123 Ignition Curve Thread

[Gregs672000]

Very good info here. It will be very interesting to see what curve my engine likes. I do have to be careful as compression is bumped to 11.7 to 1, running a higher lift and duration Isky cam, 45mm Dellortos with 37mm chokes (40s available), ported head, bigger intake and exhaust valves, etc. The megajolt has a built in MAP port and intake is providing a signal, but I have not yet changed the timing map based on vacuum, so it's input to mechanical timing is 0. You can use a TPS if desired but they suggest MAP. I should be able to pull out some advance based on pressure related to throttle position.

My timing map was set up on a dyno, but at that time we didn't realize how bad the cam timing was. With cam timing improved, the current curve has worked well so far but right now I am running a more mild cam (Racer Brown, a bit more/different than a B) than my Isky, and the Isky was better everywhere until I had to pull it due to wear. It's back in with new rockers and I got my suspension parts yesterday, so I hope to put the car back together in the next couple weeks and see how she runs with the Isky. If I feel she is good, I may play with the curve based on what others are seeing. I am also going to investigate adding a knock sensor.

[fj20spl311]

It'd be interesting to know what fuel they used as that may have necessitated more advance, much like the 85 tends to need more advance as I believe it has a slower flame front.

I don't think they had E85 alcohol blend, but leaded race gas. They had forged pistons and also did not care if the motor did not last a season.

[Gregs672000]

So I researched Megajolt and knock sensors, and clearly this is a difficult problem to work on, as each engine has its own noises, and using an off the shelf knock sensor set up for say a Chevy is not likely to work. Much of the discussion was about how to get the MJ to automatically back timing off when detecting knock and all the programing etc that would be required. All I would like for tuning would be a light or something that triggers when knock is detected. There was discussion about people using speakers etc to better hear knock, but the reality is that this is not an easy issue to solve.
What frequency is created in a U20 when it knocks? You would have to create that condition in the engine to calibrate a sensor, and I am not volunteering my engine for this study! Screening out all the other noises was the issue. Ping is hard for the driver to hear over induction, wind and exhaust noise. I can hear knock, but not before it reaches a certain volume. Ping is undesirable but ok; knock is NOT.

Mtresillian, regarding fuel octane ratings, I think our 92 octane here is about 97 or so where you are. If you're able to run it on 91, that must be closer to 87 or 89 here. Your compression ratio must be pretty low to run that without knocking regardless of timing (I would think it would detonate with that octane rating and stock compression under load), yet you're reporting about 145 whp... is that on a dyno or seat of the pants dyno? Not saying she is not making that, but it would be a considerable jump over what we often or typically see here, especially considering a possible drop in compression ratio. All the engine's components work together for sure, and it is possible everything came together for this engine "just right" and subsequently it is very happy. Maybe the EFI makes some difference there too with fuel delivery (I imagine it has a feedback loop using an O2 sensor? Maybe it's running a bit richer at times than a carburetor due to its ability to sense and adjust? A little richer would make it less likely to knock). I think Will (SLOroadster) may be experiencing this with his recent reports where his current "less than" engine configuration is making at or near the same power as his racier set up (seat of the pants dyno though, so no objective data). Keep the info coming!

[Gregs672000]

More research... found "knocklink G4" as a stand alone, self calibrating knock sensor. Reviews were scarce, but a couple people reported good results and decent accuracy. Not stupid expensive either. We'll see...

[mtresillian]

Greg
Lots to answer here.
My efi does not have an O2 sensor nor knock sensor. I have thought about an O2 sensor but understand their utility is mostly for light throttle cruise economy rather than optimising power under full load application. I know some self-learning and self-tuning ECU's read from the O2 sensor however. Because of my multi throttle setup i meter engine load based solely on TPS. This is a fine way to do it but once it's set on the dyno, thats the tune you have. It cannot measure the amount of 'air' going into the engine like a MAP sensor does so it just meters on the settings as per the dyno tune . TPS does give good throttle response however. I can compensate for air temperature and pressure using correction tables so that helps with changes in conditions that a MAP sensor would likely measure.

