5 replies to this topic

[Roger 912]

Some months back I changed the one ECU controller for another and have had some problems with fueling pre turbo.

On speaking to the installer of the ECU he says that there is too little vacuum (The ECU works on Absolute Manifold Pressure only). Having a look at the setup, I found that the inlet manifold had a balance pipe connected between ports and that the take offs to the ECU were from either end. The vacuum measured at the ECU was terrible, should be 20 inches of Mercury (HG) on idle and I could only measure 3 inches at idle. I tried one or two experiments and found that if I took off the vacuum from one port of the manifold only, the vacuum pulsed between 20 HG and Zero. I thought about how to damp this pulse and I fitted a spare Bosch fuel line filter in series with the pipe to act as a reservoir and this damped out the pulses but left me with the average of around 8 HG at the ECU. I am hoping that this will be enough, but one cannot fit a non return valve to build up vacuum as the ECU requires a swift change in vacuum with a change in throttle position.

With the ports connected with the balance pipe, I feel that the vacuum from whichever cylinder is in induction phase has to share its vacuum with the other inlet chambers in the inlet manifold causing the low vacuum.

Just wondering if anyone has some thoughts on this.

[bosbon_be]

On the Emerald ecu I have, I can change a filter factor(software) of the MAP input, so I can damp the pulsing.

[hilly]

Seeing as your ECU uses speed density to work out the fuelling requirement you really need to get a representative sample of the MAP level.

It can prove to be difficult on an engine which is fitted with individual runners and throttle bodies.
This because unlike on an engine with a plenum (which always has one cylinder pulling a vacuum on the chamber) each runner only has vacuum during the inlet stroke. This means that 75% of the time the vacuum is leaking away. If you were to use just one inlet for the ECU MAP connection you would end up with with a lumpy map signal at low RPM,which you have observed.

If the ECU MAP connection is linked across all the inlets what happen then is you still get the lumpy signal (all be it lower), but it is averaged across the 4 cylinders hence ending up with a much lower signal (remember each port is under vacuum for only 25% for each complete engine cycle) again which is what you are seeing.

The way I got around this was to use a MAP connection for each port, but to keep each pipe as long as possible before link together near the ECU via a manifold joint and then onto a small chamber, see picture below.


You may find this gets the signal good enough, if not adding a check valve (as you originally thought) and an adjustable bleed valve in parallel into the line between the chamber and the manifold joint may improve things further. Arrange the check valve so that it prevents flow out of the chamber, then use the adjustable bleed valve to control the rate at which the vacuum leaks back. Fiddle with the adjustable bleed valve until the vacuum recovery is fast enough to keep the MAP signal steady whilst still following the real vacuum level.

Sounds like a lot, but it isn't in reality. If you need more details let me know.

Hilly

[Roger 912]

Thanks for the input Hilly.

I have almost that at present with long takeoffs from No 1 and No 4 but still the vacuum is only 3 inches HG. Not enough to map properly.

My large metal fuel filter as a vacuum reservoir connected in the vacuum line to just to No. 1 inlet is giving me about 8 to 9 inches with a small amount of pulsing.

It is going to the tuners on Wednesday and we will have a look and see how many bars of tuning we can now do in the vacuum, pre boost.

I did try restricting the vacuum pipe to remove the pulsing and it worked very well, but the restriction was tiny, just like a pin prick in a solid plastic rod I mounted in the vacuum line. Perhaps using this restriction in all 4 pipes feeding the MAP sensor would do the trick.

Roger.

[Roger 912]

Took the car in yesterday to the installers and tuners and they are now going to modify the fuel injection by fitting a Throttle Position Sensor and then modify the ECU to use the TPS until the MAP input sees pressure from the Turbo when the MAP will then take over the fueling.

This is due to the low vacuum of having separate throttle bodies, a balance pipe sharing the vacuum and excessive pulsing.

Should have the car back within a couple of weeks.

[SubWoofer]

The Turbo Esprit should be a perfect candidate for a MAF based system? I don't know how many aftermarket systems support MAF yet, but it just came to Megasquirt (a happy user: http://www.msextra.c...p?f=131&t=40744 ). The OEM systems have been MAF based for at least two decades...