Manifold Design and Turbine Housing - expensive lesson learnt

[GTS-t VSPEC]

Ben, I'm running 17psi too, with overly rich fuel and little timing. But forged pistons will be going in soon.

Steve, Can you get the manifold flow tested? Be interesting to see if it is causing interference.

[Warpspeed]

Ok here is how to degree your cams.

The first and very important step is to know exactly where top dead centre is. The factory mark on the crank pulley may be o/k and it may not. But checking this is a vital first step.

Now you cannot just measure piston movement around top dead centre through the spark plug hole, because the piston slows down, stops, then changes direction. Over a very few degrees around top dead centre, piston motion is minute and unmeasurable, and things like con-rod bearing clearance can make it more inaccurate anyway.

So what you do is make yourself a mechanical metal stop that you can screw into the spark plug hole that stops the piston about half way up the stroke. This might be a bolt epoxied into a stuffed spark plug base, after you have removed all the ceramic. It does not have to be exact length, but somewhere around half piston stroke will work fine.

You then attach the degree wheel to the end of the crank pulley somehow. Strong magnets, double sided tape, bolts, children's silly putty, whatever. The zero degree TDC mark on the degree wheel should roughly line up with the factory TDC mark on the crank pulley. Fix a pointer to the engine made from coat hanger wire, or welding wire, so it points exactly at TDC on the degree wheel.

Next you slowly and carefully rotate the engine with a long spanner through BDC until the piston comes up firmly against the stop (gently does it). Note the reading on the degree wheel. It might be 98 degrees, or some other reading.

Then gently rotate the engine back the other way through BDC until the piston again goes up firmly against the stop. The reading might for example be 88 degrees.

Split the difference, and bend the wire pointer to 93 degrees (in this example).

Go back and re check a few times to ensure that the degree reading when the piston hits the stop is identical both ways. You can now be certain TDC on your degree wheel corresponds exactly to TDC at the crankshaft.

Next you make a fixture to hold the dial indicator against the flat surface of the hydraulic cam follower. This may take a bit of ingenuity, because the cam lobes always get in the way.

The end of the dial indicator will have a small ball that screws in. If you unscrew this you can fit a flat piece of metal drilled with a small hole, held in place by screwing back the ball. The flat metal sticks out at right-angles like a foot from a leg.

This flat section can rest on top of the cam follower at the side of the cam lobe, and the lobe will not strike the shaft of the dial indicator as it goes past.

Mount the dial indicator rigidly to the cylinder head, it must be parallel to the valve stem for accurate results. A magnetic stand works o/k, or you can make up something better and more rigid that bolts onto one of the cam cover studs.

Turn the engine so the valve is shut, and zero the dial indicator. Slowly turn the engine until the valve just begins to move. With a hydraulic cam, movement will be very sudden. You may have to do it a few times to get the exact spot. You can then read the exact valve opening point on your degree wheel.

Valve closing is very similar. You pick the spot where the valve just seats, and your dial indicator returns to zero.

It is not difficult, but you need patience and practice, and setting it up for the first time can be bothersome. But once you have all your bits of metal prepared, pointer, and a method worked out, it can be repeated pretty quickly.

You know the readings are accurate if they repeat.

What surprises me is how far the cams can move when you increase or decrease the timing belt tension. Also do not be surprised to find the factory cams five or more degrees away from the proper factory spec.

Knowing where the cams really are is vital, do not trust those factory dots either, they are only supposed to get you within one whole tooth.

If the cams are miles out, or unsuitable, moving them around blindly is not going to help you much. When you have everything pretty well optimum, small changes can be easily detected by how the engine goes.

[Cubes]

Your tha man Warpspeed..

[Steve]

Paul, not likely to happen in a hurry:) flow testing wont prove much in the way of interference will it? Usually just test 1 runner at a time?

Warpspeed, thanks for the novel:) I dont really have much in the way of fabrication facilities at home, so I cant see myself doing more than a very innacurate version of the above - would there be much point?

Cheers

[Warpspeed]

It can be extremely accurate !

There are a few things to watch though. First make sure the degree wheel is centered properly on the crank pulley. If it wobbles around when you turn the engine, TDC and BDC may be correct, but the ninety degree points can be a fair bit out. Turn the engine and make sure the edge of the degree wheel runs true under the wire pointer.

Also make sure the dial indicator is parallel with the valve stem. This can be judged by eye. And lastly make sure the dial indicator is solidly mounted.

A scrap of aluminium or steel angle with a couple of holes makes an excellent dial indicator mounting.

The first time you do this it will probably take most of a day to get set up. After that you should be able to get accurate valve opening and closing points repeatable within one or two degrees in typically an hour.

Plug the numbers into Dyno 2003, and see what you get. Try moving things around in the software and see the results instantly in the torque and power curves on the screen. Then go and do it on the engine. The software does not lie. I think you will be truly impressed, and you will have done it all yourself as well.

[Steve]

Thanks again Warpspeed - its a shame you live on the east coast:)

[Steve]

Sydneykid, runner length, as best I could measure is 450ish mm long (hard to measure), with a diameter of 35mm.

Whilst this is alot longer than a street type manifold that the Japs make, where they blend the runners into each other, it also looks somewhat like a plate of spaghetti to maintain the length. I dont think the lenght of itself is the main cause, just a contributing factor as longer runners will make for a more gradual build on to boost.

Where the runners go 6 into 1 they meet at an angle of about 50 odd degrees, but meet only about 30mm before the turbine flange - which is where I believe most of the problems are. This is very different from most jap designed stuff I have seen.

