15 replies to this topic

[OctaneNation]

Scratching my head on this one. Car was noticeably down on power last year. Checked ignition timing and found that it was way out. Suspected CAS. Changed it and didn't fix anything, so just reset timing and continued. Car was better but still down on power. Eventually found that the timing belt had slipped on the crank. So fixed that and thought all was fixed.

 

At the last race weekend the car still seemed slow compared to the rest of the field. Threw it on the dyno this weekend and found it down almost 17 HP from last time (2 years ago... about 5 weekends of racing).

 

My baseline 2 years ago (with tweaked AFR) was 38.5 psi and 13 degrees of timing. 

 

Things that have been checked:

-cam timing

-new plugs, wires

-swapped coils

-swapped AFM

-fresh fuel

-tried different fuel pressures from 35 to 46 psi; verified it's steady during a pull

-verified TDC with dial indicator

-exhaust is straight through and loud

 

Leakdown is maybe 15% right at TDC but immediately jumps to 5-10% on either side. Some noise past rings on two holes. Not the freshest of motors but IMO power is way down compared to leakdown numbers.

I didn't verify the correction factor being used by the Dynojet, just in case it's reading wonky.

AEM digital AFR guage reading and its analog voltage output agreed at idle, but when revved the output voltage (as measure with DVOM and Dynojet) deviated a bunch from the gauge digital readout. Not sure what's up with that.

Only other thing that comes to mind is last time I think I dynoed with a second fan blowing at a 45 angle right into the engine compartment. So maybe it's a bit low with just 1 large fan, but I can't see it being that much of a difference. Plus the car seems down on power compared to the field.

 

What else is there to try? ECU?? Injectors? (idle is fine) Temp sensor??

Attached Files

•  dyno.jpg   191.73KB   2 downloads

[SaulSpeedwell]

Do a coastdown measurement.  Run to 100 mph in 4th on the DynoJet, throw it in Neutral, start the DynoJet plot recording, wait for the over/undershoot to stop, and check the "horsepower" at some MPH that you will always use as a baseline.  I used to take the measurement from 80-60 mph, and record the midpoint.  The line will be flattish after the over/undershoot is done.  If you replicate that method, 12HP is bad/average, 9HP is good.  More is terrible, less is great.   It probably isn't your whole problem, but worth checking.

 

Keeping the car cool on the dyno is imperative.  Hood up?  Hood down?  All SMs hate hot air, 1.6s more than others.

 

The only drivers I've ever met who don't say they are down on power are selling motors, directly or indirectly. Check straightaway data with someone else, don't get tricked by the time-speed-distance relationship that makes perfectly equal cars appear to "...catch up in the corners and then he motors me on the straights!". 

 

 

Good luck, let us know what you find ... if it makes you feel better, if this was a BMW you probably would have DNFd 3 weekends ago? 

[OctaneNation]

Run to 100 mph in 4th on the DynoJet, throw it in Neutral

Didn't think to do that though I'm not sure I could find my baseline from a couple years ago...

Keeping the car cool on the dyno is imperative.  Hood up?  Hood down?  All SMs hate hot air, 1.6s more than others.

Hood up. Started runs at ~180F according to electric gauge in upper rad host. I think I ran with a second fan just blowing on the engine compartment last time. So maybe a few HP to be found on the dyno for a true apples-to-apples test but car seems slow on track.

The only drivers I've ever met who don't say they are down on power are selling motors, directly or indirectly. Check straightaway data with someone else, don't get tricked by the time-speed-distance relationship that makes perfectly equal cars appear to "...catch up in the corners and then he motors me on the straights!".

Several times I found myself with a slight run on a driver at the beginning of a straight and invariable the other car would pull on me. 3rd gear stuff--no drafting effect. It wasn't like this before which prompted me to dyno the car.

[Steve Scheifler]

Leakdown of 15% may be OK for a grocery getter but for these engines it is atrocious, so while it may not be THE problem it is certainly A problem.

