Click pic to see larger size.



.
.
.
.
.
.

IHC, Here is the first 60 seconds after starting the car from a dead cold this morning. The timeline 0.0 (X-Axis) is within 5 seconds of the engine firing-up. I fired-up the engine and then hit the record button on the logger.

Thanks for the info on the closed loop.

Also, below is the first 120 seconds after the cold start this morning. You can see that there are three main patterns. Pattern 1 is from timeline 0.0 to 20 seconds. Pattern #2 is from timeline 20 to 60 seconds. Pattern #3 is from timeline 60 and onward.

By the way, I never touched the throttle in these graphs.

 

You are right. After doing some google'ing and digging in the Service Manual, I found info supporting that it is a wideband. THANKS

"Starting in the 98-02 Accords with ULEV F23A4 engines, they used a wideband O2 sensor and a 32-bit Powertrain Control Module (computer) to cinch down on emissions by giving the computer more accurate information to adjust the fuel and spark supplies. Given that the engine is ULEV, it is probably a wideband O2 sensor."



"Volkswagen and Honda pioneered the development of the wide-band O2 sensors to provide accurate air-fuel ratios under these varying circumstances."

 

IHC,

I found that old book that I was mentioning.

"The High-Speed Internal Combustion Engine - Harry R. Ricardo". Published in 1923. I highly recommend this book. The book was divine inspiration back in 1923 (IMO), and the book is very insightful even today.


I was able to download a PDF version of the book from this link below.

The High-Speed Internal Combustion Engine - Harry R. Ricardo (click here)

As a side note - I found some original copies on the internet for sell, valued at $500.

 

Sorry for the late reply to you phee. Wow, that EE talk is past my knowledge.

Soldier, wire, relays, resistors, and maybe a diode are the limits of my electronic wizardry.

Thanks for making the suggestion however.

 

Here is an example illustrating how the ECU will shutdown the throttle between shifts.

My foot was planted heavy on the pedal from 30 to 90 mph. Nissan Titan truck drivers are always challenging me. I race more Titans truck than any other vehicle.

 

Up to this point, I have not discussed using the data logger for performance data collection. Why? I been too busy playing with it to post. Performance data collection was one of the primary reasons that I wanted this logger. With my upcoming mods, I need a method to quantify and compare my performance, before and after making mod tweaks. The DashDyno has lived-up to my hopes in this area.

When purchasing the DashDyno, I was intrigued by it's advertised ability to make dyno charts, and to record 1/4-mile and 0-60 times. But, after trying the dyno charting ability of the DashDyno, I was disappointed. You are required to entry a lot of information, such as wind drag (c/d), frontal area, weight, exact gearing including converter slippage, tire diameter, ambient temperature, elevation, barometric pressure, etc. Way too complicated. I need something simple, repeatable, and reliable.

Next, I wanted to try the 0-60 acceleration recording ability. But before ever trying it, I realized that I have something much better already. I can just have it log my MPH. Why is this better? To make a 0-60 run, you must fiddle with the unit to tell it that you are ready. You make that one run and it records the one run. Well, that's nice but I realize I have something better.

By *continuously* recording my MPH, I am able to analyze my acceleration at any point in my *entire* daily commute. I don't have to fiddle with the unit. I set it to begin recording when I start the car, and I stop the recording as I turn the car off. Everything is recorded. *Sweet*. Now, whenever I get into races, I will have data to analyze to see how my car is performing... anytime, all of the time.

As a side note. Unlike many members here, I am not interested in HP numbers. I would much rather have speed versus time data. In addition, this is raw data. Raw data is better because the data has not been "processed" thru various formulas and factors.

To me, the DashDyno is superior to the other performance meters that rely on accelaratometers. The DashDyno is identical to a "fifth wheel" like "Car and Driver" (and other car magazine) had used in the past to collect their performance data. The DashDyno just collects the MPH as raw data from the ECU. KISS is what helps the data to be simple, repeatable, and reliable.

I also use the RPM-versus-time to quantify and compare my performance. The RPM-versus-time comparison is great to see changes in the powerband between mods. Nearly identical to a dyno chart, but without the complicated formulas to derive the actual HP figures.

