philosophy of tuning

this isn't a writeup as such, but lets start a thread of nonspecific tricks that we've all picked up through trial and error, might make a good read for newbs.

here's one to start it off:

** if a small change does worse, but seems logical, try a big one! things do not respond linearly on these sketchy ol' engines, and sometimes the sweet spot is buried.

that has been my biggest mistake.

for example if it idled ok at 900 rpm and the afr was good, and got rough at 750, try 650! try 600! try less timing, more timing, less or more fuel. maybe you just hit a nasty overlap point at 750, or a rough spot in engine balance, who knows? experiment or you'll never know.

if it ran well at 30 deg adv in cruising range 'and it didn't like 34, and felt like surge, screw it, try 40! hell try 45 see what happens. remember guys ran total vac advance at 50-55, don't be afraid of it, it'll tell you if it liked it or not.

** never trust tune by mail guys

reputation means nothing anymore with wot track runs and butt dynos being the only measurement for email tune customers

I've found guys that just jack timing at wot through the roof and use KR to find the ceiling. if done right you can nail max power or pretty close to it, but damn

also seen very expensive street car tunes where they yank gobs (20+ degrees) of low end timing across the board

why? safety? cam surge?

nope. for the same reason disconnecting vac advance feels faster - just so it really snaps your neck back as it takes a short fast trip to the torque peak.

sure feels snappy when it goes from completely friggn limp in driving range to all in within 1500 rpm doesn't it? nice trick!

makes it break loose really nicely too, wow what a tire smoker

but you're leaving all your true torque band where that guy left your 200 bucks, which means a lot if you requested a 'fuel economy' tune ;)

(true story, but I'll never say which tuner..... so watch em all)

beware bench tunes, never base your tune on someone else's work! a factory cal is the only correct baseline to work from

in terms of broad, non-specific knowledge:

the engine knows what it does and doesn't like, it's up to you to find and determine good from bad. forcing your idea of what something should be is a good way of scrapping an engine.

consistent, thorough testing and changing the smallest amount of variables between tests will allow for good evaluation of if a change was good or bad. keep notes/logs around for longer than you think you'll need them, chances are, 2 weeks after you delete them, you'll want to review something that no longer exists. it will certainly take longer to finalize a calibration that way, but you won't overlook anything.

Quote:

---

the engine knows what it does and doesn't like, it's up to you to find and determine good from bad. forcing your idea of what something should be is a good way of scrapping an engine.

---

Yes.

Here's another: A wide band O2 sensor is not the only tool in the toolbox. As the tuner you should use everything you have available to make a good calibration. If the exhaust is so full of raw fuel that birds are falling out of trees, don't add more fuel just because the O2 shows lean. If your running gobs of spark advance and the WBO2 shows too rich at WOT, don't assume you can pull fuel out. Use your senses and think about a situation, especially if the clues seem to point to opposite solutions.

Quote:

---

Originally Posted by 1project2many

If the exhaust is so full of raw fuel that birds are falling out of trees

---

This is to be considered normal when tuning WOT in a big block. :happy:

Well, at least throw some alky into it so the birds don't recover.

Quote:

---

Originally Posted by 1project2many

As the tuner you should use everything you have available to make a good calibration.

---

The #1 tool is your brain. Blindly following some canned formula or instruction is a good path to failure. Understand what you are doing first or at least what you believe you are doing. Only then can you evaluate what effects your changes actually made.

The $42 TCC thread is a good example. The description in the Tunercat help file is wrong. If you blindly believe it you will be beating your head against the wall trying to figure out why your TCC clutch doesn't respond correctly.