Climbed 6000-10000' in elevation, car is E36 M3ting bricks

I can't catch a break with this berkeleying car.

Runs perfectly at sea level. As we got into Colorado, started to get weird. LEAN in almost all conditions, and often won't return to full lean when decelerating in gear, instead chooses to burble along in the 12-13:1 range.

The exhaust is getting REALLY hot due to the lean condition, which isn't making the trans tunnel pleasant.

It's also trying to overheat now. I'm unsure if this is because of the lean conditions making it run hotter than normal, or if the cooling system that almost works TOO well at sea level isn't sufficient up here? Had to pull off the road 3 times yesterday on my way up to Estes Park.

I can get it to drive OK at best by adding +15% fuel trim everywhere, but it's still not "right." It idles like E36 M3 like that, and the AFRs still aren't reacting correctly on boost/part throttle situations.

Any ideas? I'm completely confused. Trying to get some datalogging to send back to my tuner, but wondering if there's anything "normal" about all this that i'm missing. The car is VE tuned, and if that were messing up, i'd expect it to be RICH, not LEAN.

Details:
2004 mazdaspeed Miata
550cc injectors
Haltech Platinum Sport 1000
Bigass 55mm Mishimoto radiator w/ shroud and fans.

I'm sure the fans aren't the best, but it's having no problems bringing the temps down at a stop @ idle.

I'm also having some problems coming to terms with the difference between absolute and gauge pressure. My gauge is still showing 10psi max boost, though the vacuum numbers have changed with elevation. Haltech is now recording no more than 6.5psi, though is still SOMEHOW showing almost 22inhg of vacuum at idle.

Both the gauge and Haltech MAP are T'd into the same line off the intake manifold.

I really hate this berkeleying car. It's ruining my vacation/wedding/honeymoon.

Fuel pressure OK?

The thin, dry Colorado air is increasing your cooling problems as the car can't shed heat as well. Fans make a big difference at highway speed as well, far more than you think. At idle, you're not generating significant heat. You sure are when you're climbing a mountain, and the air in the engine bay is almost stagnant so a good part of the airflow is coming from the fans. The stock MSM fans would have been a lot more effective than Mishimotos, whose primary benefit is that they're cheap. I don't have a lot of good solutions for that one on the road, unfortunately. Make sure you're running a very high water percentage, ideally pure water with Water Wetter added. That'll help.

The Haltech is showing absolute pressure, the gauge is showing relative. You're making 10 psi over ambient, but ambient has dropped. Idle vacuum doesn't really change with altitude.

If it's not cutting fuel on closed-throttle decel, I'd check the TPS reading. I don't know what that would do to normal running in a Haltech.

I've seen cars with IRTBs go lean at altitude, as what used to be cruise MAP turns into full throttle. But it shouldn't happen with a turbo car. Does the Haltech run in closed loop?

Fuel setup is stock other than the injectors. I wouldn't anticipate it being a problem. I'm not entirely sure i can log it... i can check.

Fuel pressure wasn't causing a problem last time i was up here with the car. The car still ran terribly, but it ran terribly at all elevations. Now it runs perfectly anything below 5000', so i'm pretty floored that it's having problems of this magnitude.

Keith Tanner wrote: The thin, dry Colorado air is increasing your cooling problems as the car can't shed heat as well. Fans make a big difference at highway speed as well, far more than you think. At idle, you're not generating significant heat. You sure are when you're climbing a mountain, and the air in the engine bay is almost stagnant so a good part of the airflow is coming from the fans. The stock MSM fans would have been a lot more effective than Mishimotos, whose primary benefit is that they're cheap. I don't have a lot of good solutions for that one on the road, unfortunately. Make sure you're running a very high water percentage, ideally pure water with Water Wetter added. That'll help. The Haltech is showing absolute pressure, the gauge is showing relative. You're making 10 psi over ambient, but ambient has dropped. Idle vacuum doesn't really change with altitude. If it's not cutting fuel on closed-throttle decel, I'd check the TPS reading. I don't know what that would do to normal running in a Haltech. I've seen cars with IRTBs go lean at altitude, as what used to be cruise MAP turns into full throttle. But it shouldn't happen with a turbo car. Does the Haltech run in closed loop?

Thanks Keith. I actually called yesterday and spoke to Tammy, you guys have a tstat waiting for me, i'll be picking it up tomorrow hopefully. Maybe we can figure something out for getting all water/wetter in there if you guys carry that?

