Sale ends todayGet 30% off any course (excluding packages)

Ends in --- --- ---

Dyno question

General Tuning Discussion

Forum Posts



Tech Articles

Discuss all things tuning in this section. News, products, problems and results. 

= Resolved threads


If you're using a 'snapshot' type feature on the dyno to correlate engine speed with roller speed (i.e. hold the engine at say 4000rpm in your chosen gear and press a button to tell the dyno software you're at 4000rpm) rather than use and inductive clamp, injector signal, OBD or whatever to feed an rpm to the dyno; what would be the effect on the power/torque plot if you (inadvertently or otherwise) mismatched the actual engine rpm with the rpm you've told the dyno you're at when doing the 'snapshot'??

Reason I ask is I think this is what's happened with a dyno plot I'm looking at with an unrealistically high torque curve. The power curve seems sensible (albeit 'shifted' a bit lower in the rpm than I would expect) but the bit I can't work out if proves whether the plot is right or wrong, the power and torque do cross at 5252rpm on the graph.


PS this forum post GUI won't work in IE11 on my computer, had to use Chrome to get it to work (can't see the text box in IE).

It will skew the torque figure proportionally with the incorrect engine to final velocity ratio. It may have messed with poorly designed dyno software in the past (where you input gear ratios and tolling diameter) but i think they all just report based on tractive effort and roller rpm now and just swing your calculated figure back out. Without an inductive clamp i think i would just stick to the same gear for power runs and report tractive effort vs speed anyway, if the car is making decent power the tyre slip ratio won't let you directly correlate to table points in the ecu without inductive clamp or ecu comms anyway.

But if the torque curve was 'skewed' (and obviously the power curve 'shifted') it would still cross at 5252prm (assuming imperial units)?


If possible, I like to have one person operating the dyno and a different person operating the tuning laptop. If both people are experienced they can help double-check each other and catch problems like this.

I'm fairly certain the dyno is only measuring one thing (power or torque applied to the rollers, I don't remember which), and then calculating the other. That dyno measurement (tractive effort at a certain roller speed) will still be correct, even if the calculated engine RPM is wrong. Since the engine RPM is just a second calculation (roller speed multiplied by gear ratio factor), I would expect both Engine HP and Engine Tq to be off when the engine RPM is calculated wrong.

If the job was to get baseline power numbers, I'd try to fix the RPM calculation and do another test. If you're using the dyno as a tool for tuning, I don't think this wouldn't be a major problem. Even if you were logging O2 via the dyno, you would eventually figure out which fuel map cells to adjust in order to get a flat-looking O2 trace. Same for ignition timing, you might find that your changes in the ECU's 4000 RPM cell affect the '5000 RPM' measurement on the dyno, but I imagine most people could make it work eventually.

I think I've pretty much satisfied myself in my own mind that power and torque (in imperial units) will always cross at 5252rpm regardless.

AEM_SM, the dyno measures power as a function of torqe (applied to the rollers) and the rotational speed of the rollers. All else being equal, it doesn't care whether you do the run in 1st gear or top gear, it will still read the same wheel HP and the same engine HP after a coast down. Obviously in 1st gear the rollers will see a lot more torque because of the gearing effect but the rollers will not be spinning as fast. Remember, power is work done * time taken or in our case power = torque * rpm.

With the corrected flywheel power and torque graph, because (in imperal units) power = torque * rpm/5252, although our power reading will be 'correct' but at the wrong rpm on the graph, if you consider that the computer is now working back to get the flywheel torque figure from the WHP, it stands to reason that to make the 'same' power at lower rpm you'd have to be making more torque.

This isn't anything to do with locating the correct cell in a map to tune after looking at data logged on the dyno, hell unless they are both talking to each other I will always tune primarily from that the ECU is telling me.

It's just a friend/customer has brought a rally car to our shop saying he's had a VF36 twin scroll fitted (at shop A) to a freshly rebuilt engine in place of the previous VF34 single scroll and says he's lost a 40ft-lb of torque from the bottom end compared to a previous dyno plot (also done at shop A) and he wants me to find that power back for him as it "should be better with the new turbo".

I took one look at the original (vf34) graph and thought 420ft-lb on pump (99 RON) fuel with TMIC is almost 'too' impressive but only 280bhp peak is a bit more reasonable. I asked him if the car felt slower and he said "a bit". Further more, if I just move the power curve to the right it would sit pretty much on top of a similar spec' gpN car of ours (and the resultant torque curve would make a lot more sense).

I firmly believe that the original dyno plot was screwed up because of mismatched roller to engine rpm but just wanted someone to review my logic before I try and explain it to the customer.

We don't have a dyno of our own and I hire a third party rollers for tuning so I'm going to take the car there next week and get a new baseline power run on a MAHA roller I trust and know is operated properly. The previous dyno used by shop A was Dyno Dynamics for FWIW.

Either I have to convince the customer there is nothing "wrong" with it or I have to convince him to spend money for me to find more power, but besides a better intercooler than the OEM Subaru TMIC there is not going to be a whole lot I can do with it on pump fuel as I don't think it's actually badly mapped.

if the dyno has an actual load cell (can directly measure torque, not calculate it) the measured torque is the torque... it's not calculated from anything.

Sort of - they will be raw wheel torque values, just as they will be raw power values, and may be charted against 'road' speeds depending on the dyno' parameters. But without accurate engine rpm to also chart against it will be a flawed 'engine' power/torque graph.

To convert the shear force at the tyre-roller interface to a torque figure the "leverage" from the gearing will need to be known to get the at the engine value - it's much easier to work with an engine's RPM as that will bypass the calculations for the gearbox and final drive ratios and tyre size*. Remember, the dyno' does two basic things - measures the torque applied to the roller (calculated from the shear force at the tyre interface) and from that, and the roller rpm, calculates the raw power at the roller. It uses an engine rpm sender to correlate the raw power to engine rpm and hence torque.

If the engine rpm is being incorrectly reported, it will alter the power and torque figures as the correlation will be wrong. For example, if the true rpm is 6k but the dyno is reading that as 5k, the peak power figure (actually, all the power figures) will be the same but at 5/6ths the rpm, just as the peak torque values will be skewed down proportionally BUT 6/5ths of the true value.

*... and if there is any tyre growth from rpm, it REALLY gets complicated in a hurry.

Sorry if it's unclear, been up since 2:30 am-ish.

This seems to be a crossed link - it's opening this older post, not the one just made by rrodrigius (sp?)

We usually reply within 12hrs (often sooner)

Need Help?

Need help choosing a course?

Experiencing website difficulties?

Or need to contact us for any other reason?