BMW B58 & S58 Maintenance Guide

Guides – by Shopify API

Dyno Numbers Are Lying to You

I get a lot of DMs that look something like this: "Yoooo bro did you see that car made 1,100whp on a stock block with the same turbo I have! Can we do that to mine????" And I have to spend ten minutes explaining why no, that 1,200 number isn't real, and yes, your car would actually run slower if you went and got that tune. The tuning industry has a number problem. Shops post Instagram dyno screenshots that look insane, and customers shop based on the biggest number they see. The only problem is most of those numbers are a combination of correction factor math, dyno calibration variance, and conveniently-cropped condition sheets. Some shops know it and lie anyway. Some don't know it and lie by accident. Either way, you're the one paying for the tune. There are four main ways the number you see is being inflated. Here they are, in the order I run into them most often. 1. The Altitude Excuse You'll see this on every Dragy post out of Denver, Vegas, Salt Lake, anywhere with elevation. "Conditions weren't great, DA was 7,200." The implication is your car would run a tenth or three faster at sea level. For NA cars, sure, that holds. For a properly-built turbo car, it's bullshit. Your turbo doesn't give a damn about ambient pressure. It spins faster, hits the same MAP, and your engine sees the same charge density. The only thing that changes is the compressor moves down its efficiency map a few points, and the IC has slightly more heat to absorb. On a big single like a 6870, the IC eats every degree of that and post-IC IAT barely moves. I broke that down with real customer data here: same engine, same boost, sea level vs 8,800 ft DA, manifold air density dropped 2.9%. That's the total effect on a turbo car at 8,800 ft. 2.9%, not 17%. What actually hurts a turbo car is heat and humidity. The Houston summer pull where your post-IC IAT is at 140°F because the dyno cell is heat-soaked is where you're leaving real power on the table. Denver in January at 4,000 DA will make MORE power than Houston in August at 1,500 DA on the same setup. I'm not just saying that because my shop is in Houston, it's mathematically accurate. Go read the article. 2. Correction Factor Abuse This goes hand-in-hand with the DA crap except that instead of posturing "I would run faster" these shops post artificially inflated numbers, stating them as fact. SAE J1349 and STD are dyno correction factors that "normalize" your number to standard atmospheric conditions. They work fine on naturally aspirated engines because air density directly determines power on an NA car. Indeed, they were developed for NA cars. On a turbo car, they hallucinate horsepower out of thin air. We've "made 1,000whp on just a 6870 on a stock motor b58!!!!" but we knew it was bullshit and told the shop and customer that immediately. That's what inspired this series of articles. I ran the math line by line on two basically identical B58TU customer cars. Same stock 6-port, same Turbosmart 6870, same RK Autowerks intake mani, same ID 1050x injectors, literally copy-paste tunes. One at sea level on our Houston Dynojet, one at ~4k ft elevation (8,800 ft density altitude on the day). Read the full breakdown here. The corrected dyno graphs showed a 188whp gap. Same setup, same tune, two different shops, different elevations, and "the official" dyno graph for the one at elevation made 22% more horsepower than the other. The uncorrected numbers were within 29 HP of each other. The math is correct. The math is also completely irrelevant to a turbo car. Correction factors are refunding you a 17% altitude tax that your turbo already paid back. The dead giveaway: any time you see a corrected number from altitude, look at the CF in the corner of the dyno sheet. If it's a shop at elevation, just assume 10% of that is bullshit for a turbo car. 3. The Winter Uncorrected Flex This one is by far the dumbest. Shops up north love posting uncorrected pulls in January and slapping a comment underneath like "imagine what this would be with a correction factor!" Their followers eat it up because they assume corrected would mean more power. It's the opposite. Cold air is dense as hell. Your engine breathes better, your IC stays cool, charge density goes up. A 20°F day at sea level lands your SAE CF at around 0.91. That means corrected is LOWER than uncorrected. An 800whp uncorrected pull in January would SAE-correct down to 728whp. The shop knows this. They're banking on you not knowing. 4. The Hub Dyno Mystique There's a popular take floating around the tuning community right now: "hub dynos read low." Some big-name shops repeat this as gospel. It's wrong. A stock-turbo B58 Supra on E40 made 500whp on the Pro Hub and 450whp on our DynoJet 224x. That's 10% in favor of the hub, the opposite of the myth. Two things drive this. First, hub dynos run loaded sweeps with an eddy-current absorber instead of a fast inertia ramp. Slower sweep means less power gets eaten accelerating the drivetrain, more shows up as "power on wheels." Mainline publishes this in their own dyno booklet. Second, hub dynos bolt directly to the wheel hub, so they zero out tire-to-roller slip that a roller dyno like a DynoJet has to deal with. A drag radial at 20 psi can eat 3-5% of your power as sidewall hysteresis on a roller dyno. I cite the peer-reviewed work and use Mainline's own documentation against them here. Hub dynos are great equipment. Just don't compare a hub number to a DynoJet number across shops. How to spot a real number from a fake one There's a simple checklist that filters out 95% of the bullshit. Are the conditions disclosed? If the dyno sheet doesn't show cell temperature, baro, and humidity, the shop is hiding something. Don't take any number where the conditions are cropped out. What correction factor was applied? Understand based on conditions and location. If it's sea level and hot, probably fine. If it's at elevation? Ignore the correction for a turbo car. Anything claimed as "uncorrected" in winter at a cold shop, knock another 5-10% off. Is there a baseline on the same dyno? A 500whp tuned number is meaningless without a stock baseline on that same equipment. A 150whp gain over a 350whp baseline is real. A 500whp number with no baseline is just a screenshot. What dyno is it? Pro Hub, DynoJet, Mustang, portable, they all read differently on the same car. Calibrate your expectations to the dyno before you compare across shops. Does the trap speed match? A 500whp car runs certain quarter-mile trap and 60-130 times. If the dyno claim doesn't match what the car runs at the track or on Dragy, the dyno is the one lying. Trap speed is physics, dyno numbers are math. The Bottom Line Dyno numbers are a tool. Great for tuning the car, comparing before-and-after on the same equipment, and validating that your build is healthy. Outside of that, you need real context to be making any conclusions. If it's a shop like ours with thousands of vehicles tuned, and street/track data to back up the dyno, that's great. If they're a random shop that only posts dynos and nothing else? Ignore. If the numbers are too good to be true? Ignore. If you're shopping for a tuner, don't pick based on the highest dyno numbers alone. Pick the one who shows conditions, shows the baseline, shows the correction factor, and posts customers running real trap speeds. That's what we do at Racebox. We dyno-tune on our own DynoJet, we always show conditions, and we baseline before-and-after on the same dyno every single time. No dyno queens here. If you want a real number on your B58, S58, or VR30 build, come see us for an honest BMW or Infiniti tune.

