Individual Cylinder EGT Testing

[phate]

So I finally quit procrastinating on one project:





Here's a link to the entire album - EGT Album. I took quite a few pics of each cylinder with and without the thermocouples installed so you can see everything in detail.

I tried to keep the placement of the probes even between cylinders, and as perpendicular with the runner as possible. These are the measurements from inside each runner. I'd say they're within .050"

Cyl 1: Base- 1.50" || Tip - 1.50"
Cyl 2: Base- 1.45" || Tip - 1.25"
Cyl 3: Base- 1.35" || Tip - 1.50"
Cyl 4: Base- 1.35" || Tip - 1.45"

^^These measurements are to the flange mating surface.

I got it all installed and hooked up pretty easily. I dug out a serial-to-usb adapter that I bought ~6 years ago that wouldn't work with my previous computer. Works perfectly with this one, so it connected to the TC-4 and we had readings. The LogWorks program is really nice to work with, lots of things are adjustable on the dashboard view. This is what it looked like before I started the car (completely cold, had been sitting for ~20 hours):



This is what the cold start looks like in graph form. Peak occurred ~20s after it fired:

Here is the temp after ~18 minutes:

^^The orangish line on the gauge is the max temp reached. Immediately after you start the car, temp goes way up with cold start enrichment. Peak temps recorded (in order of cylinder): 1168-1065-976-999 °F.

I took it for a quick cruise. Here is the car completely warmed up, driving at 65mph with the cruise control on (had been on for a couple minutes to make sure everything was steady):

Pretty exciting, eh? 1176-1120-1086-1136 °F

No serious testing yet (it's late!). That will be tomorrow.

Any bets on how the cylinders compare at WOT?



For any newcomers to this thread, here is my build thread: Build of phate. This has my entire setup, but this section was deemed worthy of it's own thread, lol.


Update: Tips & Tricks for sensor placement and install:

-I measured, on the outside of each runner, 1.5" from the flange mating surface. I would do this @ 1-1.25" from the flange, bringing the sensor closer to the head. This would free up a little bit of space between the turbo (stock location) and sensor body.

-Cylinder 1 needs to be rotated 90° to the outside in order to clear the alternator. The other 3 cylinders should be mounted near vertically to clear the turbo and manifold studs/bolts.

-Replace the studs with bolts. The studs force you to place the mani ~1.5" from the head before it will slide onto the studs. That pushes the sensors on cylinders 3 and 4 into the turbo, forcing you to pull those studs, anyway. [If you follow tip 1, this may not be such a big deal.]

[RLam]

Awesome work Phate. I'm super jelly. I've always wanted an EGT setup, but have always been too lazy or inept to do it. I am curious on the results. How dialed in and consistent you can get your tune, most of all. Differences in temps of eth and regular, and what they look like with knock present.

[atvfreek]

Cool shit man. Do you have a stock intake manifold or ported? I would love to see what the more free flowing #3 runner would do compared to the others

Sent from your couch

[Ziggo]

Its great data, since EGTs are peak at stoich I guess cylinder 3 really is running off stoich, indicating a flow imbalance. Based on the flow bench it is probably running lean, but we will know soon enough I guess. If its lean, at WOT the EGT on the cylinder will swap from being the lowest measured to the highest.

[phate]

Originally Posted by atvfreek Cool shit man. Do you have a stock intake manifold or ported? I would love to see what the more free flowing #3 runner would do compared to the others

No, this is with the stock intake and exhaust manifold. I (we) needed a baseline to watch changes.

Originally Posted by Ziggo Its great data, since EGTs are peak at stoich I guess cylinder 3 really is running off stoich, indicating a flow imbalance. Based on the flow bench it is probably running lean, but we will know soon enough I guess. If its lean, at WOT the EGT on the cylinder will swap from being the lowest measured to the highest.