[RS500]

i am of the belief that longer runners in an exhaust manifold will actually make for a much more aggressive build in boost.

this was seen when Ford/Cosworth desinged the longer 2 piece exhaust manifold for the Ford Sierra Cosworth RS500 back in '87.

maybe Sydneykid can confirm/deny this.

only my 2c.

[GTR-Ben]

Do any other companys make twin scroll turbo's apart from HKS? I know Trust has the T88 etc, I need something smaller.

Need some specs of some of these turbos.

[Warpspeed]

All the whole Garrett T04 range optionally come in split scroll, the N flow being the smallest. Toyota have a fairly small split scroll turbo on the MR2. There is plenty of choice available.

[GTR-Ben]

I think the T04S would be a good size for what I need.

Still priced up near $3000 and non ball bearing.

[FR33ZE]

so does anybody have any idea where i can import one of these or get one made up and about how much am i looking at........

[DICKY]

INTERESTING THREAD GUY'S

so if I use a t88 manifold (designed for a twin twin scroll turbo) as a donor to modify & mount a t51r (single entry) am I going to be robing myself of horse power ???????????

I would love a few opinions on this as a dedicated hks manifold for the t51 is over 4 grand ? & the budget is already ruined !

[Steve]

DICKY, I dont see it would be a major problem, the stock RB25 manifold has a split outlet. I think the main thing is to have the gasses merged properly before they hit the turbine housing.

[ylwgtr2]

i would be dialling in the cams to the manufacturers specs first as warpspeed explained.However i think you will find that the readings you got with the cams will be at 1mm lift so you will have to let the valve push 1mm untill you take your reading from the degree wheel(make sure you take note where the small pointer is on the dial indicator as you will need to have it in the same spot on closing so you know where 1mm before closing is!!!)

[Sydneykid]

i would be dialling in the cams to the manufacturers specs first as warpspeed explained.

OK I give in, Warpspeed! Can you please explain how I would get a different answer than if I install the camshafts with the standard pulleys at TDC and the crank at TDC. Surely if I do that, "the cams are dialled to the manufacture's specs". Now if the manufacturer ground them wrong ie; the TDC is not really at TDC, then I don't see how dialling them in makes any difference, they are still wrong. Or have I missed something?

If I then fit a pair of adj pulleys and tune the camshaft timing for the power curve I want, then I automatically fix any "wrong" timing by the manufacturer. Don't I?

Since I have to do that to get the best power curve, then haven't I wasted my time dialling them in in the first place? After all I am going to change it anyway.

Maybe I am missing something, help me out here...........

[Warpspeed]

He-he, hello again Sydneykid.

I am sure the cams and all the pulleys are manufactured extremely accurately, so the stock cam timing SHOULD be spot on, but it never is.

The problem is the cam belt.

There will always be some permanent stretch in an old belt, and even a new belt is going to stretch a bit as you tension up the adjuster. Now the adjuster is on the slack side, so as you take up a bit of extra belt length on the slack side, this will retard both cams with respect to the crank. And it does not take much stretch to shift the timing by rather a lot.

Think about it. One full turn of the cam is 720 Crankshaft degrees, and the cam pulley has 48 teeth. That is 15 crank degrees per tooth. The belt pitch is 8mm per tooth.

There is a fairly long length of belt between the crank and the exhaust cam (not sure exactly how much) and if it stretches only 4mm, that would retard both cams by a huge 7.5 degrees.

So you might have a situation where you have four different dead stock engines, but the individual cam timings all over the place by a few degrees because of different belt ages, different non Nissan belts being fitted, and different belt tension settings. And what about head gasket thickness..... ?

So everyone retards their exhaust cam by exactly four degrees. Two guys get a power increase ) one guy says it makes no difference :0/ and the fourth guy gets a slight power drop ( I wonder why ?

And why do Nissan bother to make the cam position sensor adjustable if the cam position can NEVER vary away from stock?

I prefer to set my cams up using a degree wheel and check the exact opening and closing points with a dial indicator on the cam followers. Adjustable cam pulleys allow you to always put it back to what it was when you did that really fantastic dyno run. If you do not do this, in reality you have no clue where those cams actually are, and after a major rebuild there is no chance it will go back together exactly the same.

I just believe there is no substitute for measuring and checking things, and if anything is going to change, that very hard working rubber timing belt should be treated with the greatest suspicion.

If you really heave down on the belt tension adjustment, you can actually see both cams move slightly !......... Try it and see.

While I am sure YOU do all this, a lot of guys just fit the variable pulleys, and back off the exxhaust cam from the marked line, not bothering to actually measure anything. All they really care about is the dyno figure.

[GTRgeoff]

But don't leave it over tensioned when you run it. Makes a nasty whizzing noise.

Gotta support Warpspeed here, as this is one area of engine building that is often overlooked and is critical to maximising torque, all other elements being 100%. Once you have determined the crank/cam relationship you can maintain the optimal angles throughout the engine life, if you are that keen to do so. Really only needs to be done once or twice though for streeting as the belt stretch should not be enough to be critical in itself, but you do need to establish the baseline. Possibly just as easy to spend a little longer on the dyno and compare some variations on a common tune, then fine tune the best torque setup.

If the belt is stretching that much, or you are chasing enough power to really be at the sharp edge, it's likely more frequent belt changes are warranted.

[driftaR32]

hey all

im coming in a little late i now, for that im sorry

im looking to purchase and install a new HKS 3037PRO s with internal gate and .87 rear end

im really hoping for gull booost by around 4000-4500 rpm on an rb20det

other modifications will be power fc, gtr injectors, and obvious support systems (fmic exhuast etc)

i just wanna know...am i dreaming??

thanks in advance

[b0neR]

Alot would depend on manifold design and the quality of tune but it does seem like a bit of an ask.