Simple performance measurement…

Posted on September 21st, 2007 in Opinion,testing by Julian Edgar

stopwatch.jpgNothing, but nothing beats timing the car on the road to find if performance improvements are really that. 

Over the years I’ve made plenty of modifications that resulted in no gains. One of the first was to my Holden VL Turbo, which I owned when it was only a year or so old. I fitted a new aftermarket exhaust and found that the ‘before’ and ‘after’ times had not changed. That’s right: there was no performance gain.

More recently, when Frank the EF Falcon had his new cam fitted, on-road stopwatch testing again sorted fact from fiction. Despite feeling stronger, the stopwatch showed that the gains were trivial.

My favoured timing split is a rolling 60-90km/h in second gear. I hold the car at 40km/h, floor the throttle, and then click the watch at 60 and 90. Irrespective of fancy timing systems good to 1/1000th of a second, if there’s less than a tenth of a second change, you’ll never feel it on the road.

With practice, getting times repeatable to 1/10th of a second can be achieved, and – if they can’t – noting the spread of times is almost as indicative. For example, splits before modification of 2.2, 2.3 and 2.2 seconds are obviously slower than an after-modification 1.9, 2.0 and 2.1 seconds. 

If there is no noticeable change when making these measurements, generally the power gain isn’t worth it. Obvious? Not so! Some quite high-priced changes will yield an unmeasurable change in performance!

The important aspect is to believe your figures and not be swayed by the effort and/or money spent on the modifications. It’s hard to do but otherwise you’re wasting your time….

Another way of measuring on-road performance is with an accelerometer. Again nothing very fancy or expensive is needed – as we showed you here, you can use a boat clinometer very effectively.

Compared with stop-watched runs, the clinometer will show you the shape of engine torque development. If torque is dying away at the top end, or in a turbo car, not coming on hard enough at low revs, the clinometer will show it. All you need is a passenger to read the numbers (and write them down) as you call out ‘now!’ every thousand rpm. (If things happen too fast, select a higher gear and/or do repeat runs.)

The readings of a clinometer, when plotted on a graph, also show you the optimal gear change points. Changing gear at the right point can give a major improvement in acceleration times. This is because holding on to a gear too long will sometimes result in the rate of acceleration being less than if you were to change up. Conversely, changing too early can also cause a drop in acceleration greater than is necessary. Some people try to work out the best gear change points from the power and torque curves of the engine, but this ignores important factors like aerodynamic drag. Other people think that the change point should be at the same revs in each gear, but this also isn’t always the case.

The best way to work out the speed at which to change up is to measure the acceleration in each gear and then plot the different curves on the one graph. It’s then obvious when changing gear will result in a drop in acceleration.

If you aren’t close to a dyno or can’t afford its use, a stopwatch, clinometer, empty road and an assistant will give you a huge amount of valuable tuning information.

One Response to 'Simple performance measurement…'

Subscribe to comments with RSS

  1. Matt King said,

    on September 21st, 2007 at 8:12 am

    Here’s another idea (untried by me, but it should work) – a video camera recording of your speedo or tacho. Given that most mobile phones can now record video, it’s a cheap method.

    You can dump it to a PC, and it shouldn’t be hard to find some video playback software that can count the time of the frames to suitable accuracy.

    Plotting speed vs time is straightforward, and differentiating to calculate acceleration is also easy..

    This method also has the benefit that you don’t have to focus your vision on the speedo, you can keep looking at the road (though the govt. would have us convinced that staring at your speedo is safer…)

  2. Ben said,

    on September 21st, 2007 at 11:45 am

    Another thing with the video, some cars actually have an inbuilt stopwatch, so you don’t ned to count frames. The trip computer on EF/L Falcons has such a feature. It isn’t a ‘normal’ display, but the right buton presses brings it up. It measures in 0.1 second increments.

  3. Scott said,

    on September 25th, 2007 at 2:45 pm

    One issue with monitoring the speedo is that some speedometers (particularly older cars) lag behind the actual wheelspeed during (very) hard acceleration. Probably less of an issue on newer cars.

    But for a personal example, my speedo still shows 75-80kph when I do the 100kph-wheelspeed shift from 2nd to 3rd, if I’m gunning it hard. Presumably Toyota never intended for the AW11 to move quite as fast as this one does now. 🙂

    The basic G-tech accelerometer is available on eBay for $30 or so these days too, of course.

  4. M Peters said,

    on September 26th, 2007 at 10:54 am

    After your many comments about cornering speeds of motorcycles….
    I feel a bit better now. Having been passed by a utility around ‘Chain of Ponds’ in the Adelaide Hills, and many times pulling over to let cars past, a well driven car will go thru the twisty bits better then a motorcycle. (well perhaps might be different if ridden by Casey Stoner!)
    I think the real reason if that a car might be able to operate with smaller margins, but a motorcycle needs larger margins because with only two wheels, a wet patch or a gravelly patch or spilled diesel can have disastrous outcomes.