I started custom burning chips for a friend's 91 Caprice with an L03 and LT1 cam last night. Aside from the fact that I think I have VE dialed in well at stock fuel pressure (at least initially, until I really mess with the timing), I noticed it seems to require a lot more timing advance than stock. Stock was already a lot, at 46* max advance at WOT at high rpm. With a few corrections, I have a calculated 55* max advance at 5000rpm and 95kpa with no knock counts up there. Does this sound strange to anyone? Is the ESC perhaps not advancing as much as I think it is? He is running 93, btw...I know the L03 heads are terrible and they require a lot of advance to get an efficient burn, especially at higher rpms, especially with a bigger cam, but the number "55" is really bothering me.

Over a stock 91 Caprice LTZ L03 calibration, I have modified the timing as much as +12* in the no load to 80kpa, 400rpm to 2400rpm area, with still no sign of knock counts. Judging by how it drives, it still wants more. Is there a maximum advance for these "HEI" setups? Should we move the base timing to give the ECM more headroom?

Attached is a comparison table showing how much timing I've added over stock.

Comments?

 

There is a limit in the code, With a distributor you never going to get 45d, without miss firing.

Too much timing, and too rich often seem fast when in fact it's slower then being correct. You need to measure performance with a timer, not SOTP to get the calibration CORRECT

 

Without knowing what you are running, it is hard to comment on a "Comparison Table". You may be able to get away with 46* on a flat road at 45 kpa while cruising. But the moment you put ANY LOAD on the engine, you may find a massive amount of engine knock. (And don't go just by your ears, by the time you are hearing knock, it is already too late...engine damaging knock starts occurring even before you hear it).

Low RPM & high Load is often encounter when going up a hill before the transmission downshifts. This the most common time you are going to encounter knock. If you are lucky, your knock sensor will pull timing out but it MAY pull more out than you expect and actually end up with less "real spark advance" once you subtract all the knock retard the ECM pull away.

About the only time it is safe to get close to 45* is when you get into Highway Mode and a fully functioning EGR. The "weak mixture" results in sluggish performance and the 45* spark advance is more to compensate for the sluggish performance than anything else.

Lastly, as Bruce eluded too, there is only so much spark advance you can give before you start to cross fire across the distributor's terminal another cylinder. The distributor does have a little "advance" built into it, but with a V8 you can't go much past 45-47* before you begin to experience cross-firing in the distributor cap.

Be prepared to clean the cap on a regular basis, as high spark adavnce does promote "arcing". And, you may find "surging" when going up a hill (even if no knock occurs). This is usually the first sign of too much timing. And you better monitor your EGR's duty cycles. Once the EGR's DC starts to drop, you will have a mixture that is really prone to pre-ignition.

 

Grumpy: the GM factory stock calibration for a 91 Caprice L03 calls for a maximum of 46d under the following conditions:
95 or 100kpa, PE, 4200+ RPM

I have checked two L03 calibrations with tunercat (a Caprice Classic with 2.56 rear end and a Caprice LTZ with 3.23 rear end) and BOTH have 46d max advance. Considering this is the case, do you still think needing another 8d over this for an LT1 F-Body cam is too much?

Glenn: see above. Also, I NEVER EVER saw a single knock count. So I don't think there is too much timing.

56d is now called for by the ECM at 100kPa, PE, 5000+ rpm. This is my doing. And, again, no knock counts. I am using WinALDL, not just my ear or something...

Another tidbit of info to support the lots-of-timing thing: On the stock calibration, the ECM was throwing code 43. I thought about it for a while, and realized when it comes on and put 2 and 2 together - the ECM was failing the knock sensor test that it conducts in PE. Why? Because it was advancing the timing and never finding knock, so it sets DTC 43 and pulls tons of timing out (it drove even worse after the SES came on).

After adding the timing I mentioned adding in my original post, I have received confirmation from the owner of the car that DTC 43 is no longer being set (or any codes). This confirms that adding timing WAS what was needed. I only burned 3 chips, and I think it needs a lot more in the low rpms. LT1 cam in an otherwise stock L03 is an odd combination...

Anyway, thanks for the responses, and let me know what you guys think.

 

I'm having a hard time believing that the motor is seeing 55 degrees of timing. That's amazing! Either TC isn't taking into account the main spark bias OR your base timing in the eprom is off by a bit OR your actual dizzy is retarded by ~5 or more degrees base. I can't see it being possible that there isn't detonation or maybe the AFR is just that far off....but it feels really good like you said then that pretty much eleminates that as a possibility. I don't think the motor is seeing as much as is being thought. Check your base timing, zero out the main sa bias constant. Might as well zero out your initial sa constant and set the dizzy to 0 as well. Then output final calculated SA into the aldl stream in place of eprom id or battery volts. Then use a timing light and winaldl (raw values * (90/256)=degrees) to figure out what is really going on.