 

What's up with the dyno AFR?  Did you not use their sensor and instead connect your AEM output to their analog 1?  That's excellent in theory, but they should be able to calibrate it accordingly to read AFR or Lambda, and you would do well to use theirs in the tailpipe for a sanity check on yours.  Looking at the 3rd graph, either your sensor is bad, or the output calibration has changed, or your AFR is WAY off.  Given the lack of power, I'm guessing option 3.  This is why using theirs for a sanity check is a good idea.  If it is extremely rich you may have washed down the cylinders, which in turn wears the rings, which might explain your leakdown.  If it is extremely lean you may have burned valves or the hot side of the pistons, which could explain the leakdown.

 

Forget AFR at idle for a moment, what did the AEM read under load from 5k-7k?

 

Along the lines of Mark's comments, did you check rear brake temps on the dyno?  I've seen 7hp from a single sticky caliper.

 

Edit: I see now that the 3rd graph is calibrated to AFR, I was looking too far right.  Way rich. If correct, that's the problem so fuel pressure, AFM, injectors are the place to focus. Look at all the plugs, are some blacker than others?

[Steve Scheifler]

Oh, and yes, intake air temp is critical to repeatability.  The extra fan will help with that by supplying air that is more consistent and closer to ambient temp than you otherwise get, assuming the dyno SAE correction is using ambient temp and not a probe stuck in your intake.  But as you said, the difference is typically low to mid single digits, not double.

[OctaneNation]

I'm not sure what to make of the AFR readings. My AEM gauge was showing say 13:1 but its analog output voltage equivalent was more like 10:1 or 11:1, and that's what was logged by the Dynojet. I confirmed the Dynojet transfer function and voltage reading with a volt meter and it didn't agree with the gauge display. So unfortunately all the latest AFR readings on the dyno are suspect. 

 

With that said maybe I'll check the CLT resistance at the ECU plug...

[Steve Scheifler]

Checking the CLT sensor is a good idea but check it at normal operating temp.  You could just pull it and test in a pan of water on the stove. The impedance curve should be easy to find. 

 

Need to sort out the AFR, but putting that aside for a moment, did you use a timing light to watch the timing during a pull?

 

Are you ABSOLUTELY sure that the cam timing is correct?  Very unusual from my experience to have a belt jump a tooth unless it is very badly stretched and loose or the tensioner spring is missing, broken or weak.

 

Too many things to rely on the "last thing you touched" rule so it's time to start over and check everything.

[OctaneNation]

I'd like to test the CLT signal to the ECU just to test all of the wiring/connector at the same time. Plus removing it isn't much fun for the knuckles

 

Didn't check timing under load as I was working by myself. Checked with two timing light against verified TDC mark, which BTW is out by 3*. So I set it at 16 BTDC with advance light set to 0. Both advance lights I have get confused by the waste spark so I leave them at 0.

 

Pics attached of cam timing. 19 teeth between the marks. And I'm pretty sure the power/torque curve would have a different shape vs just being moved down 9 ft-lbs everywhere.

 

And just got off the phone with AEM. The analog out ground needs to be tied to chassis ground for proper reading... They said this usually happens in the ECU but that's not my case where I'm using an external logger (Dynojet or voltmeter).

 

Attached Files

•  IMG_3045.jpg   86.01KB   1 downloads
•  IMG_3046-1.jpg   84.14KB   0 downloads

[Steve Scheifler]

Good point about the shape of the power curve.
I've never seen a 1.6 make best power at a true 13 BTDC, but again, not going to cost many HP. But we have multiple things that collectively could add up to quite a bit. I think you need to get the AFR sorted with certainty and watch timing live, then install a new plugs and immediately do some pulls before they get sooty from a rich idle or whatever, and have a look at them.

[OctaneNation]

I'll retest with a second ground on the AFR analog out and retest. Along with heat gun on rear calipers, check CLT and maybe new injectors. Is there a WOT signal from the TPS that should checked?

