Go Back   EcoModder Forum > EcoModding > EcoModding Central
Register Now
 Register Now
 

Reply  Post New Thread
 
Submit Tools LinkBack Thread Tools
Old 05-04-2021, 06:18 PM   #31 (permalink)
Master EcoModder
 
Ecky's Avatar
 
Join Date: Dec 2011
Location: New Zealand
Posts: 5,005

ND Miata - '15 Mazda MX-5 Special Package
90 day: 42.54 mpg (US)
Thanks: 2,866
Thanked 2,501 Times in 1,547 Posts
This afternoon I flashed a new experimental lean burn tune. From 1000-2500rpm, 0-500mbar, I'm running 1.2 lambda (17.62:1), and from 5-600 it's 1.1 lambda (16.17:1), with 5° and 2.5° of ignition timing added respectively.

I found the software's soft limit to be ~1.2 lambda. I can "trick" it into allowing me to flash a much leaner tune, but if I save and reopen my calibration in software, it resets back down to 1.2 lambda.

Even at that lean AFR, cruising, I'm seeing only around 360mbar, and I can climb pretty steep highway grades at <500mbar in 6th - even with gearing that puts me around ~1700rpm @ 50mph.

On my way home from shopping I averaged a hair over 64mpg, with a target speed of just over 50mph. I'd say in exactly equivalent conditions, this first attempt at a leaner tune is worth very roughly 3-5mpg, or approximately 6-7% better fuel economy.

  Reply With Quote
The Following 2 Users Say Thank You to Ecky For This Useful Post:
Isaac Zachary (05-04-2021), Juho (05-07-2021)
Alt Today
Popular topics

Other popular topics in this forum...

   
Old 05-06-2021, 08:52 PM   #32 (permalink)
Master EcoModder
 
Join Date: Jan 2012
Location: United States
Posts: 1,756

spyder2 - '00 Toyota MR2 Spyder
Thanks: 104
Thanked 407 Times in 312 Posts
Quote:
Originally Posted by Ecky View Post
I found the software's soft limit to be ~1.2 lambda. I can "trick" it into allowing me to flash a much leaner tune, but if I save and reopen my calibration in software, it resets back down to 1.2 lambda.
What software are you using? You can probably take the raw binary and jack up the values there. If it lets you go to 1.2, it probably will go higher. (as opposed to e.g. an early Bosch ECU, where it's encoded in 8 bit and 127 = stoich)
  Reply With Quote
Old 05-06-2021, 09:52 PM   #33 (permalink)
Master EcoModder
 
Ecky's Avatar
 
Join Date: Dec 2011
Location: New Zealand
Posts: 5,005

ND Miata - '15 Mazda MX-5 Special Package
90 day: 42.54 mpg (US)
Thanks: 2,866
Thanked 2,501 Times in 1,547 Posts
Quote:
Originally Posted by serialk11r View Post
What software are you using? You can probably take the raw binary and jack up the values there. If it lets you go to 1.2, it probably will go higher. (as opposed to e.g. an early Bosch ECU, where it's encoded in 8 bit and 127 = stoich)
Hondata's KPro.

I can trick the program into letting me enter a higher value into the field, and then I can upload it, but I can't save it.

Anyway, I think I figured out (one of the reasons) why Honda's lean burn engines all have a separate "lean burn mode" the car enters under certain conditions, rather than simply having it run lean at part throttle and get progressively richer at higher loads:

On a cold start, my engine stumbles for a minute if it's much leaner than 14.7:1. Meaning, I needed to set up a lambda limit curve, based on temperature. However, there is no way to progressively increase timing in the lean burn section of the map, without doing it to the entire map. Therefore, until the engine has been running long enough to lean out, it's running too much timing advance for the air fuel ratio. I imagine this isn't an issue when it's only 5 degrees ignition timing at <60% manifold pressure, but if I tried to go much leaner, and needed, say, 10 degrees of timing, that might be another matter.
  Reply With Quote
Old 05-07-2021, 12:43 AM   #34 (permalink)
Master EcoModder
 
Join Date: Jan 2012
Location: United States
Posts: 1,756

spyder2 - '00 Toyota MR2 Spyder
Thanks: 104
Thanked 407 Times in 312 Posts
I'm sure the ECU has such a parameter (since usually you delay ignition during warmup), it's just probably not one that people usually modify. Maybe if you dig or ask around you can find it.
  Reply With Quote
Old 05-07-2021, 02:14 PM   #35 (permalink)
EcoModding Lurker
 
Join Date: Apr 2021
Location: Finland
Posts: 29
Thanks: 3
Thanked 6 Times in 5 Posts
Quote:
Originally Posted by Ecky View Post
This afternoon I flashed a new experimental lean burn tune. From 1000-2500rpm, 0-500mbar, I'm running 1.2 lambda (17.62:1), and from 5-600 it's 1.1 lambda (16.17:1), with 5° and 2.5° of ignition timing added respectively.

