|
|
Joined: Oct 2000
Posts: 3,867
Hard-core CEG'er
|
OP
Hard-core CEG'er
Joined: Oct 2000
Posts: 3,867 |
The Skinny:
Tested the effect of the MSDS Y-pipe on 0-60 times using a G-Tech. Result was a 0.193 second loss under these particular test conditions.
The Details:
Test vehicle is a 96 Contour SE with a dirty ITG (foam) open element air filter and Borla Cat-back exhaust. Car also has 16" Kosei K-1s with Falken Azenis 205/55-16 with low mileage. Suspension is Koni Yellows with Roush springs, all Konis set near to full "soft." Car has about 73,000 miles on the odometer, and is currently sporting a CEL light with code P0155 (O2 sensor heater circuit). Mobil 1 synthetic 5W-30 in the crankcase.
Test "instrument" is a G-Tech "classic," mounted to the windshield using the center rearview mirror attachment point as reference, and calibrated to level with the car parked in the garage.
Baseline test conducted on 12 Jul 05, with outside temp of about 70 degrees F, full tank of gas, and Traction Control off. Conducted two practice runs prior to beginning of test. Test consisted of 5 passes on (mostly) flat road segment. Planned for 6 passes, with 3 in each direction, but aborted last run when G-Tech suffered a user-induced error (I bumped it completely off level, and had no way to recalibrate). Best and worst run times were scratched, and best 3 averaged to arrive at the result of 9.207 seconds.
After MSDS Y-pipe installation and at least one full drive cycle, returned to same location on 10 Aug 05 to conduct follow up test. Weather conditions were about 73 degrees F outside air temp, and humidity conditions very similar to the baseline test. No other changes were made to the car during this time, and fuel was topped off prior to test. No practice runs were conducted prior to this session. Test consisted of 6 good runs, but with 4 in one direction and 2 in the other. Plan was 3 and 3, but one run yielded the wrong kind of results from the G-Tech (quarter-mile data, instead of 0-60). Best and worst times were scratched, and remaining 4 averaged to yield a result of 9.400 seconds.
Observations/Notes:
Why 0-60? For one thing, I'm testing on public roads. I originally intended to use the G-Tech's hp feature, but I found that the speeds required to achieve a good reading were somewhat ludicrous given this constraint. Even 60 mph is speeding in most places suitable for test, but it's a safe speed. My second justification is that in autocross, you almost never get above 60 mph anyway, so the 0-60 performance measurement made more sense than peak hp or quarter mile times, anyway. It doesn't matter so much if I gain 4 peak hp at insane rpms, if I can't get any pull out of a corner. At least, that's my theory.
About a month passed between test sessions. Obviously, during this time the G-Tech was stowed. So, some error could be induced in the placement and calibration of the unit.
Conclusions:
What conclusion can I (we) draw from this? Does the MSDS Y suck performance? Does it sacrifice some low-end for top-end? Is 0-60 the wrong test? What sort of hp loss does this equate to?
In the future, I had hoped to play around with the intake configuration, and maybe even get on a "real" dyno. At this point, I'm not sure it's worth it to do any more engine work, unless I go "whole hog" with a 3L or FI. Engine modifications allowed in SCCA's STS Solo class are pretty limited, and there isn't much left--pullies, intake tweaking, ADC tuning...
I'm hoping some of you will have other ideas or data, or at least find this helpful.
Function before fashion.
'96 Contour SE
"Toss the Contour into a corner, and it's as easy to catch as a softball thrown by a preschooler." -Edmunds, 1998
|
|
|
|
|
Joined: Aug 2003
Posts: 4,140
Hard-core CEG\'er
|
|
Hard-core CEG\'er
Joined: Aug 2003
Posts: 4,140 |
Roger, thanks for that tech! With so much BS on the boards these days, real scientific evaluation is a nice find. My take on the situation is this: I feel you did a very good job, but can we see the raw data? Without knowing what the raw numbers were, I don't have a good feel for your margin of error.
|
|
|
|
|
Joined: Oct 2000
Posts: 3,867
Hard-core CEG'er
|
|
OP
Hard-core CEG'er
Joined: Oct 2000
Posts: 3,867 |
Originally posted by Auto-X Fil:
Roger, thanks for that tech! With so much BS on the boards these days, real scientific evaluation is a nice find. My take on the situation is this: I feel you did a very good job, but can we see the raw data? Without knowing what the raw numbers were, I don't have a good feel for your margin of error.
Good point.
Here goes.
