RomRaider

Hello everyone,

I found KNeuer's M52B25 thread and I liked the way he documented his findings of mods for his engine.
So I thought I do the same, I hope the information in this thread can be useful for others who like to play with mods on their engines.

Few words on my engine/setup: (This is the baseline or what I call stock in my case)
- M52B28 from 728i
- powers my E30
- around 300tkm on the clock
- Cheap TA-Technix Exhaust manifold. Probably does nothing. I only installed it to ease the installation of the m52
- cheap "bandel-automotive" catalytic converter (For TÜV). Probably inferior to the BMW cats
- stock E30 325i rear silencer
- stock E36 328i Airbox with stock filter
- 520i ECU with a m52b28 map I found on this forum
- Vanos has been refreshed

I've created logs and diagrams for the following setups:
- stock m52b28
- m52b28 with m54b30 intake cam
- m52b28 with m54b30 intake cam and m50b25 intake manifold
- NOT MY CAR: M52B28 with M50 Intake and 257/250 Cams

In the future, I will log the following setup:
- m52b28 with m54b30 intake cam and m50b25 intake manifold and m50b25nv intake as exhaust cam

My way of evaluating the logs is the following (not sure if this smart or not):
- Try to do as many WOT Runs as possible.
- only consider data points where the Throttle has been fully opened
- group the data points by RPM
- average the data points over a sliding window
- to limit the influence of air temperature, I try to normalize the Load by converting it to a volume using the measured IAT at each data point. I think this makes sense when I try to gauge if a mod helped air flow yes or no
- create diagrams using this data

I do this in a simple Interactive Python Notebook

Here are some diagrams:

The car with the Catcams belongs to a friend of mine, we are currently trying to find out if his MAF has aged since the data does not seem to plausible.
We think the catcams should outperform my M54B30 cam, especially in the high RPM range.
Otherwise, I think it's interesting to see the influence of different mods.
I thought the M50B25 manifold would perform better, I'm not sure if I liked the combination of M54 Intake cam and Stock manifold better.
Maybe I will try playing around with the cam timing and see if I can regain some lost load in the midrange


Here is a comparison of my VANOS tuning efforts. I think it is not perfect yet, but you can clearly see the effects of the delayed vanos off point.
It also shows the MAF value

That should be it for now, I will add more Diagrams as I make modifications to my engine.
Please take these results with a grain of salt. I actually have little understanding of tuning and engines your milage with these mods may vary greatly.

Very interesting findings and comparisons of different setups and referring to my thread. Thanks.

However, the load values seem kind of low for b28, or maybe it is your scaling and IAT compensation that you did?

also kind of wondering why the green setup is so much worse at lower rpm than the stock red one.


Looking forward for more results and maybe you can post some real csv logs as well!

Thank you, I've attached the relevant logs, using the filename it should be possible to find out what is what.

If I don't scale for temperature, it is actually even worse. The stock log was driven when it was cold and for the m54 cam it was already a bit warmer.

This is unscaled

But I do not have an explanation on why the m54 cam is so much worse in the low revs. I guess nothing is for free and what you gain on top you loose down low

did you go to single lambda channel? Because in your later logs only one lambda channel is working. Also, do you feel the loss of torque now below 5500 rpm with the M50 mani? it is clear that the powerband is shifted updwards above 5500. I still dont get it though why the m54cam with m52 mani is so much worse than stock.

Some suggestions though, or what I found works somewhat better in comparing logs. Always do the pulls in third gear from like 1500 rpm up to redline. It kind of feels slow in the beginning, but this is where you get good data points. Use the exact same stretch of road (obviously somewhere where you know it is safe and straight and preferably with no cops around due to the speed you will reach in thrid gear at redline) and do the pull after the engine just warmed up without it being heatsoaked from long urban drives. Gearing has big influence on the load levels, so cant compare 2nd gear to 3rd and so forth.

Also, do you have wideband lambda probe? This is a must for tuning WOT fuel enrichment specially with cams.

I did not activate single lambda mode, but I know that I had a few times a DTC for the lambda. Looks like it finally let go, and a new one is in order.

With m54 cam and M50 Manifold, loss of torque below 4500 is very noticeable, but it's manageable. Above 5000 it's quite nice and the engine pulls hard. When the rev limiter engages, it feels like stepping on the brakes.
With the m52 Manifold and M54 Cam, the loss of torque was barely noticeable, but the top end improved alot. But this kind of contradicts the data I gathered, so I'm not sure what is going on. Something to keep in mind is that for the M50/m54 curve the vanos point has been already optimized. I'm certain that for the m54 cam alone the vanos has to be already quite delayed
This makes wonder why there is this hype around the M50 manifold, If I had to choose between m54 Cam and M50 Manifold, I would go with the cam any day.

