(I didn't see that Pete had posted while I was futzing around with this; it was just in reply to unplugging the MAP sensor)
Try unplugging the Baro sensor to see what it does?
I'm absolutely flailing in my attempts to either figure it out or be sane about reasoning it out, but I'm still having trouble seeing why a barometric sensor that doesn't see vacuum should ever read under half atmospheric, as it appears to in Analogue Values Engine 5, second image from top of page, KOEO.
I'm definitely getting confused reading through the varying KOEO, smooth, rough, WOT states viewed with Engine Values 2, Engine Values 5, or both. I haven't quite gotten fed up enough to make a spreadsheet. 
But I want to say we're also seeing a big change in nominal pressure with a small change in voltage, so does that tell us anything about the car or INPA? If the pressure changes and voltage doesn't really, does that mean it's being interpreted differently, or just that there's a disconnect there? How can 3.83V = 469 hPa while 3.67V = 1103 hPa? I've searched a bit, but haven't found an INPA user guide... Just lots of how-tos on getting it downloaded/installed/working.
Despite the word "point," I've found one reference indicating that this is a range indicating more or less sane, with values outside that range indicating an issue. Not sure why baro would be sane down to 0, but I guess that's just the range the sensor's capable of reading, rather than a sanity check on operating range?
Pete. (l33t FS) said:
I'd plug everything in, clear the codes, start it and see what it does.
If it runs like crap until it starts ratcheting up the fuel trims, start flowing propane into the engine to see if it will run well AND drop fuel trim back down.
If BMW had their hands in the engine controller, it may run poorly until you clear codes (and therefore adaptive memory).... AND you may need to go into manufacturer specific, and generic OBD-II, and clear everything out, because BMW engineers were obstinate and had "their" computer controls AND "American" controls in the same computer and they would get kind of schizophrenic. Like when the corpus collosum gets severed in a brain.
Everything is plugged in, no codes showing, does the same thing. I have cleared all adaptives multiple times, including the most recent when I plugged all the MAP sensors back in (tried it with upstream only, downstream only, and both unplugged with no change).
By flowing propane do you mean around possible vacuum leaks? Or just into the intake stream to see if it makes it run better?
In reply to Jesse Ransom (FFS) :
I also wish there was a better user guide or at least list of what different definitions mean.
I'm not sure if some values freeze or don't reset from running to KOEO. I could try running them again after not running the engine and opening INPA fresh.
In reply to EvanB (Forum Supporter) :
Just into the intake. See if it will run smoothly and with sane fuel trims if you are adding fuel.
Propane is nice because it is readily available, easy to control, and because it is a gas, it will flow evenly through intake plumbing/manifolds not designed for wet flow. It won't puddle.
MAPP is probably a bad idea, but set up the video camera first if you try it.
regarding the computer... Again, not sure if it got translated to MINI, but I have had BMWs with no codes in the BMW side and codes in the OBD-II Generic side, and vice versa. And they will cut injectors and/or lock fuel trims if there are certain codes present, which will fight you for trying to diagnose. (Vacuum leaks on E46s are horrible for this!!)
One reason why I am not thrilled with generic-only tools.
In reply to Pete. (l33t FS) :
That last bit did remind me to wonder whether we might get a less weird-looking view into MAP/Baro (and other?) normal-OBD values through a generic tool. Even if all it does is give us vanilla MAP/baro values that don't leave us wondering about the INPA glossary.
I'll try the generic adapter and see what I get.
I did get some screenshots of values from a R53 that actually runs. Target for the manifold pressure is 300-500hpa at idle (which I get when it runs well on initial start) and 800-1100hpa KOEO (which I get when it is running terrible).
No codes with a generic adapter. Here are the screenshots of data from the Torque app. When it is running rough and propane is added to the intake duct it smooths out a bit. I did not run it long enough to go into closed loop and see how STFT was affected.
Engine off, shows pressure as 14.1, atmospheric.
Running rough.
These don't really provide any insight since it's that same thing that INPA shows.
Video of what it does when started for fun.
I moved the MINI out of the way yesterday to get the Volvo truck in the garage. Revving it when it was first started produced some smoke from under the hood. I tried to pinpoint where it was coming from but couldn't narrow it down any more than "exhaust area". I'm still thinking the aftermarket header just has a lot of leaks. I'm also not convinced enough that that will fix the issue to spend the money on a new OEM style header/cat.
