August 26, 2005
Anatomy of a drag race: PT Cruiser vs Hemi Charger
One of the great things about drag racing is that it gives you a safe venue to get the speed bug out of your system. Another is that it lets you know how fast your car really is, which means the temptation to constantly prove it at the stoplight is lessened considerably. But perhaps the nicest thing about drag racing is that it is so much fun.
For those who've never raced, hobby drag racing is simple. If you're lucky enough to have a local dragstrip, head on over. There are often weekday events for regular guys (and girls) with regular cars. At Sacramento Raceway it costs just $15 to race on Wednesdays. What happens is that you get to do a few time trials. Then you "dial in" a time and it is marked on your windshield and side window. The dial in time is fast a bit faster than you think you can go, and it is used for "bracket racing" competition. Bracket racing uses handicaps. So if you dial a 14.8 and your opponent dials a 13.7, you get a 1.1 second headstart. However, it's only 1.1 seconds if both of you have the same rection time. If your opponent sleeps at the wheel and has a 0.7 reaction time and you have a 0.2, you just added half a second to your headstart. All this as a prelude to what was one of my most satisfying races ever.
I had lost in the first round of eliminations to a truck, crossing the finish line 2/100th of a second too late. So I was headed to the "loser lanes" 1 and 2 where eliminated cars could continue racing, without handicap, but still with reaction time included.
As luck would have it, my 2004 PT Cruiser GT was paired with a 2006 Dodge Hemi Charger. An all-Chrysler race. A Mopar race. A race between what is sold as the muscle car successor to all the great muscle car Dodge Chargers of the past and Chrysler's small retro paddywagon. A race between the almighty Hemi V8 and turbocharged 4-cylinder. As I drove up to the staging lane I felt history weighing on my shoulders. The Hemi Charger did some burnouts. I gunned the PT's turbo engine to treat the crowd to some blow-off valve hisses. I knew everyone considered it a mismatch. PTs aren't fast. Hemi Chargers are. Then again, the 2006 Charger is a big 4-door beast that's based on the Chrysler 300 platform. But it has a Hemi.
So my heart was pounding as I staged. Then the lights counted down and we were off. I'll include the numbers from the timeslip in what happened next. Makes it more dramatic. I caught a very good light, with a 0.176 second reaction time. The big Hemi Charger was no slouch either. He got off the blocks in 0.317, which means he was 141 thousandths of a second behind me. Though the Hemi Charger has an autostick and my PT a 5-speed, we did almost identical 60 foot times. I had a 2.268, the Hemi a 2.274. We passed the second timer with virtually identical times: I had a 6.317, the Hemi a 6.318. I power shifted from 2nd to 3rd, which means I never lifted my foor off the gas. The Stage 1 computer in my PT does a "soft rev limit" and so I can do that.
Despite the powershift, the Hemi now inched closer. I passed the third timer after 9.605 seconds, he did it in 9.576 seconds. So now he was just 112 thousandths behind me. At that point my speed was 75.884 mph, his 75.657. I powershifted from 3rd to 4th and felt the PT's engine straining harder as the big and not very aerodynamic body encountered ever more wind resistance. I passed the fourth timer at 12.438. The Hemi in 12.391. I was now just 94 thousandths ahead. The big Hemi inched closer yet, and we passed the final light, the end of the quarter mile. Who'd pass the finish line first?
I did. With a 14.878 to the Hemi Charger's 14.788. I won by 51 thousandths of a second.
The Hemi Charger did not return to race again. It must have been embarrassing to lose to a PT Cruiser. He ran a very good race. We both did. The small difference in reaction time meant I won. And it felt really, really good.
Posted by conradb212 at 2:05 PM
August 23, 2005
Car back at Comptech for injector tuning
Yesterday I brought my RSX back to Comptech in Eldorado Hills. I had been running a set of standard (read: inexpensive to buy) Honda injectors that Comptech had modified to flow about 530cc instead of the stock 240cc or so. Comptech is using those as a reliable, inexpensive alternative to expensive third party injectors for their Stage II supercharger upgrade kit. Another advantage of using Honda injectors is that they fit without having to splice wires. Anyway, Comptech had some customers experience Check Engine lights with these injectors and so we wanted to see why. Turns out that the modified injectors had a different fuel delivery curve than stock injectors. Under light load they delivered too little and under full load too much. That's no big deal as in closed loop, the ECU will compensate via short term fuel trim. However, it may be enough to trigger a Check Engine light for a "lean condition" that customers without programmable ECUs or scan tools cannot diagnose or eliminate.
