[TDIwyse]

Originally Posted by iaknown

... I'm curious to see what effects this has on performance for you.

It might be helping it a little. Although I'm mostly happy to see the MAF readings are now agreeing with calculated values and Matchbot values (Matchbot just slightly lower than calc and measured).

Did a 4th gear pull this morning (I've left the EWG setting such that it's clamping exhaust pressures at 40psig, which is a little lower than I like... but trying not to change too many variables at once). Info attached. OBD MAF readings peak at 4604 rpm, which is right where PerfExpert shows peak power. Plugging the logged numbers into matchbot show very good agreement on what the MAF should be, and its predicted engine power number is in the sanity realm of PerfExpert's predictions (MB ~527 vs PE ~497).

[DWR]

Originally Posted by TDIwyse

oops, DWR doesn't want me to tweak in public :-)

Oh, so now you are going to listen to me, ha! In that case, having learned and shared more about the MAF, I restore all tweaking privileges. And furthermore, certifiy that your standing as an exemplary model for the community has only been enhanced ... who the f knew I had so much power?

This has really been a serendipitous turn in this thread that sets me up for a subject I wanted to explore with the forum - measurement and adaptive control. One of the features of a programmable piggyback is the ability to manipulate multiple inputs to the DDE, in a coordinated manner. We sometimes complain about the adaptive algorithms interfering with manipulations/changes we would like to make. But why don't we exploit these algorithms? ... more to come, taking a break.

[DWR]

Originally Posted by TDIwyse

Also can confirm that the max fuel rate is tied to the MAF readings.

Can you share the data? Would really like to see that effect in numbers. That may be a possible tuning exploit!

Great stuff, TDIwsye!

[TDIwyse]

Originally Posted by DWR

Can you share the data? ...

Absolutely. Always appreciate it when we all work together to progress the knowledge base.

I've attached a couple plots showing the relationship for 4th gear pulls that have been posted earlier in this thread. This is from the run this morning and then the one on Monday.

Interestingly, I'm not sure what to believe in these sensor outputs. The power differences measured by PerfExpert don't seem to reflect the significant differences in reported air/fuel. I now know the MAF data with the cone was wrong, so I'm also wondering if the reported fuel numbers are correct? The EGT's were basically the same with peak temps of ~1050-1060F.

Another question related to your comments about calibrating the MAF response of the cone filter vs OEM setup. I'm attaching another image of the MAF responses of these same runs where I'm scaling the cone filter MAF response by ~15%. It kind of lays on top of the OEM response, although it still looks like there might be some issues fitting the whole response with a simple scaling factor. Curious as to your thoughts regarding that.

[TDIwyse]

sorry for the huge images ...

also, this is now how I feel about the cone filter:

[DWR]

OK, so lets start with fuel rail pressure sensor signal manipulation. This is the method used by many of the commercial tuning boxes that are single channel. Single channel has come to mean only affecting one input. In this case, it is the signal from the fuel rail presure sensor.

Fuel rail pressure directly affects the amount of fuel that is injected, for a given duration of injection. The relationship between pressure and the injection quantity (IQ) is not linear. It is a square root relation, whereby it takes a quadrupling of pressure to get doubling of fuel flow. For example, if we increase the pressure from 1600 Bar to 1800 Bar (a Bar is used to represent atmospheric pressure) we will get sqrt(1800/1600)/100% = 106%, or 6% more fuel. The components in our fuel system have been used in 2000 Bar max pressure systems. Our system is set for 1600 bar max pressure. So, raising the pressure has been an accepted method of increasing horsepower.

In order to keep injection duration short for maximum efficiency, the DDE exercises a wide range of fuel pressure. In fact, the voltage signal from the fuel rail presure sensor looks like the output from a "Volumetric Efficiency"/load sensor. In stock form, that sensor maxes output in the vacinity of 4.2 volts. The sensor can output 4.8 volts, if the presure is high enough. The voltage is linear with the pressure. However, zero pressure does not equal zero volts, so we have to be careful when calculating pressure increases and fuel flow. See the graph below.