I only have one gallon of coolant in here, and about 2 gallons of water. There's also the possibility that there's a bubble in the system i suppose.

I know the stock MSM fans are better, but they weren't fitting on this radiator. Unfortunately, "cheap" turned out to be a necessity when i found out that my stock radiator was leaking due to the a/c stuff up front being jammed into it due to an accident in the past. This whole setup wasn't my first choice, but it's what i could afford and get quickly. I feel like a large portion of my cooling problems are due to the lean conditions, though... But better fans are never a bad thing.

TPS values are still fine. It's going all the way to 0.0% on decel. Cut is still set to activate when decel anything 1700rpms and above, and turn off at 1400rpms.

Can you explain why idle vacuum isn't changing much? This is where i'm getting confused with the Haltech MAP vs. the FM guage i have of yours. The gauge has shown a reduction in vacuum around 6-8inhg. MAP is still showing something similar to what it sees at home. The MAP readings are still behaving relatively "normally" when monitored, but the fueling just doesn't seem to be reacting as quickly as it should. If i'm dumb and DO stay in the gas when i see the Powerdex max out as the turbo starts spooling, it WILL eventually richen out to where it's supposed to, but it takes a really long time.

Closed loop is a go for Haltech.

As for the fans.... your Stage 1 airflow setup would fix that in a hurry, yes?

As for your gauge - again, it's absolute vs relative. Switch the Haltech over to show absolute KPa and it'll make more sense than inHG/psi. Your engine is still pulling the same absolute pressure at idle. I'm guessing you have a mechanical boost controller, so your boost is fixed to 10 psi over ambient. If you had an EBC that was tied to absolute pressure, you'd see higher boost levels on the gauge, although the MSM turbo would have a bit of a struggle reaching 13 psi.

We do have water wetter. That'll help. It's not a cure-all, but it'll help the system by increasing the heat capacity. A higher pressure cap might help as well. Heck, we might want to check yours to make sure it's holding what it's supposed to.

And yes, Stage 1 fans would likely make a big difference. One of my own cars used to have trouble overheating on the highway and long climbs, and the Stage 1 solved that.

You'll have to ask the Haltech guys why you're not seeing decel fuel cut. I'd try to find out, it may be an indicator of what else is going on. A closed loop MAP based system should have no difficulty with higher elevation.

I presume the Haltech has a barometric correction table somewhere - it sounds as if it needs to be retuned to add more fuel at altitude.

What does "eventually" mean in this case? Seconds?

I assume you're running a wideband? Is that feeding the Haltech?

From all your descriptions it sounds like you have a fuelling problem, not necessarily a cooling problem.

Assuming a fresh or very recent fuel filter, I would try monitoring the fuel pressure first. To me, if the fuelling comes in with a delay, it sounds to me like either the pump is struggling to build up the necessary pressure or the FPR doesn't work right.

Does the Haltech allow you to look at the injector duty cycles by any chance? They're not maxing out shortly before the mixture corrects itself?

<thinks>

You don't have some sort of vacuum/boost leak by any chance, do you?

Keith Tanner wrote: As for your gauge - again, it's absolute vs relative. Switch the Haltech over to show absolute KPa and it'll make more sense than inHG/psi. Your engine is still pulling the same absolute pressure at idle. I'm guessing you have a mechanical boost controller, so your boost is fixed to 10 psi over ambient. If you had an EBC that was tied to absolute pressure, you'd see higher boost levels on the gauge, although the MSM turbo would have a bit of a struggle reaching 13 psi. We do have water wetter. That'll help. It's not a cure-all, but it'll help the system by increasing the heat capacity. A higher pressure cap might help as well. Heck, we might want to check yours to make sure it's holding what it's supposed to. And yes, Stage 1 fans would likely make a big difference. One of my own cars used to have trouble overheating on the highway and long climbs, and the Stage 1 solved that. You'll have to ask the Haltech guys why you're not seeing decel fuel cut. I'd try to find out, it may be an indicator of what else is going on. A closed loop MAP based system should have no difficulty with higher elevation.

Ok, i think i'll pick up the Stage 1 while i'm there if you guys aren't crazy booked or anything. I seem to have this habit of walking in your door last minute, dont I?