So Comptech is now doing a new baseline for my car on their Dyno Dynamics chassis dynamometer, and will then re-tune the car so that the fuel curves are okay. I will then road-test the new configuration and check it via K-Pro datalogs.
I also plan on replacing my current 3.6-inch pulley with a 3.3-inch pulley later this week. That should increase boost from 6 psi to maybe 8 psi. Comptech also has a 3.1-inch pulley for the new MP62 snout assembly, but I'll switch to that once the aftercooler is installed, to make up for the cooling-induced lowered boost. As for the aftercooler, it is still not finished. The two test cooling cores are still sitting on a bench, without all the other parts required to install them in my car for testing.
After experiencing a total lack of traction at the drag strip last week with my Kumho Excta 712s I am switching to Falken RT-615s which have a much stickier rubber compound. Should be interesting to see how much of a difference it makes. The Kumhos are not bad tires, especially for the money, but they simply do not provide good traction in powerful frontwheel drive vehicles.
Posted by conradb212 at 2:01 PM
August 18, 2005
At the track
Yesterday I took my supercharged RSX to Sacramento Raceway. I had never raced it before and was wondering how it'd do. Based on experience on the street I expected to have plenty of power, but not enough traction. And that is exactly how it turned out. Needless to say, I took my notebook with me so I could datalog and study my runs on the K-Pro's K-Manager software.
On my first run I basically just sat there at the starting line, spinning my wheels. I had launched at perhaps 5000 rpm and it was like the wheels were spinning in butter. I must have just stood still for a good two or three seconds. I still managed a trap speed of over 103 miles per hour, but the time was miserable, high 14s. I analyzed the datalog, found that at least my air-fuel ratio was perfect and that the engine had no knocks at all, and that even my intake air temperature wasn't too bad. I also found that the IAT can drop by as much as 20 degrees Fahrenheit over a single 1/4-mile pass.
I did six runs in total, each with a different launch technique. Nothing worked as my 225/45-17 Kumho Excsta 712 tires were simply overmatched by the 260 wheel horsepower of the rather light car. My trap speed was consistently in the mid 103s, but I never managed better than a 14.1 through the quarter. In general, a trap speed of almost 104 mph should be good enough for mid 13 runs.
Before my last run I all of a sudden had a Check Engine light. Normally that would be cause for serious concern. However, with the notebook attached to the ECU I could quickly determine that the error code was for a lean condition that did not worry me. I am running experimental fuel injectors that are lean in closed loop where the ECU corrects via short term fuel trim, but run rich enough in open loop. Thanks to the Hondata K-Pro, I could diagnose that MIL on the fly and erase it as I was driving up to the starting line.
This morning I ordered a set of Falken 615RT tires. They are far stickier than the Kumhos and should make a difference. I could have simply gotten a set of true drag radials on stock wheels, but I don't like changing wheels and I do like traction on the street as well, so the Falkens may be a good compromise.
Posted by conradb212 at 2:01 PM
August 9, 2005
IPS-Ksc -- too good to be true?
After all the cam horror stories (or at least the "well, it'll always be a compromise" stories) about cams, the news so far on the Intrinsic Performance IPS-Ksc cams sounds almost too good to be true: great drivability, plenty of power, idle, quiet operation, apparent promise of acceptable emissions, decent gas mileage. Where's the hitch, the weak spot, the compromise? There's got to be one if the no-free-lunch theory holds.
And yet, consider the reports by the initial IPS-Ksc "test pilot," OsideJimC:
"First was a little run around the neighborhood, felt goooood. Lots of torque, nice throttle response. Then a short sprint on a nearby highway, WOW, the car responds like never before. Massive torque, instant throttle response, OK, this is FUN."
