Map Signal Source


On a suggestion from Ken Hill at FM, I moved the source of the MAP signal for the Link from the tee that is in the brake booster line to the line that feeds the FPR on the intake manifold. He suggested that this might provide for a smoother signal source. I have datalogged the car with the MAP sourced at the brake booster and then at the FPR port on the manifold. There is a marked difference in the smoothness of the MAP plot from the datalog. I have put up two of the graphs. One is the before and one is the after. You can find them at

I've been trying to persuade folks to do this for years.  With the map data variations reduced you can improve accel/decel fueling by reducing z700.  It works even better with the map sensor mounted on the intake manifold, but extending the wires is a bit of a pain.

There appears to be longitudinal (front-to-back) pulsations in the plenum, similar to those in the intake runners but at different frequencies.  The magnitude of the resulting pressure variations seems to be maximum at the rear of the plenum, and also maximum at the throttle plate when the throttle is mostly closed.  The middle of the plenum seems to be a minimum when the throttle plate is mostly closed, and still a good compromise when it's open.  I think that Mazda added an extra hose nipple at that point for the FPR map reference for a good reason.

> While on the subject of replumbing vacuum lines, does relocating the vac source for the BOV offer any advantage? I have a hot-side FM BOV, BTW, sourced to the brake booster.

It reacts to vacuum more slowly when sourced at that point because the brake booster check valve is often open during vacuum conditions.  The preferred source is the nipple close to the TB.  There may be two of them there, so make that the one you use has normal vacuum present when the TB is fully closed as well as when partially open.

>While I'm at it, I'm going to change everything. Come off the nipple near the throttle body into one of those one-in X three-out 3/16" plastic manifolds.  And then run a separate 3/16" line each to MAP sensor in Link piggyback, to FM BOV, and to boost gauge. Dropping to 1/8" line close to each.

It's not good to run the map sensor off of a tee when there's something else with a substantial vacuum volume also connected to the same nipple.  The bov's vacuum chamber volume is large enough to cause a vacuum signal delay.  But if the M2 doesn't have another vacuum nipple somewhere mid-manifold for the map sensor, then the front nipple even with the bov teed in is still better than the brake booster hose.

>M2 also has a vacuum nipple at the rear of the manifold. Should I put the BOV there?  Accuracy of the MAP signal is more important than the BOV? Correct?

The TB end should be better for the MAP sensor than the back end, and the back end is OK for the BOV as long as the nipple is not shared with the brake booster. 

>IIRC, Jess posted that he found that the vacuum line between the manifold and the fuel pressure regulator was steadier than other locations...
Just to clarify:
Should a tee be put in the line between the manifold and the pressure regulator solenoid valve, or in the line between the pressure regulator solenoid valve and the fuel pressure regulator?
Also, is it better (for a steady map signal) to use a tee with an ID of 1/16", or one with an ID of 1/8"??

Tee in very close to the manifold.  Bigger is better.

>Ray, you mean just for the tee, correct? The actual line itself should be small dia., correct?

If the vacuum source is reasonably "quiet" (free from induction pulsation) then a larger ID hose will reduce the delay induced by the hose itself.  But if you're stuck with a "noisy" vacuum source then a small ID hose will filter out some of the "noise".   The ideal is a "quiet" vacuum source and a short, larger ID hose.

>Any idea how "noisy" the PCV valve port is on the intake manifold?  That's where mine is currently.

No idea.  Look at the MAP reading in your datalogs (or on the keypad) during smooth, steady cruise and see if it fluctuates.  If it holds steady, change only +/- 1 occasionally, then it's a good spot.  And when you have found a good spot, change Z700 to a value of 2 to get better accel/decel response from a 1.6 (1.8's get a heads-up from the TPS).


Keep in mind that 10" of 1/8" ID vacuum line provides about the same signal
delay as 62" of 5/16" ID tubing or 90" of 3/8" tubing.

Tony Schreiber wrote:
> So the "best-possible-solution" is to mount the MAP sensor somewhere
> next to the intake manifold and use a short vacuum line from the
> intake (and by just-behind the throttle body, you mean the very first
> nipple? Not the one just behind that and off to the driver's side?)
> and lengthen the wiring need from the harness...
> I think I'll make that an after-I-start-the-car project.
> > FWIW, for Miata applications, I feed the MAP sensor using a 10 inch
> > long thick-walled 5/16" ID rubber hose plumbed into the nipple just
> > behind the throttle body.  Electromotive specifically warns against
> > using hose longer than 12 inches.  You'll find that most modern OEM
> > applications have the MAP sensor integrated *into* the manifold for
> > best response

Ideally the MAP sensor shold be connected to the intake manifold via
>> a very short hose, but the sensor wires aren't usually long enough to
>> allow the hose to be short.  That being the case, the next best thing
>> is to have most of the hose length be large ID, like the brake
>> booster hose.  Just keep the small ID section of hose from the brake
>> booster fitting to the sensor as short as possible without the sensor
>> being directly exposed to the exhaust heat.