Help fine tuning emu black tune for MOT

[jvanzyl]

Hi all- would it but possible to get some time with anyone on hear to try and fine tune the map on my ecu master black so that it stands a better chance of getting through an MOT?

Basically a screen sharing call via zoom or something where you can see the map and provide input on what I can/shouldnt alter to try and improve things?

Cheers!
John

[Gaz2405]

Have you tried running it in auto tune once it's warm?

Let it idle, run auto tune, and rev the car up to where they test for emissions.

Do it for 5 minutes or so.

Then apply the auto figures to the fuel table. (I do it at the 50% value so it's small increments)

Then highlight the surrounding areas of the map,right click and interpolate to smooth the map out.

Rinse and repeat.

Always save your original map down before making any changes.

[jvanzyl]

Have you tried running it in auto tune once it's warm?

Let it idle, run auto tune, and rev the car up to where they test for emissions.

Do it for 5 minutes or so.

Then apply the auto figures to the fuel table. (I do it at the 50% value so it's small increments)

Then highlight the surrounding areas of the map,right click and interpolate to smooth the map out.

Rinse and repeat.

Always save your original map down before making any changes.

I have not tried this (didn't know about it) but it sounds ideal- I shall do some googling on how to actually do it.. might pop back into your PMs with a question or two on the actual implementation of this exercise when you say things like "at the 50% value level".. thanks very much again for the pointers!

[Gaz2405]

I have not tried this (didn't know about it) but it sounds ideal- I shall do some googling on how to actually do it.. might pop back into your PMs with a question or two on the actual implementation of this exercise when you say things like "at the 50% value level".. thanks very much again for the pointers!

No worries, yeah drop me a pm when you've had a play.

It's pretty intuitive to use.

[jvanzyl]

So first attempts at tuning started this morning! Watched a few videos on it last night.
The clear point of failure with this approach is if the initial target tables are not setup sufficiently well enough... I noticed for example that the injectors are setup to be 500cc even though they are the 600 odd whatever's..
Any way the car is significantly closer to the target lambdas than it was before at least at the idling to 3.5k rev range at stand still..

[simonrobinson]

Don't change that injector CC value as it will totally put the whole thing out of tune. It does mean that all your VE values are total skewed which is very poor practice. There is only 2 areas of the table that matter for the MOT. Look for the cells that it uses when you hold the throttle to get it to rev between 2000 and 3000 rpm at standstill. And secondly the area where it idles naturally. The aim is to get your wide band to hit Lambda 1.0 to 1.01. Make sure you have no exhaust leaks, put your hand over the tail pipe and listen for leaks, leaks are your enemy and will total screw your emissions targets and test result. You don't need to use the autotuner for this, just manually set your target lambda correctly for these areas, then adjust your VE table to suit, watching your lambda value and your correction value. These cells will not effect your driving tune as the load is nothing.

[jvanzyl]

Don't change that injector CC value as it will totally put the whole thing out of tune. It does mean that all your VE values are total skewed which is very poor practice. There is only 2 areas of the table that matter for the MOT. Look for the cells that it uses when you hold the throttle to get it to rev between 2000 and 3000 rpm at standstill. And secondly the area where it idles naturally. The aim is to get your wide band to hit Lambda 1.0 to 1.01. Make sure you have no exhaust leaks, put your hand over the tail pipe and listen for leaks, leaks are your enemy and will total screw your emissions targets and test result. You don't need to use the autotuner for this, just manually set your target lambda correctly for these areas, then adjust your VE table to suit, watching your lambda value and your correction value. These cells will not effect your driving tune as the load is nothing.

Thanks for the pointers- didn't realise you were so close! I'm over in Kings Langley. My exhaust is v-banded together but will check when cold.

After a fair bit of puttering around today below is the summary of changes I've made to the fueling table:



And here is the target lambda table



Anyone got any comments on what level to set the transient to? Mine's currently on 500ms.. most videos I see have it down to 100ms..

[simonrobinson]

You cannot view this attachment.

Which is the basic cat test and MOT testers are not very good at hitting target rpms.