"OK, drove about 90 miles today out in Temecula wine country. About 40 on the freeway, 10 in the city, the rest out in the country. Wow I love my car. All my earlier impressions remain intact. Great throttle response, outstanding part throttle driveability. inital fuel economy impressions are favorable, drove all day on 1/8 tank (indicated), but as stated before real mileage numbers will come when I go back to a 9psi / 91 octane setup. A good example is at one time I as behind a truck going up a steep hill at 45 (5th gear), got a chance to go by, just moderate throttle and next thing I know I'm doing almost 90, no downshift to 4th, try that N/A!!"
"OK, it's Tuesday. Not much new to report. Still love 'em!! I did run through a talk of gas. Got 23mpg, not bad for a 343whp car! But going back to 9psi will yield more realistic results since that's my "normal" setup."
"The only issue I found was in heavy traffic yesterday, going 30 in 2nd I switched from a slow lane of traffic to a faster moving one. Changed lanes and punched the throttle, tire smoke filled the passenger compartment, oops. Reminder to self, no need to hammer the throttle to get results."
Posted by conradb212 at 2:00 PM
August 8, 2005
The Hondata k100--9 months later
Almost nine months after its November 19, 2004 introduction, Hondata's k100 remains perhaps the least understood and most perplexing of the company's Honda ECU upgrade products. Intended to offer a less intimidating, less expensive non-userprogrammable version of the Hondata K-Pro, the k100 really never established a niche and perhaps there really wasn't a need for it.
Let's revisit what the k100 is, what it can, and what it cannot do. The idea behind the k100 is to provide the exact same ECU programming capabilities as the K-Pro has, with the sole difference that only a dealer can actually make programming changes. In order to facilitate that, a stock Honda ECU is prepared in the exact same way as an ECU that is being upgraded to K-Pro. However, the k100 daughterboard is much simpler than the K-Pro daughterboard and does not have its own programming interface. The ECU that's being upgraded to k100 status also does not receive a USB connector cutout in its metal housing.
How is the k100 programmed then? The procedure is not complicated but rather cumbersome. The dealer must remove the customer's k100 ECU from the customer's car, open the ECU, remove the motherboard and install the motherboard in a "true" K-Pro ECU that has the USB cutout. He then removes the k100 board and replaces it with his K-Pro board. This "Franken" ECU then goes back into the customer car for tuning. Once tuning is finished, the dealer then removes the ECU from the customer car, replaces the K-Pro board with the customer's k100 board, puts them back in the customer's ECU casing, and then puts the now reprogrammed k100 back into the customer's car. Quite an ordeal.
The theoretical benefits to the customer are that s/he never needs to worry about tuning. The dealer does that. The beenfit to the dealer is that he has a captive customer who needs to come back for re-tuning or software updates. Manufacturers of power adders may decide to bundle their products with a specially tuned k100. To the best of my knowledge, only Comptech has done that so far, with its street supercharger kit for the 2002/2004 Acura RSX Type-S.
There are problems. Judging by the number of questions asked in forums, customers seem very confused about what the k100 actually is. Dealers have been reluctant to accept the k100 because Hondata requires them to do the board upgrades themselves, which requires training at Hondata. The soldering involved is not trivial and the procedure is somewhat complex. Once a k100 has been programmed, it cannot simply be tweaked if it doesn't work right in a customer's car. This means that k100 calibrations must be quite conservative. Even then, a customer's modifications may not agree with it. Further, Hondata releases frequent software updates that often include bug fixes and important enhancements. They can easily be loaded into the K-Pro. Loading them into a k100 requires a trip to the dealer (theoretically, a dealer need not be involved to program a k100; a K-Pro owner willing to perform the above described procedure can do it). Customers have also been irked by the fact that pricing for a k100 to K-Pro upgrade has been left up to dealers. It's a simple procedure (making the USB cutout and replacing the k100 with a K-Pro board) but there are less than a dozen listed Hondata dealers worldwide that handle the k100.
There has been one big success for the k100. Since it is a non-programmable system, Comptech has been able to receive CARB approval for k100-equipped supercharger kits. Those kits perform much better than competing kits without Hondata functionality. I also hear that foreign Hondata dealers may be happy with the k100 as there is no need to send the entire ECU to the US.
