MAP bouncing around

[BigD]

Dynoed the E30 last week and it's delivering quite deliciously - 648whp/531wtq at 18psi. But before going any further there are some issues to sort out. One of them being what we see on the MAP log. It helps explain why there are power dips at the same time, with the ignition table being explored a bit too much.

First we thought that maybe it's an interference issue but I have an AEM boost gauge with a sender in a different location than the nipple of the VEMS MAP sensor. The same kind of jumping around was visible on the gauge too. I don't know but feels like the odds of two sensors in different places seeing the same interference seem low - am I wrong?

It just seems too fast to be mechanical - if it was a boost leak it would just be a steady bleed, like an MBC. If it was the wastegate, it should react slowly.

Any ideas anyone?

[BigD]

Also worthwhile to note that I'm overboosting badly. I don't think my WG is opening or not opening much. The spring combo is only 13psi, wg is MV-R.

I'm starting to think maybe I didn't plug one of the bottom ports of the WG or the diaphragm is torn - that would probably both cause an oscillating boost leak and overboost.

[mattias]

I would start  by leaning out the VE table. Hit lambda target first, was this not fixed at the dyno? I would be nice to know what it would make if it ran leaner.

It looks like you have lean mis-fires from running too rich, below 0.80 lambda is never  a good idea.
That can cause the MAP reading to fluctuate wildly.

Could be your ignition system mis-firing. What plugs are you running? Either way - gap the plugs to 0.40 mm.

[BigD]

I looked through the individual power for the duration of the pull, there were no anomalies and it sure sounded good! Also I don't think I want to go any leaner as I'm already hitting over 900c egt

[NOTORIOUS VR]

That is an earlier pull...

I tuned this car, maybe because it's not zoomed in enough but there are no misfires happening at all here, and I have yet to see a car misfire unless you're below 0.74 and even then it's only with a weak ignition system.

There were absolutely no misfires happening while logging/on the dyno.  This is a boost issue 100%.

Running regular NGK copper plugs, but again there was no issue with spark.

Here is a later run zoomed in to see AFR's better:

[Sprocket]

Not being too experienced with turbo motors, and perhaps teaching egg sucking, has anyone considered the possability the dump valve could be lifting, wastegate actuator weak, or worst case rotating stall (pre curser to full on surge) as boost rises?...............

[BigD]

Not being too experienced with turbo motors, and perhaps teaching egg sucking, has anyone considered the possability the dump valve could be lifting, wastegate actuator weak, or worst case rotating stall (pre curser to full on surge) as boost rises?...............

This is actually my first kick at the turbo can. I don't think the BOV is opening, it's got a pretty strong spring to account for the good vacuum these motors pull and it's got a 1/4" reference nipple and line. The wastegate is a TiAl MV-R which should be plenty. Surge is actually something I suspected. I don't hear any but one theory I have is if I had a major enough boost leak, the turbo, as big as it is (GT4094r) is hitting the surge line - we see only 18-20psi but it's actually spinning its ass off.

I didn't have a chance to do a full soapy spray test but I did a basic test for hisses on the weekend and I had a huge leak at the BOV, but I redid the TB side of the intake pipe for a few reasons so it may have been me putting it on poorly when putting it back together (pinched the o-ring). Will never know now.

Another thing I suspect is the wastegate. I mean, it's definitely in play here, something's wrong. It's got a 13psi spring and we see spikes to over 20 and sustained 18. I'm going to pull it tonight and look. In my surfing for answers, I found a diagram of the MVR and I saw that it has two reference air ports at the bottom. If I didn't plug one, that would pretty much explain everything. Will report back.

[BigD]

No smoking gun...

The ports are plugged tight, there are no obvious leaks, certainly none major enough to cause a 40r to surge I would think. The signal to the WG is air tight, the top port is vented through the EBC valve.

Pulled the WG, didn't see anything obviously wrong.

Can anyone confirm visually from experience if these springs look right for 13psi? No that there's a 3 spring combo which should cause this boost but still. It came with a 0.5 bar combo, which should be Brown/Gold (the 2 smaller springs). I later bought the big spring which should be the "Black", from the new color scheme, Grey by the old one, which should add up to 0.9bar.



Diaphragm seems ok, doesn't look like it was pinched



Valve looks like it opened at some point...

[Sprocket]

With the data you have from the dyno runs and the logging you should be able to work out roughly the peak air consumption. With that and the pressure ratio, you should be able to roughly plot on the compressor map and see where it is in relation to the surge line. If it is close, the pressure spikes could be linked to rotating stall.

The thing with rotating stall is that it generates a pulsing discharge pressure but just about keeps pushing air. More worryingly induces loads on the compressor impeller that the shaft bearings were never designed for. The net result is failure. The other worrying thing is that rotating stall is less noticeable than a full on surge. If it surges you bloody well know it does.

I know somewhat about rotating stall and surge on centrifugal compressors with our machines at work, they are only about 100x the size of your turbocharger lol. Rotating stall and surge on these causes 10s of thousands of ££££s damage in a very short space of time Good for us, bad for the customer


Just one other thing to consider is the compressor intake tract. If there is enough restriction here at max airflow the compressor eye can be subject to a vacuum, artificially increasing your pressure ratio, thus pushing it closer to surge. The discharge pressure is only one part of it, the suction pressure also needs to be considered, and why when talking about centrifugal compressors, pressure ratio is used.

PS. I hope you get it sorted. this sounds awesome especially with 600hp+ lol. Here's me with a poxy 1.3l pushing 200hp

[BigD]

It would be nowhere near surge if there are no leaks. The turbo is capable of much more. The possibility is that if it was a giant boost leak then while I was seeing 20psi in the manifold, the turbo was actually churning out 40... but it's a pretty wild theory, I doubt it's the case.

The intake is no problem, it's very smooth, 3" pipe and 4.5" intercooler.

[Sprocket]

Your intercooler is before the compressor inlet?.................

By compressor intake tract I was meaning everything before the compressor, including the air filter (Cotton gauze I hope and not that shitty foam of an excuse for a filter )

[BigD]

Oh sorry, this was with no filter at all. I actually went to great lengths to beat filter makers into telling me the CFM ratings of their filters and almost none of them have anything that works - I don't understand how so many guys with turbos run these tiny filters. I ended up with this K&N:

http://www.knfilters.com/search/product.aspx?prod=RD-1460

Going to require some creative and liberal use of an angle grinder to get it to fit.

[BigD]

Looking more at the logs, it doesn't seem to depend on anything but RPM. The same behavior is there regardless of throttle or MAP. Once the RPMs go over 3000, the MAP signal gets jagged. Here's one at low throttle and MAP:

[BigD]

OK this latest discovery made me wonder if it's actually even related to the turbo. I got VEMS working naturally aspirated first. Dug up a log. Sure enough, it did the exact same thing. Going to try to add a filter:

[mattias]

Where exactly is the MAP fitting on the intake? If it's near an intake port opening then it's possible to get real bad pulses going. And no signal sharing with anything else. FPR and MAP must have their own (they can share), and not take the same source as wastegates or blow-off valves.


You seem to have a bad indication at 4k, with old and the new firmare. Those indications of a problem have to go before you tune an engine for any kind of power.  It's probably some kind of mis-fire or maybe just an exhaust leak. Who knows. No details. Ignition system is still suspect if you ask me.
 You do have a software filter you can try, I'd use 2+10 as the settings if you're not already set up right. You'll find it in the ECU calibration dialog. It doesn't really solve anything here.

In the log viewer, bring up the "isTriggerError" flag instead of "isWarmup". Why that is indicated in the log is in itself really bad.