The results may be a little surprising...I may have done a couple pulls last night, already

When I have some time, I'm going to create a set of tunes that alter fueling without other changes, and a set of tunes that alter timing with no other changes. We should get a good feel watching those changes in each cylinder, so we can get an idea of what is happening between the cylinders.

[GODspeed7]

So you are telling me that at least now I will know EXACTLY why I am going to shoot rod #3 through my block?


Sent from my iBro 5 using tapatalk, Bro!

[Dano]

Badassness Clint !!

Give us a teaser of what # 3 looks like compared to the others. We've all suspected it is horribly lean but you can remove all speculation right meow !


Tappin

[Nliiitend1]

Good shit, Clint!

[phate]

Originally Posted by Dano Badassness Clint !! Give us a teaser of what # 3 looks like compared to the others. We've all suspected it is horribly lean but you can remove all speculation right meow ! Tappin

Remember way back in *I think* the intake manifold thread (JMF) where we talked about flow benches vs reality? And that what happens on a flow bench may not necessarily be what happens when the car is running? Judging by the pulls last night, it seems like it might be one of those cases.

But, I really do want to get those multiple tunes knocked out so we can identify what controlled changes do to each cylinder's EGT's. I just figured out how to export into .csv, so I'll be able to package AP and LogWorks logs together for simplicity.

Side note: I've wondered about this toggle and table for a while, now:

^^Curious if it actually works, and if there is a similar by-cylinder fueling compensation.

[Tokay444]

this thread fucking delivers.

[ms3blackmica]

If the ignition compensation control works, your EGT logs will show it for sure!

[phate]

I just finished the first session of afr testing. I'll try to compile all of this in something manageable, but there's a lot to work with. Afr's @ WOT from 10:1 to 14:1 (lol) in steps of 1 full point, 4 4th gear pulls per tune, both ap and egt logs.

[TiGraySpeed6]

damn phate, 14:1?

Balls.
very large ones too....

Thank you for doing this testing on your shiny new monster engine!

[phate]

I'm running approximately 50% E85 and 50% 93 octane. Stock intake manifold with stock exhaust manifold.

These tests were done by changing only target AFR's between the tunes. The ignition tables were set up to target the same advance value at each rpm range through the entire load region. The BAT vs ECT tables were zero'd to disallow compensation in case of temperature swings.

I'm only going to post up a few charts, but I'm attaching the "master" file, which has every run from every tune. It's the roll up file. I'm also attaching the tune roll ups for each tune; you can find individual runs in separate worksheets in these files. I'm not attaching the individual run files, there are 40 of them between the AP and EGT logs. If you want them, I will email them to you.

Anyway, on to the testing. Like I said before, I only wanted to alter AFR's and see the corresponding shift in EGT's. I started with 10:1 and worked up to 14:1 in steps of 1 full point. I won't post up charts of every parameter, but nothing significant changed between any of the pulls (bat's, ect's, iat's, etc, etc). This is what the average for each tune looks like (Load in blue, spark advance in red, and AFR in green):



Here, I've broken out each tune's EGT data into separate runs, charted together.

Blue=Cylinder 1
Red=Cylinder 2
Green=Cylinder 3
Purple=Cylinder 4

.68l (target 9.99AFR):


.75l (target 11.02AFR):


.82l (target 12.06AFR):


.88l (target 12.97AFR):


.95l (target 14.01AFR):






This is what the EGT's look like, side by side (lol):






We can see that cylinder 1 runs the hottest in every run, and the coolest is usually cylinder 3, and sometimes cylinder 4 (3 and 4 are very similar, though). This is the ΔT between cylinder 1 and cylinders 3 & 4 for each run:







Sooo, what I've learned today:

• For every 1 point change in AFR, we are looking at 50-125°F difference in EGT's. This difference increases the closer you get to stoich.
• Cylinder 1 runs the hottest, and presumably the leanest
• The ΔT between the coolest and hottest cylinders increases as you lean the engine out. ~80° @ 10:1 and ~160° @ 14:1

Subectively:
-14:1 AFR does not pull cleanly in this engine. No significant KR, but it was definitely hesitating in the mid range (felt like blowout).
-The 12:1 and 13:1 tunes seemed to pull the smoothest and strongest.

