Cause of fluctuating MAP signal ?
 

Hi All,

Wondering what is causing my MAP signal to fluctuate like it is in the picture of a third gear pull below ? MAP is in blue, RPM in red, throttle % in yellow, AFR in green, AFR target in white, purple is Lambda FB %.

http://forums.pelicanparts.com/uploa...1567950874.PNG

It seems clean up to about 5000 RPM then starts to fluctuate at greater RPM after that. Wastegate is set at 7 psi, turbine housing AR is 0.63, stock 3.2 heat exchangers and 930 J Pipe and zork, if that helps. Does this have something to do with exhaust back pressure ?

Thanks

Sandeep

What's your recirculation/BOV setup? BCV? Wastegate?

Looks likely the spring alone may be too weak to hold shut and is bobbling around at higher gas flow/RPM - opening under pressure/dumping boost, closing again.

Maybe a PWM-controlled BCV to apply gas to the top of the wastegate would help hold it stable?

Tial F46 on the 930 J Pipe with side port teed just above the 3.2 throttlebody. I plumbed it according to the instructions. It holds 150 KPa up to 5000 rpm, after that it jumps around between 135KPa and 160 KPa at 6500 RPM and hits those extremes building greater than 5000 RPM as in the log picture.

Sandeep

Uhhm,

http://forums.pelicanparts.com/uploa...1567974698.jpg
Without boost control, this doesn't make sense unless you mistyped "bottom port" or "side port" - the top port should be left open to atmo for boost regulation by spring pressure only, like the factory wastegate.

If you are using closed-loop boost control, I'd guess an issue with the PID values, or incorrect Hz/duty cycle settings for the valve. What does the duty cycle log for that output look like?

A 7 PSI spring is comparatively wimpy (I could quite easily move mine with my thumb, but with the .8 bar took both hands to "just" move it) - and an F46 has a lot of surface area for exhaust to act on the valve face.

My Bad .... Here is a pic of how its plumbed. SIDE port to manifold pressure above the throttle plate, TOP port to atmosphere ... no solenoid valve controlled by ECU. WOW ... so many things wrong with my first post :eek:

Sorry for the confusion and i've edited third post.

http://forums.pelicanparts.com/uploa...1567976535.jpg

Sandeep

Well, to my mind, only 2 things it could be:

• MAP sensor giving dodgy readings over 5K, perhaps interference?
• Valve blowing open under exhaust pressure against a weak spring.

My money would be on the second. This: https://en.m.wikipedia.org/wiki/Boost_controller

I'm guessing the valve is cracked off the seat at lower pressures (which will be hurting spool quite a lot), but it's "OK" until there's a lot of gas flow - at which point the valve starts bouncing and can't maintain pressure

This "early opening" can still be an issue even with higher spring rates in the WG, but usually not as bad.

Seem to me that an MBC would help considerably by applying pressure to the top of the diaphragm until set pressure reached, rather than relying on spring pressure alone.

Or, you obviously have an ECU; drive a $40 solenoid directing boost to the top of the wastegate for the same effect.

With a zork? I don't see - oh, wait, a joke right? You're a funny guy :D

Really appreciate the feedback on this issue.

This is my first Porsche and I brought most of my air cooled knowledge from my beetle days. My '74 Beetle (long since sold) had a 2.5L Type 4 turbo with megasquirt and this same turbo but different external wastegate.

Looking over those logs, the MAP signal was pretty flat all the way to redline. I thought that the turbine housing on this engine may be too small so may only pass a certain amount of flow, before the excess starts shooting out the wastegate and causing the valve to flutter ?

I do have an AEM 3 port boost solenoid. Was planning to wire that up this winter if I needed to.

Sandeep

Can you adjust the smoothing factors on an AEM for the MAP signal?

Of course, doing so would only make you feel better, and not correct any underlying issue (if there is one).

There's the rub. Is more data better, or would you be better off with your head in the sand? It's a real question, not a dig.

Another way to answer this is, "what's the operational impact of the fluctuations you are seeing?" ie does it matter if you have a wiggly line on a graph?

I'm not sure thats the issue as the signal is fine up to 5000 RPM. When the car was NA before the turbo mods the signal was fine as well, looking back over old logs.

Would it make sense that the turbine wheel is a restriction to exhaust flow over 5000 RPM and causing the wastegate to dance ?

Thinking that the gate opens at 7psi as seen in the log at 2900 rpm, holds fine to about 5000 RPM, and then is overwhelmed with exhaust flow causing the bouncing ? I'm not sure if this gate is sort of half open at 7psi and then the exhaust flow causes the valve to bounce against the spring causing the oscillation ?? :confused:

It sure messes up my VE tuning on the upper right of the grid as it looks like a roller coaster. I had to add a few extra rows to handle this. I was also thinking that if the turbine housing is too small (AR too small) then this is how you would determine that on the log.

Sandeep

Well, my only thought would be if the MAP signal is "correct", the compressor is choking...

What turbo are you using and how much power is it making?

Suby guys who see this (17-21 PSI or 21-25PSI fluctuations) say it absolutely kills all their power. They, almost to a man, report that some form of boost control sorts it.

If the wastegate is uncontrolled and erratic, my thinking is that it'll be off the seat and leaking boost long before it starts to flutter and shows up on the MAP reading. That could be very significant, IMHO - on a config with a target boost of 7 PSI total...

What does the boost gauge show?

The only guarantee you have with a .5 bar spring is that it will (or "should") be fully open at .5 bar.

In practice, it will probably open earlier than that. And a .5 bar spring is VERY SOFT. you can move the wastegate with your thumb against a .5 bar spring. Unless you control the wastegate in some way to hold it shut until you reach target, it'll open against exhaust pressure. Try blocking the exhaust with your hands when the car is idling. Pushes your hand off with each exhaust pulse, huh? Now imagine the volume of gas with the motor revving hard - and think about that soft spring again.

If you want to test other theories about compressor etc, the traditional way to do this is to apply the side-port line to the top port(!!! danger Will Robinson) and then go an CAREFULLY drive the car, keeping a very close eye on the maximum boost. Check your boost cut settings...

If the MAP signal is smooth, you just proved that the wastegate needs help to stay shut.

Spuggy and Tippy

You guys seem to be pretty well versed in this. It sounds like a boost control solenoid can be a big help in making power when as as desired even if the same max boost is set as the wastegate spring. ie not using the solenoid to allow more boost than the spring's documented pressure.

I had not even considered that the WG would be opening significantly below its rated pressure.

Would the solenoid need to be in closed-loop for that? Or could you do the same thing in open loop with a duty cycle table or whatever the ECU supports along those lines?

I've heard of issues with closed loop boost control, and can think of at least one potentially damaging failure mode with that.

What is the general consensus? Thoughts?

GT3076R with 0.63 AR turbine housing ... 300 RWHP at 0.5 bar

Thanks for this and appreciate the feedback.

I dont have a boost gauge but the logs show an oscillation between 135KPA and 165KPa.

My boost cut comes in at 170 KPa but have not hit that yet.

Thanks

SAndeep

I changed my post. I put surging, I meant choking....

Doh..