If we take the signal coming from the fuel rail pressure sensor and lower it, the DDE will in turn increase the pressure in order to create a higher voltage reading, in line with what it was targeting (pressure). This technique can run into 2 issues. First, the increased fuel will be detected by the DDE from the the O2 sensor, as the air to fuel ratio (AFR) will lower. If this happens within the operating ranges of adaption, the DDE will learn its way around this manipulation and return over a period of time towards 'normal' fuel flow. It does this by adjusting the injection duration. The trick is to do the manipulation outside of the adaption operating ranges. Staying above @ 3,000 rpm is the simplest method. The JDB, at times, works within the operating range of adaption. Stories of decreased performance over time or greater immunity to limp mode (the second issue) are based in fact.

What if we purposely manipulate the fuel rail pressure sensor signal to cause the DDE to adapt? What if we raise the voltage and hence the targeted pressure? Then the DDE will have to compensate by lowering the pressure. But now the O2 sensor will tell the DDE there isn't enough fuel being injected. The DDE will correct this by increasing the duration of injection. If done correctly, this kind of manipulation will fool the DDE to increase injection duration throughout the rpm range.

Now, if we are clever (and we are) we can use the full breath of DDE adaption to both increase injection duration and injection pressure. iaknown and I have been able to gain clarity on the limits of pressure maniuplation (more to come on that). The next step is to understand the limits of duration manipulation.

Ok, I'm going to take another break.

[DWR]

Originally Posted by TDIwyse

sorry for the huge images ...
also, this is now how I feel about the cone filter:

Dude, that is the thing I most admire about you. You fight hard as hell. You test the logical, push the limits, (push my buttons sometimes) stand your ground. But in the end, you are steadfast to the truth - whatever that maybe.
I, for one, would be happy if you deleted that post. It isn't warranted ... OK, it is funny.

P.S. do I have to apologize for the huge images too? I've attached some whoppers lately.

[TDIwyse]

Ha.

I wish we were neighbors, or at least lived close by.

I'll raise a glass of distilled spirits your way in a few hours :-)

[DWR]

Attached is a document that TDIwyse shared with me a while ago. While not specifically about adaptions, it gives information on the plausibility checks performed by the DDE. The DDE doesn't just let input signals get out of line, it takes corrective action to keep from triggering DTCs, etc. Thus, you can think of the plausibility limits as the bounds for adaptions.

[DWR]

Originally Posted by TDIwyse

I wish we were neighbors, or at least lived close by.

OK, with me. Who's going to move? Tough choice ... maybe not,
my wife just told me you are moving.

[DWR]

OK, back to work. Enough of this frivolity.

As was shown, the fuel rail pressure sensor can put out a signal up to 4.8 volts. 4.8 volts is roughly 1900 Bar of pressure. 4.1 volts is roughly 1600 Bar. That means in theory we could obtain fuel flow of sqrt(1900/1600)/100% = 109%, max pressure (the gain % can actually be greater at lower pressures). Using a base of 265Hp, we could gain 24Hp by exploiting fuel rail pressure manipulation. This is where I have to admit, it isn't that easy. The DDE will make up to a 0.8 volt correction to the rail pressure, however, it won't do it at the 4.0 volt level. 4.6 volts is the max we can drive out of the sensor before the DDE starts to balk. Other factors start to move out of the range of plausibilty. Anyway, the DDE may clamp its adjustment to pressure, but it does not stop there. In order to compensate, it starts to increase the duration of injection.

This is how we get more than 24Hp gain. Many have said there isn't much gain from a JDB until you crank the setting, this is the reason. The DDE is fooled into increasing both the rail pressure and injection duration. We get a multiplicative effect but it isn't the 20% more HP that gets claimed. iaknown and I did not run a JDB but we were able perform the same kind of manipulation with the piggyback (and more). If you see our results, the increase in acceleration did not backup the 20%Hp claims made by the tuning box suppliers. We gained about 11%, 34Hp on the 312Hp baseline - we'll take it.