[Steve Scheifler]

Don't know, but if you have compression, spark, fuel and air in the correct amounts and at the right time you are close. Almost everything else impacts one of those.

[SaulSpeedwell]

I'll retest with a second ground on the AFR analog out and retest. Along with heat gun on rear calipers, check CLT and maybe new injectors. Is there a WOT signal from the TPS that should checked?

 

Yes.  The 1.6 TPS indicates 3 states:  Idle, WOT, and "Other". But the reality of the fuel/timing map is that dyno pulls and track duty all end up at the same open loop place on the map based on "big" AFM angle and "high" RPM.  So I strongly doubt the TPS is your problem.

 

17HP is a lot.  You can run 9:1 or 15:1, combined with running 10BTDC or 18 BTDC, and not lose that much.  Unless the dyno is just crazy wrong, I agree you are looking for something very fundamental - a seizing drivetrain part, a jumped cam tooth, or the motor may simply be failing/tired. 

[Steve Scheifler]

Add AFR way off plus timing way off plus 15% leakdown (if that's "new") and 17HP is not such a stretch. If the original 119/104 would be considered pretty strong on that dyno then I'd guess it didn't have so much leakdown at the time, but that really depends on exactly where the leakdown is. Another check while on the dyno, have it log intake manifold vacuum. It should be near zero under load and perhaps climb gradually but never get beyond a couple inches hg. Pull the air filter to see if that makes it drop. Pull the cam cover. Are the cam lobes all in good shape? Are any of the HLAs real soft? No broken springs? On the dyno, with the engine already warmed up, it might be worth pulling the exhaust manifold heat shield and cooling the tubes with a fan. Then fire it up and do a quick pull before using an IR temp gun to check each pipe right off the port. Is one much hotter or cooler than the others?

If it isn't just a tired engine it should show up with methodical testing.

[Bench Racer]

Pics attached of cam timing. 19 teeth between the marks. 

I viewed your pictures, I read your 19 teeth.

 

Is your 19 belt tooth count from the "U" notch directly above the "I" intake gear mark to the "U" notch directly above the "E" exhaust gear mark when the crank "U" notch is perfectly aligned with the upside down cover arrow mark?

 

https://www.bing.com...F971D&FORM=VIRE

[OctaneNation]

 

Add AFR way off plus timing way off plus 15% leakdown (if that's "new") and 17HP is not such a stretch.

AFR isn't way off despite what the graph shows. AEM requires a second ground on the analog out of the gauge so it wasn't logging correctly. The gauge read similar to the baseline run. I did a real quick leakdown to look for a smoking gun so I wouldn't take the numbers as gospel. Several times I was past TDC which unseats the rings and hurts the number. The previous engine had 50%+ leakdown on all the holes and ran suprisingly well. (It was an ex-rental car for who know how long). The leakdown was so bad that we thought the gauge was broken.

Pull the cam cover. Are the cam lobes all in good shape? Are any of the HLAs real soft? No broken springs?

Wouldn't hurt to check. Engine isn't noisy but maybe there's something there. As an aside, 2 years ago I installed a 'professional' race head that dropped a valve the first weekend out, so I put the previous head back on. The race head was worth ~1.5 HP.

On the dyno, with the engine already warmed up, it might be worth pulling the exhaust manifold heat shield and cooling the tubes with a fan. Then fire it up and do a quick pull before using an IR temp gun to check each pipe right off the port. Is one much hotter or cooler than the others?

Something else I'll check...

 

Cheers,

Matt

[Steve Scheifler]

You are probably right about the AFR, it's just that from your initial post all three factors seem to have question marks after them without a secondary way to check them. So I was just saying that if by some chance all three really were issues, it *could* explain the power loss without additional parasitic losses or the like. I know from painful experience that once you mentally rule something out it can be a long painful trip back to reexamine it. Just repeating a prior test isn't necessarily confirmation, when possible find another approach that should provide the same answer, and as Mark was suggesting, question each assumption and input along the way.