[wilf leek]

I only put my car onto a chassis dyno to tweak air-fuel mixture, and I was glad that I did, as the original settings were far too close to lean-detonation for comfort.

The dyno operator was extremely open - he said his rig wasn't calibrated worth a damm, could not be compared with any other dyno, and no, he would not "warrant" any output figures. What he did say he would do, and did, was extract around 40bhp more from the engine than it had when it went in. Shows how far out the air/fuel mixture was, plus we recalibrated the advance curve and timing ( I can't believe my engine starts without kickback, with around 20 degrees of static timing advance, but it does, and runs well, with 36 degrees total all in at 3000).
The fact that we had such trouble with wheelspin on the rollers (finally had to do the top end run in 5th - running over 190mph) showed that we had some serious ponies around.
All the dyno operator would say was "well, I run a lot of Ferraris on this dyno, and none of them ever got near the figures we saw today").

This was enough for me. I still do not have any BHP/Torque figures I could bandy about in the bar, but I do have an engine running as efficiently as it can.

From an engineer's viewpoint, (sorry, I am one of those troublesome kind) I can easily see how unscrupulous operators could frig their dyno to flatter to deceive. And I also fully understand why you should not compare figures from one machine to another. These things are not calibrated to "absolute" standards - only comparitive.

The only dyno cells I ever saw that made a reasonable attempt to produce "absolute" figures to rigorous standards belonged to Lotus cars. And they wouldn't take my engine in!