Zsolt

<br>Hello All,<p>A couple of weeks back MichaelA posted his Gtech measurements before and after the installation of Neuspeed K04 w/extrude honed manifold. See link below. (http://www.a4.org/bbs/a4/9912/messages/282.shtml)<p>Well, his stock 1.8tqms clocked in at 155 hp as stock. This elicited much flame, understandably. His measurement for the K04 kit clocked in at 212 hp. There was much debate if these numbers were meant at the crank or the wheels. THESE WERE MEANT AT THE WHEEL. <p>However, both he and I felt that there is something fishy here, so last weekend we got together to see how his K04 does on my "personal dyno" (my well calibrated stretch of road I always get consistent results with my Gtech).<p>We hooked up both his and my Gtech one under the another, computed our total weight (came to some 37XX pounds). I handled the measuring equipment, he drove. So here are the results from my personal dyno (sm on his Neu K04:<p>His Gtech My Gtech<br>HP:<br>212 195<br>208 192<br>216 195<br>209 198<p>0-60 (took it easy, purpose just to compare the Gtechs):<br>6.28 7.20<br>7.25 7.84<p>Interpretation:<p>1. DON'T YOU EVER COMPARE NUMBERS FROM TWO DIFFERENT GTECHS UNLESS YOU FIRST RUN THEM SIDE BY SIDE AS WE DID IT. Either Gtech QC sucks, or Gtechs suck in general. However, other than the 4th HP measurement, both Gtech behaved pretty consistently. <p>Seems like the only use of Gtechs is that you buy one and compare the the effects of the mods you do on your car, comparing numbers from before and after installation from the same Gtech. <p>2. If you take the average of the first 3 HP runs you get 212 hp vs. 194 hp on my Gtech. This is a 9.3% difference. If you adjust Michael's original 155 hp result with this factor, you get: 141.8 hp, which is just what I got on my stock car (you see, now you can compare Gtech to Gtech!). <p>I believe a ~140hp on a stock car is still a little high, but pretty close to real.<p>3. The 194 hp average indicated by my Gtech on a Neu K04 kit approximately means 225 hp (= 194/.99/.87 to account for air and driveline losses, per Alan A.) at the crank. This is about the figure Neu advertises.<p>Subjective comments:<p>- K04 is BRUTAL. It is loud, it is whiny and it is STRONG. It has a pretty uneven power curve, with slightly more power than Neu .8 bar chip at 1000-2500 and an incredible kick in the butt coming as late as 3000. Then it keeps pulling way high up. <p>- Neu should work with Michael on his chip, it pings. Tests were with Exxon 93 gas, in 70 F, so this only gets worse (in Texas). <p>- Due to the brutal kick coming at 3000 this setup is not as driveable as the chip only. If I was to get this kit I would request Neu to do something about the low end (wett is supposed to pull strong from 2300 - per donp) and smooth out the kick at 3000 (take it back a little). <p>But I guess it only takes time to learn how to handle the fast go pedal. If you learn, I am sure you can get used to the dual personality of the beast. <p>Hope this helps all Gtech users and K04 wannabes.<ul><li><a href="http://www.a4.org/bbs/a4/9912/messages/282.shtml">Neuspeed K04 before and after measurements</a></li></ul>

Alan Adamson

The figures for calculating back from the numbers on the gtech are as follows.<p>FWD<br>gtech/.99(air resistance, no science here, determined after many tests)/.87 (13% loss on the drivetrain)<p>AWD<br>gtech/.99 (same as above)/.85 (15% loss on the Q).<p>I'm pretty depressed that there is that much difference from gtech to gtech. I would suggest an email to the author. They do state, accuracy numbers in their literature!<p>Now I'm really curious, anyone in the Atlanta area with a gtech, care to do some side by sides.<p>BTW, Zsolt, your gtech numbers are much more in line with what I'd expect. The ones from Michael, seem pretty suspect all around!<p>Alan

Lucas

<br>I am sorry about my this posting.<br>I am sick to death of Serbs and Yugos.<br>Especially Now the world has seen them systematically murder Croats, the Bosnians and now Albanians.<br>Tell Clinton to send a Cruise Missile into G-Tech!!<br>I will Never buy a G-Tech, or any product from people or companies associated with yugoslavia or serbia!!<p>My advice send those G-Techs back for a refund!!

Zsolt

Why do you say Gtech is made by Serbs? Are you SURE about this?<p>If I am not mistaken, It says "Tesla Electronics" on it, which is a reputable CZECH company for the better part of the century. <p>Both Czech and Serbs (and Russions, Slovaks, Slovenes, Croats, Polish etc.) are slavic nations (therefore it is easy to confuse names sounding similar), but the Czech live in the middle of Europe and are about 2 million light years more civilized than Serbs living/killing in the Balkans.<p>Otherwise very constructive followup you have here, Lucas. Apologies accepted.