CO up to 0.2% at fast idle (2,500rpm to 3,000rpm)
HC up to 200ppm at fast idle (2,500rpm to 3,000rpm)
Lambda between 0.97 and 1.03 at fast idle (2,500rpm to 3,000rpm)
CO up to 0.3% at idle (450rpm to 1,500rpm)

Your lambda targets are too rich 0.94 & 0.95 in the MOT tested area in your table.
Remember the MOT is stationary, you can fix this tuned area parked up.
It needs to be targeting and achieving a proper stoichiometric mixture = AFR 1.00
This means you get a complete burnt, with no left over HC's (fuel) and no CO (monoxides)

Use the log graph in ecumaster, graph you lambda and rpm. Do yourself a dummy MOT test. Idle for 30 seconds, then hold 2500rpm for 30 seconds. You want to see the lambda graph holding a rock solid 1.0 without bouncing about.

[Gaz2405]

Don't change that injector CC value as it will totally put the whole thing out of tune. It does mean that all your VE values are total skewed which is very poor practice. There is only 2 areas of the table that matter for the MOT. Look for the cells that it uses when you hold the throttle to get it to rev between 2000 and 3000 rpm at standstill. And secondly the area where it idles naturally. The aim is to get your wide band to hit Lambda 1.0 to 1.01. Make sure you have no exhaust leaks, put your hand over the tail pipe and listen for leaks, leaks are your enemy and will total screw your emissions targets and test result. You don't need to use the autotuner for this, just manually set your target lambda correctly for these areas, then adjust your VE table to suit, watching your lambda value and your correction value. These cells will not effect your driving tune as the load is nothing.

Good comments there Simon, I'd only suggested using auto tune, if Jay wasn't confident in making changes directly to the fuel map himself. (In terms of the figures required)

But like you say, it's pretty self explanatory once you get going.

Easiest way like you say, change the cells highlighted whilst idling, I like to highlight the surrounding cells too to stop any interpolation from moving one cells too much and to try and smooth the transition out that way.

Good work, J keep going.

[jvanzyl]

Hi both @Gaz2405 & @simonrobinson i really appreciate the input!
What you are saying makes sense with regards to what the theoretical process should be to achieve the right values for MOT... what I'm wondering is WHY the target values have been set this way?

Surely there must be a reason why RRR did this, I even specifically stated that I needed this as a daily driver and would be looking to get it through an MOT.

As it stands I think I'll be fine tuning the current map and then building a map that I will call "peoplescar" that has the correct lambda targets and looking to load that upon arrival at the mot Center..

[Gaz2405]

Hi both @Gaz2405 & @simonrobinson i really appreciate the input!
What you are saying makes sense with regards to what the theoretical process should be to achieve the right values for MOT... what I'm wondering is WHY the target values have been set this way?

Surely there must be a reason why RRR did this, I even specifically stated that I needed this as a daily driver and would be looking to get it through an MOT.

As it stands I think I'll be fine tuning the current map and then building a map that I will call "peoplescar" that has the correct lambda targets and looking to load that upon arrival at the mot Center..

RRR, are known for giving a good WOT tune, but sometimes not getting the drivability spot on. As that takes a little more time.

[simonrobinson]

To get a reasonable all round calibration I think is at least 3 full days work. Doing all the WOT stuff to achieve MBT via ignition, fuel and vvt is a couple of hours dyno time if hardware is working good, its the fun stuff that gets people excited. Now the reality! Cruising areas, tip in, transients are almost endlessly time consuming, but that separates a good tune from the rest, and is best done on the road. What makes all this more viable is accurate calibration of intake air temp sensor, coolant senor, tps, crank trigger drift calibration, wastegate boost control and injector flow cc's which all in can be a days work in itself. The last icing on the cake is the cold start and warm up, you only get one shot in the morning at tuning that, and its seasonally different, I admit mine isn't great below 5 deg C.

