Britbike forum Forums British Motorcycle specific Brands Bsa bulletin board The what if A65 67hp challenge.

Thanks John.
I've recently stripped it to sort out barkes and check the frame over etc. The head brace is being moved as the old featherbeds need one.
They like to crack around there. New suspension is also on the cards.
Just a note to say the 44mm 'something' valves Mark mentions are trimmed Velo ones similar to ones i used a while ago.

Hi Nick

Thanks for the information on the source of the 44 something valves, worth adding to the note book for future reference.

Thanks

John

I am utterly amazed by your A65's output figures. 97 BHP and 71 ft/lb is incredible, as is 110 bhp/litre. If these figures are possible from and A65, how much might be possible from a triple?

What have you done to the bottom end to make it hold together under that kind of punishment? You must have made very significant main bearing & oiling changes?

44mm valves with nominal 5/16" stems:
Volvo 2.1L 8 valve '76-'85 1.732" dia 4.35" long 0.3135 stem
Volvo 2.3L 8 valve '85-'95 1.732" dia 4.35" long 0.3135 stem
GM 2.2L 8 valve '90-'02 1.732" dia 5.457" long 0.3136" stem
Toyota 2.8L 12 valve '82-'88 1.732" dia 4.235" long 0.3138" stem
Nissan 2.8L 12 valve '75-'83 1.732" dia 4.500" long 0.3139" stem

Has anyone thought about using valves with slimmer stems? Or is this impractical because of the size of the head of the valve (38mm and up) increasing the chance of the valve getting bent because a smaller diamter stem will be too weak? I know some engineers will make their own valve guides as opposed to buying them in. It will reduce a lot of obstruction in the direct flow of the port....

Race engines often have the inlet guides cut flush to the port roof

Hi All


Alan, Have you looked at "Nucleus" valves ? they have a smaller diameter section between the guide end and valve head
I fitted a set to my A10 Super Rocket and have done 25k + miles trouble free without removing the head
It goes really well and sounds sharp
Valve clearance settled down after about 5k. inlets gaps were closing up a touch until then
Dave Degens is an advocate of thin stemmed valves, I seem to remember him talking about 5mm stems

john

The Jaguar 44.6mm valves have a slight stem reduction between guide and seat. Guides can be shortened a mm or two but cutting them off flush with the port wouldn't be good because they are already pretty short. The valves need to land on the seats. Finding a shop that can fit Kliners is worthwhile rather than replacing guides.

Most of the high performance valves sold now have a reduced stem for 20mm or so before the head.
Black Diamond, Nucleus etc. They run standard 5/16 guides. I've tried the old 'cut-off-the-guide' trick
but with the ludicrous cams i used the shortened guides lasted about 30 minutes before being clapped.
That may be worth while on an engine with better valve geometry or with more sensible cams. Standard
guides had a short enough life anyway. We used to make our own guides from Albronze at one time but
then went to either CHS colisbro or PM ones as they were as good if not better, just had to turn the tops
of the inlets for seals. Never tried 6mm stems for the same wear related reasons.
The reason for using those Velo valves was because my older brother bought a 40ft container full of gear
and amongst that was a box of valves, some of which were those ones, so we used them. They were a
good shape and seemed to last well although we had to trim them and cut collett grooves to suit.
There was no scientific reason like a lot of the stuff we did when racing, time and budget was important.

Just repeating this in case it wasn't noticed first time.

The lower end of Marks' bike is an offset norton crank which is end fed.
I ran a similar lower end on my race bike, that was getting 80+bhp at the rear wheel.
Whilst these would probably not make a touring bike in terms of vibration etc they
can go very well and be reliable. Mark has progressed with cylinder head work and
has approached the flow side with more emphasis than tuners years ago when these
motors were being used on the track. It shows that with later techniques the engine
potential was huge. To achieve similar things the factory would have needed to have
been interested in racing, which it wasn't.

There were several triples out on the track in the 90's that were producing over 90bhp
as were some of the the 900 weslake twins. Classic race regulations restricted them somewhat
in the uk though. A few of the imp motors could get well over 100. This is all on petrol and
using regulation carb sizes etc.

Thanks NickL - it would be very interesting to see an in-depth illustrated article on engines like these. The timing side main bearing must, presumably, have been very much altered in Mark's A65, I'm guessing?

Unless I’m mistaken, he’s using an RNA4907 needle roller with an outrigger on the drive side. The 4907 is what I’m using on mine, the common end feed conversion used a NKIB5907. The 4907 is similar but isn’t annular contact and doesn’t control the end float. The bearing width is narrower to accommodate this.

NKIS 30 I think it is, 30x52x22, I haven't had it apart in a long time, these bearings don't seem to wear out. I use one with c3 clearance so it can be tighter in the case. The Norton size drive side can wear over a long period.

100 HP per 1000 cc's is no longer the benchmark for highly modified 2 valve engines as Mark has shown....5700 cc Nascar OHV V8's make in excess of 825 hp at 9000 rpm.Without Nascar imposed restrictions, 950 HP ....$3000-4000 of bolt on parts will get 120 HP at the rear wheel on Evo 1200 Harley Sportsters that can be street ridden with decent reliability..Many modern 4 valve sport bikes with full emissions equipment get over 200 HP per 1000cc's....Hell, Mark could probably squeeze 67 HP from a rattly Triumph...