Lucas

I checked their web site.<br>The Owner is From serbia, yugoslavia and educated in beograd.<br>

MichaelA

It was fun working together to test our ideas. <p>A side note: We used a dual oulet plug for the cigarette lighter in order to run both G-Techs at the same time...but it didn't want to stay plugged in. I'm sharing this because we owe Zsolt a special thanks; in addition to insuring the G-Techs were leveled for each run, he had to stay leaned forward and hold the plug in with his fingertips throughout the tests (for about 45 minutes?)...I know he was hurting by the time we finished.<p>As was stated in my original post, I was interested in the degree of improvement rather than the absolute numbers generated by the G-Tech. But many people seemed to fixate on the individual numbers and that was unfortunate. FWIW, I've worked with chassis and engine dyno's in the distant past and experience then was that two dyno's seldom give the same numbers. In that regard, the G-techs are no different. On the other hand, I am concerned that mine seems to be more erratic, but that's why multiple runs are necessary.<p>In any event, I'm reluctant to trade-in my G-Tech because the baseline would be meaninless on a new one. Moral to the story: Never mix G-tech numbers!<p>In any event, the %change from before to after should be valid...and that was what I really intended to measure from the very beginning. <p>Mike

James R.

You've made an interesting analysis…but I feel compelled to pose a question / comment or two. I've have a fairly good understanding of accelerometers and their applications and have to wonder if some of the error is attributable to the alignment / misalignment of the accelerometers sensitive axis with respect to the angle of applied force…in this case the forward movement of the car. <p>Most accelerometers are designed to have one or more well defined axis of sensitivity…in the case of the G-Tech I can say with confidence this is a single axis part…most likely one of the Analog Devices or Motorola parts, single axis accelerometers are the most simple variety. In the case of the accelerometer that is most likely used in the G-Tech there is one well defined sensitive axis which MUST be perfectly aligned with respect to the applied force in order to make a meaningful measurement. The problem is that all accelerometers have a certain amount of transverse (transverse would be 90 degrees either way from the design axis…i.e. straight forward pedal to the metal, one could also call this lateral acceleration) sensitivity. This is usually stated as a percent reading of the actual force as a function of the sensor orientation with respect to the design axis…so basically if the thing is designed to read forward acceleration and you stick it on your window sideways it will still take a measurement…but it will be off because the designer of the accelerometer designed it to measure on only one axis…but some of the force will still be translated to the output terminals of the part as a voltage, current, frequency….whatever.<p>Anyhow what I'm getting at is if you put two accelerometers (which is what a G-Tech is) on a car one facing perfectly forward and one that is off by a degree or two…the one that is perfectly aligned will show a higher level of acceleration than the one that is off a degree or two. Therefore I make the following logical assertion…if two G-Tech meters are on your windshield and one is even a little sideways it will read lower Hp figures and slower 0-60's ect ect.<p>The only way to get around this problem is to use a two axis part which can take into account the lateral acceleration and note its contribution to the calculated figures for the display. Not to mention such a device could allow you to make all kinds of cool measurements…like G pad stuff and what not. Lateral acceleration in the slalom and with even more axis like 3 for instance you could calculate how much roll you have when cornering hard ect. Maybe I should start work on a new G-Tech like product…Hmmm.<p>Anyhow I'd be weary of how these devices are used…they must be applied consistently or you will get wildly varying results. I can tell you with a great measure of certainty that G-Tech probably uses a reasonably accurate part…and I would imagine the unit are trimmed at the factory on a shaker table or the like.<p>Hope you found this informative…if not ohh well what can I say. If you've got any questions or comments….bring it on <br>

MichaelA

First, neither Zsolt not I are that unaware of the potential problem. Second, they were mounted vertically, one directly above/below the other. Third, they were longitutdinally aligned by eye (which is extremely accurate and easy to do when they are mounted vertically and close together) And fourth, they were leveled before each run using their internally generated levelling indications.<p>IMHO, there is simply no way that the mounting/alignment can explain the differences.<p>Mike<br>

Alan Adamson

Ever notice the nose dive, rise that an A4 with stock or sport suspension has? That will influence an accelerometer as well. <p>Not that it has any bearing on this topic. It sounds like you guys did the scientific thing, glad to hear they were stacked and side by each.<p>Anyway, if I knew of another one here in ATL, I'd provide you with the same feedback in my car. BTW, my numbers are more closely aligned with Zsolt's Gtech. I don't have them right in front of my, but that's more in line with what i've seen on mine. <p>I'll keep you posted if I find another one around here. I'm about ready to head to the "real" track and see just what's what.<p>Alan

Ned

Nose dive or any other perpendicular motion does not adversely affect the Gtech or any accelerometer. If the car is moving in any direction except forward, that motion is not being translated into forward speed, and the Gtech accurately compensates for that. A car that has less nose dive, will be faster because it will transfer this movement into a forward motion instead of an upward motion. Get it?