Just had a flick though your daily commute smile build thread, looks good, well done, needs an MOT asap for daily boost smiles while the weather is better! When you have rich lambda and excessive monoxides it points toward excessive fuelling - looking at your target lambda table v's your MOT test result sheet its hitting its target very well. And that is your problem, the target is too rich. This is not a Cat fault as your HC's are passing OK - well only just within the 200ppm but that is due to running excessively rich target AFR. Truth be known pretty much any car tuned correctly will pass the standard emissions gas test without a catalytic converter even present, only to really fail due to cat being visually not present.

This is my target lambda table on my ecumaster black, but realise its a naturally aspirated engine. Only the two right columns are used to make any power with an NA engine.



I would suggest something like this should be where you need to be for MOT requirements, but have a little test to sanity check it obviously, and for new players never assume a wideband sensor reading is real world accurate. Injector fuel quantity delivery differences and exhaust leaks skew wideband readings along with duff sensors.



Keeping your target AFR's to 1.00 in areas where you are cruising the car about in traffic and steady speeds will wildly gain you better MPG's, possibly beating a stock NA 1zz as the turbo yields you better thermal efficiency. The OEMs target cruising at 15.5:1 being 1.05 lambda for fuel economy and CO2 emissions. Meanwhile running excessively rich in lower load areas cause excessive fuel on the cylinder walls leading to oil dilution and ring wear, along with quickly fouling your wideband sensor which destroys its accuracy. Daily driver realities v's race car WOT tuners! You've got this far with the build, now its time you get hands on with the tune and you'll be wiser for it. Studying logs is very easy on ecumaster and is an investment for reliability. P.S Complement to the charge cooler, welding and flip flops.

[Alex Knight]

@simonrobinson

That's a fantastic post.

Chapeau 🎩

[jvanzyl]

@simonrobinson

That's a fantastic post.

Chapeau 🎩

That is an indeed excellent post @simonrobinson ! I'm looking forward to mapping this further early next week when I get a chance! Your post was really helpful.

[jvanzyl]

Tried again with the MOT.
Basic result is that the CO is too high despite the wrapping of the mid pipe, and despite it sitting at about lambda 1 (according to my Wideband).

Additionally it seems my Wideband sensor is off by about 0.06.. compared to the MOT test centre sniffer which doesn't help things.

I welcome comments on what to do.

Bah.. will post the actually readings when I have them in my hand.

[jvanzyl]

before it was wrapped cat and changed tune:


After wrapping and tune change:

[simonrobinson]

Lambda measured at the MOT is 1.269 = Which is 18:1 AFR! That by oxygen measurement is insanely dangerously lean. But your HC's are very high, which is unburnt fuel. Those two together is impossible if the combustion/exhaust is sealed and true, how can combustion have unburnt oxygen and unburnt fuel, think about it. Maybe you have an exhaust gas leak which is causing oxygen ingress in to the exhaust? Maybe you have poor ignition/misfire which is leading to unburnt fuel and oxygen? Was the car tested with a cold or hot engine? Ideally you want to get it to full temperature then have it tested immediately as cold start enrichment will be against you. Either way you need to solve this as I wouldn't trust your ECU's closed loop corrections until you get trustworthy lambda values.

[jvanzyl]

Lambda measured at the MOT is 1.269 = Which is 18:1 AFR! That by oxygen measurement is insanely dangerously lean. But your HC's are very high, which is unburnt fuel. Those two together is impossible if the combustion/exhaust is sealed and true, how can combustion have unburnt oxygen and unburnt fuel, think about it. Maybe you have an exhaust gas leak which is causing oxygen ingress in to the exhaust? Maybe you have poor ignition/misfire which is leading to unburnt fuel and oxygen? Was the car tested with a cold or hot engine? Ideally you want to get it to full temperature then have it tested immediately as cold start enrichment will be against you. Either way you need to solve this as I wouldn't trust your ECU's closed loop corrections until you get trustworthy lambda values.

Right - good points all round. Car was up to temp. I've ordered some high temp silicone gasket sealer which will be liberally applied to the v-bands as they are where the leak is likely to be at. Thanks for the guidance!