In reply to EvanB (Forum Supporter) :
I thought I read somewhere along the digging about these things being prone to a significant split somewhere, but damned if I can find it again.
I'm at a loss (obviously), but it just seems like there are a few things that are clearly dubious enough to pick an order and swap them out when you have the enthusiasm.
Junkyard manifold? OEM MAP sensors? O2 sensor locations? O2 sensors themselves?
All of it is a bit parts cannon without enough evidence, but on the other hand, none of these is without some suspicion.
I might see if I can find a manifold from a junkyard that hasn't just cut the cats out. The OEM ones were prone to splitting before the flex joint. The one I have is an ebay stainless aftermarket that could be prone to splitting in any number of places.
At this point I think "troubleshooting" should involve a Mossberg.
In reply to eastsideTim :
Cam timing appears to be correct, fuel pressure is spot-on. Throttle body is clean. Disconnecting the converter made no difference. Unplugging the manifold-side MAP made no difference. (did not try the other one). Crank pulley is not slipping on the damper hub.
It feels like a classic fuel starvation issue. Adding fuel makes it smooth right out.
I think my next plan of attack is to scope the injector power and grounds to see if the ECU is able to properly open them. I did bring my scope and other toys, but it is dark and cold.
If that proves fruitless, I'll open a help request on iATN and let ten other people tell me to check the things I already have checked.
Partially clogged injectors, maybe?
In reply to eastsideTim :
Who the hell knows.
Fuel pressure is verified. Any reason to suspect fuel volume?
mdshaw
Reader
2/6/21 9:48 p.m.
If extra good used injectors are cheap (since it's a BMW I doubt it), get some spares & send them off to get clean & checked, or get some good ones already cleaned/checked. May not be the issue but at least you'll know they are good.
The JDM K24 I just swapped in the TSX had bad injectors from sitting in the motor after being removed in Japan who knows how long ago. They wouldn't even fire they were so stuck from the bad/dry gas. Swapped in the ones from the US motor & it fired right up. I'm thankful that Honda's are so reliable, easy to work on & parts are cheap. Did a motor & transmission swap & was very easy really & is running great.
well, that's one more thing it's not...
In reply to Pete. (l33t FS) :
I'm assuming that's injector, and... shows it opening and shutting with flyback indicating actual motion? Does that rule out gummed up passages even if the pintle moves?
In reply to Jesse Ransom (FFS) :
Was checking to make sure the injectors were getting full battery voltage, and that the DME was capable of pulling them fully to ground. Did not check current. Probably should have, but getting an amp clamp on an individual wire would have been problematic.
In reply to Pete. (l33t FS) :
Would any of that tell you whether they were actually moving fuel well? I mean, do they always gum up in a fashion that restricts movement, or do they sometimes just block the fuel passages so the pintle can do whatever it wants and the fuel flow will be insufficient?
A quick search suggests that I wasn't misremembering too badly about that showing a spike from the injector generating voltage as it falls shut, so while we don't know exactly how much it's moving, it looks like it's moving. (?) Would a current check mostly tell us whether it was shorted/dead?
I'm liking the notion that the injectors are so filthy/half-plugged that they're not moving enough fuel, and especially that what they're moving is in a terrible spray pattern. So you still get sooty plugs from having extra fuel in the chamber by the time there's enough atomized fuel to run smoothly?
But that's easy for me to say without spending time or money or freezing my ass off.
In reply to Jesse Ransom (FFS) :
The only issue with that is that the injectors were replaced, presumably in order to solve this issue. They don't look particularly new but that doesn't mean much either. (Just as being new doesn't mean being good)
I'd like to run a manual injector test but the bits to do that aren't here right now.
Well, techhelp request has been posted to the hive of scum and villainy.
I thought about running out to get my injector actuator harness to run a test on the injectors. Then found out my tractor wouldn't start to plow out the snow piles around the trailer. And if I'm running out to my friends house I might as well take the trailer to retrieve the E30 that is going to occupy the garage for a bit.
So instead I spent my time making new battery cables for the tractor to put a car battery in it and putting the tire chains on. Because shoveling snow sucks.
sounds like it needs to be pushed out of he garage and set on fire.