MadScientistMatt wrote: I presume the Haltech has a barometric correction table somewhere - it sounds as if it needs to be retuned to add more fuel at altitude.

It does... but it's disabled right now because i only have one MAP sensor. From what i understand, the "correct way" to do baro correction on this unit is to use the onboard MAP sensor for baro, and use an external MAP for fueling/ignition requirements.

I could be wrong with that. Honestly, Ludwig would be able to set me straight with that, but i don't have phone service right now, and won't til tomorrow.

In reply to BoxheadTim:

+1

BoxheadTim wrote: What does "eventually" mean in this case? Seconds? I assume you're running a wideband? Is that feeding the Haltech? From all your descriptions it sounds like you have a fuelling problem, not necessarily a cooling problem. Assuming a fresh or very recent fuel filter, I would try monitoring the fuel pressure first. To me, if the fuelling comes in with a delay, it sounds to me like either the pump is struggling to build up the necessary pressure or the FPR doesn't work right. Does the Haltech allow you to look at the injector duty cycles by any chance? They're not maxing out shortly before the mixture corrects itself? You don't have some sort of vacuum/boost leak by any chance, do you?

Sure, i can log duty cycles. I've been monitoring them, and it never really makes it past 60%. There's a lot of fueling available on this car for a turbo setup that will likely never breach 250whp.

I don't THINK i have a vacuum/boost leak... i did think of that earlier, and ended up replacing all the vacuum line from the manifold to Haltech.

Even if i did have a vacuum leak, boost/pressure is referenced off the MAP, so all that would do would decrease my MAP values, and fuel SHOULD adjust to suit.

Definitely something i can poke around with, though.

You said it's running lean almost all the time, not only during acceleration, right? Or did I misunderstand something?

Re boost/vacuum leak, my guess would have been that it'd be somewhere around there if it doesn't run a fixed fuel pressure, not anywhere near the ECU itself.

BoxheadTim wrote: You said it's running lean almost all the time, not only during acceleration, right? Or did I misunderstand something? Re boost/vacuum leak, my guess would have been that it'd be somewhere around there if it doesn't run a fixed fuel pressure, not anywhere near the ECU itself.

It'll cruise about where it's supposed to. Everything else is pretty wonky. Accelerating lightly from a stop yields no power because it's lean for a good 2 seconds every shift.

It's fixed fuel pressure. Returnless system. Believe it's somewhere around 65psi, but Keith would have to keep me honest on that one.

55, actually. We can check pressure tomorrow.

I think the running hot/overheating is a symptom of the lean condition. I would concentrate on the fuel trims primarily before spending too much time on the cooling system.

It's a long shot, but effective VE changes with elevation as a result to less exhaust pumping losses....
Basically, more air pumping through the motor at a given load. Without barometric correction: lean at altitude. Once it goes to closed loop (cruising) it adds fuel based on O2 feedback. I'm pretty sure that ecu isn't adaptive, so the open loop will run lean indefinitely at altitude until re-tuned or baro correction is on.

Or it could be something else.

Ok, I couldn't remember if the MSM had a similar system to the regular NB or not.

The Haltech has acceleration enrichment tables, doesn't it? Are they turned on and set correctly?

Does the behavior change with altitude change? I'm assuming you're varying your altitude by 4-6,000' as you drive around the mountains. That might help indicate if the problem is a mechanical thing that went south on the trip (such as a failing fuel pump) or if it's a tuning problem that's related to altitude compensation.

Just how lean are we talking?

We've run ECUs without barometric correction in the past, and they don't seem to exhibit massive changes in AFR. Certainly not enough to create " accelerating lightly from a stop yields no power". So the VE change is likely not enough of a factor to cause this kind of behavior.

All i really know is that it was fine pretty much all the way through Kansas, but started getting weird once i got off of 70 to go to Colorado Springs.

I researched on the VE issue last night, and it didn't seem like it should be this extreme. Certainly not bad enough that i couldn't fix it all by adding 15% across the board.

I knew I should have rented you the Forte for this trip!

Bobzilla wrote: I knew I should have rented you the Forte for this trip!

snicker

Wetter Water is stocked at OReillys.

sounds like the larger FI injectors might be stressing the stock system and especially if a component is failing ie fuel pump. I dont have any experience with MSM but I used to make the switch of injectors on 1.6 cars and it made some cars run leaner than others. Can you switch back to stock injectors?