edekker
05-03-2000, 07:05 PM
I think it's a great device too. Ther are a number of us who use it.
But like boxerman pointed out, it has to be used carefully.
I've got some pointers on its use, which someone might see as useful.
I invite anyone to please add to this (or even correct me please) - to help me or anyone get the most out of it.
I find that the power cord puts a strain on the unit, moving it, causing it to loose calibration in short order. What I found helpful is to wedge the cord between the utility tray and its cover while ensuring the cord is not taught, at least not between the unit and the tray. This also reduces the cord's effective mass - the forces enacted upon it (under hard acceration) do not cause as much stress on the unit - its effect on the unit's calibration is reduced.
I suggest not to put too much credence on its 1/4 mile trap speed calculation - it is generous to say the least.
When performing g-force runs (skid and instantaneous) make sure it's done on a flat, very level area. If the area is not level, your zero cal will be off. There's away to check that. Once the meter is zeroed, drive slowly around the circle by 180 degrees and stop. The reading should be still be zero. If it's off by a significant amount, say by 0.03g or so, your measurements cannot be trusted - to that level of accuracy anyway. The area has got to be clean - no dust, sand or dirt of any kind. As you make the turns, try not to make the tires screech - at that point already, the tires may very well have lost alot of its grip.
When entering the vehicle's weight, make sure you enter the total weight. That includes the curb wt (all fluids topped up, including the gas), yourself, mods, stereo stuff, everything. And remember, it measures all consumed HP at the wheels including power to overcome wind drag, which becomes substantial approaching terminal velocity.
When you are making concerted runs - to compare a mod for example. Great care has to be taken in your choice of test methodology, to ensure an accurate, dependable result. We may be talking differences of a few horsepower here. Fortunately, while the g-tech may not be accurate in the absolute sense, it could be depended upon as accurate in the relative sense, as long as the test methodology is sound and reproducible.
The key is, to remove sources of error and uncertainty as much as possible. Reducing uncertainty is crucial - the nature of the car's performance (with which the mod affects) needs to be determined and its test methodology tailored to that end.
Possible sources of uncertainty are -
Clutch engagement, throttle control.
Gear shifting, shift points.
Surface gradients, textures, location.
Wind direction and speed.
Barometric pressure, humidity and ambient temperature.
The list goes on...
Instead of accelerating from a dead stop, you can make a rolling start (clutch already let out), say from a steady state 1000 RPM or just rolling on idle (set the g-tech for 'ready' during this steady state) and mash it. A lot of measurements can be made without ever shifting gears - just leave it in first and mash the throttle to redline. These two methods alone can reduce errors alot. If you make several such runs and take the average, the erreors are reduced even more. If you are consistent, taking care of wind, temperature and other effects, I wouldn't be surprised if changes in a couple of HP can be measured with certainty.
Remember, you needn't be concerned with the actual HP, just the changes in HP the mod makes.
Have fun.
Ed.
[This message has been edited by edekker (edited May 03, 2000).]
But like boxerman pointed out, it has to be used carefully.
I've got some pointers on its use, which someone might see as useful.
I invite anyone to please add to this (or even correct me please) - to help me or anyone get the most out of it.
I find that the power cord puts a strain on the unit, moving it, causing it to loose calibration in short order. What I found helpful is to wedge the cord between the utility tray and its cover while ensuring the cord is not taught, at least not between the unit and the tray. This also reduces the cord's effective mass - the forces enacted upon it (under hard acceration) do not cause as much stress on the unit - its effect on the unit's calibration is reduced.
I suggest not to put too much credence on its 1/4 mile trap speed calculation - it is generous to say the least.
When performing g-force runs (skid and instantaneous) make sure it's done on a flat, very level area. If the area is not level, your zero cal will be off. There's away to check that. Once the meter is zeroed, drive slowly around the circle by 180 degrees and stop. The reading should be still be zero. If it's off by a significant amount, say by 0.03g or so, your measurements cannot be trusted - to that level of accuracy anyway. The area has got to be clean - no dust, sand or dirt of any kind. As you make the turns, try not to make the tires screech - at that point already, the tires may very well have lost alot of its grip.
When entering the vehicle's weight, make sure you enter the total weight. That includes the curb wt (all fluids topped up, including the gas), yourself, mods, stereo stuff, everything. And remember, it measures all consumed HP at the wheels including power to overcome wind drag, which becomes substantial approaching terminal velocity.
When you are making concerted runs - to compare a mod for example. Great care has to be taken in your choice of test methodology, to ensure an accurate, dependable result. We may be talking differences of a few horsepower here. Fortunately, while the g-tech may not be accurate in the absolute sense, it could be depended upon as accurate in the relative sense, as long as the test methodology is sound and reproducible.
The key is, to remove sources of error and uncertainty as much as possible. Reducing uncertainty is crucial - the nature of the car's performance (with which the mod affects) needs to be determined and its test methodology tailored to that end.
Possible sources of uncertainty are -
Clutch engagement, throttle control.
Gear shifting, shift points.
Surface gradients, textures, location.
Wind direction and speed.
Barometric pressure, humidity and ambient temperature.
The list goes on...
Instead of accelerating from a dead stop, you can make a rolling start (clutch already let out), say from a steady state 1000 RPM or just rolling on idle (set the g-tech for 'ready' during this steady state) and mash it. A lot of measurements can be made without ever shifting gears - just leave it in first and mash the throttle to redline. These two methods alone can reduce errors alot. If you make several such runs and take the average, the erreors are reduced even more. If you are consistent, taking care of wind, temperature and other effects, I wouldn't be surprised if changes in a couple of HP can be measured with certainty.
Remember, you needn't be concerned with the actual HP, just the changes in HP the mod makes.
Have fun.
Ed.
[This message has been edited by edekker (edited May 03, 2000).]