 

Are you sure that initial is set to 6 degrees?
Maybe its way retarded causing you to run high advance where you may actually only be getting 40 total...

 

its easy to get the numbers in those tables confused.

You most likely arent seeing anywhere near that probably closer to half. I like to zero out all that spark bias BS and set the initial timing at the distributor and in the chip to the same setting. I also zero out the pe spark advance. That way, what you see is what you get in the main table. I pretty much guarantee that spark bias crap is screwing you up.

Ive run into the same thing a million times

 

Guys,
The main spark bias is set at 20d, and the initial spark advance is set to 0. I timed the car and it may be 1 or 2 degrees ATDC, base timing. But not any more than that (91 Caprice calls for 0d BTDC, btw).

Stock, 46d is the max timing, including the main spark table, + the 20d spark bias + the PE added spark, plus the extended RPM spark advance (which is 7.7d/1000rpm and terminates at 4600rpm stock). This assumes no knock is detected.

Using the compare function of TC, I added 6.6d at 100kPa, 3600rpm (the last cell in the main spark table). I also changed the extended rpm spark advance to terminate at 5000rpm, instead of 4600. (5000-3600)/1000*7.7 = 10.78d added spark at 5000rpm, above 3600rpm. 100kPa 3600rpm spark = 23.2 + 20 (main spark bias) + 2.1 (PE) + 10.78 = 56d

The only thing I changed to affect max timing at 5000rpm was the one cell at 100kPa, 3600rpm.

Stock settings were:
16.9 (main spark, 100kPa 3600rpm) + 20 (main spark bias) + 2.1 (PE) + extended spark advance (4600-3600/1000*7.7) is 7.7 = 46.7d max spark advance on a stock L03. Am I adding wrong?

 

Yes, the main spark bias gets subtracted:

(16.9 - 20) + 2.1 + 7.7 = 6.7 deg

Unless the 16.9 deg is already compensated for by the main bias (which TunerCAT does):

16.9 + 2.1 + 7.7 = 26.7 deg

(and don't forget the coolant comp table & bias either).

Even if you crank in 57 deg of timing (via ECM tables) the distributor advance is limited in the code to 42 deg. With the initial & base set to 0 deg BTDC the max crankshaft timing will be 42 deg BTDC.

RBob.

 

26.7d doesn't seem like enough, does it?
I read the TC $42 calibration help file and you're right. Both initial spark advance and main spark bias are subtracted. And yeah, TC appears to pre-compensate for the main spark bias in the main spark tables. I'm just not sure why it's there. And 26.7d of max advance in PE doesn't seem like enough. I don't see any other tables that could add spark.

 

How do the coolant bias(all the time?) and coolant comp table work(only on light loads?)? Add or subtract?

 

I should clarify, the 91-93 B-Body uses $62, not $42. But TC has written a calibration help file only for $42...

 

When I back off the main spark bias in tunercat then open the main spark table, everything is higher by the amount I lowered it. So I took out the 20d of main spark bias, then opened the table of a stock L03 calibration and saw 36.9d at 3600rpm, 100kPa. When the PE adder (roughly 2d) and the extended rpm advance is added (7.7d/1000rpm) to 36.9, we get 46.7. I still don't understand why it's not 46.7....anyone?

 

I struggled with this question as well. The confusion stemmed from the way tunercat reports the main spark table. Rbob's example shows what is a happening. When using Tcat leave the MSB alone (stock) and don't use it in your calculation. This will eliminate any confusion.

In your last example you have done the math correctly. By lowering the MSB from 20 to 0 you force Tunercat to show the real value in the Main Spark Table (36.9). The problem is, you have just added 20 degrees of timing to the entire curve except where it exceeds the 42 degree cap. Running this bin would most likely be disasterous.

If you lower the main spark bias from 20 to 0 you have to lower all the values in the Main spark table by 20 also.

You may find this isn't possible because some values may become negative. There aren't any negative values in Hex, hence the reason for the Main Spark Bias.

 

Then I guess I'm pissed because TC must not have discovered all the spark adders for PE. An LT1 runs 29-30d of max timing at high rpm, high load, PE. An L03 is going to need more than 27d at high rpm, high load, PE - I am 99% positive about that.

I guess I'll fire off another e-mail to TC and see what he says.