To 'dial in’ a knock sensor on the dyno, with the operator using ear phones, all they need to know is the bore diameter and that can be used to approximate the frequency of the ping. A nice solid spot on the block close to the cylinders is handy (my tuner used a smog mount bolt hole). Agreed i don't know how one would go about deciding which knock sensor to use for ongoing running in a car. Perhaps getting one from a car with a similar bore and a cast block? Most aftermarket EFI units have auxillary inputs which could be used to pull timing out of a map if knock were detected. It would be a good protection device on a highly strung engine.

Regarding the power output and compression. The exact ratio is hard to know as the engine was built by the previous owner. Lou Mondello said the CR is low on a u20 with dished pistons, even with 3mm taken off the head. It's 112.5mm thick, I got it wrong above. The cam was a good 12° retarded as a result but Lou fixed that for me. As you know, cam timing is imperative on these engine and the white dots are useless as mine were lined up! I so often hear ppl say their cam timing is good because they lined the marks up. A dial gauge tells the truth.

My engine has a pretty nice tune, the EFI probably helps with peak HP. My intake is 45mm all the way, no chokes, long intake trumpets. Mine were 110mm long trumpets. It also has the Long pipes that are made here. On Lou's engine dyno (Heenan and Froude) they give 20% increased power at 3500 rpm tapering to 8% at 6500. They transform the engines in a way i could not imagine. The standard 'headers' often used are poor scavenging pipes. So it's not a completely stock engine and all those little bits add up. The engine has not been on an engine dyno however. On a rolling dyno (really not cross comparable at all for numerous reasons) it showed 110rwhp. Nobody can really say what the driveline losses are so estimating flywheel hp figures is an imprecise activity.

Lou drove my car last week at the Nationals we held here and said he said it feels around 145hp. I bet he's close. He would have a better idea than me and has driven numerous 150 fwhp engines that he originally made many years ago. So yes I am guilty of proffering a 'feels like’ figure. My original point was that it's not a performance engine. Indeed Lou said to me..” I can tell the intake and exhaust are l excellent, it's the bit in between that is the problem”. Having driven a close to 200 fwhp engine of his just prior to that, i understood what he was saying. The u20 is a nice motor but overweight in many ways. The way his engine 2065cc pulled with gusto from 1500 rpm despite 50mm solexes and it kept hauling harder as it got to 7000. I didn't go further as it was not my engine, but it does. My car has a tiny little more zest on the road than solex engines cars with better comp and the same pipes. There are a few cars in my club with that setup and we drive together frequently. The difference is minimal though.

Back to knocking tho, my car was tuned to an AFR of around ~12.7:1 under load and ~14:1 on light throttle cruise. My timing is in many places 5 degrees below ping so there is leeway for fuel variations. The u20 likes compression but the longer intakes and proper scavenging pipes gives much more bang for buck.

Michael

[Gregs672000]

Thanks for the detailed post Michael! Lou certainly knows these engines, sure wish he was here in the STATES and I had unlimited Funds! Bottom line for me is that I just enjoy messing with things and seeing what comes. I've gone forward, I've gone backward, but it's always a learning process! I would love to have a custom header for the car. Ive talked to my friend Steve (cylinder head developer) and he was saying that you can guess at what will work but ultimately you have to spend the time on the dyno to know how well it's working... pretty much like cam timing! Lou has spent that time. Its all a system! Thanks for sharing.
I'm headed into the garage now to put more of the front suspension back on the car...

[Gregs672000]

Posting the ignition timing map for the Megajolt I'm running. High compression engine (11.7), Isky Z 197 cam (246 duration, 536 lift at .50 with our rocker ratio), 37mm choked carbs, header, 2"-2.25" exhaust. So far no significant ping below 3000rpms unless being stupid (hit it full hard in 4th gear at 2000rpms for example). This is at 60KPa manifold pressure, so at idle. Mainly copied Michael's earlier posted timing map.
RPM advance
1000 18
1500 23
2000 27
2500 30
3000 34
4000 37
5000 38
6000 38
6500 37
7000 36
I pull 2-4 degrees out based on load and rpm, or add a couple on deceleration (doesn't pop as much). No ping down low unless I'm really stupid, can stomp on it at 3000 with no problems. Not sure it's optimal, but it doesn't give me any trouble. I'm going to replace the coil as I noticed some differences in spark color the other day when testing. After that, some carb jetting and I'll be ready for a dyno pull.