RPM lags MAP by 0.4 seconds

[turbo conversion]

billr, the OP is complaining about the oscillation and also the engine dying as soon as it fires unless he opens the throttle about 12%.

The map signal has been mentioned several times and he says he has checked it and everything checks out okay.

David

EDIT: When I look at the rpm oscillation it reminds me of the way some blown alcohol engines idle.

[jsmcortina]

That "idles at 25" Hg vacuum" is another clue that I think needs further discussion. I have never seen an engine idle with that low a manifold vacuum, only see those levels in over-run. 21 or 22"? Sure, but that is quite a ways from 25" when talking about vacuum.

Why are we talking in inches of Mercury? The MS2 measures in kPa.

Jim - please talk in standard units of kPa, otherwise we'll all get confused.

James

[billr]

Sorry James, I was just trying to not add a burden to the OP as we try to help. I "grew up" with manifold vac as the only way, but agree that kPa seems simpler to both use and understand.

[Blown88GT]

That "idles at 25" Hg vacuum" is another clue that I think needs further discussion. I have never seen an engine idle with that low a manifold vacuum, only see those levels in over-run. 21 or 22"? Sure, but that is quite a ways from 25" when talking about vacuum.

Why are we talking in inches of Mercury? The MS2 measures in kPa...

Vacuum gauges are in inches of Mercury.

[jamies]

also noticed all the lag factors in the tune are set to 50%, crank them up..

[Jim_Blackwood]

Hey guys, first chance I had to get back online. I'm a couple hundred miles away from the car.
Thanks for the suggestion about the MAP delay. (Lag Factors) I noticed that just as I was shutting down and wondered if it could help.
I really did mean to run the VE table generator but I ran out of time. Had to go to a funeral, a very unexpected one.
So I do intend to do that as soon as I get back. And I'll get you more info about the plenum volumes but just from memory, the intercooler chamber sits between the blower (Eaton M112 from a Ford Lightning truck) and the intake runners and is roughly 13" x 8" x 5" with a 3" tall evaporator core inside. The runners are horizontal across the base of the intake and are about 1-3/4" tall. The scoop is an Enderlie style and extends past the front of the blower body, is wider than the blower, and probably 3-1/2 or 4" tall. From that it might be possible to get a rough estimate of plenum volume. All those elements are behind the throttle body.

Also I'm not saying there isn't a vacuum leak but that would make the MAP reading settle faster, not slower, and the reading drops a lot faster when the car is running as I explained before on Aug 30 @10:33 "I'm showing a change rate in the other datalogs as high as 4.11 ms/kPa (smooth curve)" so the only thing that leaves is the plenum volume.

Sorry about the vacuum comment, it was based on the in-car vacuum-boost gage. Eventually I plan to change that to straight MAP but all this takes time. A more accurate reading was probably in the range of 22-24" but I was estimating based on a quick needle observation, you are right, I'm sure it wasn't as high as 25 except on decel. But the point was that the engine is tight and the cam isn't particularly radical. Like most, I don't have an intuitive feel for the conversion or I'd have put it in kPa. Anyway it should be there in the datalogs and that should be dead accurate. I wouldn't expect to see anything beyond normal expectations, but from my gage it does appear to be on the high side. Also, since the car won't idle smoothly below 1200rpm that's the speed I'm taking the vacuum reading at, meaning it's bound to be a little higher.

I'm very appreciative of everything all of you are doing to help me sort this out. Any perceived slight is completely unintentional and I'm doing all in my power to be cooperative. I just can't do anything until I get back and it'll probably be Wednesday before I'm able to try the VE table generator and the MAP delay setting.
==============================================
Those (Lag Factor) numbers in that dialog were the ones that were there when I upgraded the firmware and I don't know what they do so I didn't change them. I think I had numbers in about the 80-90 range across the board before I updated the firmware. Any pointed recommendations or any explanations about what they do would be most welcome.
==============================================
Sorry for the long posts, I'm trying to get a lot of info in but I don't want to burden you with too much.