All in all, one can sort of see the reason behind the k100, but Hondata really never made a clear and convincing case for it. It's hard to accept the idea of a programmable ECU that is rendered unprogrammable while having the same altered functionality compared to a stock ECU that makes programming/tweaking/tuning a virtual necessity. (Consider also that the K-Pro board is only the interface; the Honda ECU is inherently programmable, and it is the Hondata modifications to the motherboard, combined with the K-Manager software, that make the Honda ECU programmable).
At some point, well before the launch of the k100, there were ECU forum discussions about the business aspects of the K-Pro at clubrsx.com. A lot of interesting ideas were offered, including many that addressed issues that I believe the k100 seeks to address. Hondata did not, as I had hoped, participate in that discussion. In my opinion, that exchange of ideas and opinions could have led to interesting alternates to the current k100.
Posted by conradb212 at 1:59 PM
August 7, 2005
Intrinsic Performance FI cams - first result
Ron from Intrinsic Performance posted the first dyno results for their supercharger-specific cams on Oside JimC's Acura RSX Type-S equipped with an 11psi Comptech SC, Jackson Racing race header, 650cc injectors, Hondata K-Pro, Injen CAI and Comptech 2.5" exhaust. The dyno comparison was done at Church Automotive in Torrance, CA, on a DynaPack. Power went from 324 to 343 at the hub, and torque from 218 to 225 ft/lb. Most interestingly, the IPS cams almost completely eliminated the boost drop usually seen at the VTEC point. This means up to 3 psi more boost between VTEC and midrange, and additional boost even at high rpm.
These results seem to affirm Ron's earlier prediction: "As you may know already, overlap allows boost and fuel to blow right through and ignite a good portion of the mixture well out of the combustion chamber through the exhaust. This causes high EGT's. This is very bad. So, for a proper FI camshaft to perform well - overlap must be reduced or eliminately completely. This is what IPS-Kpsi's were designed and engineered to do. Moreover, the ideal FI cam profile will have significantly larger lobe area conducive to boosted application. And because there no longer exists an overlap condition, the pressurized intake charge will see less restriction at the valve due to the greater amount of valve lift realized with IPS-Kpsi cams. Lastly, the result from meticulously analyzing valve motion and harmonics, IPS-Kpsi cams allow Volumetric Efficiency to scale exponentially under boost, with superior valvetrain stability. We'll see torque curves and HP powerbands hold very strong past their peaks, unlike what is seen with stock camshafts. FI amplifies the engine's natural ability to make power." Ron further stated, "The benefit from overlap in an NA engine helps with cylinder filling at mid to high rpm. I think by the time you hit this portion of the powerband while SC'd, the blower will be doing what it does best. Therefore, I feel it will be an non-issue and that part-throttle driving (SC bypassed) will not incur any downsides other than the inherent parasitic losses seen with a SC installed." OsideJimC, the car's owner, reported, "I'm pleased to report no loss of driveability. Under light throttle the cams are very tame, no bogging and again partial throttle response is excellent. One thing my G/F mentioned, the car is quieter under light/medium throttle. Even WOT seems a little quieter, my theory is without the overlap less intake charge is burning in the exhaust, a very interesting by-product."
So this is a very interesting development. An earlier test with special Crower supercharger cams did not yield good results, and the general recommendation is to use the stock cams with supercharging unless the exact right cams are available. Looks like we now may have those.
Posted by conradb212 at 1:58 PM
August 6, 2005
Entering the world of blogging
Okay... so I installed the demo version of Movable Type on our main Linux server. The install went pretty smoothly, although I did have to look up a few things. For example, I wasn't sure whether I had the MySQL database running on the system or not. Turns out it was there, but the server wasn't running. Then the usual "Google to the Rescue!" in order to find how to start SQL, how to add users to SQL, how to give them proper privileges, and so on. Nothing is ever entirely trivial, but within an hour or so I had remotely installed Movable Type and configured it on the pencomputing.com Linux server.
What am I going to use it for? Initially just as a test to see how Movable Type works and how it can be used, accessed, customized and configured. After that I may figure out if I want dedicated blogs for some of the topics I am interested in, or perhaps none at all. Candidates for blog activity are my work with mobile technology and digital cameras, my work with automotive ECUs and superchargers, or perhaps some of my personal interests, though I doubt I'll pursue the latter.
Posted by conradb212 at 1:57 PM