The 883 with the 51mm mufflers. Trying to illustrate the midrange and bottom end this engine has. What I like is the acceleration it has in a higher gear at lower rpm in corners, especially ones where you cannot use full power, or at least I cannot because I'm not Troy Baylis. This is with stock primary ratio and 21-36t final drive and C/r 5speed. It would be nice to experiment on a circuit where no one minds how fast you go. But my guess is there may be better drive and traction banked over if rpm is lower. Like if a long corner required you to be banked over and accelerating between 55 and 120mph, it may be better just in 3rd, rather than using 2nd and 3rd.

https://www.youtube.com/watch?v=ZTvWCw33BAw&feature=youtu.be

When i started using the big offset lump it took a while to get used to changing up as early as we did but it did improve lap times.
After using a buzzy short stroke 750 the drive out of corners was bloody fantastic, 6800 was the bloodline really, a lot different to the
8000 of the 750 parallel short stroke motor. Red mist sometimes meant 7200 but it didn't make us go much faster. So i totally agree
with you there Mark.

That's what i like about the early 90's 900 2 valve belt cam drive Ducaits...They will rev freely to 9000 but theres that abundent hydraulic torque from 4000 on up so 7500 rpm is good enough...They have about 75 RWHP in stock tune but it puches like 100 HP
If I had 3000 bucks to throw around I would do a 76 degree crank in my T140...

So I had some time to get back onto my Thunderbolt head yesterday and this morning for a little while. So I just hacked out the R/H port thinking I'll just do it all including building up the floor and patching where it breaks through into the pushrod tunnel, except this time I JBwelded a thicker plate in there so I could go wider. Its wider by maybe 1.5 or 2mm. I didn't test it till it was finished. And straight away I have 167cfm @ .350" with the squashed AMAL trumpet bell mouth. This bell isn't as good as the manifold plus short bell. And I need to finish the manifold.

For a factory it would be simple to change the casting slightly and do these ports by hand. I was interested to see if I could get a good result just copying the first port and relying on the fact it was flowing well, so that can be done successfully, and surprisingly very easily compared to the struggle of doing the first. It becomes quite a different shape to the BSA port.

Widening the R/H port a little more into the pushrod tunnel improves the flow and the right port will probably end up over 170cfm. So I'll have to try coaxing a bit more from the left.

For interest this is the 750 90degree, A65/RGV, the head on this has bigger valves and carbs flowing around 190cfm about 10 more than modern Triumph 8valves.

I think at one stage he takes it to around 8,000rpm.

https://www.youtube.com/watch?v=8OO...Jdd9E9HhSRfsHOZs-xvOHD4u4UdmL5bad24JcuyQ

Me this morning having a conversation with myself I welded the little manifold so I could bolt it on, yesterday I'd held it in place and it looked like I'd have over 170cfm. So the oval port in the head is a little smaller than the oval end of the manifold. It was only a couple of mm why not chew the head out and match it rather than filling in the manifold? But that's flowing great I don't want to touch it! But it's just a little bit and it will be bigger and should flow more. But it changes the shape and makes more of a bend. Just blend it. But the air will be aimed across the port. But this is easier, what can it do?

Found out, it can lose more than 10cfm. If I'd left the step and done nothing but bolt it on it would have had 170cfm or more, now it's 162 and I'm waiting for JBweld to set.

What I've learned, don't listen to idiots. Second thing; fast moving air is heavy and doesn't like any extra curves to turn, esp curves that might send the air across air going down the other side. The port has to turn onto the valve, so we are stuck with that curve and can make it smooth as possible, and the manifold needs to place the carb where it needs to be and curve a little to connect but it needs a smooth connect with no compound curves of compromises. And no variation in cross section. This became like a bulge in one side of the port that is unnecessary just for the sake of matching manifold to head rather than putting a small patch of JBweld in the manifold to do the same thing.

Fixed it enough to get [email protected]" and just need to fix a couple of small irregularities by filling rather than taking it out bigger. If I put the carbs on longer manifolds they could be straighter. Though the Thunderbolt ports have a bias, which may be a benefit if it creates swirl.

Edit: Smoothed it and retested; showing 170cfm @ .400 and 160cfm through the carb after checking the calibration. The difference is the width into the pushrod tunnel. Very cool. 85hp should be possible from a high compression 750 with a good exhaust.



Welding metal into the pushrod tunnel would mean no JBweld except for raising the floor and if it breaks through between the fins, both low stress areas.

Using JBweld on its own on a break into the pushrod tunnel results in it cracking, sucks in oil, but JBwelding an alloy plate right over the break through is durable, but welding should be possible.

Next experiment will be a 34mm Lightning head using the same oval type entry and having the manifold transition from round to oval like this one. The 38mm port on the 883 and 750 use that set up but I haven't done a smaller port on the Lightning head that way yet and it allows a better alignment and shape.

Oko carb heresy. Did you make a plot of port area versus length? Or have any idea what it is? What size carbs are those?

Carbs are 34s, the ports are about an inch longer than a Lightning, I'll measure the areas sometime and take off the extra length for the comparison.

I was just testing a stock lightning head and it's a bit of a wakeup call after stressing over coaxing a couple of cfm out of the Thunderbolt's 167-170cfm ports; So stock Lightning is 109.4cfm bare, 112.6 with bell on the port, 100.9 with 30mm Concentric, 102.7 with 30mm Concentric with the intake blended, on my gear. So this head with carbs is about +60% in flow. The port shape is quite different and I probably need to leave it alone and just prepare it for use.

These are nice looking heretic carbs, VM Mikuni s do not look much like Amals and are bulky (though they work well) these look a bit more like the compact AMAL. Just need to find out how they jet up and function.

Mark, I'm sure many of us who have been following this from the beginning would appreciate if you could sum up your findings in a practical way to benefit the A-65 owner who wants to do some porting improvement.
When the testing has reached it's conclusion that is.
Some of this is over my head, I hate to say.

All the Keihin PWK tuning parts fit the Oko. The JRC site has some tuning information and I think CBS. Heresy is a good alternative.