Home built Hot Rod

[Peter M]

5 minutes ago, MFX said:

I will have another look and do some minor tweaks to start with to make sure I don't mess anything up  

Jeff,

If you post a screen shot of your current VE table, I reckon the PFA community could suggest the changes required.

Also gives you a record of the current table so you can always reinstate it if need be! 

[MFX]

1 minute ago, Peter M said:

Jeff,

If you post a screen shot of your current VE table, I reckon the PFA community could suggest the changes required.

Also gives you a record of the current table so you can always reinstate it if need be! 

Haha, go back and have a look at my last post  . You posted a couple of seconds after I did.

[Peter M]

13 hours ago, Peter M said:

Then adjust the VE cells around the above cells to blend it all in so it's a smooth transition both across and down at least 2 or 3 cells as I expect the adjustments above would be all in the top row of the table.

Give it another ramp run and see if that gives you a flatter Lambda curve at WOT across the rev range that is closer to your target of 0.82ish.  If it's still a bit lumpy, try some other minor VE table adjustments based on your data log until you are happy with it.

After seeing your VE table and realising you are running Throttle Position rather than MAP and the rows you need to change are at the bottom rather than the top because of how your table is formatted, I've attached a spreadsheet showing the cells I would change to flatten the Lambda curve below. 

EDIT: Just noticed the cell at 100%/6000rpm should be shaded as well in the table as it was also a peak used to inform a change in the table.

Edited 18July, 2021 by Peter M

[Pork Chops]

Peter would you consider these changes ‘significant’? Numerically they seem that way e.g. 111 to 123. Not exactly splitting hairs with a change like that IMHO.

[Peter M]

However I would suggest the best priority would be to sort out the CHT sensor and figure out why the trouble with your O2 sensors first as you'll have to sort these matters anyway to allow you to get to the best tune.

As far as learning how to tune, I notice HP Academy have some videos on Link tuning and a good series on introduction to tuning.  There is also an assortment of stuff on Youtube that may help.

As for professional help, Uncle and I have used Benchmark just south of Newcastle with success.  Benny is a Link dealer so prefers/specialises in Link (and Motec!).  I know a fair distance from you but I don't know anyone in your region I could recommend that isn't hamstrung with Covid.

I'm sure there are some PFA'ers in the area who would be happy to put you up for a night or two!

 

[Peter M]

47 minutes ago, Pork Chops said:

Peter would you consider these changes ‘significant’? Numerically they seem that way e.g. 111 to 123. Not exactly splitting hairs with a change like that IMHO.

I've just done a mental calculation on the % change required for the Lambda correction and added that % change to the appropriate VE cell in the table.  Assuming a VE of around 100%, a 10% increase in VE table cell, to better model the engine's actual airflow at that particular speed and loading, should result in a similar increase in the Lambda as there will be a similar increase in fueling.  ie 0.9 Lambda should go to a more acceptable (at WOT running condition anyway) of around 0.81 Lambda which is consistent with Jeff's target Lambda.

Not a mathematically perfect solution as adding 10 to a VE cell of say 75% gives a greater proportional increase than adding the same amount to a VE cell of 125.  But I expect Jeff will need to do a few ramp runs to get the Lambda reasonably consistent over the WOT rev range and the changes suggested are just to get him started and confident we are going in the right direction.

So not splitting hairs for the large change but this is in a safe "increasing fueling" direction while the reduction in the fueling are limited to about 5%   

[Redracn]

36 minutes ago, Peter M said:

I've just done a mental calculation on the % change required for the Lambda correction and added that % change to the appropriate VE cell in the table.  Assuming a VE of around 100%, a 10% increase in VE table cell, to better model the engine's actual airflow at that particular speed and loading, should result in a similar increase in the Lambda as there will be a similar increase in fueling.  ie 0.9 Lambda should go to a more acceptable (at WOT running condition anyway) of around 0.81 Lambda which is consistent with Jeff's target Lambda.

Not a mathematically perfect solution as adding 10 to a VE cell of say 75% gives a greater proportional increase than adding the same amount to a VE cell of 125.  But I expect Jeff will need to do a few ramp runs to get the Lambda reasonably consistent over the WOT rev range and the changes suggested are just to get him started and confident we are going in the right direction.

So possibly not significant change in numbers but potentially significant at WOT for engine longevity, increasing power or avoiding knock.    

What Peter is doing is correct. Work out the %error in lambda and apply that to the fuel number at that point. It assumes that the injector latency is properly calibrated but if it is not that is not usually a big issue for full throttle.  

New_fuel_num =  old_fuel_num x actual_lambda / target_lambda.

If everything is perfect the new fuel will be spot on but there are usually some errors so a couple of passes are required. 
As you are targeting a very safe 0.81 rather than the 0.85 to 0.88 that I would be targeting I wouldn’t worry about the size of the changes required as they are reasonable.

[Redracn]

On 15/07/2021 at 22:12, Peter M said:

I think your logs should look more like this.  Looking at a similar run on a similar time scale shows:

Greater acceleration enrichment

Less variation in the actual lambda to target plus none of the wild cycling from rich to lean.  The target lambda below would be in the 0.85 to 0.81 range with 0.81 at WOT at 6,000rpm  

Not shown on this screen but another, the IAT on this ramp run was 27 degrees and the cylinder head temp went from around 130 to 152 degrees over the run.

 

Given the discrepancy between Peter M head temps which are correct and the very unrealistic ones from Harry it would pay to look into this further. Before proceeding it also needs to be established if the ECT table has been modified to reflect the low readings or is it still setup for water cooled or real head temps otherwise it could be still supplying cold running enrichment which is very bad from a tuning perspective. 

[MFX]

1 hour ago, Redracn said:

Given the discrepancy between Peter M head temps which are correct and the very unrealistic ones from Harry it would pay to look into this further. Before proceeding it also needs to be established if the ECT table has been modified to reflect the low readings or is it still setup for water cooled or real head temps otherwise it could be still supplying cold running enrichment which is very bad from a tuning perspective. 

23 hours ago, MFX said:

 

@Redracn your point is exactly one of the issues I was looking at. As in the image above, my ECT table looks like it has some correction across the entire heat range, which I never understood. I obviously have very little idea about tuning, but I was under the impression that it should give some help at start up, but I would have thought it should transition to zero once it was up to temp?

As for the tuner, to be honest he has a good reputation at least in the import side of the scene, but he has no experience with Porsches. Although I do realise that the basics of tuning would be mostly the same regardless of the make. I really don't think he has put the time into the map. That said, time on the dyno is expensive so I think his tune, aiming for .815 lambda was being super conservative, to just get a safe tune, not necessarily the best tune.

I will give some of @Peter M's suggestions a go and post the results. 

[Pork Chops]

ECT is Engine Coolant Temp? And in the case of an aircooled engine it’s head temp? And if head temp, how is it possible Jeff’s getting 40-50 degrees. Sensor broken or incorrectly installed or a software setting issue? I’m probably playing catch up here so sorry if this has been covered.

[Redracn]

15 minutes ago, MFX said:

@Redracn your point is exactly one of the issues I was looking at. As in the image above, my ECT table looks like it has some correction across the entire heat range, which I never understood. I obviously have very little idea about tuning, but I was under the impression that it should give some help at start up, but I would have thought it should transition to zero once it was up to temp?

As for the tuner, to be honest he has a good reputation at least in the import side of the scene, but he has no experience with Porsches. Although I do realise that the basics of tuning would be mostly the same regardless of the make. I really don't think he has put the time into the map. That said, time on the dyno is expensive so I think his tune, aiming for .815 lambda was being super conservative, to just get a safe tune, not necessarily the best tune.

I will give some of @Peter M's suggestions a go and post the results. 

I’m pretty sure that the screen shot doesn’t show the ECT fuelling table. Looks to me like throttle drive by wire and idle tables. Is the engine using Electronic Throttle  Control aka Drive By Wire? 
I also see a “mixture map” menu at the top of the screen. Click on that and see what you get. 

[Peter M]

Jeff,

From what I can see the Idle Base Position table on your ecu assists with engine idle speed management anly.

If you look in the Mixture Map tab you should find something called Warm Up Enrichment or very similar.  What you want to do is make sure the enrichment value is zero'ed after say 30 degrees (with the current head temp sensor setup) so that it isn't adding to your fueling.  If the current enrichment value is still significant at 30 degrees and above, you may have to adjust the VE table to compensate for this change before you do your next ramp run. 

[MFX]

Ok I have delved into the Post start enrichment and the Warm up enrichment tables. The link transitions from post start enrichment to the warm up table. The warm up table looks like it still has 5% correction at 40 degrees which is obviously an issue if the ECT sensor only gets to 40 when it is at full temp (at least on a cold day). 

I have changed @Peter Ms correction in the table and I will take it out for a run after lunch and get another log to see how it likes the changes.

 

[Pork Chops]

Same day updates 💪. This is fun 🤩.

[Peter M]

17 minutes ago, Pork Chops said:

ECT is Engine Coolant Temp? And in the case of an aircooled engine it’s head temp? And if head temp, how is it possible Jeff’s getting 40-50 degrees. Sensor broken or incorrectly installed or a software setting issue? I’m probably playing catch up here so sorry if this has been covered.

The ECT or CHT give a very accurate indication how efficiently the injected fuel will burn in the engine.  Since there is this strong correlation between the temperature of the head and the intake port in particular and fuel atomisation, low ECT/CHT's are used to trigger temporary enrichment for starting and initial running.

Appears Jeff's CHT sensor isn't mounted close enough to the action to model the actual temperatures that the fuel mixture is seeing in real life.

 

[MFX]

I am just delving into the Link's ECU and it actually has a quick tune function that can tune the fuel to the target lambda signal on the fly. I am reading more up on it, but that may be the way to go getting it all much closer to a better tune.

Having a look at my target lambda table, it looks a bit too conservative. Should I blend it down to .85 at 100%? Leave it at .95 at 0?

[Peter M]

Jeff,

You will need to sort your O2 sensors before first before you can use auto tune.  The CHT sensor issue also needs to be fixed.

Certainly go for a run after lunch (but don't change your Warm Up Enrichment table just yet as it appears the VE table has been set up with WUE table unwittingly "activated" and adding supplementary fuel) but I think until you resolve the o2 and CHT issues you are wasting your time fiddling with the tune.

 

'Aint this fun! 

[Peter M]

28 minutes ago, MFX said:

Having a look at my target lambda table, it looks a bit too conservative. Should I blend it down to .85 at 100%? Leave it at .95 at 0?

Jeff,

I wouldn't change it until you sort the VE table out and are on a dyno and can better monitor some other critical factors like CHT and knock.

By comparison my target Lambda table varies from 0.97 to 0.81. 

Also be aware that a touch too rich has less impact on torque than a touch too lean so don't sweat it. 

[Redracn]

While the CHT reading is obviously bogas the Lambda reading from the latest video did look ok. Would be better to graph both left and right banks and not the average as that can hide issues related to the setup of each bank.
Auto tune functions can be helpful in certain limited situations but often require more effort to setup and monitor than just doing the job. They are also of no use in overrun situations or any situation that the possibility of a misfire exists or with bigger cams. .i.e. All air that enters the engine must be fully and completely combusted before it reaches the lambda sensor. So as others have said leave it off for now but certainly read up on it and  give it a go in the future. 

[Pork Chops]

Jeff, are you running dual wide band O2 sensors?  May last set up was 1 on the left bank.  New set up is 2.  

[MFX]

16 minutes ago, Pork Chops said:

Jeff, are you running dual wide band O2 sensors?  May last set up was 1 on the left bank.  New set up is 2.  

Yep. One per bank.

I just went and did another log with @Peter M's corrections. At first glance the shape looks similar but I think it has actually flattened things out between 3500 and 7. Still needs more tweaking though. Looks like when I initially hit the throttle it goes really rich for a second before flattening out?

[Peter M]

I was expecting better!  You really need to compare with your last run by overlaying and ideally keeping the same scale if possible, to see the changes.

The first hump after the acceleration enrichment trough looks better.  The next trough is better too but the next peak looks like we have gone too far.

It's a bit weird that there is over 0.15 Lambda difference between the banks before the ramp run.  There is something not right here.  I would expect their trend lines should be bang on each other.  I think you should investigate this before going on further.

[Redracn]

Yep some improvements. For the next run get on the throttle very slowly to reduce the accel enrichment used or disable it all together it certainly seems to be excessive and holds on for a long time. There should be a channel you can log that shows the amount of additional fuel the accel function is supplying. 

So before going any further you need to sort out the ECT (head temp) so that there is no compensation being applied at normal operating temperature. As it stands you will need to add 5% (or whatever is at 39c in the table) to the entire fuel table after fixing head temp to get back to where you are. Also check all other compensations such as IAT and MAP etc. Are the injectors a matched set or have you had them flowed?

While ideally each bank should be identical it is seldom the case. A quick check of throttle blade angles and linkages as they progress to full throttle. A quick compression test would be good just to confirm no issues otherwise it is what it is. Worst case you should be able to tune each bank individually. After that if you are chasing the absolute most the engine can make is individual cylinder tuning. 😁

 

[MFX]

I will have a further look into where I can better mount the ECT as I drove around a fair bit before I did the run and I couldn't get it any higher temp than that. 

With the throttle, I suspect that there must be a tiny bit of play in the linkage between the two banks, so it looks like they ramp the same when I accelerate, but when I back off they are a long way apart. 

I will have a play with the software and see if I can overlay the old log with the current one.

[Redracn]

Some more hopefully useful info regarding Lambda measurements.

Back in the early 90’s when I was designing my first Lambda Meter we did extensive testing of the sensor location in the exhaust. What we discovered is that the exhaust gas flow from a cylinder after say a 4 to 1 would not fill the entire pipe but remain as a smaller diameter flow in the larger pipe. It could also spiral along the bigger pipe. This means that often the lambda sensor if placed within a meter of the collector will mainly read the lambda of a single cylinder and not necessarily the one you think based on any spiralling which can also change with rpm and load. 
So without seeing the installation it is highly likely that the so called “ Bank “ sensors are actually mainly  reading the lambda of individual cylinders.