P.S. almost forgot. Fuel is compressible, not much, but a little. Increased pressure will increase the density of the fuel. At maximum rail pressure distortion that gives about 1% more power. Isn't science cool

[TDIwyse]

Hey DWR, I never heard back after sending you the link on May 1 for where I had measured this duration effect with the JBD when I first started stacking a tune with the sensor distortion back in 2012:

http://www.bimmerfest.com/forums/sho...8&postcount=49
http://www.bimmerfest.com/forums/sho...3&postcount=50

Are you guys also seeing similar increases in the start of, and duration of, injector opening?

P.S. For those interested, the first link also has a cross link to where I had investigated a used 335d HPFP to verify it had the needed internal hardware to support much more flow. I still have that pump and never sent it off to be reworked because I haven't hit the limit of its fueling capability yet...

[TDIwyse]

Originally Posted by DWR

... If you see our results, the increase in acceleration did not backup the 20%Hp claims made by the tuning box suppliers. We gained about 11%, 34Hp on the 312Hp baseline - we'll take it.

Interesting, that's also in the realm I measured for JBD increases with the Gtech Pro device: http://www.bimmerfest.com/forums/sho...9&postcount=27

[DWR]

Originally Posted by TDIwyse

Are you guys also seeing similar increases in the start of, and duration of, injector opening?

At some point, we started to get pragmatic and traded parameters for sampling rate. To keep the number of logged parameters to a minimum, we did not look at it this time. Saw it in previous sessions, but didn't fully investigate.
Truth is, I could follow it with AFR and knew the extra fuel was flowing. Only one way that could happen.
Still it would be nice to document the whole story.

[DWR]

Originally Posted by TDIwyse

Another question related to your comments about calibrating the MAF response of the cone filter vs OEM setup. I'm attaching another image of the MAF responses of these same runs where I'm scaling the cone filter MAF response by ~15%. It kind of lays on top of the OEM response, although it still looks like there might be some issues fitting the whole response with a simple scaling factor. Curious as to your thoughts regarding that.

Well, I think what you are asking me is, "how do I really know how much air is flowing?" Off the top of my head I can think of 3 different techniques. In all cases, you must have some reference standard to leverage eveything else. You've seen that I use pressure difference across a fixed restriction to gauge relative differences (my intake tract thread). If you know flow at some pressure difference, you can calculate others. So, if you think a MAF is accurate in some region of flow and not another, use pressure difference measurements to scale into the flow region you want to calibrate. Or you may have to start with a known restriction, pressure difference and flow that you calibrate the MAF to (bench test). Sometimes you can't easily do any of that, in which case you use knowledge of AFR, from an accurate O2 sensor, calibrated fuel flow, and work backwards.

Happy to supply additional details and assistance, at your request.

[Chief Orman]

Originally Posted by DWR

OK, so lets start with fuel rail pressure sensor signal manipulation. This is the method used by many of the commercial tuning boxes that are single channel. Single channel has come to mean only affecting one input. In this case, it is the signal from the fuel rail presure sensor.

Fuel rail pressure directly affects the amount of fuel that is injected, for a given duration of injection. The relationship between pressure and the injection quantity (IQ) is not linear. It is a square root relation, whereby it takes a quadrupling of pressure to get doubling of fuel flow. For example, if we increase the pressure from 1600 Bar to 1800 Bar (a Bar is used to represent atmospheric pressure) we will get sqrt(1800/1600)/100% = 106%, or 6% more fuel. The components in our fuel system have been used in 2000 Bar max pressure systems. Our system is set for 1600 bar max pressure. So, raising the pressure has been an accepted method of increasing horsepower.

In order to keep injection duration short for maximum efficiency, the DDE exercises a wide range of fuel pressure. In fact, the voltage signal from the fuel rail presure sensor looks like the output from a "Volumetric Efficiency"/load sensor. In stock form, that sensor maxes output in the vacinity of 4.2 volts. The sensor can output 4.8 volts, if the presure is high enough. The voltage is linear with the pressure. However, zero pressure does not equal zero volts, so we have to be careful when calculating pressure increases and fuel flow. See the graph below.

