Spark retard above 4K

[jckrieger]

Hey guys, I was wondering if you have had problems with the computer pulling timing due to false knock above 4K or so. I'm using the stock timing curves for my 2.5L and the cal is showing knock at all boost levels above 4K. The car has new PT lash adjusters and valve springs and the valvetrain is very quiet. Is this normal? What's the solution? I'm going to try 100 octane fuel and see what it does, but if it's still knocking at 14psi I'm going to unplug the sensor or relocate it.

[MiniMopar]

My motor has horrible piston slap and used to have a bum cam and soft lifters. Long story short, it made all kinds of racket. Never had a problem with false knocks that I am aware of. It does detect knock sometimes and if I keep pushing it, it rattles.

Is it a 2.5L cal?

[jckrieger]

It's the 87 T2 blueberry cal with 2.5L spark tables (advance vs map cold, warm, and WOT) and advance vs RPM (governor). I know I'm overfueling, but the engine itself is rather quiet and the car seems to run ok. The check engine light comes on at high rpm's (it lights when any knock is detected). I'm making good power, but I just can't figure out why the check engine light is coming on at 16psi. Is there a timing table I missed that is throwing things off?

[mw6886]

If you were to compare the spark knock detection tables vs RPM on say a 2.2 and a 2.5 you would see that at 4000 RPM on the 2.2 the volts that the computer says is knock is 1.38v. On a 2.5 at 4000 RPM the volts are 1.09v

You might want to copy the knock vs rpm table from the 2.5 89 cal over to the 87. You'll have to adjust it, as the 87 is based on 16384 RPM, and the 89 is based on 8192 RPM.

[ShelGame]

You should probably also copy the "RetardMaxFromMAP" table - the 2.5 allows a little more retard based on MAP. Maybe even add a couple of degrees to this table, since the '87 code doesn't have the adaptive boost retard that the '89 T1 code has...

[jckrieger]

Ok, I added a few degrees to the RetartMaxFromMAP table and I had already copied over the knock volts vs RPM table.

My new question is:
Where do I start with the fuel tables? The 3 fuel tables (no throttle, part, and full) are on completely different scales and don't really make sense.

[Frank]

Sounds like you also need to copy the pumping efficency tables over from a 2.5L 89 cal. This is key as it defines the engine as a 2.5L.

[MiniMopar]

I just eyeballed it.

[jckrieger]

I eyeballed the pumping efficiency table as well since I don't want the fuel leaning out from 5-6K by almost 20%! I tested the car out some tonight and I'm still getting knock at 12psi, so something is definitely up. Maybe it's time to buy some urethane motor mounts?

[MiniMopar]

I'd agree that something is up. Try retarding the base timing to see if that helps. If it does, then the cal may still be giving you too much timing. If you have a real 2.5L T1 car, you could hook up an OTC and observe the total timing that it is getting at various boost vs RPM levels and compare them to the 87. If they are close, then you can at least say that the cal is working.

[glhs875]

I eyeballed the pumping efficiency table as well since I don't want the fuel leaning out from 5-6K by almost 20%! I tested the car out some tonight and I'm still getting knock at 12psi, so something is definitely up. Maybe it's time to buy some urethane motor mounts?

You may need the fuel to lean out some in the upper R's. As VE goes down so does the fuel needed.

[jckrieger]

You may need the fuel to lean out some in the upper R's. As VE goes down so does the fuel needed.

With this turbo, head, and intercooler, I actually need to add more fuel in the upper RPM's and boost levels. The car is a monster that just needs a little more tuning. I currently have half a tank of race gas and the car is still "knocking", so I'm not sure what to do next. The motor does vibrate like hell because of all the preload I have on the bobble strut (solid with urethane), so I'm going to loosen that up a bit and see what happens. I thought the knock sensor would only generate voltages when shaken at a certain frequency, so I never though "low" frequency vibrations from the engine mounts would cause false readings.

[jckrieger]

I set the base timing at 0 degrees and ran the car at 15psi and it didn't set off the knock sensor... so I'm going to pull 4 degrees of timing from the advance vs rpm table above 3500rpm and see what happens with the base timing set back at 12 degrees. Even with the timing retarded 12 degrees, the car still spun the tires like crazy in 2nd gear :)

[MiniMopar]

There's all kinds of crazy harmonics going on in the intake, besides the noise from combustion. Like a piezo buzzer, these sensors (a piezo mic, if you will) will ring the best at a certain frequency, but they can pick stuff up all over the band.

[jckrieger]

Ok, here is what I've found so far:

-If I retard the timing by 12 degrees, the car sees no knock at 15psi
-If I retard the timing 4 degrees and set the knock vs rpm table back to the original 87 2.2L spec, the car sees no knock at 15psi
-If I retard the timing 2 degrees with the above knock table, the car sees knock in 3rd gear at 5K rpm at 15psi

In the top case, the highest EGT recorded was 1570F
In the bottom case, the higest EGT recorded was 1530F

What do you guys think? Should I really run 4 degrees less than the stock timing curve? This is all with about 97 octane fuel and good A/F ratios.

[MiniMopar]

Yeah, I think Rob may have the right idea. There may be more to the advance calcs in the 89 cal than just the tables. You would have to know what the final advance is with an OTC or datalogger to know for sure.

[jckrieger]

At least now I have the knock out of the way... I'm glad I didn't decide to just ignore it and unplug the sensor! I had no idea a couple degrees at such a low boost level would make that much of a difference.

[jckrieger]

Could running pig rich on fuel cause the car to read false positives on knock? I fixed my O2 gage and the car is pegged at 10 lights at all boost levels! I thought I was lean at .8V before I moved from the sensor ground to the battery ground!

[MiniMopar]

Yeah, the 4-wire sensors are nice, but if the guage doesn't have a sensor ground then you are not getting a valid reading. Anyway, I don't think running rich will generate knock. If anything, it should help quench it by keeping your EGTs lower.

[jckrieger]

I don't think grounding to the battery was a good idea... now the stupid meter always reads 10 lights whenever I'm at WOT... and I know I'm not THAT rich, as the car never misfires or cuts out... and the plugs aren't black. I'm going to go back to the sensor ground and use that as a baseline again until I get a wideband.

Speaking of, is there a way to calculate pulsewidths for a baseline calibration? I mean, how do you even start with an aftermarket kit like the Megasquirt?

[Frank]

They have a calculator which gives you a rough table estimate that is specific to Megasquirt... however I am the king of the internet...

http://www.stealth316.com/2-calc-idc.htm

Just run thru the white boxes after the "input parameters" header...


Frank

[MiniMopar]

Whether you ground at the battery or at the sensor ground, the key is to ground both the meter and the sensor to the exact same point.

[jckrieger]

That's exactly what we had decided when we talked about the conversion to the 4 wire O2. I'm sure my gage is reading over a volt now that I'm grounded at the battery! I'm going to go back to 15psi with the stock timing curves and see where I can go with the knock sensor unplugged. If I don't hear any knock at that boost level, well, I'll assume things are ok. Hell, maybe I'll log the sensor's output with my DMM and see if it's peaking at funky times!