As another side note, the DashDyno has a tire-diameter correction factor that can be set. I used my Escort Passport 9500i radar detector, which has a GPS-based speedometer, to calibrate my DashDyno. The DashDyno was calibrated to read 100 MPH as the Passport was concurrently reading 100 MPH. So, I hope we don't have that "tire diameter" discussion again.

I have been doing some magazine racing with my data. Fun. I just beat a 2007 BMW 335i "on paper".




The 2007 BMW 335i did 40-70 mph in 3.9 seconds (according to Car & Driver).
(click here for source)









I did 40-70 mph in 3.67 seconds (according to the DashDyno).

By the way, in real life, I was actually racing a newer Corvette (Z06 ). He spanked me . But, he had to put his foot into it hard. At first, he thought he could just grunt ahead of me. Nope. A second later, I easily hear his monster roaring as he pulled ahead of me.

 

I somehow missed this the first time around. Thanks for sharing. Sometimes going back to the roots can help you discover things you might have overlooked. I've got some reading to do.
Damnit! I've wanted to see this for a while now and I have to wait a couple hours till I get home. The red X doesn't show much.

 

Inaccurate, I think you're going to have to go to the track eventually. Once you install the nitrous and wheelspin becomes a problem in the first two or three gears it may throw your readings off a little.

This is OT but I'm interested to see what happens once your car gets into major wheelspin territory since you're datalogging. There was one thing I was never able to solve in my car and that was knock during wheelspin. All the numbers looked fine, load was reduced with wheel spin, it never made sense. I verified that it was real by adding race gas to my pump gas tune and it went away.

 

IHC,

I sent the following to you a few days ago in a PM. In case you overlooked the PM in your over-active inbox , here is my message again -

1923 Book by Harry R. Ricardo

I think that you will really find this book intriguing. I was just now reading over the chapter devoted to fuels and a chapter devoted to detonation. He talks a lot about the magical properties of methanol and why methanol is able to produce more power. The detonation chapter is just as interesting.

Please let me know if the link that I gave is not working properly. I will gladly walk you thru the process to get this book downloaded. It's free too.

 

That should be some good reading. Have to wait till I get home though.

I'm sure he will touch on this but does he mention the detonation characteristics of different fuels. I've always noticed that if I push the limits of methanol, once the detonation starts, it's violent. There's no 1-2 degrees of knock at first, it maxes out the knock guage right away. For this reason I always keep the meth tune a little more conservative. With C-16 race gas I can tune this thing to the ragged edge and if I get too close it will show up usually in a couple tenths of a degree of knock retard. To put it in perspective on my car, you can't hear it with your ears until you're around 13 degrees of retard.

 

IHC -

Yes. The methanol is known to produce a more severe level of detonation when the methanol does detonate. I have read this several times during my researching. This tendency for methanol to violently detonate is also mentioned in one of my nitrous books. A damn good book on nitrous. I will even venture to say that it is the best written and most knowledge filled and most accurate book that is available.



Nitrous Oxide Performance Handbook, By Jeff Hartman (click here)

In the Hartman book, he also warns the nitromethane will produce extremely damaging detonation when it does detonate.

In the 1923 Ricardo book, he discusses the influences of engine design and ignition timing upon detonation tendency in the chapter devoted to detonation. However in the chapter devoted to fuels, he does discuss each fuel's detonation tendency.

Below is a pic of one of those pages. Sorry, I know that you are "pic challenged" until you get home tonight. So, here is the key quote - "when used under very high compression, methyl alcohol in particular is liable to pre-ignite without warning." But, I don't see him mentioning the severity.

 

That's unreal. I always wondered, while it's easy enough to do on a DBW car, if any manufacturers made use of it. I had considered trying it for myself, but could never muster up the cajones.

Thanks again for all the great information here, Inaccurate.

 

This is so interesting. It's nice to see some literature back up what I found years ago about the methanol. If only I had that knowledge before I blew things up lol. Can't wait to do some reading.

The pic of the port nitrous injection on the LSx (I'm guessing) showed up but just a red X for the second one.

 

The throttle position closing to 39% is not exactly correct. It is misleading. It is *at least* 39%. But, it could be closed even more, perhaps completely closed.

All OBD2 scanners and loggers operate by polling the ecu. If you have polling, then there is a polling interval. Each car manufacturer and model has a different polling interval. The parameters being recorded (example: ATP, RPM, MPH) are polled in a "round robin" fashion.



