought it...
It came with 40mm Webers that I could never get right, so swopped them for these 50mm PMO's.

[PeterK]

How about converting to CDI. By using one of Jonny Hart’s boxes you can programme / optimise your advance curve, plus set both soft & hard rev limits (doesn't help much with mechanical over revs though ).
If you accidentally recalibrated the rev counter, it might read 3.5k instead of 4k during the noise checks
Peter



Sent from my iPhone using Tapatalk

[911hillclimber]

Going for an engine that will tolerate mechanical over revving would be the 2nd advantage to this conumbrum, the first being able to hang on to a gear rather than change then very soon after changing back,
Hill climbing is a very tight affair and there can be lots to do in a very short period and get it wrong...

On Impact Bumpers, someone has suggested fitting a WEVO gate into the box to help define the gate.
This was something Mike Bainbridge doubted would really help, so not sure.
I've driven a 911 with a we've equipped 915 on a hill and it was very very precise.

[fetuhoe]

Graham,

The WEVO Internal Gate Kit partially replicates the internal gate used in a G50 Box and traps the selector into a fixed H Pattern.

Unlike the G50, however, it doesn't have any springs to locate the selector and control the movement of the gear stick.

When you use this kit in conjunction with the WEVO Shifter it does work well and we first used set up in a Rally Car we built in 2005. It is the shifter that controls the 'feel' of the stick and keeps it centred in the gate.

Without the Shifter it will still be difficult to define the side to side movement of the gear stick to ensure it positions correctly in this plane but it will define fore and aft movement.

[911hillclimber]

Then it will not do me much good.
The car has no gate in the gear lever at all so I need to guide it in with a delicat feel, you can't just thrust the lever to the next gear position...

[wadsworthi]

Graham,

I have run the Wevo shifter along with the Wevo PSJ coupler and billet clamp for about 5 years on my Hillclimb car and it is brilliant. My son used to struggle with the 915 box, but has had no issues since. They are expensive, but do work providing the box is good to start with.

Ian

[andytat]

Thinking of the gearshift, but maybe you already have done this.
Would it be possible to fit a kind of blanking mechanism that prevents
the gear lever moving into the 5th/rev plane? This would allow you to
really go for the 2nd to 3rd change without thinking you might hit 5th
and loose momentum.
The plate would need to be able to move if you wanted to select reverse
but shouldn't be too difficult to engineer.

Andy

[911hillclimber]

Andy, have that already.
This is a tough spring that takes quite a force to compress using the side motion of the stubby lever to allow the access to the 5th/Rev plane.
There is no 5th gear in the box, Mike removed it as it is not needed in the car so 5th slot is not there, just right and back for reverse.

The grave danger is slotting 1st from 'red-line in 2nd' instead of 3rd. This box will engage 1st at that speed easy, that resulted in the Big Bang a few years ago and I've nipped it a few times since.

3rd to 4th is good, 4th to 3rd is good, 3rd to 2nd is bulky or you go back to 4th in the attempt.

If I take my time the box will behave, but that is changing gear slowly and I loose hill time every change which can lead to a poor climb time.
A totally fluffed change will loose a good 1 to 2 seconds!

There is no gate defined by the lever, I only have the definition inherent in the gear positions in the box itself to work with.

[andytat]

I can see your problem now Graham.
There was a chap who posted on pelican parts forum some years ago
who raced a 911 with a 915 box and he converted 1st to 4th to dogs
in place of the synchros. I just had a quick look but cannot find the
post. Would that be a possibility?

Andy

[fetuhoe]

Wevo used to partially convert to dogs but no longer offer this conversion but there is an Australian based company that supply Tuthill with Dog boxes at around £16K they still use the basic 915 box and haven't changed the selector design.

The shift issue, however, is not really associated with the synchromesh which is just a bit on the slow side the problem is that there is NO gate definition inside either a 901 or a 915 transmission.

The gate is defined entirely by the shiter and the springs inside the gear lever mechanism.

Changing to dogs won't influence the lack of gate definition which is the real cause of Graham's shifting problems.

Installing dogs will improve the speed of the shift but won't impact on selecting the wrong gear.

The G50 uses a shifter which has no springs and is quite 'floppy' ( ) with the gate definition and springs being installed inside the box so selection is much improved.

The Wevo shifter does work well as it improves the spring system used in the shift tower and in a 911 is worthwhile but I don't believe that Graham can install a shift tower in his Lola.

A box with a sequential shift or a box with a G50 style selector system is the only real way to solve this problem and if you look at a Hewland LD200 4 or 5 Speed Box you can see the difference in the selector arrangement compared to a 915 and these boxes work well with the type of shifter fitted to the Lola.

[christaylorsound]

Graham, sorry to side-swipe your thread, but it has got me thinking..... Next week my newly built 3.4 gets dyno tested and tuned. It has a Canems ECU so the rev limiter can be set anywhere I choose, but where should I set it?
Spec is Std 3.2 crank and rods with ARP rod bolts. Bored to 98mm with JE 10.5:1 pistons. 993SS cams and race valve springs, std valves, std retainers. I also have a 993 inlet and SSIs but the stuff above is what should determine the safe rev limit.

So, what should I set it too?

Thanks, Chris.

[911hillclimber]

If specialist built, ask this of the builder!
If diy hen 6400, the stock red line.
Start safe!

As to the gear shift. This issue has haunted me for years, ever since building this car up.
Due to issues, one box rebuild lasted 17 gearchanges, yes, 17 operations.
Destroyed synchro rings resulted in 110 pieces of steel in the box.

Chris has the situationsussed well.
I have a totally un defined H pattern in the lever and only the 915 box internals to feel for all while diving to a corner and braking for dear life.

The car system is typical of its era and with a Hewland all would be well, but you try justifying the cost and getting the adaptor made et.
It seems nobody has ever put a classic Hewland onto an air cooled engine.

The closest I came to a fix is to adapt a Subaru gearbox, itself converted to 2 wheel drive. I have good experience of these boxes do they shift well and fast and are bomb proof.
They are big due to the clutch diameter and the starter motor is all wrong, then there are the drive shafts. I have one in my garage in perfect condition.

The AUDI gearbox from a 1.8 turbo is a good box, the same as the early Boxster box, but the shift input shaft is on the wrong side!

Paul Howells used to have a rare 915 box with sequential dog gears in his RSR hill climb car, I think this is one of 2 ever made.

The Audi boxes are dirt cheap and a rough one could be bought to be cut and shut to sort the engineering out next Xmas.
Nothing with this car is simple.

[Bruce M]

Make a shifter gate or remote gate (close to shifter) with adjustable stops?
That would be an interesting project.

Engine wise... modest turbo or spend big for revs.