1MeanZ

iTrader: (9)

I have a 91 TA 5speed with a swirl port headed TPI 305. I have tinkered with timing on the car for a long time and Ive gotten where I think it should be for the most part. I started with the TPI timing map and made changes as needed.

My dad just bought a 91 RS 305 TBI 5speed car. I'm going to play with his tune so I pulled the stock L03 5speed timing maps up and they are MUCH more aggressive than the TPI tune appears to be.

What confuses me is that I've heard so many people on here say how lame the L03 timing is. In reality it buries the TPI tune at nearly every point. I've even read in the TBI stickys where guys routinely advance their base timing 6-8* with no problems.

I thought the TBI swirl port heads were supposed to be good mixers and less timing would be needed.

I guess I'm just confused about the vast difference and how it compares with what some of you have found in your tuning experience with TBI and TPI vehicles.

I'm trying to post graphs on here but it's being a huge PITA....

1MeanZ

iTrader: (9)

1MeanZ

iTrader: (9)

alvanwie

I believe the tbi table you posted is the table value before subtracting out the main spark bias. The main spark bias is normally 20. So.... the actual spark advance would be the values shown -20.

Fast355

iTrader: (2)

The only $61 stock calibration I have in front of me is a 350 cop car. But that table is WAY WRONG....The actual WOT timing would be generally around 20*BTDC on a TBI engine. The offset is not being accounted for in that graph or table. The TPI setup gives more timing, sooner.

Here is what the true (very rounded)WOT (100 KPA) timing curve looks like with this calibration

RPM-------SA
400------ (-1)
600-------(-1)
800-------(-1)
1000------(1)
1200------(6)
1400------(6)
1600------(8)
1800------(9)
2000------(10)
2200------(11)
2400------(11)
2800------(11)
3200------(12)
3600------(13)
4600------(17)

1MeanZ

iTrader: (9)

Fast, this is what I thought was happening. I wondered if my .bin or my .xdf was corrupt or wrong. It appears my .xdf was wrong. For the stock 91 TBI 5speed .bin I'm using AXKW. I was using the generic 61.xdf I got from moates' site. I'm now using Bens 61 for 8746_TBI_A.xdf and it gave me the timing values you suggested. THey are posted below. I deleted the TBI timing maps above because I didnt want wrong info to come up of someone does a search.

1MeanZ

iTrader: (9)

As a further reality check, do these look right to you guys? (I'm going to have lots of other questions after this but a search will likely answer them I hope.) This is the first TBI car I've tuned, I've done 4 TPI cars now.

My primary concern is the main VE table at the top right. As you can see the max values are only like 51%. Is that correct?

RBob

iTrader: (1)

For the VE there is an adder table. Goes from 0 to 6400 RPM. This gets added to the value from the main VE table (shown). The max is 100% VE, but some calibrations add up to more then this. But it stops at 100% inside the ECM.

For the SA versus which XDF to use. Many years ago we didn't have calibration editors. We hand edited the BIN in a hex editor.

Eventually calibration editors were created. These early editor(s) had limited capability. They would show the main SA table without taking into account any bias value. Which was OK as long as you mentally accounted for the bias value.

Then along the way better/updated editors came along where items such as bias offsets could be taken into account. But it was up to the person that created the XDF file to do this. So, some do and some don't. Just need to check whether they do or not then account for it.

Note that in the '8746 ECM there are two spark bias values. One for the main SA table and another for the coolant compensation SA table.

RBob.

1MeanZ

iTrader: (9)

So if VE table #2 has a value of 50% at 2400rpm, that means you add 50% to ANY value in the 2400rpm column in VE table #1? So then if you have 60% VE at 90kpa at 2400rpm in table 1, and you have 50% at 2400rpm in table 2 this gives you a total VE of 110%. However the ECM won't recognize any value over 100%?

I don't intend to throw the author of an xdf under the bus. I'll openly admit that I rely on the good will and talent of people that frequent this site as well as the software and hardware developers (magnus and moates) to do any tuning. If I had to come up with this on my own I'd be lost as a ball in tall weeds. There are places where I can help others and I make effort to do so. This is where I come to cash in on the favors I do for others on this site. My point is, I know how well I have it. And the reason I have it as well as I do is because lots of people have spent their time and money to move the DIY community forward. I want to take a minute right now to acknowledge those folks and show appreciation for what they've done. (Rbob, this means you, Grumpy RIP, and many others too numerous to list.)

I noted that the spark bias value in the stock tune is 20.04. If I subtract that value from the tables I originally posted, I do get the real timing values, so that problem is solved in my brain now. Can anyone explain why the spark bias values are there? If I could learn to read code I'm sure I'd be able to find it, but thus far I've not gotten very far trying to read the code myself.

alvanwie

"Can anyone explain why the spark bias values are there? If I could learn to read code I'm sure I'd be able to find it, but thus far I've not gotten very far trying to read the code myself."

Without getting to a lot of detail, it has to do with how computers handle negative numbers and the desire to maximize the resolution available from an 8 bit number. So basically by using the offset (spark bias value) to set the range of the numbers represented to account for the negative values you can still maximize the resolution of the 8 bit number by storing all the values as positive values.

Hope that makes some sense to you.

RBob

iTrader: (1)

That is correct. The ECM firmware knows that the value overflowed and sets it to 100%.

RBob.

1MeanZ

iTrader: (9)

Absolutely. That is a good explanation that does not get in over my head.

Thanks for your help guys. This is why this is the best DIY tuning forum on the 'net...