I found the software's soft limit to be ~1.2 lambda. I can "trick" it into allowing me to flash a much leaner tune, but if I save and reopen my calibration in software, it resets back down to 1.2 lambda.

Even at that lean AFR, cruising, I'm seeing only around 360mbar, and I can climb pretty steep highway grades at <500mbar in 6th - even with gearing that puts me around ~1700rpm @ 50mph.

On my way home from shopping I averaged a hair over 64mpg, with a target speed of just over 50mph. I'd say in exactly equivalent conditions, this first attempt at a leaner tune is worth very roughly 3-5mpg, or approximately 6-7% better fuel economy.
I'm running about 1.2 lambda without problems, but on E85.
It's quite difficult to get a stable 1.2 or 1.25 lambda, because injector pulse width maps are quite bad for optimizing air-to-fuel ratio without use of lambda sensor.. so it's around 1.15-1.22, would like to try leaner but then it'll start bogging when sudden throttle changes.
But I have tried to run it around lambda=1.50 on low load but was prone to missfiring (sudden load/throttle increases).. and another problem was my wideband lambda which doesn't show leaner than lambda=1.50

It's a load x rpm map.
E.g load=200 and rpm=2500, then the next cell for load is 130 or 250.
  Reply With Quote
The Following User Says Thank You to Juho For This Useful Post:
Ecky (05-13-2021)
Old 05-09-2021, 07:19 AM   #36 (permalink)
Master EcoModder
 
Join Date: Jan 2012
Location: United States
Posts: 1,756

spyder2 - '00 Toyota MR2 Spyder
Thanks: 104
Thanked 407 Times in 312 Posts
Quote:
Originally Posted by Juho View Post
I'm running about 1.2 lambda without problems, but on E85.
It's quite difficult to get a stable 1.2 or 1.25 lambda, because injector pulse width maps are quite bad for optimizing air-to-fuel ratio without use of lambda sensor.. so it's around 1.15-1.22, would like to try leaner but then it'll start bogging when sudden throttle changes.
Have you tried ~1.05-1.1 at low load, and 1.25 at medium load? I think at low load, you have a lot of residual exhaust gas diluting the mix, so it becomes quite sensitive and combustion variance goes up. At medium load, you have more oxygen to get faster combustion.

I recall trying to run 1.2 at idle on my old car and it would stumble a little when hitting the accel pedal. 1.15 was OK.
  Reply With Quote
Old 05-09-2021, 08:13 AM   #37 (permalink)
EcoModding Lurker
 
Join Date: Apr 2021
Location: Finland
Posts: 29
Thanks: 3
Thanked 6 Times in 5 Posts
It goes to around lambda=1.12-1.14 at low load, because of the load cells.
But maybe with changing the load values I can get it work and try lambda=1.05-1.10 at low load.. and a bit leaner at medium load.

Edit: looks like I'm already running now lambda=1.25 at 100km/h ~flat roat.
Only luck, since I had at some point modified the throttle body map to open 0.5% at 10% TPS. There is some pedal movement allowed, only thing that changes is how much air the engine gets -> afr goes leaner slightly (from about lambda=1.20 to 1.25) if I press only slightly more throttle.

At lower speeds I have some work to do, so I could achieve the lambda=1.25

I'll try to find the limit where it starts missfiring or is too sensitive for missfiring.. but before that I drive the whole week with current ecu file so I can actually see if it's any good.
Did some minor afr and ignition timing optimizing.
And if my thermostat arrives next week, I test how it works with higher temperature.

The ignition timing maps do look quite different than with the catalytic converter headers.
Quite a lot less advance required for the ~same fuel consumption.

