I've done some searching and can't find an answer to this question. Anyone know?
I've done some searching and can't find an answer to this question. Anyone know?
9:1 up to 17-18:1 roughly.
Obviously stoich is 14.7:1, most cars tuned for power are in the 13-13.5:1 range with turbo cars a bit richer for extra cooling/det resistance. Lean cruise is typically 15-15.5:1 area. And I think the old Honda lean-burn engines ran even more lean, but that sends NOx through the roof.
Hopefully Alfa will chime in.
9:1 is the rich limit? While working on tuning the Haltech in the RX-7, I was seeing AFR's dip into the single digits, like 9.8-9.9. It wasn't blowing black smoke or anything. Seemed pretty happy with all that fuel, although I do need to change oil often.
I'm pretty sure that old engines, like late 1800s-early 1900s, basically before they had everything down to a science, ran much, much richer than just 9:1.
In my experience, once i pegged my wideband below 9:1 the engine would start to 'flood out' and just lose power until it acted dead. I also own a Honda Insight that supposedly will run at 26:1 afr under very specific conditions. If you're talking about being able to run it at 'normal' power levels throughout the rpm range you're basically talking about ~9.5-17:1. The leaner end will be down power but if your ignition system can consistently ignite it it will still make a decent percentage of what it would make at a more normal AFR. It's just harder to light.
freetors wrote: I'm pretty sure that old engines, like late 1800s-early 1900s, basically before they had everything down to a science, ran much, much richer than just 9:1.
I didn't think we were talking about a Model T regarding this question.
1988RedT2 wrote: 9:1 is the rich limit? While working on tuning the Haltech in the RX-7, I was seeing AFR's dip into the single digits, like 9.8-9.9. It wasn't blowing black smoke or anything. Seemed pretty happy with all that fuel, although I do need to change oil often.
9.9:1 is less rich/more lean than 9:1
Yeah, it will run richer than that but you aren't making any more power and are wasting fuel and washing down the cylinders.
z31maniac wrote: 9:1 up to 17-18:1 roughly. Obviously stoich is 14.7:1, most cars tuned for power are in the 13-13.5:1 range with turbo cars a bit richer for extra cooling/det resistance. Lean cruise is typically 15-15.5:1 area. And I think the old Honda lean-burn engines ran even more lean, but that sends NOx through the roof. Hopefully Alfa will chime in.
That's the range I would have put out.
9:1 is in the range where you get into rich misfire, and it seems to be pretty common across a lot of engines. The lean misfire limit is pretty variable on engine design- but most will run pretty well up to 18:1.
(anything north of 15:1 will spike the NOx pretty badly)
Peak power range (aka LBT) may be 12.5-13.5, but if you are recording a production motor, you are more likely to see 11ish range. Peak power fuel only happens for a very short time, then it will go into temperature protection.
freetors wrote: I'm pretty sure that old engines, like late 1800s-early 1900s, basically before they had everything down to a science, ran much, much richer than just 9:1.
Actually, I disagree with that. You can hear and feel the engine running well- and there's always been some kind of tweaking in combustion fuel systems. It's more likely they were running in the very robust 12-13:1 range- you can hear and feel that's the best.
Thanks, the reason I'm asking is that I'm trying to figure out if the apparently false AFR readings I'm getting (last post with a video on it) could be plausible, and it looks like the answer is that, unfortunately, they could.
Honda has been making their lean burn engines for quite awhile now- the 99-00 d16y5 HX with the 4 valve to 8 valve on the intake vAtaak Yo setup is still being used and 22:1 in optimal lean burn mode pops up in the literature lots. As stated NOx is the bugaboo and on the older y5 Honda pumps exhaust into the intake at the head under specific conditions. I think they still do that with the new motors. I always wondered what that does chemically...
ah ha https://en.wikipedia.org/wiki/Exhaust_gas_recirculation
Well...
That's an interesting problem. Do you have any exhaust leaks? Or any fault codes for the sensor? Better yet, do you have a basic O2 sensor to drop in there? If it's actually 16:1, it will show .2 V all day long. But black smoke is usually in the 9:1 range.
