kpr's starlet

[Snoozin]

Too many valves. No one wants that.

[kpr]

20v's and strokers,  not the one 

[BiTurbo228]

There's an interesting thread on the Sideways Triumph forum called 'How to measure a cam in hand' where a sadly departed excellent chap by the name of Alan was measuring various different cams for old Triumphs using a lathe and dial gauge.

He found, much like you have, that the manufacturer numbers for cams have practically zero resemblance to the actual cam profile. Not even if you factor in the various different geometries for rockers, or tweak the numbers around to account for the fact that the follower rides up the cam non-linearly as it sweeps across it.

As far as I can tell, they're only there so punters can think 'bigger number better', and if anyone tries to grind their own camshaft to the paper specs they'll end up with scrap to make forgeries harder.

[xsspeed]

It's why 3/4 race cams are so lumpy

[kpr]

Yeah pretty much at the mercy of the camshaft manufacturer, unless can design your own. 
I've just gone full retard and tried countless cam combo's.  Have ran through almost every off the shelf kelford cam  and mixed and matched them. 
Every time i settle on a combo its always some weird mix and match deal that works best. rather than an off the shelf set.   

[xsspeed]

Digressing a little deeper, besides what's off the shelf and reprints, who actually makes cams from scratch to a spec. Is that a kelford or Franklin thing they do, thinking say for some engine that's otherwise got nothing available aftermarket wise

 

[kpr]

For something that doesn't exist.  I assume not many places would do something from scratch.  unless you had a big stack of cash.  or they can sell a bunch of them.   since will be a bit of r&d working with the valvetrain etc of said unknown engine. 

 

[xsspeed]

thats what i'm coming up with (this is only a hypothetical turd polishing excercise at this stage), I have enquired on some supposed "race cams" on alibaba lol

[Hyperblade]

These guys can do custom (not sure on costs but may be less then you think)

https://www.kennellycams.co.nz

 

[BiTurbo228]

From speaking to people mucking about with Rover 2600 engines (which practically no-one ever did anything with), cam manufacturers will fairly happily grind the profiles for other cams onto them.

I know at least one chap who's got a 'Piper 270' profile ground onto his, and another that raced one in a Group A tribute series that had a 'works Austin Healey' profile ground on.

[kpr]

Assuming they would be regrinds?   not new cores  cast or machined from billet  

[Hyperblade]

7 hours ago, kpr said:

Assuming they would be regrinds?   not new cores  cast or machined from billet  

Kennelly can do from billet.

[BobbyBreeze]

Kennelly likely won't do a custom profile without big money changing hands. PK does know how to design a lobe (I replaced him at Kelford), but he has no easy way of getting that design onto a cam easily. 
They may have done some of their own profiles, but only having manual grinding machines means designing the profile, wire cutting a single lobe, then cutting a master off that, then grinding your cam > not cheap.

Most of the time a cam shop like Franklin, Kennelly, Auckland etc will look at your factory lobe, and match it to a master they have on hand. 
Kelford are the only CNC grinders in NZ who can take a profile design straight from software to cam in one go.
They won't touch one-off's these days though unfortunately. Last I talked they said a batch of 10 minimum. 

My CNC grinder is still unfinished, silly me got too engaged in a casting side quest.. 

[BiTurbo228]

Ah that's why custom cam profiles are expensive. It's a tooling challenge. Rather than having a bunch of masters you can have your machine copy onto a workpiece you have to have a different machine that can create the master itself (and/or a bunch of iterations of it until it works). Either that or a steady hand on an angle grinder, as I've seen on various 'tutorial' shorts from places that major bigtime in backyard engineering (e.g. South East Asia).

On 29/11/2025 at 06:15, BobbyBreeze said:

My CNC grinder is still unfinished, silly me got too engaged in a casting side quest.. 

Bench grinder bolted to the bed of a 3D printer with a rotary table geared on the sleds with the cam in it?

[Roman]

Love seeing progress on this. 

so is the end goal to take the learnings from the bigport head and apply to one of the better ones later on? 

Imagine how good it would be if it worked out that chopping up silver top 20v heads and converting them to 16v worked out well. 

[kpr]

Yep,  once its proven itself for a little bit,  will get straight into doing the same to the big valve head that come off it.   hence why im testing it on this engine/car.  same deal with the manifold setup,  will go straight back on the big  valve head. 

At the moment its holding torque/cylinder fill pretty flat till just after 8k.    roughly should work out another 5kw  every 500rpm, if can make it hang on flat.    So if can get it to hang on till 9k would be amazing, but anything would be a win at this point    

[Roman]

Flat till after 8 is crazy! 

Look forward to seeing the end result

[Roman]

Gonna be awesome to see how this goes with the good motor. 

Even if it's a tiny fraction of the gains seen on the other one, it's going to be crazy

[kpr]

was one part of the port that was borderline being big enough for the valve throat,  so possible was a little more in the head as it was.  but was also not much left in the port walls, in the best place to take it out.  

 

[Rhyscar]

Welcome to the bendy trumpet club 😂😂. 
 

Loving seeing this coming together. Anticipation to see the results is project binky spec 🍿📺