Question I have:
-Why does the EGT plateau when running it leaner (and hotter)?
-The EGT's in the 4 14:1 runs aren't consistent. I don't see anything in the logs that I can attribute this to. I'm presuming blowout has an affect here, anyone care to speculate?
-Cylinders 1&2 have the longest runners in the exhaust manifold, and are combined pretty early - anyone think this has to do with them running hotter?


Near future testing (I want to do it on the same tank of fuel) -
-test ignition advance at a constant AFR.
-test by-cylinder ignition comp tables to see if they work.

[Ziggo]

Awesome testing, thanks for sharing the results.

Zigatapatalka

[ms3blackmica]

What I always expected to see was cylinder 1 to have the highest EGT Clint as the runner has the end cap casted into the runner for the VCTS rod to rotate and be suspended in the center which reduces flow. I assume this would cause cylinder 1 to actually run richer since injection is commanded equally between cylinders showing a higher EGT for that cylinder in reference to the other 3.

Thanks so much for doing these tests.

[Etipp98]

Wow <3 You don't see any other car forums doing testing like this shit.

[BrianB6]

[Boosted Beluga]

After this do you have a ported intake mani to test? Then a jmf mani?

[phate]

Originally Posted by Boosten Ya(josh) After this do you have a ported intake mani to test? Then a jmf mani?

I personally don't have either. But, I'll do what I can to borrow one for testing.

[Boosted Beluga]

Then maybe an exhaust manifold? Or a series of exhaust manifolds? Dnp, cpe, fullrace, pg, etc...

[silvapain]

Originally Posted by phate I personally don't have either. But, I'll do what I can to borrow one for testing.

I have ported my IM and can trade you for your stock one for testing.

Originally Posted by Boosten Ya(josh) Then maybe an exhaust manifold? Or a series of exhaust manifolds? Dnp, cpe, fullrace, pg, etc...

Testing different EMs would require drilling and tapping four holes in each EM for the EGT thermocouples. I doubt many people would want that done to their manifolds.

[phate]

Originally Posted by Boosten Ya(josh) Then maybe an exhaust manifold? Or a series of exhaust manifolds? Dnp, cpe, fullrace, pg, etc...

Exhaust manifolds are a bit different. The tubular manifolds need an 1/8" NPT bung welded onto each runner for the probes. I doubt many people want that (I'm not buying them because I definitely don't need them). I have a DNP, though, and planned to get the bungs welded on at some point.

[Dano]

I think this topic deserves its own thread.

Will look at all the charts when on my PC.

Awesome data.


Tappin

[phate]

Originally Posted by Dano I think this topic deserves its own thread. Will look at all the charts when on my PC. Awesome data. Tappin

Will do tonight. I'm hoping to get the ignition change results after work.

[Spectrix]

Originally Posted by phate We can see that cylinder 1 runs the hottest in every run, and the coolest is usually cylinder 3, and sometimes cylinder 4 (3 and 4 are very similar, though).

Firstly, many kudos to you for completing such thorough logging and putting it all together into graphs. Seriously

Um.. I am the only one thoroughly confused by the results?

One would think Cylinder 1 to be the richest/coolest due to the poor runner air flow, though it has tested the highest EGT.

One would also think Cylinder 3 to be the leanest/hottest due to best runner air flow, and it's the most common to fail. Though it has tested the lowest EGT.

I'm wondering if the results are somewhat to do with exhaust manifold flow but I'm really at the point of guessing now.
Is anyone is able to have a stab at an explanation?

[Tokay444]

more fuel means more fire.
he's not measuring combustion chamber temp. it's exhaust temp.

