Crank hp or wheel hp?Build Ladder blog

Two places to measure

A car maker rates the engine on its own, at the crankshaft. A tuning shop usually measures the whole car on a chassis dyno, at the wheels. Between the two sit the gearbox, the driveshafts, the differential and the tyres, and each one turns a little power into heat. So a wheel figure (whp) is always lower than a crank figure for the same car.

• Nissan GT-R R35. Nissan rates the 2017 US car at 565 hp. Cobb measured a stock 2017 car at 501 whp on its own dyno, about 11% lower (GT-R guide).
• Litchfield's LM1000 GT-R. Litchfield published both numbers from the same car: 1,047 bhp at the engine and 824 bhp at the wheels, about 21% lower. At high power the gap grows in horsepower, not only in percent.
• Honda Civic Type R FL5. Honda quotes 315 hp at the engine. A stock FL5 measured 303 whp on Cobb's dyno and 306 whp on TSP's, only 3–4% lower (FL5 guide). Either the car makes more than Honda says or those dynos read generously; the numbers can't tell which.

So the loss on real cars runs from a few percent to about a fifth. There is no single correct number, which is why our crank to wheel calculator lets you pick one and shows a range.

What the official ratings mean

A maker's figure follows a test standard that sets what is fitted to the engine and which weather the result is corrected to.

• SAE J1349 (US net power). The engine is tested with its production accessories, air cleaner, emissions equipment and exhaust. The result is corrected to 25 °C and 99 kPa of dry air. US makers moved from "gross" figures, measured without accessories, to net figures in 1971–72 (Wikipedia: horsepower, DDI: correction factors).
• DIN 70020 (German PS). Also measured at the engine, with the ancillaries and exhaust as fitted in the car. The DIN correction in dyno software uses 20 °C and 101.3 kPa.
• Older SAE J607 ("STD"). Corrects to about 15.6 °C and 101.3 kPa. Its cooler, denser reference air gives a higher number from the same pull than J1349.

So two sheets from the same car can differ only because one uses SAE and the other STD or DIN. Ask which correction the shop used.

hp, bhp, PS, kW and whp

• hp and bhp are the same unit in practice: horsepower at the engine. UK tuners such as Litchfield write bhp.
• PS is the metric horsepower European and Japanese makers use: 1 PS is 0.986 hp, so 600 PS is about 592 hp.
• kW is the official unit on European papers: 1 kW is 1.341 hp, so 300 kW is about 402 hp.
• whp is measured at the wheels. The guides on this site label every figure hp or whp, and say which one each tuner published.

The exact values come from NIST: one horsepower is 745.7 W and one metric horsepower (PS) is 735.5 W (NIST SP 811). The 1.4% gap is 8 hp on a 600 PS car. Our hp, PS and kW converter does the sums.

Why two dynos give two numbers

Even at the wheels, two dynos rarely agree. Dyno types measure differently, shops apply different weather correction, and the same car reads differently on a hot day. On the Audi RS3 8Y, APR measured a stock car at 412 hp on its own dyno, above Audi's own 401 hp rating. That doesn't make either one wrong: it is why a single peak figure tells you little on its own.

A worked example: converting the right way

Say a shop measures 340 whp and you want an engine figure to compare with a tuner's claim. With a 15% loss:

• Right: 340 ÷ (1 − 0.15) = 340 ÷ 0.85 = 400 hp at the crank.
• Wrong: 340 × 1.15 = 391 hp. Adding 15% to the wheel figure undercounts, because the 15% was taken from the bigger engine number.

Then check it against a real pair from the same car. For the GT-R, 501 whp against 565 hp is a loss of about 11%. Using 15% instead would overstate the engine figure by around 25 hp.

Peak power can hide what changed

The GT-R shows why the peak isn't the whole story. Cobb's Stage 1 map on a 2017 car measured 505 whp against 501 stock, only 4 whp more. Torque at the wheels rose from 453 to 531 lb-ft, a gain of 78 lb-ft (GT-R guide). On a sheet that shows only peak horsepower, that tune looks like it does nothing. On the road the car pulls much harder in the middle of the rev range.

Heat can move the number more than a tune does. In PRL's test, a stock Civic Type R FL5 fell from 302 to 237 whp after repeated pulls with the factory intercooler (FL5 guide). Cobb's Stage 1 for the same car adds 19–35 whp. One cool pull and one hot pull from the same car are 65 whp apart, about twice the tune's gain.

How to compare tunes honestly

• Compare the gain, not the peak. Stock and tuned runs on the same dyno, on the same day, show what the tune really adds.
• Compare like with like. Crank with crank, wheels with wheels, and the same fuel: a figure on E85 or race fuel is not a figure for Israeli 95 or 98.
• Ask the shop three things. Is it at the engine or the wheels, on which dyno, and with which correction? A good shop will show you the before-and-after sheet and the logs.
• Watch for heat. One cool pull flatters every car. If the power drops over several runs in a row, the car is heat-soaking (see Do you need a bigger intercooler?).

Reading a dyno sheet, step by step

1. Find the label: hp or whp, and the correction (SAE, STD or DIN). If neither is printed, ask.
2. Check that the stock and tuned curves come from the same dyno and the same day.
3. Look at the whole curve, not just the peak; mid-range torque is what you feel on the road.
4. Look at the air temperature and how many pulls were made. A first pull on a cold morning is a best case.

Quick answers

• How much power is lost through the drivetrain? On the cars in our guides, from about 3% to about 21%. Our calculator starts at 15%, but treat any single figure as an estimate.

Sources

• NIST Special Publication 811: conversion factors for power
• DDI Software: SAE J1349, J607 and DIN correction conditions
• Wikipedia: horsepower (SAE net and gross, DIN 70020)

Keep reading

• MHD, bootmod3, Cobb or EcuTek?
• Which tuning stage should you start with?