Bill, the engine does start OK and runs very well. The idle oscillation is the single glaring issue. I think if I can resolve that the rest might be pretty easy.
The M112 is capable of producing much more boost on a 350 cid engine than I am using. It's main purpose here is to help extend the rpm range by overcoming the port shortcomings. 5psi is adequate for that purpose. My target is in the 400-500 hp range and I think I'm there.

David, your fuelie comment isn't too far off point. Except that those generally seem to ramp up very fast and come back down much slower. They also have MUCH more radical cams. My grind does have about .500 lift, which is not that much in a Buick, more than stock duration, but the overlap is small. If it's relevant I can get the specs Tuesday.

Jim

[billr]

Where, specifically, is the MAP ported into the intake? How about posting some pictures of the intake system? That lag factor may be an issue, but I have never changed mine from the default 50 and response is reasonable, even with fairly large plenum/runner volume.

[Jim_Blackwood]

The MAP sensor is connected to a line at the rear of the intercooler housing. Don't recall at the moment if it is above or below the core but there wouldn't be much of a pressure drop there anyway. I can look when I get back.

I'll try to give you a youtube link:
https://www.youtube.com/watch?v=wkoF2yl ... e=youtu.be

Jim

[Blown88GT]

Where, specifically, is the MAP ported into the intake? How about posting some pictures of the intake system? That lag factor may be an issue, but I have never changed mine from the default 50 and response is reasonable, even with fairly large plenum/runner volume.

...More seriously though, at 5.012s in the log, your engine has stalled, but MAP is showing 52.2kPa, it then takes 1.4 seconds to return to 93kPa?! It should be instant. I can only assume you have a kinked or blocked MAP line. Ambient appears to be 97kPa...

He has a vacuum leak, probably at the blower flange or possibly a cracked blower housing. Nothing else would cause the sinusoidal oscillation he has. If he removes the belt from the blower, it will take a lot longer than 1.4 sec to return to 93 kPA.

[Jim_Blackwood]

http://forum.britishv8.org/file.php?2,f ... tachment=1

http://forum.britishv8.org/file.php?2,f ... tachment=1

http://forum.britishv8.org/file.php?2,f ... tachment=1

[Blown88GT]

http://forum.britishv8.org/file.php?2,f ... tachment=1

http://forum.britishv8.org/file.php?2,f ... tachment=1

http://forum.britishv8.org/file.php?2,f ... tachment=1

Link give this message.
Linking to files in this forum is not allowed from outside the forum.
BritishV8 does not allow linking outside their forum.

[Jim_Blackwood]

I think if I removed the belt it would make no difference. Not only will the suction pull air past the rotors but there is a bypass valve that opens at lower manifold pressures. In either case the pressure will equalize between the manifold and the scoop unless the blower is making boost, which at idle it does not.

Could you please explain how a vacuum leak is going to cause the MAP reading to take more time to return to atmospheric? Seems backwards to me. Anyway I'm quite certain I have flooded the area of the intake system with various hydrocarbons to check for leaks and found no indication of any.

James' suggestion seems like a good thing to try. If this has an influence on the algorithm then it should have the potential to eliminate the oscillation entirely, by reducing the gain in the feedback loop. With simple systems that typically also reduces the response but it seems that for this application the algorithm would be complex enough to also use history and prediction to improve the response rate. Not knowing what it looks like I couldn't say, but typically the proportional, integral and derivative gains would be involved. Tuning that can be tricky and it has to be tuned for the specific system. Here I'd expect it to be even more complex, having additional input factors. I'm no mathematician but I can appreciate the fine balancing act that is going on there.

Jim

On the links, it was the best I could do from here. BV8 has no fee to be a member.

[turbo conversion]

"David, your fuelie comment isn't too far off point. Except that those generally seem to ramp up very fast and come back down much slower. They also have MUCH more radical cams. My grind does have about .500 lift, which is not that much in a Buick, more than stock duration, but the overlap is small. If it's relevant I can get the specs Tuesday."