[phate]

Originally Posted by Spectrix Firstly, many kudos to you for completing such thorough logging and putting it all together into graphs. Seriously Um.. I am the only one thoroughly confused by the results? One would think Cylinder 1 to be the richest/coolest due to the poor runner air flow, though it has tested the highest EGT. One would also think Cylinder 3 to be the leanest/hottest due to best runner air flow, and it's the most common to fail. Though it has tested the lowest EGT. I'm wondering if the results are somewhat to do with exhaust manifold flow but I'm really at the point of guessing now. Is anyone is able to have a stab at an explanation?

That's the burning question, lol. I don't have a good explanation for it, yet.

I'm almost to the point where we can change a variable - presumably the intake manifold, first - and we can track the changes. We should be able to narrow down issues as we go.

Another thought, and Ziggo pointed it out in a different conversation, is that the factory may have by-cylinder fueling compensation in the tune. [It looks to me like we don't have any by-cylinder ignition compensation, so I think we can rule this out.] We don't know if these tables even exist, let alone are being utilized, but it is a possibility. I don't think it's a stretch to say the OE's would have tested for cylinder variations and done some compensating with the tuning...Maybe @David@COBB has more info about this.

[Lex]

Yes, we need a scope with 4 channels to measure PW at each injector. I might be able to get to this in a week. This will tell us if the the injectors are commanded the same.

I know that in some cases manufacturers also "match" injectors by flow using a bar code but I haven't seen that here.

These are great findings. Also knowing that there is no power to split between 11 and 12 AFR this further points to running richer = cooler/safer and now we can quantify it.

[Yatta]

so comparing this with AP logs (fueling and MAF) should show exactly how much air is being pulled into each cylinder, would widebands on each runner show more (or at least avoid more math) or can you do all of this with just EGTs?

KUDOS on the rig, real time data FTW beats the hell out of my laser pointer IR gun....

[Spectrix]

Originally Posted by phate [It looks to me like we don't have any by-cylinder ignition compensation, so I think we can rule this out.]

Oh.. did that variable in ATR not lead anywhere? That's a shame.

Originally Posted by Lex Yes, we need a scope with 4 channels to measure PW at each injector. I might be able to get to this in a week. This will tell us if the the injectors are commanded the same.

THIS!

Originally Posted by phate I'm almost to the point where we can change a variable - presumably the intake manifold, first - and we can track the changes. We should be able to narrow down issues as we go.

...and THIS! ...is why I love this forum so much.
I'm really excited to see how this testing pans out, wondering what EGTs for the ported intake manifold are going to be like.

Many thanks @phate; and @Lex;

[BlueStreak]

Originally Posted by Lex Yes, we need a scope with 4 channels to measure PW at each injector. I might be able to get to this in a week. This will tell us if the the injectors are commanded the same. I know that in some cases manufacturers also "match" injectors by flow using a bar code but I haven't seen that here. These are great findings. Also knowing that there is no power to split between 11 and 12 AFR this further points to running richer = cooler/safer and now we can quantify it.

Should also scope the coil packs as there could be inherent spark advance compensation affecting EGTs.

[phate]

Originally Posted by Yatta so comparing this with AP logs (fueling and MAF) should show exactly how much air is being pulled into each cylinder, would widebands on each runner show more (or at least avoid more math) or can you do all of this with just EGTs? KUDOS on the rig, real time data FTW beats the hell out of my laser pointer IR gun....

If Lex can scope each cylinder's actual IPW, we will be able to narrow things down further without the need for O2 readings in each cylinder. Individual O2's would get pricey I'm sorta hoping for equal fueling between the cylinders just for the sake of simplicity, haha. [But I would love to have that option available.]

Originally Posted by Spectrix Oh.. did that variable in ATR not lead anywhere? That's a shame. ...and THIS! ...is why I love this forum so much. I'm really excited to see how this testing pans out, wondering what EGTs for the ported intake manifold are going to be like. Many thanks phate and Lex

Sorry, I didn't mean I had tested it yet. I just meant the by-cyl ignition comp table doesn't show any variation between the cylinders.

