Topic: Turbochargers

[FourbangerYJ]

Came across this site http://www.turbominivan.com/ Thought it was pretty interesting. Dyno's 400+ ft of torque and 192+ HP. Check the dyno fun on the left side of the page.
Can a east-west motor be put in a North-south Jeep?

[sharpxmen]

Came across this site http://www.turbominivan.com/ Thought it was pretty interesting. Dyno's 400+ ft of torque and 192+ HP. Check the dyno fun on the left side of the page.
Can a east-west motor be put in a North-south Jeep?

that dyno chart is very unusual

[jfrabat]

that dyno chart is very unusual

X2; all the ones I have seen come up first then down...  both for HP and torque.

[chrisjsmith4]

haven't done anything in a while. Running without the fuel controller right now. Sends my jeep into limp mode or something. my mechanic at work thinks its because I have the tach wire hooked up to the cam pos sensor and the draw throws the computer off. Anyone know if the controller can be used on the 2.4l the split sec guy said that it should. If it doesn't should I be able to get a full refund.

[sharpxmen]

haven't done anything in a while. Running without the fuel controller right now. Sends my jeep into limp mode or something. my mechanic at work thinks its because I have the tach wire hooked up to the cam pos sensor and the draw throws the computer off. Anyone know if the controller can be used on the 2.4l the split sec guy said that it should. If it doesn't should I be able to get a full refund.

i don't see any reason for which it would not work - you should tune it, that's my guess why you are having issues. you could try to skip the timing adjustments and see if that makes it better (don't connect the crank output and just leave the crank input connected to the crank signal wire)

dunno about the refund, you should ask SSecond, but i doubt it won't work. You should really have a way to log your engine parameters and AFR to understand what is happening.

[chrisjsmith4]

I remember the guy saying something about hooking the tach wire to something else. I'll find out and give that a try. new pcm will be ordered on monday. that could be a huge part of the issue too. we'll see.

[sharpxmen]

I remember the guy saying something about hooking the tach wire to something else. I'll find out and give that a try. new pcm will be ordered on monday. that could be a huge part of the issue too. we'll see.

i really don't think the cam signal is an issue. on the 91-95 PCM there is a tach signal wire that can be used but there isn't one in your case - you could try one of the Injector command wires, that should give you the same signal timing as the cam sensor signal and should not affect the injector at all (the input impedance of the tach sense on FTC would most likely be over 1 MOhm so should not draw any significant current) 

[chrisjsmith4]

I'll play around with it.

[Sherpa]

X2; all the ones I have seen come up first then down...  both for HP and torque.

That dyno plot looks the way it does because of torque multiplication.  While the converter is slipping at lower RPM (and the turbo was already starting to build boost), the torque output is artificially raised a bit.  As the converter begins to catch up and the multiplication effect begins to go away, the torque plot gradually reduces.  By the time the converter completely catches up and the multiplication effect has completely gone away, the max boost output has already been reached (in this case, 20 psi) so the torque curve is already on its way back down.  The result is a steadily-decreasing curve across the plot (especially when the engine's top-end horsepower "corks" are still in place--notice the horpsepower peaked at a very low 3200 RPM).

[sharpxmen]

That dyno plot looks the way it does because of torque multiplication.  While the converter is slipping at lower RPM (and the turbo was already starting to build boost), the torque output is artificially raised a bit.  As the converter begins to catch up and the multiplication effect begins to go away, the torque plot gradually reduces.  By the time the converter completely catches up and the multiplication effect has completely gone away, the max boost output has already been reached (in this case, 20 psi) so the torque curve is already on its way back down.  The result is a steadily-decreasing curve across the plot (especially when the engine's top-end horsepower "corks" are still in place--notice the horpsepower peaked at a very low 3200 RPM).

but in that case how do you get a relevant power/torque/rpm curve - you usually get a dyno graph based on the rpm increasing from low to high, in what you describe would be backwards until the converter locks up (the rpm would decrease during the dyno run), not sure if that is relevant for the engine itself.