Baseline
9.15 (West)
10.23 (East)*
8.97 (West)*
9.39 (East)
9.08 (West)
Follow On
9.33 E
9.64 W*
8.96 E*
9.54 E
9.21 W
9.52 E
* - Not figured in final result.
Last edited by RogerB; 08/10/05 09:40 PM.
Function before fashion.
'96 Contour SE
"Toss the Contour into a corner, and it's as easy to catch as a softball thrown by a preschooler." -Edmunds, 1998
|
|
|
|
|
Joined: Aug 2001
Posts: 5,810
Hard-core CEG'er
|
|
Hard-core CEG'er
Joined: Aug 2001
Posts: 5,810 |
Roger,
Less than .2 seconds is too small to really say one way or another. Other factors, temperature and humidity may be close to equal but not necessarily the same and as I've learned, barometric pressure can be signficantly different due to air masses and still have same temp and relative humidity. A lower pressure mass in the area would mean worse driving conditions for power.
The only thing you might be able to say is that the product doesn't appear to have helped you or significantly hurt you.
Former owner of '99 CSVT - Silver #222/2760
356/334 wHP/TQ at 10psi on pump gas!
See My Mods
'05 Volvo S40 Turbo 5 AWD with 6spd, Passion Red
'06 Mazda5 Touring, 5spd,MTX, Black
|
|
|
|
|
Joined: Aug 2003
Posts: 4,140
Hard-core CEG\'er
|
|
Hard-core CEG\'er
Joined: Aug 2003
Posts: 4,140 |
Yes, unfortunetaly, those numbers aren't that good. Your margin of error is large enough that while I suspect you really gained ~.1 sec, I wouldn't call it a sure thing. G-techs just aren't accurate enough for this kind of testing. With 10-20 runs per session you could safely determine your errors and get that kind of accuracy, but that's probably not worth it.
Thanks for the data, though: I find this extremely useful ammunition for "discussions" with people claiming interesting numbers from accelerometer-based timing devices. I'm also curious how much more accurate the new ones are. Enjoy the y-pipe, in any case!
|
|
|
|
|
Joined: Oct 2000
Posts: 3,867
Hard-core CEG'er
|
|
OP
Hard-core CEG'er
Joined: Oct 2000
Posts: 3,867 |
Originally posted by Auto-X Fil:
Your margin of error is large enough that while I suspect you really gained ~.1 sec, I wouldn't call it a sure thing. G-techs just aren't accurate enough for this kind of testing. With 10-20 runs per session you could safely determine your errors and get that kind of accuracy, but that's probably not worth it.
Yeah, as soon as you asked for the raw numbers, I knew what you were thinking. I'd need 12 runs, at least, to establish a preliminary baseline, and then I'd have a better idea of whether the follow-on results were real or random.
Still, I find that the G-Tech is as good as you make it. If the statistical control rules are true for the G-tech, they're true for the dyno, but nobody follows them. You're lucky if most people make 3 pulls. It's a lot easier to control or correct for variables on the dyno, though.
For what it's worth, my median 0-60 time (using all data points) went down from 9.6 to 9.3. I was more consistent on today's runs, so there is probably a skill variable involved here, too.
Originally posted by Auto-X Fil:
Thanks for the data, though: I find this extremely useful ammunition for "discussions" with people claiming interesting numbers from accelerometer-based timing devices. I'm also curious how much more accurate the new ones are. Enjoy the y-pipe, in any case!
Two things, here. First off, my numbers from this test stink. In 2003, I managed an 8.3 using the same methodology. I had stock 15" wheels with half-worn all-season radials, stock suspension, a relatively clean air filter, cool weather, and (IIRC), no CEL. A lot of variables between then and now, so I'm not really worried. But if anyone looks at my numbers and thinks they are crappy, well, I agree but I'm not concerned about it. I can pretty much explain that.
Secondly, I doubt the guts of the G-Tech have changed. The accelerometer itself is extremely accurate. Trouble is usually the way it's mounted, the movements of the car, the undulation of the terrain under even the best circumstances. Other trouble is understanding just what you're measuring. You might get a very accurate quarter mile time and trap speed, but it's only good for that moment and that place. It's very difficult to make real comparitive measurements. The newer devices have more features, and manage the data better. Some of them might have sturdier or more permanent mountings, but those are for more dedicated racecars.
Anyway, enough prattle from me. Thanks for your input.
Function before fashion.