Originally Posted by BlueStreak Should also scope the coil packs as there could be inherent spark advance compensation affecting EGTs.

Confirmation would be ideal.

[TiGraySpeed6]

Some old school injector stuff, from @djuosnteisn; back in 2009

Have you measured when the injectors...

[Ziggo]

Afr probes don't tell you anything that the EGTs do not. It gives the ratio of fuel to air. If one cylinder gets more air, but also gets more fuel than another, then it will look balanced on both the EGTs and AFR probes and you will still need to rule out cylinder to cylinder compensation in the tune.

For a two cylinder motor, with either 3 wide bands or a wide band and two EGTs you would be able to infer the flow imbalance between the two cylinders by comparing the average reading to the two individual readings, the average reading will be closer to the individual reading of the cylinder that flows more.

For a four cylinder where you have one average, and four individual readings, there are too many variables and you end up with an infinite number of possible solutions.

Zigatapatalka

[Fatguy729]

Damn this thread rocks... But keeps getting more and more WAY over my head LOL

[phate]

I'm 99% sure the individual O2's aren't going to happen. Not only do I need to use 4 LC-1's, because they will daisy chain with my TC-4, but you also have to compensate for exhaust back pressure at each sensor. The back pressure "stretches" the reading away from stoich...rich reads richer, and lean reads leaner.

That, and I'm pretty sure these high of temps would dramatically shorten the life span of an O2 sensor, let alone the space consideration of having an egt probe, O2 sensor, and pressure sensor in every runner.

Then you have everything Ziggo already mentioned. Just doesn't seem practical or worth the effort.

[Enki]

Ditch the turbo, new custom manifold with for individual open pipe runners right to the ground.

Because science.

PS
The flow imbalance would be magnified due to lack of boost, right? Or does it not matter? If this is the case, then this might be a contributor to people blowing at low/no boost.

[Tomas]

Originally Posted by phate Question I have: -Why does the EGT plateau when running it leaner (and hotter)? -The EGT's in the 4 14:1 runs aren't consistent. I don't see anything in the logs that I can attribute this to. I'm presuming blowout has an affect here, anyone care to speculate? -Cylinders 1&2 have the longest runners in the exhaust manifold, and are combined pretty early - anyone think this has to do with them running hotter?

I think the EGTs plateau because it comes a point where running leaner does not necessarily translate into higher temps. I.e going from 15:1 AFR to 17:1 will not show a proportional increase in EGTs. John from PTP did some tests on this going up to 18:1 AFR and EGTs where the same from 15 to 18. That also leads me to believe that the second question you hav is not blow out related. I'd say it is likely that cyl 1 and/or 2 are actually running too lean. A 14:1 AFR might mean cyl 1 is running 13:1 AFR and cyl #4 a 15:1 AFR, From the tuning for fuel economy experiments we know the DISI can run 17:1 and higher AFRs without knocking the only problem being bugging or hesitation at those ratios.

As to EM log length I would say Cyl #2 definitely has the most restriction and could affect EGTs for that cylinder. The exhaust gas pretty much hits a wall when it exits the manifold.

Originally Posted by ms3blackmica What I always expected to see was cylinder 1 to have the highest EGT Clint as the runner has the end cap casted into the runner for the VCTS rod to rotate and be suspended in the center which reduces flow. I assume this would cause cylinder 1 to actually run richer since injection is commanded equally between cylinders showing a higher EGT for that cylinder in reference to the other 3. Thanks so much for doing these tests.

^ This.
You can grind the end cap for the VCTS rod and still maintain full VTCS functionality. I would also bet that restriction is the reason the high EGT on cyl#1.

[Ziggo]

I think you guys are seeing cylinder 1 backwards. If the hump was restricting flow, and it was not balanced by the tune, it would have less air, and more fuel, meaning richer AFRs and lower EGTs