I try to use third gear as often as possible, but there are very little roads here which are straight enough.
Why does gearing make a big difference on load? Wouldn't that mean that a given RPM the engine would consume more air in a higher gear?
But I do get the point that more data points are being collected for a given RPM Range.

I do not have a wideband lambda probe yet, but I plan on investing into one. Not sure which one to get, so recommendations are welcome. (Romraider Logger compatible of course)

You can even get the S52 cams US M3 fairly low price new from bmw dealer here in Germany. This will improve your whole power band and not just on top.

I am using the AEM Uego wideband controller and probe, which works perfect with romraider. No issues here.

I dont get why the loads are so low ? can you confirm this is a B28 ? other than that good work.

As it turns out my intake cam was timed quite a bit late. After fixing that some of the lost down low load has returned.
This also means that you have to consider the diagrams with the m54b30 cam I posted before differently.


(Vanos not optimal, need to retune)

@328i-sport-rossi yes it is a B28, checked the casting on the block. What numbers should I expect for a B28? Keep in mind on some diagrams I scaled for air temp.

I now also have a M50B25 Intake cam which I plan to use as a exhaust cam, courious what that will do to load.

when looking at the last graph posted it looks to me like the MAF is maxed at 5V from 6250 rpm and upwards!
The plateau looks too clean. I am pretty shure MAF_Voltage is at pinned 5V. BUT B28 MAF is maxed at 725 kg/h not 550 kg/h.
I had this when using the B20 MAF on my stroker. What MAF are you running? B25/28 one? OEM Sensor element?
When trying a cheaper (non-original) MAF the readings were all over the place for me...this meant retuning everything...

A modded B28 could max a "real" MAF at 725 kg/h, maybe rossi can comment more on what is a expected range?

Also what is an Airmass in "Vol/h" ?! You mean kg/h?

Did you do the Full Flash? Which Software Version are you on?
I found stock B20 Software has a lower load limit hidden in software.

Thank you for your input!



I only flashed the partial, so there might be some kind of limit? I though and maybe I thought wrong that the only difference is the map it self. But this is good info and I will investigate
How would the load limit show it self?


On some diagrams, I try to remove the influence of the air temp. This is excatctly that. For load mg/stroke and maf kg/h I do:


This is the same diagram with out correction:


I'm using the original B28 MAF original BMW part. Maybe the maf has aged?
I think the B20 MAF is not really suited for use a with a B28, if I remember correctly the diameter of the MAF housing is also different? So if you put the Sensor in the B28 housing your are asking for trouble

Next time I will log the MAF Voltage but I don't think it is maxed out. As far as I understood the MAF Voltage is just being multiplied by some factor, if it would be maxed out we would see ~750kg/h in the log

max should be 725kg/h=!5V

I didnt check your math but it makes no sense that the corrected MAF-Values are lower, if you are correcting a 40°C IAT run to something like 20°C (reference temp). The engine at 20°C IAT will log higher Load/MAF Values as the same engine at 40°C IAT!

okay, I will flash a full B28 bin thank you for the insight.
And you are right, I think there is something wrong with my air correction approach.
Until now I only looked at the relative position of the lines in the diagram which made sense to me.

My stock log was driven at a average IAT of 23°C, the others around ~36°C. Because the stock log is much colder, the line gets pushed down while others get pushed up relatively.
As for the numbers I don't even know which unit they are in lol so I was not too concerned


Edit:
I flashed a full EU 328i tune and then applied my partial again. Curves look similiar, still have the plateau in the high revs. Max kg/h was again 640 and max MAF Voltage was ~4.72VDC

Second Edit:
So I thought about the math some more and in my opinion it does make sense.
https://en.wikipedia.org/wiki/Density_of_air
What I do is take the mass of the air and divide it by its density (kg/m^3) at a given temperature. mass / density = volume
So at this point we are not talking about mass anymore but volume, therefore it also makes no sense to compare the values from a 'corrected' chart with a 'uncorrected' one. I guess the term corrected is also misleading since the unit of measurement changes.

But at this point I'm fairly confident that the math checks out

...your math is correct. I missed that you are using m³/h (Airflow) in one graph and kg/h (Air-mass-flow) in the other! You wrote Airmass on both axis descriptions
Maybe stick to kg/h as its what this forum is used to. I bet Rossi thought you were showing Air-mass-flow also
But i get why you are doing it that way...


I wouldnt mix a b20 partial with b28 software. Following Pazis instructions it should be full flash, then read-back the partial, do the changes in RR, flash the modified partial.
Otherwise you could be changing stuff without knowing it. Its not guaranteed that every map/Value is represented in RR-definitions.

I'll label the axis better next time but I will keep doing the correction for temp because imo it eliminates or at least reduces a variable

My partial has been always a b28, so its a full b28 tune with my changes.

My M50B25 Intake cam is ready for installation but I still need to get dial indicators for installation. Will post logs/diagrams then