mechanical fuel supply system and turbo for M10's

[Guest Anonymous]

I know the factory turbo cars all had a mechanical K-fisher pump that was tuned and enhanced for the KKK turbo system. So I am fairly positive this can be achieved today still without EFI or ECU controls like with an aftermarket device such as Megasquirt.

Some my say WHY? It is a hassle, hard to control, impossible to tune, etc... I say you are probably right, although there remains a certain sport to pulling it off.

I was thinking of doing this using the tii system and turbo plus a boost metering fuel delivery assistant. The Bosch K-Jetronic fuel distributor looks to be the perfect part.

I'm roughly figuring an electronic secondary pump at 45 + PSI and bypass return lines to the tank. The idea would be to tune the airmass meter in the K jet to operate only when the boost comes on. Otherwise the regular tii mech injection would run the motor at lower bar pressure and feed the air intake through with a little extra psi provided by the spooling turbo until the full boost in effect. My thought is this might overcome the fact that the normal air volume would be choked down as a result of the longer and narrower passage from the air filter. The whole system would be mechanically pumping as usual and only aided by the electronic K jet system under boost.

I have not been able to decide if tii injectors would need to be tuned for higher volume (possible?) or are there other Bosch injectors off the shelf that flow a higher rate available.

Another concern is to add a check valve in series from the K-Jet to the K-Fish at what point? Closet to the injector or at the beginning of flow at the K-fish pump?

Can or should, the plastic injector lines be tapped into with a splitter and interlaced with the secondary fuel supply line? In not, then how to run two injectors at the same port? Or is that redundantly stupid?

I have been reading a little about diesel engines having elaborate mechanical fuel pumps. Just wondering if there is anything there worth investigating further.

Any of you serious tuners have experience in turbo’s with non EFI turbo systems?

[snowflake]

Sounds very complicated, but here's what i think.

I installed a fuel pressure gauge in my m20 conversion on the dash.

Fuel pressure in ignition on position is exactly 55psi, i'm using a walbro 255.

i have yet to test running because i'm still waiting on some small parts, but it should remain constant.

I don't see the reason for a second set of injectors, you need to deliver more fuel through the first set once boost comes on(assuming around 10psi boost) this is the hard part, i'm not a k-fisher expert but what comes to mind is fuel delivery.

Meaning that when no boost is present the tii injectors deliver regular amount and when boost is on deliver more than usual.

Have you thought about incorporating a rising rate fuel pressure regulator into the system somehow? Maby right before the K-fisher unite?

Having 55psi at all times, The (RRFPR) would adjust and allow fuel to flow into k-fisher unite, according to vacum level in the intake manifold.

I don't know just an idea

[PatAllen]

Dont want to jump over snowflake but heck, k-fish pumps are well unknown by many...no offence.

It wont ever know you are pushing boost, since it only rely on rpm and throtle position.

and pushing 100psi of fuel through it wont cure it either since the k-fish pump build much more presure than that (around 500psi iirc) to feed the injectors (exactly like a diesel pump/distributor)...so in order to "help" the pump to throw more fuel youd need to pump like 800psi to it, not a good idea.

A

id say...

1-why not getting a real turbo k-fish system (with boost sensitive cone/mixture control)...not that much of a crazy idea...but still not realy "tunable" by tosays standard.

2-get a stand alone single injector controler to inject more fuel@boost. since you seems to plan to use modest boost, one may try to use the "5th" injector (cold start injector) as the fueling aid...

[snowflake]

Thanks for the correction pat,

i just remember the pump is belt driven off the crank.

What's a boost sensitive cone/mixture control? is this some kind of sensor you plug into the intake somewhere?

I'm just curious about this because in the 70's Formula 1 Turbo era, i don't think they had complex engine management systems,

just carbs, loads of lag and big power

i could be very wrong though.

[PatAllen]

the k-fish pump meter the fuel from a moving three dimensional cone.

as the rpm increase, the cone rotate and as the throtle open, it slides. So the shape of the cone dictates the fuel curve.

There is a lever on the cone and the height of the lever dictates the amount of fuel to inject, almost like a cis injection.

turbo 02 has a boost sensing diaphragm that raises the lever more according to boost, and is part of the heat sensing device to enrich under cold condition. the lever actualy control the stroke of 4 litle pistons that pump the fuel for each injector. So its all mechanical, nothing fancy or exotic here.

have a look at this...nice and clean.

http://www.bmw2002faq.com/content/view/66/32/

[tomorrowcalling]

A bit out in left field but,

If you changed the 3D cone to a 2D cone such that it only adjusted mixture as it moved in and out. Then if you connected the cone to a solenoid that was driven by a pulse width modulation scheme. Then it may be possible to use a MS to control the mixture by varying the cone position.

It's probably more of a research project than I'd like to tackle and you'd have to fit a map sensor, TPS, coolant sensor & rpm sensor but most of the changes wouldn't be that visible and maybe you could get it past tech inspection as still being "as fitted".

"How hard can it be?"

"What could possibly go wrong?"

[peteinjp]

A bit out in left field but,

If you changed the 3D cone to a 2D cone such that it only adjusted mixture as it moved in and out. Then if you connected the cone to a solenoid that was driven by a pulse width modulation scheme. Then it may be possible to use a MS to control the mixture by varying the cone position.

It's probably more of a research project than I'd like to tackle and you'd have to fit a map sensor, TPS, coolant sensor & rpm sensor but most of the changes wouldn't be that visible and maybe you could get it past tech inspection as still being "as fitted".

"How hard can it be?"

"What could possibly go wrong?"

This is exactly what I was thinking of doing on my car. One thing is that you would need a larger head and pistons for the kugelfischer if you were planning on getting tons of power out of the car. I kinda gave up on the idea for now because I have no experience in electronics and no idea where to start. My plan was to make a separate box that translated the signal to the injectors into a linear motion via a servo motor. IF anyone has any idea how I might approach this idea I'd love to hear it.

[florida2002m3]

sorry to bring up an old thread

just probing to see if anyone has transformed their tii to a turbo tii using a modified k-pump. From what ive found the only difference is the fuel cone on the tubo k-pump is setup to deliver more fuel under boost.

my question is: Is the boost advance vacuum controlled off the intake? Is it as simple as buying the cone with the advance mechanism? Has anyone found a place to pick this kit up, or have PN's?

this is gonna be for a low boost system similar to the original turbo.