The more parameters being recorded, the longer time it will be before that same parameter is polled again. In the example immediately above, I am recording 4 parameters (MPH, ATP, RPM, Timing). Each poll interval for the TL requires approximately 0.146 seconds. So, to complete a series of four items will require 0.584 seconds. It actually is closer to 0.535 seconds in practice however.



In the chart above, the blue dots represent each discrete polling. The black line represents a the theoretical throttle position that I drew for our discussion. The green line represents nothing more than "connecting the dots" for each polling value of the ATP.

As can be seen, the 39% can be misleading. The 39% was just the value at that exact time when the poll occurred. But, we don't know for sure what the throttle position was between each polling.

The line that we see on the DashDyno chart that I post is nothing more than "connecting the dots" between each polling value. The green line in my illustration is how the DashDyno draws lines too. Just connect each discrete polling value.

 

Very true. One thing that could be done is to eliminate all other inputs except for TPS and RPM in a controlled environment and make duplicate runs. The RPM's over time would help you synchronize the multiple logs (as closely as possible anyway).

On some data loggers (or at least my LM-2, anyway) you can set the priority level for each of the inputs. So I typically set RPM, TPS and timing advance to high priority and poll things like IAT, ECT, etc... much less often. Doing this still does not net you real-time data, but it's potentially a lot closer than what might otherwise get logged.

BTW, you're the king of the MS Paint visuals. It's freakin' awesome. I'm a very visual guy and it always helps to get some sort of visual on the subject, so keep it up!

 

bmeyer,

I use Photoshop CS2. Doing those things in Paint would be challenging.

I run as few parameters as possible. Provided that I am not looking for anything special, I run just the minimum parameters for performance recording. For my daily performance logging, I log just RPM, ATP, and MPH. The RPM and MPH does not change abruptly, so it is safe to extrapolate (via the line connecting the dots) between each discrete polling to collect my performance data.

If our car's ecu had CAN Bus protocol, the polling would be extremely quicker (I think 6x). I enabled the "CAN Bus 6x mode" support on the DashDyno, but the polling interval did not increase. I thought that our TL ecu had CAN Bus Do you know if our ecu is CAN Bus? Mine is a 2006.

 

Inaccurate,

I was using the "MS Paint" term more as a verb than anything.

On the topic of CAN Bus, no, our cars do not have it. Possible that the 07+'s have it though? I tried a friend's DashHawk which was CAN Bus-only and it could not make any sort of connection to the ECU. You're right about it being a much faster interface. Comparing his data logs to mine nets a significant amount of more data with more fine grained details.

 

Thanks bmeyer. At least now I don't feel like that I am missing-out on the CAN. I thought that maybe I had the unit configured wrong.

Paint me as

 

Had some rain this morning. The chart shows how the ecu was modulating the throttle to control the wheel spin.

The red line is the ATP (Absolute Throttle Position) as was being delivered from the ecu.

The dashed green line is my best guess of how my foot was feeding the throttle.

To give you a feel of what was happening in real life, I felt that the car was accelerating fantastic for the wet conditions. The ecu was modulating the throttle very quickly and allowed full throttle as soon as the traction became available. At no point did I feel like the ecu was holding me back unnecessarily. On the contrary, I feel that there is no way that I could had accelerated anywhere near this quick if it was just me trying to control the wheel spin without the computer assistance.

 

Here is another log showing the Traction Control System in action.

This is from the same day as my previous post. Previous post was from my morning commute. This post is same day but commute going home that evening. Yes, still raining.

A Camaro was next to me at the red light. I was there first and he pulled up later while I was waiting. Looked like it was fast just based on appearance. We are waiting for the green light and he keeps nudging (rolling) ahead while the light is still red. By the time the light actually does turn green, he is one length ahead of me just from nudging during the red light. I remain stopped. "Fine. I will give that to you buddy. Pay back is just moments away."

I go wot fairly aggressively for a wet road. I really felt the TCS cut the power at that 7 mph position. But, I know that it was not the ecu's fault. The traction simply was not available at that point. I never saw that Camaro again until I quite at 52 mph and looked in my rear view mirror. Later, I did see it was a base Camaro with single exhaust tip = V6.

 

Just placed my order for the Innovate Motorsports OT-2.

I'll be ordering the knock detector next. I'm going to need a SSI-4 to interface it to the OT-2 but this will get me started.