Last edited by Juho; 05-09-2021 at 12:44 PM..
  Reply With Quote
Old 05-09-2021, 06:35 PM   #38 (permalink)
Master EcoModder
 
Join Date: Jan 2012
Location: United States
Posts: 1,756

spyder2 - '00 Toyota MR2 Spyder
Thanks: 104
Thanked 407 Times in 312 Posts
Quote:
Originally Posted by Juho View Post

The ignition timing maps do look quite different than with the catalytic converter headers.
Quite a lot less advance required for the ~same fuel consumption.
Yea that's expected, the residual exhaust gas fraction is reduced with different headers. +10% EGR is on the order of +5 degrees of timing, so seeing several degrees of change at low load is completely normal.

Last edited by serialk11r; 05-09-2021 at 06:41 PM..
  Reply With Quote
Old 05-13-2021, 01:52 AM   #39 (permalink)
EcoModding Lurker
 
Join Date: Apr 2021
Location: Finland
Posts: 29
Thanks: 3
Thanked 6 Times in 5 Posts
I reached 7.9l / 100km based on the average fuel consumption meter on highway at 100km/h (real speed) and a little bit of city driving.
I reseted the avg. consumption meter after filling tank so the engine was already warm.

Heat helps a lot, works much better overall (kind of expected... ) ; about 25-26C outside temperature.
Engine could keep the speed at as little as 6l/100km momentary consumption for some time on flat road sections.
The fuel consumption meter is on the 'happy side' so it's out almost 10% (based on calculated fuel consumption when filling the tank).. true fuel consumption would have been about 8,7 liter per 100km, which is still quite well for this engine running on E85 since usually it's around 10l/100km, or more. Or gasoline ~8l/100km.

I got a few days ago new thermostat (Meyle).
What a failure, temps were even lower by few degrees. New OEM, here we go..

While revving the engine to reach thermostat opening temperature, I noticed intake and exhaust vanos making bad noises around 2500...3000rpm.
It's quite clear the vanos is not working well, which explains some random power losses I've experienced recently.

Should be expecting some fuel economy improvements with vanos service.
  Reply With Quote
The Following User Says Thank You to Juho For This Useful Post:
Ecky (05-13-2021)
Old 05-15-2021, 02:52 AM   #40 (permalink)
EcoModding Lurker
 
Join Date: Apr 2021
Location: Finland
Posts: 29
Thanks: 3
Thanked 6 Times in 5 Posts
Useful links related to lean burn:

Lean burn mode rules 70+mpg
https://ecomodder.com/forum/showthre...8mpg-9801.html

Lean Burn Success
https://ecomodder.com/forum/showthre...ess-36110.html

lean burn tuning ideas
https://ecomodder.com/forum/showthre...s-33565-2.html

E85 tuning, Subaru:
E85 Freeway fuel efficiency (300-400 miles per tankful)
https://www.romraider.com/forum/view...p?f=15&t=10725

____________________________

Increasing engine operating temp to increase MPG
https://ecomodder.com/forum/showthre...g-37356-2.html

BMW N43, N45 and N46 engines atleast are running on 105 to 112C coolant temperatures.
112C being an "eco mode", while the temperature will drop to 95C when going high load.

So it's very possible to run without major issues at higher temperatures.

I could possibly try at some point +100C coolant temperatures, if the thermostat just allows it.
Because the M52TU engine already has a MLS head gasket, and I'm running on E85 so knocking is not an issue = safer to go further than on gasoline.. within limits of course.

I checked yesterday my injectors, while doing other service (oil change).
They were very dirty and probably that was the cause of random missfiring at cold start (not getting enough fuel).. should replace my fuel filter, and maybe the injectors too.. I did attempt cleaning them properly, but failed to pressurize the fuel rail with injector cleaning fluid (whatever it really is).. so only cleaned them outside.

If there is too much difference on the injectors flow, I cannot get any accurate enough lambda readings (since it's only average of 3 cylinders) and probably getting random missfiring what may not be noticeable other than worse fuel economy and power output. Just another problem to be fixed.

  Reply With Quote
The Following User Says Thank You to Juho For This Useful Post:
Ecky (05-15-2021)
Reply  Post New Thread


Thread Tools




Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2024, vBulletin Solutions Inc.
Content Relevant URLs by vBSEO 3.5.2
All content copyright EcoModder.com