[Sherpa]

but in that case how do you get a relevant power/torque/rpm curve - you usually get a dyno graph based on the rpm increasing from low to high, in what you describe would be backwards until the converter locks up (the rpm would decrease during the dyno run), not sure if that is relevant for the engine itself.

The engine RPM never decreases throughout the run; it's just that the converter causes a multiplication effect on the torque output at low engine speeds.

To clarify: this dyno phenomenon is common when you're running moderate boost levels (15+ psi) on an engine with an automatic transmission, particularly when you set up the turbo to spool as rapidly as possible (which was indeed the case with that minivan, but I doubt anyone would ever try to set up their stock-engined 2.5L Jeep to spool as hard).  Here is a similar dyno plot from another Mopar four cylinder engine running big boost on an automatic transmission:




In contrast, here is the dyno plot from a Daytona which was mechanically similar to the minivan but had a manual transmission instead (and hence no torque multiplication):

[sharpxmen]

what i'm saying is that if the converter is not locked up, flooring the gas pedal on a dyno would result in the rpm to shoot high and then go down (not sure why you say it won't happen) - a dyno run should be performed with no converter slip or the result is not relevant for the engine itself (would be the same as slipping the clutch on a manual).

[Sherpa]

what i'm saying is that if the converter is not locked up, flooring the gas pedal on a dyno would result in the rpm to shoot high and then go down (not sure why you say it won't happen)

I say it won't happen because it won't, and it doesn't... at least not in my own real-world experience.

In case you haven't already figured it out: www.TurboMinivan.com is my web site, the dyno plot originally linked above is my dyno plot, and the van that generated it was my 1989 2.5L turbo Caravan.  Also, in case you missed it, my minivan had a reverse-pattern manual valve body.  This means I could put the van in 3rd gear at any road speed (including a mere crawl), floor the throttle, and there would be no kickdown into a lower gear.  This is what the dyno operator decided to do for my pull: he generated about 20 MPH of simulated road speed, shifted into third gear, floored it, then started the dyno's recording.  (This means he was generating the largest slippage/multiplication scenario possible--not what I was after, but I didn't know it at the time.)  The engine speed quickly began to rise (due to converter slippage/stall), of course.  This initial spike of artificially-accelerated RPM increase slows down a bit as the slippage decreases, which means the RPM rate of increase slows down for a moment but it never reverses--certainly not while remaining at WOT.

Let me give you another example: my usual dragstrip launch routine.  I bracket raced this minivan for six years.  Like any bracket racer, I was all about developing a consistent launch technique.  After some experimentation, I really hit the sweet spot for launching my particular setup.  For those of you out there who are interested in long-term transmission durability, do NOT try this at home!  Here is what I would do for every pass (this was after I installed a larger turbo that spooled later than in the dyno plot).  Pay particular attention to steps 6-9:

1: pull up to water box, wet tires, switch boost control to LOW, then apply parking brake.
2: brake with left foot, floor the gas with right foot, wait until boost starts to build.
3: let off foot brake so that tires instantly begin spinning; shift to second gear around 4500 RPM.
4: once tires start to "bite" release parking brake and then gradually close throttle.
5: pull up to the line and pre-stage; switch boost control to HIGH.
6: brake with left foot, gas with right foot until RPM levels off at 2800.
7: while holding 2800 RPM, carefully "bump" footbrake until van creeps forward and is fully staged, then reapply full foot brake.
8: still holding 2800 RPM, focus on the tree's countdown lights; when first yellow illuminates, floor gas the rest of the way.
9: when the third yellow illuminates, release foot brake and hang on.

I would hold the engine at 2800 RPM in order to generate some boost at a standstill; this means the converter was stalling heavily.  Yet when I finally released the brake, the engine speed would never drop--it would only go up since I was still at WOT.

a dyno run should be performed with no converter slip or the result is not relevant for the engine itself (would be the same as slipping the clutch on a manual).

I agree completely... but that's impossible to do on a chassis dyno when the vehicle is backed by an automatic transmission with a non-lockup converter (or even a lockup converter that isn't locked).