I was going to use the F/IC as the datalogger to start but I always have my Iphone on me so this will work out perfectly. If I need anything with more storage, I'll just use my laptop.

 

KN_TURBO_TL,

Thanks for the update.

I'm looking forward to seeing your data from the OT-2.

Maybe we can compare our MPH-vs-Time data

Which knock monitor are you getting? Where will the knock sensor be mounted?

 

KN_TURBO_TL - LOL.

I'm undecided but it's between the Phormula and J&S leaning toward the J&S running 2 sensors. I am going to look under the intake in the area of the stock sensor.

TL money is a little tight but I wanted to be able to see for myself what I had in DD mode, so that was my reasoning on getting the OT-2 before the knock sensor. I wish someone would develop enhanced obd-2 code for the J Series so I could use the onboard data, but I will keep an eye on the timing to start. If I see timing getting pulled on a regular basis, it will expedite my decision to monitor knock.

 

KN

It is extremely hard, at least for me, to determine if the ecu is pulling back the timing by reviewing the data logging. The normal timing (without having knock retard) jumps all over the place as normal practice. So without knowing where the timing should be at that split second, you can't say if it is or isn't experiencing knock retard.

Not truly knowing if I have knock retard has bugged me ever since I started this knock sensor endeavor. The only true way to know is to check the OBD2 Enhanced parameter for Knock Retard.

IHC mentioned in his Pinging thread, that he used a Snap-on brand OBD2 scanner that allowed him to see the Knock Retard parameter of his TL. I checked the Snap-on website and was not able to fully determine which Snap-on scanner might have the ability to see the TL knock Retard Enhanced parameter.

 

overpriced pieces of shit (especially there yearly updates that are like $500) , don't get me wrong i love there hand tools and such (there 3/8 impact is awesome btw), but there are much better scanners out there

 

Try this: http://www.autotap.com/

 

I just discovered something strange and need opinions please.

For Bert, I was developing a math formula that gives mph in 3rd gear (5AT) such as

mph x factor = rpm

I was surprised to see that the data logger was capturing a difference in this ratio. In the illustration above, please notice the widening gap from Point "A" to Point "B". Please notice that the mph does not increase proportionally to the rpm.

Without the torque converter, the 5AT behaves just like a manual trans. That is, the rpm is directly tied, mechanically and mathematically, to the mph because the gear ratios can not vary.

Thus, this leaves just the torque converter causing the varying mathematically ratio between rpm and mph.

I had always thought that a converter will "tighten up" (have less slippage) as rpm's increased. But, the logging is clearly showing the opposite.

My only theory is that the increasing wind drag is progressively working against the converter's fluid coupling. Opinions please.

 

I don't see any mention of supporting manufacturer enhanced data.

The only place I have seen so far is AutoEnginuity. It's around 250 for the base and another 250 to support the Honda enhanced codes.

It appears as though access to this data is very expensive depending on the manufacturer. It must be GM, Ford, Toyota aren't as expensive since you see a lot of enhanced parameter packages for these guys.

 

I have come to grips with the need to get a better data logger. From my research of this thread and the loggers laid out in it, the Auterra Dash dyno seems to be the best. Is anyone running other data loggers that are working well for them? Although I would love to be able to work with the enhanced codes ( Knock ) I am not sure its worth $500 for the Autoenginuity+ enhanced codes. I like that the dash dyno comes with a screen to display some real time info.
I also found a super cheap logger($50) called the davis car chip, but the sampling rate is slow and only 4 sensors can be logged at a time out of 20 or so available sensors. Super small and not a bad deal for people with a less serious need for data logging.

Bert how is your logger working? Are you logging the knock sensor?

 

The AutoEnginuity's Scan Tool (click here) is *not* a viable performance logger based on the logging that I have seen from Bert's AutoEnginuity.

Although the AutoEnginuity is a great scanner for diagnostics, I believe they are over-selling it's ability as a performance tool. The AutoEnginuity appears to lack the ability to accurately collect real-time performance data and/or lacks the ability to present the collected data in a format suitable for evaluating performance (Speed versus Time).

My opinions are based on Bert's graphs presented on page 104 of the turbo thread (click here).

My reservation of the AutoEnginuity's abilities as a performance monitor were presented on page 104, post # 4130 (click here).

 

Thanks for the heads up Inacc. I was considering this product but this has changed my mind.