I don't think this is the case here, just making a comparison.

David

[ashford]

The MAP sensor is connected to a line at the rear of the intercooler housing. Don't recall at the moment if it is above or below the core but there wouldn't be much of a pressure drop there anyway. I can look when I get back.

I'll try to give you a youtube link:
https://www.youtube.com/watch?v=wkoF2yl ... e=youtu.be

Jim

with that much volume behind the throttle body you will need closed loop timing setup. post your latest msq and your target idle speed then ill give it my best guess

[jamies]

He has a vacuum leak, probably at the blower flange or possibly a cracked blower housing. Nothing else would cause the sinusoidal oscillation he has. If he removes the belt from the blower, it will take a lot longer than 1.4 sec to return to 93 kPA.

wouldnt be a vacuum leak, bit will be like james said a blocked/restricted/kinked vacuum hose, or possibly a bad map sensor. the map sensor will read instantly, so when the engine stalls etc it should immediately go to atmospheric kpa.
id first air test the vacuum hose to the ecu, then test the map sensor and find where the response is lagginng.
Assuming the map sensor is connected correctly and directly to the plenum below the blower (ie some T-pieces have restirictors built into them so ensure nothing like this is in the path)
Tee in a mechanical gauge right near the ecu.
Also need to check the map calibration is correct as it appears to have a lot of vacuum (25-28kpa)

[Jim_Blackwood]

Got back quicker that I expected.
So the first thing I did was run some starts at different lag factors from 10 to 100. Wasn't a whole lot of difference in the way the engine ran but at 10 it did actually keep running where at all the other settings it eventually died. This is the first datalog below. Note that it stabilized at 1660rpm, dropped to about 1300 and then began oscillating again, finally stabilizing in oscillating mode between 1800 and 400 rpm.with factors set at 70 it shows this same pattern except that the engine died at 26-1/2 seconds. Most of the rest of the tests were the same except for not recovering from the last 400 rpm dip. So it died one cycle earlier.

I did these tests first before changing anything else so I could get a valid comparison, and could post the entire series if anyone wanted them, but they look similar enough to me that I don't really see the need. Sadly it wasn't enough to validate the lag factor changes but perhaps that is because I changed all factors across the board at the same time. I can try other settings if anyone can recommend them.

So with that done I'll pick back up tomorrow.

Jim

[Jim_Blackwood]

.msq file attached. I'd like to see a stable idle at 600 rpm but really anything up to about 800 would be fantastic.

A very rough estimate of the plenum volume would be about 1000 cu. in. based on my very early dimension estimates.

During stable idle at 1650rpm the datalog shows ~26kPa with a high of ~62 and a low of ~18kPa during oscillation. At 1204rpm it is 25.5kPa.
The MAP calibration could be off. Using the standard on-board baro sensor, but do not see it in the drop down list.

The vaccuum/boost gage appears to be accurate.

Jim

[turbo conversion]

Have you tried unplugging the map signal line at the plenum and seeing how the signal responds, if not maybe try this.

With key on engine off pull a vacuum on the line (to the kpa it is oscillating at) then release the vacuum and see how it responds in TS.

You could even T a mechanical gauge in the line (at the plenum end of the hose) and see if they sweep together.

That way the gauge will vent to atmosphere and the map sensor will vent through the line.

David

[jamies]

Tee your vacuum / boost gauge into the line going to the ECU.
start and idle the engine, then shut it off, does the vac/boost gauge slowly drop the same as the map gauge on the MS, or does it drop instantly.
you need to determine the cause of the lag on the MS for engine load sensing before you bother doing anything else.

[ashford]

give this a shot i have setup adaptive timing, also had to setup closed loop idle for this to work correctly. you may need to change the frequency of the idle valve depending on what generation you have. should be close to 80 hz( factory) for the earlier 2 styles by hitachi. th newer ones use 300-400 hz.