If we take the signal coming from the fuel rail pressure sensor and lower it, the DDE will in turn increase the pressure in order to create a higher voltage reading, in line with what it was targeting (pressure). This technique can run into 2 issues. First, the increased fuel will be detected by the DDE from the the O2 sensor, as the air to fuel ratio (AFR) will lower. If this happens within the operating ranges of adaption, the DDE will learn its way around this manipulation and return over a period of time towards 'normal' fuel flow. It does this by adjusting the injection duration. The trick is to do the manipulation outside of the adaption operating ranges. Staying above @ 3,000 rpm is the simplest method. The JDB, at times, works within the operating range of adaption. Stories of decreased performance over time or greater immunity to limp mode (the second issue) are based in fact.

What if we purposely manipulate the fuel rail pressure sensor signal to cause the DDE to adapt? What if we raise the voltage and hence the targeted pressure? Then the DDE will have to compensate by lowering the pressure. But now the O2 sensor will tell the DDE there isn't enough fuel being injected. The DDE will correct this by increasing the duration of injection. If done correctly, this kind of manipulation will fool the DDE to increase injection duration throughout the rpm range.

Now, if we are clever (and we are) we can use the full breath of DDE adaption to both increase injection duration and injection pressure. iaknown and I have been able to gain clarity on the limits of pressure maniuplation (more to come on that). The next step is to understand the limits of duration manipulation.

Ok, I'm going to take another break.

This is really terrific information and research. That said, I think I understand at best 50% of what you are talking about. What is the upshot of all this research? Are you guys capable of creating a new tuning box? In short, how do we implement all this terrific work in our cars?

[DWR]

Originally Posted by Chief Orman

This is really terrific information and research. That said, I think I understand at best 50% of what you are talking about. What is the upshot of all this research? Are you guys capable of creating a new tuning box? In short, how do we implement all this terrific work in our cars?

That's a fair question. But I need to know is the understanding issue due to my explanation? If so, I'm happy to try and make things clearer. If not, then the upshot of all this is, this is not for you - respectfully. You maybe looking for a tune, instead of a tuner. The programmable piggyback is a tuner.

Since there are many vehicles in different states of tune on this forum, one size does not fit all. For example, if you have a stage 2 remap, you can definitely push too much fuel. That won't happen with a stock DDE, so there is a little more leeway. Point is, we are showing a method to optimize to the specifics of your vehicle. For that reason, not interested in creating a new tuning box. I have enough trouble keeping up with what I have on my plate now

There are more aspects to piggybacking that can be helpful, such as O2 sensor signal manipulation. Just trying to gauge if this is useful to more than a couple people.

[Chief Orman]

Originally Posted by DWR

That's a fair question. But I need to know is the understanding issue due to my explanation? If so, I'm happy to try and make things clearer. If not, then the upshot of all this is, this is not for you - respectfully. You maybe looking for a tune, instead of a tuner. The programmable piggyback is a tuner.

Since there are many vehicles in different states of tune on this forum, one size does not fit all. For example, if you have a stage 2 remap, you can definitely push too much fuel. That won't happen with a stock DDE, so there is a little more leeway. Point is, we are showing a method to optimize to the specifics of your vehicle. For that reason, not interested in creating a new tuning box. I have enough trouble keeping up with what I have on my plate now

There are more aspects to piggybacking that can be helpful, such as O2 sensor signal manipulation. Just trying to gauge if this is useful to more than a couple people.

DWR there is nothing lacking in your explanations. I am not a trained engineer so some of the terminology/science is a bit over my head. That said, I'd say I am actually understanding more like 80% of the posts and appreciate the clarity of your writing.

Unfortunately, as a CA resident removing the DPF is not in the cards. Moreover, while on CPO warranty I don't want to reflash the DDE. That said, once my warranty is up and I start driving my car only off road I will be seeking your advice on a remap.

[charlie fairmont]

Originally Posted by DWR

...Just trying to gauge if this is useful to more than a couple people.

Yes it is. Please don't stop posting your findings!