Desktop Dynos, How accurate are they?

[Jeff Chambers]

For my combination, DD was about 55 hp off from what we measured on the engine dyno and it predicted the peak hp & tq points to be about 800rpm less than what we measured. And that was with a carbureted combination which you'd expect its math models to be more accurate at compared to a long runner EFI model.

This doesn't mean that DD was worthless though. You've got to use it parametrically to examine the effects of a component change. Start with your current combination and get a feel for how the DD output varies from what you know about your combination. Then start making changes ONE-BY-ONE to examine how it affects the output.

I used the solver to literally run through thousands of cam combinations over the winter. In the end, it was useful in narrowing in on the cam that I did have ground but I also had to sort through some very bad cams that DD actually said made more power. It'll converge to a poor solution but the computer can't recognize the fact; that's where the user's brain power has to take over.

Bottom line, DD is qualitative not quantitative but it can still be a very useful tool.