And on a more topic-friendly note: I'm already thinking about turbocharging my Wrangler... just as soon as I can get my Miata project completed.

[sharpxmen]

I say it won't happen because it won't, and it doesn't... at least not in my own real-world experience.

In case you haven't already figured it out: www.TurboMinivan.com is my web site, the dyno plot originally linked above is my dyno plot, and the van that generated it was my 1989 2.5L turbo Caravan.  Also, in case you missed it, my minivan had a reverse-pattern manual valve body.  This means I could put the van in 3rd gear at any road speed (including a mere crawl), floor the throttle, and there would be no kickdown into a lower gear.  This is what the dyno operator decided to do for my pull: he generated about 20 MPH of simulated road speed, shifted into third gear, floored it, then started the dyno's recording.  (This means he was generating the largest slippage/multiplication scenario possible--not what I was after, but I didn't know it at the time.)  The engine speed quickly began to rise (due to converter slippage/stall), of course.  This initial spike of artificially-accelerated RPM increase slows down a bit as the slippage decreases, which means the RPM rate of increase slows down for a moment but it never reverses--certainly not while remaining at WOT.

[...]
I agree completely... but that's impossible to do on a chassis dyno when the vehicle is backed by an automatic transmission with a non-lockup converter (or even a lockup converter that isn't locked).

And on a more topic-friendly note: I'm already thinking about turbocharging my Wrangler... just as soon as I can get my Miata project completed.

since it's yours you know better how it was ran and it makes some/more sense when explained, the discussion that started around the dyno graph is pretty much what you just explained as there was already turbo buildup and that's what we were missing (or didn't understand for a better term), but in all fairness i didn't read the whole thing so it could have been explained right there.

since you're here now i'll pick your brain a bit: can't remember what the max boost you mentioned (20psi?) - what CR are you running and how did you get it there? this is for my personal reference if you don't mind sharing.

oh, and just so you know my knowledge about autos is limited to what i gathered from driving my '86 Camaro quite limited

[Sherpa]

since you're here now i'll pick your brain a bit: can't remember what the max boost you mentioned (20psi?) - what CR are you running and how did you get it there? this is for my personal reference if you don't mind sharing.

The Mopar 2.5L turbo engine had a lowered compression ratio of 8.0:1 for durability purposes.  By comparison, the 2.5L AMC-derived engine as used in the TJ had a 9.2:1 compression ratio.  Needless to say, running high boost levels with a high(er) compression ratio is a very delicate balancing act; I certainly wouldn't recommend it for any "casual" wrenchers unless you don't mind rebuilding an engine after you've blown it up.

The old school FWD turbo Mopar engines have horrible cylinder head flow.  For example, the non-turbo TBI version of that 2.5L engine only made 100 HP peak at the crankshaft.  (Even the TJ 2.5L--itself no gift to head flow--manages 120 HP.)  To make serious power on the turbo Mopars, we have to run serious boost; 14-18 psi is pretty common and typically easy to do on pump gas (even for me where I'm limited to 91 octane here in Utah).

There are a number of guys around the country running 28-35 psi on their Mopar 2.5L engines, though I personally never went higher than 24 psi.  I usually could only run about 18 or 19 psi max on 91 octane; I would have to mix in some 100 octane in order to go higher (which is why I had a dash-mounted HI/LOW boost switch, with LOW set to 18 psi).

At one time I ran an old cylinder head which had been planed a couple times.  Fearing a higher compression ratio, I used a copper head shim to replace some of the lost chamber volume.  While this worked okay, lowering your static ratio will soften your off-boost power and response; I wouldn't really suggest doing this unless it was absolutely required due to head milling or you were dead-set on running very high boost pressure.

Anyway, I hope that helps answer your questions.  I don't pretend to know every last little thing about turbocharging, but I do have lots of experience and have done years of research about it.  I'll be happy to field any questions brought up here, and I'll let you know if any particular topic is outside my area of knowledge.