[simonrobinson]

Exhaust leaks at idle create far less of a venturi effect so the air ingress is minimal thus the wideband measurements are less skewed. But as you rev it the gas speed inside the pipe vastly increases so the scavenging of air/oxygen into the exhaust is far greater. Dual tailpipe exhaust systems can screw with the lambda reading on MOT test equipment but will not cause your excessive HC's.

Certainly put your hand over your tail pipe while its running and listen, do this test to a few other cars of friend and family in your life. Listen to how they behave, a good exhaust builds pressure quickly and the hissing leaks are very minimal. If the exhaust is sealed properly the engine will eventually stall with your hand over the tail pipe.

If V bands are leaking you can get annealed copper gaskets for them but there is no standardisation of V band sizing. Warped ones can sometimes be ground flat too with effort and patience. Silicon sealants tend to burn away pretty quickly.

[jvanzyl]

Exhaust leaks at idle create far less of a venturi effect so the air ingress is minimal thus the wideband measurements are less skewed. But as you rev it the gas speed inside the pipe vastly increases so the scavenging of air/oxygen into the exhaust is far greater. Dual tailpipe exhaust systems can screw with the lambda reading on MOT test equipment but will not cause your excessive HC's.

Certainly put your hand over your tail pipe while its running and listen, do this test to a few other cars of friend and family in your life. Listen to how they behave, a good exhaust builds pressure quickly and the hissing leaks are very minimal. If the exhaust is sealed properly the engine will eventually stall with your hand over the tail pipe.

If V bands are leaking you can get annealed copper gaskets for them but there is no standardisation of V band sizing. Warped ones can sometimes be ground flat too with effort and patience. Silicon sealants tend to burn away pretty quickly.

Right. Massive exhaust leak confirmed at the turbo v-band. Will try and resolve this one and then see if there are any others.

[simonrobinson]

If it has been tuned with an exhaust leak which was causing air ingress on the tuning session, once the leak is fixed you may look at your logs to find that the whole tune is overly rich as the lambda readings used while tuning contained rogue oxygen. Once you fix the leak, take it for a drive then attach your log files and tune file as I would give you an uneducated and unqualified opinion if you desire. After all the infinite room full of monkeys with typewriters would write the perfect tune eventually.

Attached is a log from one of my cars showing the purple line "EGO Correction (%)" value though gear pulls. 100% is zero correction from the VE tables tune, while say 105% is adding 5% correction thus the VE table is too skinny, and if it were 90% correction then the VE table is too fat. Looking at the 4th gear pull highlighted in blue, getting that green Lambda line to sit at flat as possible at the target is the aim of the game while the corrections are as close to 100% as possible, preferably sitting below 100% because its better to take away fuel than rely on closed loop corrections to add it in with delayed response.

[jvanzyl]

Thank you- the only thing I could find at Halfords this evening was their Holts Fire gum stuff... might work.

I get what you're saying with the chart, it's a very helpful description.
Hopefully I'll get a chance tomorrow to see what has actually happened at the V-band..

[simonrobinson]

You just need your V band to seal, use a grinding stone to get them flat as they warp pretty badly when you weld them. If you use any fire sealants that cure like a ceramic they will not only crack to bits but also the excess ends up trapped in the catalytic honeycomb where it erodes it away remarkably quickly.

[jvanzyl]

You just need your V band to seal, use a grinding stone to get them flat as they warp pretty badly when you weld them. If you use any fire sealants that cure like a ceramic they will not only crack to bits but also the excess ends up trapped in the catalytic honeycomb where it erodes it away remarkably quickly.

Got it. I do have a nice big grinding stone, however my v-bands are the male female type so I'll see what I can actually affect.

I'm hoping it's a case of them just not being lined up properly..

[Call the midlife!]

Got it. I do have a nice big grinding stone, however my v-bands are the male female type so I'll see what I can actually affect.

I'm hoping it's a case of them just not being lined up properly..

Did you tack them in place before welding in the bends etc? Could be that the bends are putting stress into the V band faces as the system moves with the engine, especially if the flexi isn't big enough or not in the right position.