I bought the OT-2 and I am wondering if these OBD II scanners are really worth it. I haven't attached it to a laptop and have only used my iPhone with it, but so far I am not that impressed.

Looks like I am back to deciding on a knock sensing solution. Leaning back toward the J&S unit.

 

Inaccurate,

Some random thoughts as I read this thread...

- Honda OBD-I systems had ability to put ECU into base ignition timing mode. 90-93 Accords had them in passenger foot wheel area, 2-pin connector you shorted with paperclip. If you found out how to do the same on your TL, it may help for your ignition timing baseline perhaps?

- Has anyone verified since we have ignition coil per cylinder, does each cylinder receive same ignition timing?

- If you plan to read AFR, will your equipment measure from each primary oxygen sensor independently?

- Your planned nitrous/methanol mods will no doubt increase torque. I wonder at how the ECU will react and hinder peformance on aggressive launches. What are your plans for this?

 

Honda,

Yes, the individual O2 sensors (one per bank) can be view independently when using a OBD2 scanner. In the pic above, the blue and green lines represent the two oem O2 sensors (one per bank). Note how closely the afr is between each bank.


Thanks for the heads-up on the 2-pin connector. I doubt that the OBD2 system would allow such an over-ride. However, I will google it just to make sure.


Regarding the Nitrous Methanol Injection (NMI) - I will not intentional spray thru 1st gear (5AT). I am at the threshold of traction thru 1st already. On anything less than a concrete road, I can flicker the TCS light all thru 1st. On concrete, it grips perfect but it wouldn't tolerate much more. I had to struggle to get my 4.7 second 0-60 in my video. I had fresh tires and used a new concrete road.

I will be using just a plain wot pedal switch. I plan to leave the pedal slightly away from wot thru 1st as to not trigger the switch. Then go completely wot after the 1-2 shift to activate the nitrous. It will be interesting to see if the TCS will work when in 2nd gear. According to my calculations, I don't need to worry about the nitrous breaking my traction once I am in 2nd gear........ provided that the tires are not bald and that the road is concrete. But, I am sure that I will have the opportunity to see how the TCS handles the nitrous when I do hit it on an asphalt road. We'll see what TCS does.

I will be using a window switch for sure. The window will be set for 3500-6300 rpm.

 

The logger is working well. The tech support for autoenginuity is excellent. I was having a problem connecting to the enhanced powertrain option so I sent an email to the head engineer and he went out a got hold of a 04 TL and was able to simulate the problem. He then redesigned the software and sent me a beta version to try out and everything works. This was all within one week.

There are a few different knock options in the enhanced powertrain but I have yet to pick up any knock on my logger. I find that extremely hard to believe that I have no knock. Although, when I was testing the knock options, I had a few gallons of leftover 100 octane in the tank. I have yet to try again. I honestly dont know if that option works.


I never got a chance to call the company and explain your findings with them to see if they could explain why everything was logged in 1 second intervals. I will do this soon though.

Inaccurate, Im very tempted to send you my scantool and usb cord so you can use the software yourself and see if maybe Im the problem and Im not doing it correctly. What do you think?

 

Have you tried doing an ecu reset? You may have no detonation because the ecu has already learned how to avoid it. Whereas with a ecu reset, you would have a much better chance to see the knock before the ecu learns to pull the timing in those areas of the map.


I recall IHC saying that he used a enhanced parameter which tells you how much timing is being pulled-out from the map. Any chance that you can log that parameter to see how timing is being pulled?



Let's see what the tech support says about the issue first.

 

I can do it (not sure if you were talking to Bert or me lol) It's a snap-on scanner, I've never had to hook it to the printer but I know it has the capability. I'm going over to the shop and borrowing a rack to do my timing belt in 6 weeks and I'm sure I can do it then. If you need it sooner I can probably get over there later today or tomorrow since I'm off. It's been so long, I'm curious where my car is at now. If you were talking to Bert, maybe we can compare the stock tune vs turbo tune just to see the differences.

 

IHC,

Sorry for the confusion. The suggestion was for Bert. Bert is unable to get his scanner to show any Knock Retard.

 

I did a ecu reset last week so I could watch my fuel trims re-adjust themselves. The ltft's go to -18.0 in one day. However i have not hooked up the data longer since then. Ill try it again with a fresh reset