Individual cylinder tuning via EGTs

What is the best way to approach individual cylinder fuel trim based on data received from EGT probes in each runner? How much deviation between cylinders for EGTs is acceptable?

How do you determine which cylinders to trim up or down.... take for example cylinder 1- 2 are running hotter than cylinder 3 and 4. Seems like you can end up with a cat and mouse chase between AFRs and EGTs

Do you have individual cylinder O2/Lambda sensors? If not, then I would first get the mixture average in the target ballpark. Then try individual cylinder trims looking at the EGT balance. Assuming you are lambda < 1.0, you will find that a richer mixture for a given cylinder will lower the EGT, and a leaner mixture raise it. Another thing that affects EGT is ignition timing, and ignition timing is affected by cylinder temperatures -- and not all cylinders are the same temperature -- so advancing timing for an individual cylinder should lower EGTs and retarding it will tend to raise EGTs.

Once you have the EGTs as balanced as you can get them (how good is too good?) Then revisit the mixture average to match your target.

Im running 4 EGTs probes close to the head and then a single probe in the collector. I don't have individual O2 sensors. just a single o2 sensor in the down pipe. Haven't done much testing as yet.. new engine/ set.... just ran its first heat cycle but ive noticed that at idle the EGTs at the head can be as far apart by a difference of 80 -90 degrees C. and im seeing a drop of over 100 degrees C between the probes at the head and the collector with what I would consider to be a relatively short exhaust manifold. Im rather surprised at the temp drop.

In the situation you have I try and match the EGT within about 20 deg C. You will never get them 100% exact and also remember that compared to the likes of a lambda sensor, the EGT is a little slower to respond. I'd also suggest only trying to match EGT in the high load/high rpm area. At idle and low load the discrepancy is hard to eliminate and also a little irrelevant.

When trimming the cylinders I personally add fuel to the hot cylinders rather than removing fuel. There isn't a right or wrong way of approaching this but that is my preference. Make you changes in small increments of 1-2% at a time and creep up on the correct trim.

Always use the lambda sensor for your actual AFR tuning and just use the EGT's to guide you in a cylinder to cylinder trim. I'd also mention I've seen a huge variation in EGT readings between exposed tip and enclosed tip EGT sensors. The exposed tip read hotter, react a little faster but also don't last as long.

Thanks ever so much guys... That approach makes lots of sense.

The 4 probes up at the head are enclosed tip with the 1 in the collector being an exposed tip.