Tip in knock is a pain, and given how MS works, you should either have code written to fix it, or use some kind of calibration patch.
Isn't there a case where you go so far in one direction that it starts reading the other? Like, so lean it presents itself as overly rich. Or am I thinking of when you are reading EGTs and not A/FR?
No errors from the sensor except there's one I can't test for right now, which is the sensor temp too high/too low error. I have to get an LED wired up to see that one, which I'll get done probably next weekend (not driving the car much now and I'm trying not to push it hard until this is sorted out). There doesn't seem to be any exhaust leaks, all the gaskets involved are fairly new. I do have a narrowband sensor with its own port and could hook that up, that's something I've been planning on doing at the same time as the temp LED.
I've made a little progress with the tip-in knock by blending in a MAPdot-based table similar to the TPSdot one. I made that table enrich more at low MAPdots just now, I'll see how that works on the way home from work.
If it seems to be running rich and the O2 is showing lean, you almost definitely have a misfire problem. At high RPM, if it's from weak ignition, it may not feel like an obvious misfire, but it'll be down a bit on power.
Misfires make the O2 sensor read leaner, as there's extra O2 getting into the exhaust from the un-burned cylinder charge.
I know the Haltech has a supplemental map called "Accelerator pump" or something like that which will give an extra bit of gas on reading sudden throttle opening. Is there something similar for megasquirt?
rslifkin wrote: If it seems to be running rich and the O2 is showing lean, you almost definitely have a misfire problem. At high RPM, if it's from weak ignition, it may not feel like an obvious misfire, but it'll be down a bit on power. Misfires make the O2 sensor read leaner, as there's extra O2 getting into the exhaust from the un-burned cylinder charge.
Interesting possibility. It's running Toyota CoPs at the recommended 2.5ms dwell, but I had the same problem when it was running 3.2ms~3.5ms dwell (which I'm lucky didn't fry the CoPs). It also has some fairly new iridium plugs that are still in good shape.
Isn't there a case where you go so far in one direction that it starts reading the other? Like, so lean it presents itself as overly rich. Or am I thinking of when you are reading EGTs and not A/FR?
EGTs will do that as you go lean. Hotter at first and then cooler as you start not burning all the fuel or the total thermal energy goes way down. Usually if you're at WOT and any kind of decent rpm it's going to be broken right about the time you notice that, i think.
1988RedT2 wrote: I know the Haltech has a supplemental map called "Accelerator pump" or something like that which will give an extra bit of gas on reading sudden throttle opening. Is there something similar for megasquirt?
Yeah, MS has all kinds of accel enrich features including a pump-style mode, but no burst knock feature, which works a lot like accel enrich but retards the timing instead of adding fuel.
rslifkin wrote: If it seems to be running rich and the O2 is showing lean, you almost definitely have a misfire problem. At high RPM, if it's from weak ignition, it may not feel like an obvious misfire, but it'll be down a bit on power. Misfires make the O2 sensor read leaner, as there's extra O2 getting into the exhaust from the un-burned cylinder charge.
Really slow or partial burns will do the same.
Another idea- SEFI, right? Real sequential- how's the injection timing? We typically play it safe and end it right before the valve opens. Could be that it's crossing over/through and burning terribly.
I've tried all different injection modes from batch with 2 squirts per rotation to semi-sequential with 4 squirts (I don't have a cam sensor so can't do full sequential). Injection timing from 360deg everywhere to a proper table that ranges from 400-452deg. No difference in the mysterious AFRs from any of it.
Did some logging today and WOW this is the longest sensible full-throttle run I've recorded so far:
The AFR fell to 11.1:1 at about 6.6krpm just before I had to lift for some slowpoke. The AFRs normally start to divorce from reality around 5.5-6krpm. This one matches up with the sensible data I got at the last autocross. I'm going to adjust my VE table based on this.
As a note, an easy way to identify a false lean AFR reading when tuning is when the thing is showing leaner than you want, so you add some fuel. And you end up having to add a ton of fuel to change the AFR a little.
You'll need to log in to post.