'96 Contour SE
"Toss the Contour into a corner, and it's as easy to catch as a softball thrown by a preschooler." -Edmunds, 1998
|
|
|
|
|
Joined: Oct 2000
Posts: 3,867
Hard-core CEG'er
|
|
OP
Hard-core CEG'er
Joined: Oct 2000
Posts: 3,867 |
Originally posted by warmonger:
Roger,
Less than .2 seconds is too small to really say one way or another. Other factors, temperature and humidity may be close to equal but not necessarily the same and as I've learned, barometric pressure can be signficantly different due to air masses and still have same temp and relative humidity. A lower pressure mass in the area would mean worse driving conditions for power.
The only thing you might be able to say is that the product doesn't appear to have helped you or significantly hurt you.
Thanks. I hadn't thought about barometric pressure, but of course it makes sense.
Function before fashion.
'96 Contour SE
"Toss the Contour into a corner, and it's as easy to catch as a softball thrown by a preschooler." -Edmunds, 1998
|
|
|
|
|
Joined: Aug 2001
Posts: 5,810
Hard-core CEG'er
|
|
Hard-core CEG'er
Joined: Aug 2001
Posts: 5,810 |
Originally posted by RogerB:
Originally posted by Auto-X Fil:
Your margin of error is large enough that while I suspect you really gained ~.1 sec, I wouldn't call it a sure thing. G-techs just aren't accurate enough for this kind of testing. With 10-20 runs per session you could safely determine your errors and get that kind of accuracy, but that's probably not worth it.
Yeah, as soon as you asked for the raw numbers, I knew what you were thinking. I'd need 12 runs, at least, to establish a preliminary baseline, and then I'd have a better idea of whether the follow-on results were real or random.
Still, I find that the G-Tech is as good as you make it. If the statistical control rules are true for the G-tech, they're true for the dyno, but nobody follows them. You're lucky if most people make 3 pulls. It's a lot easier to control or correct for variables on the dyno, though.
For what it's worth, my median 0-60 time (using all data points) went down from 9.6 to 9.3. I was more consistent on today's runs, so there is probably a skill variable involved here, too.
Originally posted by Auto-X Fil:
Thanks for the data, though: I find this extremely useful ammunition for "discussions" with people claiming interesting numbers from accelerometer-based timing devices. I'm also curious how much more accurate the new ones are. Enjoy the y-pipe, in any case!
Two things, here. First off, my numbers from this test stink. In 2003, I managed an 8.3 using the same methodology. I had stock 15" wheels with half-worn all-season radials, stock suspension, a relatively clean air filter, cool weather, and (IIRC), no CEL. A lot of variables between then and now, so I'm not really worried. But if anyone looks at my numbers and thinks they are crappy, well, I agree but I'm not concerned about it. I can pretty much explain that.
Secondly, I doubt the guts of the G-Tech have changed. The accelerometer itself is extremely accurate. Trouble is usually the way it's mounted, the movements of the car, the undulation of the terrain under even the best circumstances. Other trouble is understanding just what you're measuring. You might get a very accurate quarter mile time and trap speed, but it's only good for that moment and that place. It's very difficult to make real comparitive measurements. The newer devices have more features, and manage the data better. Some of them might have sturdier or more permanent mountings, but those are for more dedicated racecars.
Anyway, enough prattle from me. Thanks for your input.
Amen to that brotha! I too am glad you are attempting to quantify something as opposed to most of the bullcrap floating around. 
I too had a G-Tech once and found it to be pretty reliable and accurate once I found a consistent place to run it. Unfortunately a car jockey at the Ford Dealership STOLE mine !  They wouldn't take responsibility for it though.
I agree with what you say especially about the dyno's and the crap associated with it. I just want to reassure you that I myself have done over 45 dyno pulls on my own $ to determine the horsepower numbers that I quote, information that I post, and influence the advice/opinions that I give.
Former owner of '99 CSVT - Silver #222/2760
356/334 wHP/TQ at 10psi on pump gas!
See My Mods
'05 Volvo S40 Turbo 5 AWD with 6spd, Passion Red
'06 Mazda5 Touring, 5spd,MTX, Black
|
|
|
|
|
Joined: Oct 2000
Posts: 3,867
Hard-core CEG'er
|
|
OP
Hard-core CEG'er
Joined: Oct 2000
Posts: 3,867 |
I trust and value your input. Believe me. Thanks again.
Function before fashion.
'96 Contour SE
"Toss the Contour into a corner, and it's as easy to catch as a softball thrown by a preschooler." -Edmunds, 1998
|
|
|
|
|
