A New Easy Way To Really Dial In A Tune?

I think you are diving much farther into this than required.
I'm enjoying the deep dive :)

As you have stated, never run egt's in the collector, always in the primary tubes. I personally see no benefit to one egt and would opt for all 8 if you decide to go that route.
I can't run that many through the BS3 and would have to opt for some other 8 channel setup. The short amount of reading I've done shows guys that put them in their "lean" cylinders.

The thought of individual cylinder tuning is a very daunting task. You can with most stand alone dfi's that run in sequential adjust fuel and timing on a per cylinder basis. This is tricky, different plenum designs typically lead to different cylinders receiving more air or less air than others. With a carburated setup where the fuel is atomized with the intake air the same general fuel/air mixture is distributed to each cylinder. With an injected setup varying amounts of air are fed to different cylinders due to flow inside the plenum while typically the same amount of fuel is fed to each cylinder. This is where the benefit to individual correction can be realized and typically the only benefit is seen at WOT.

That's a great point about carbed motors. They do a better job atomizing and distributing AFRs equally to the cylinders. Makes you wonder if port injection is the right approach.

I'm unclear why you wouldn't notice a benefit at partial loads. If you're going up a grade and you start getting detonation in a cylinder, you can adjust it and see a benefit too, right?

With out EGT's you must rely on reading the spark plugs

This is why I'm interested in the knock sensor. If it shows it's pulling timing on an individual cylinder, you don't have to rely spark plug reading, and with datalogging, you're going to know exactly where it measured detonation.

and make small corrections to your individual cylinders for a base tune up and if you do any sort of adaptive corrections based on AFR they must be global corrections to the maps.
So adaptive corrections are made to the entire map, but that's ok, right? I mean if the individual cylinder tuning has leveled the AFRs and detonation threshold across the cylinders, then you would only want adaptive correction to work in this manner, I'd guess.

Honestly I doubt I will ever do individual cylinder corrections with mine. If I was class racing and needed every last ounce of power from my combination than yes I would go down this road.
I'm kinda thinking it would just be fun to do. I didn't see going down this road either, but if the knock sensor will help me identify differences in individual cylinders, then I think it'd be cool to be able to tune these diffs out.

I personally can't afford to have my stuff on KILL all the time running ragged edge amounts of timing and lean AFR's. I prefer to keep things on the safe side and that is why I run one timing map and one fuel map for everything. When we pull plugs we check 5 & 7 because they seem to be the leanest two on my combo for whatever reason. All in all running the same map for everything there are only small variations in the way the plugs read. Almost all of them read the heat mark in the same spot on the strap.
Yeah, I can see why you wouldn't want to do individual cylinder tuning if it were already so close. I run one timing map, but 2 fuel maps (AFR and VE). I wouldn't want my stuff on kill/ragged edge either. Even with this system, I'd give it a buffer. That way, I'd know something was actually wrong if the knock gauge lit up. I do not want to trash a $10k shortblock, so I'm not going to be playing on the ragged edge, either. I just kinda think it would help me understand where that ragged edge really is and that I have a little safety net in case something is wrong.

As far as a knock sensor goes, I don't see this being a very successful device in a high horsepower application.
At some point, as I read more, I'll start posting some of the combos this system seems to have been successful on. I won't buy it if I don't think it's reliable for a 1000+ rwhp combo like I intend to use it for.

There is a lot going on inside on an engine that makes a lot of horse power. I remember years ago fighting with some of the GM stuff we played with, while showing no signs of detonation we would have knock sensors pulling the timing down from whatever was going on inside the engine.

Yeah, you know I was very skeptical when I started reading about this system. All I've ever heard about knock sensor systems is that they're very restrictive and unreliable in tuning. It doesn't seem to be the case with this system, unless you've got some bad piston slap. Some guys are saying if you're at or under about .006 piston-wall clearance, you should be in good shape with this system.

I personally feel rather than re inventing the wheel here, we stick to a good a/f gauge and some old fashioned reading the plugs. I like to stick to the KISS(Keep it Simple Stupid) principal to the best of my ability on everything I build.
I hear ya. Nothing wrong with that, especially since it's working so well already.
As far as the O2's going bad, we have a limit set to the amount of correction that the wideband can control. I know the tune up is good and depending on elevation/weather there will always be a small correction in the tune up. If the correction begins to go to far one way it will be limited. I think my correction is limited to 4% currently. I get an alarm in the laptop dashboard if the correction is trying to go past that. Had it happen earlier this year, was calling for 11% which wound up being a bad sensor.
Great point. I didn't realize you could do that. I wonder if the BS3 can. I haven't tried to use corrections, yet. I was still working on the base tables.
 
  • Sponsors (?)


Regarding the dart block, it sounds like this is the way to go:
[URL='http://forums.corral.net/forums/members/125808-andrewtac.html' said:
andrewtacI have a J&S knock sensor, using one sensor on my Dart block. There was a threaded hole one the passenger side on block, close to the starter (above it). No need to drill another hole.

A couple other ideas:
[URL='http://forums.corral.net/forums/members/140327-tomr.html' said:
TomRFor two, I machined a bolt that threaded into a threaded water jacket plug on each side. I used a pipe taper die on a hard brass rod. I used hard brass because it was easy to machine, soft enough to seal with minimal sealer, and would not corrode and lock into the hole.

I machined a large step or flange, and outside beyond that I machined to the inside diameter of the sensor with a threaded area to nut the sensor against the flange.

I found this really worked no better than mounting a single sensor on that wide flat area at the back of the block behind the intake. I drilled through the block into the area inside the bell housing, and tapped that hole. I now have a single sensor bolted down to that hole.

Other helpful installation or informational threads: http://reutterwerk.com/forums/showthread.php?p=205802
http://www.efi101.com/forum/viewtopic.php?t=8257
 
Last edited:
So after more reading, I don't think tuning off of a knock sensor is a good idea. From another forum:

I don't believe this to be an acceptable method as I have personally witnessed a vehicle on a dyno that made MBT at about 24-25 degrees. To satisfy our curiousity, we advanced it until knock occurred, at about 41 degrees. The engine configuration and fuel used made this engine very knock resistant. So, tuning as suggested would have the engine running somewhere well beyond the MBT timing point. You won't see that without the dyno output.

In the end, I guess it would still be much appreciated insurance against motor destruction and it would be a backstop to keep you from advancing timing or leaning fuel too much. But, I was under the mistaken impression that leaning and advancing would usually increase torque until you reached detonation, and sometimes even into detonation. Since this isn't true, the premise/subject of this thread also is not true.
 
Last edited: