how-to: ECU-based direct boost control

[mchuang]

ok nm, i see the axis on the opposite side my bad

[mrfred]

Originally Posted by mchuang

Wow that is nice, what do you use to plot graphs mrfred?

Its a template I made up in Excel. I have an older version posted here somewhere. Its setup for easy cut and paste from my EvoScan files. I'm thinking of automating it so that it can plot data from anyone's EvoScan file. Its pretty far down the list at the moment.

[RedV]

I think I'm getting there mrfred...



... There's a little dip between 4500-5000 RPM so I'm gonna give it a little bump there to see if I can correct that. Otherwise I'm thinking it looks good.

[nj1266]

Can anyone translate these formulas into language that Logworks can understand?

<DataListItem DataLog="Y" Color="" Display="Boost Error" LogReference="BoostError" RequestID="8A" Eval="0.0241*x-3.087" Unit="psi" MetricEval="" MetricUnit="" ResponseBytes="1" GaugeMin="-5" GaugeMax="5" ChartMin="-5" ChartMax="5" ScalingFactor="1" Notes=""/>
<DataListItem DataLog="Y" Color="" Display="WGDC Correction" LogReference="WGDCCorr" RequestID="8B" Eval="0.5*x-64" Unit="WGDC%" MetricEval="" MetricUnit="" ResponseBytes="1" GaugeMin="-50" GaugeMax="50" ChartMin="-50" ChartMax="50" ScalingFactor="1" Notes=""/>

[mrfred]

Originally Posted by RedV

I think I'm getting there mrfred...

...

... There's a little dip between 4500-5000 RPM so I'm gonna give it a little bump there to see if I can correct that. Otherwise I'm thinking it looks good.

Yep, its getting there. Looks like another 5-8% BWGDC can be added from about 4250-5000 rpm? A bit more fine tuning after that maybe. I attached my latest spreadsheet if you want to use that.

[nj1266]

Tonight I implemented direct boost control on my car. But the logs left me with a few questions. Here is one of the logs. I did them in Evoscan, but I am reading them in Logworks.

1. I am not hitting the boost that I entered in the baseline boost tables. I attached the tables. I am using the same WGDC setting that I used with the Load system. Is that to be expected? Do we have to run higher WGDC with the direct boost system than with the load boost system?

2. Could someone explain to me how boost error relates to my actual boost? Does it work this way: At 5656 rpm my actual boost is 17.22, but the boost error is -2.70 psi. If I add them together I get 19.92 psi which is close to 20.1 psi in my baseline boost table. If that is the way this works, then how can I counteract it from happening. What do I do to hit my baseline boost of 20.1 psi?

[mrfred]

Originally Posted by nj1266

Tonight I implemented direct boost control on my car. But the logs left me with a few questions. Here is one of the logs. I did them in Evoscan, but I am reading them in Logworks.

1. I am not hitting the boost that I entered in the baseline boost tables. I attached the tables. I am using the same WGDC setting that I used with the Load system. Is that to be expected? Do we have to run higher WGDC with the direct boost system than with the load boost system?

2. Could someone explain to me how boost error relates to my actual boost? Does it work this way: At 5656 rpm my actual boost is 17.22, but the boost error is -2.70 psi. If I add them together I get 19.92 psi which is close to 20.1 psi in my baseline boost table. If that is the way this works, then how can I counteract it from happening. What do I do to hit my baseline boost of 20.1 psi?

The BEC table I recommend is very conservative in the sense that if the actual boost is more than 1.4 psi below the target boost, the system will not try to increase the WGDC to compensate. In the BEC table, this is represented by the zeros at -1.4, -2.0, and -3.1 psi.

In your situation, you were below your target by almost 3 psi, so the BEC system didn't try to fix it. You'll need to increase your BWGDC table in those areas where the boost was off by that much. Alternatively, you could change the zeros at -1.4, -2.0, and -3.0 psi to 1's or 2's, but I don't recommend this because it can lead to boost overshoot during spool. I think its better to have a BWGDC curve that provides the desired boost with minimal corrections. If I'm reading your logworks plot correctly, the boost was more than 1.4 psi below the target boost for the entire run. I'd recommend increasing the BWGDC curve by maybe 5% across the board. Maybe increase it even more in the areas where the boost was 3 psi below the target.

The direct boost control system does not affect the BWGDC tables in any way. I'm guessing what happened is that with the load based setup you were using before, the ECU was correcting the WGDC upward during spool because of you were non-zero values in the upper right of the TBEC table.

[nj1266]

Originally Posted by mrfred

The direct boost control system does not affect the BWGDC tables in any way. I'm guessing what happened is that with the load based setup you were using before, the ECU was correcting the WGDC upward during spool because of you were non-zero values in the upper right of the TBEC table.

In the load based set-up I was using the stock TBEC table. Could that be the reason for the lower BWGDC in the load set-up?

also, could you tell me if my understanding of the way the BEC relates to actual boost is correct? Is it as I stated before: At 5656 rpm my actual boost is 17.22, but the boost error is -2.70 psi. If I add them together I get 19.92 psi which is close to 20.1 psi in my baseline boost table.

[mrfred]

Originally Posted by nj1266

In the load based set-up I was using the stock TBEC table. Could that be the reason for the lower BWGDC in the load set-up?

also, could you tell me if my understanding of the way the BEC relates to actual boost is correct? Is it as I stated before: At 5656 rpm my actual boost is 17.22, but the boost error is -2.70 psi. If I add them together I get 19.92 psi which is close to 20.1 psi in my baseline boost table.

Yes, and Boost Error is the difference in psi between the target boost and the actual boost as you were guessing.

[razorlab]

nj, you need more cowbell.

[tephra]

this is great stuff guys - congrats Mr-Fred I'm tempted to get rid of my boost controller But alas I need a gauge + MAP ... hehe

[nj1266]

Originally Posted by razorlab

nj, you need more cowbell.

Explain please

[nj1266]

mrfred,

Why did you set-up the range on the BEC table from 3.1 to -3.1? Why not use the stock range? Just curious

When I modified the xml file, I noticed that the TBEC table (the one that I use for load based boost) also changed its range. I had to make two xml files/base files and swap them around to get around this issue.

[MR Turco]

Originally Posted by razorlab

nj, you need more cowbell.

LOL

[mplspilot]

I think i need more cowbell as well