Guides – by Racebox

Dyno Correction Factors Are a Lie (for boosted cars). Here's the Math.

Y'all know the Instagram dyno flex. Some shop posts a screenshot of their Turbosmart 6870 build hitting 1,000+whp, the comments are full of fire emojis, and we inevitably get customers in our DMs asking why their car didn't make 1k whp. Some of them switch, get a new tune claiming that beautiful 1,000, and run slower on the street. I'm tired of saying it over and over again so I'm just going to write an article that we will link you every goddamn time. So if you were sent here via Mo or Alex in an email or DM, enjoy. If it's your second or third time that we link this to you, you're the problem and should seek help. You're getting fed corrected dyno math that doesn't apply to your car. I've been saying this in passing for years - the SAE and STD correction factors were built for naturally aspirated engines, not turbo cars. I get on my soapbox about it every chance I get, but people keep getting tricked by altitude dyno screenshots. I pulled data from two customer cars - same setup, basically twins - and ran the math line by line. One was dyno'd at sea level on our Dynojet in Houston. The other was a remote dyno session at around 4k feet elevation. Same B58TU, same turbo, same intake manifold, same fueling, same target boost. The dyno graphs as presented show one car makes 22% more horsepower than the other. It doesn't. Two identical builds, 188 HP apart Build details for both cars: Both cars Engine Stock 6-port B58TU Turbo Turbosmart 6870 @ 40 psi Intake / fuel RK Autowerks Intake Manifold w/ ID 1050x injectors Transmission Stage 2 built Tune E80 fuel, 40 psi, lambda 0.85, 14° timing. Literal copy-paste between cars. One car at sea level (~96 ft elevation, 88°F cell, baro 30.21 inHg). One car at elevation (~4,000 ft elevation, 86°F cell, baro 26.02 inHg = roughly 8,800 ft density altitude on test day). Same Dynojet model at both shops. Here's the graph: The difference is STARK. You can clearly see the blue line (car at elevation) trumping the other car. But here's what happens when you turn off corrections: Now that looks much better. A variance of ~28whp. Easily understandable given regular variations (3-5%) in dyno cells. Here's a summary of the differences: Correction Sea Level Car Altitude Car "Difference" Uncorrected 834.0 whp 862.9 whp +29 whp (+3.5%) SAE J1349 821.4 whp 1,009.2 whp +187.8 whp (+22.9%) STD 842.6 whp 1,010.0 whp +167.4 whp (+19.9%) As you probably assumed, one car doesn't make nearly 200whp more than the other on the same setup and same tune. This is also why a lot of you fall into the marketing of "oh my tuner is better than yours because I made xxxx whp at only xx boost!!!" That's bullshit. Your tuner knows it (and if they don't, they're a shit tuner). You can have a variance in power between cars with identical tunes based on the health of the motor and a turbo, but at a given boost, fuel, and timing on the same exact motor, there is only so much power a car can make. Tuners that say otherwise are just lying to get your money, which is honestly a good portion of the industry. What are dyno correction factors? Dyno correction factors are just calculated factors to multiply the raw power output of a car to account for atmospheric conditions. There's quite a few, but most commonly you will see shops use STD and SAE corrections. Both corrections take your actual rolled number (what would be displayed if the run showed the "Uncorrected" numbers) and try to "normalize" it to a standard atmosphere. They use different reference conditions and slightly different math, but on a turbo car they're both wrong for the same reason. SAE J1349 references 25°C (77°F) and 99 kPa dry. It includes a friction multiplier: CFSAE = 1.176 × [(99 / Pd) × √(Td / 298)] − 0.176 Where Pd is dry barometric pressure in kPa and Td is dry-bulb temp in Kelvin. Source: SAE J1349 standard. STD is the older standard. References 77°F and 29.92 inHg. Pure atmospheric ratio, no friction multiplier: CFSTD = (29.92 / Pd) × √(T / 537) Where Pd is dry barometric pressure in inHg and T is dry-bulb temp in Rankine (°F + 459.67). Reference conditions per Dynojet's published STD spec: 77°F, 29.92 inHg, 0% RH. Plug in real numbers from both pulls: Sea level car (88°F, 30.21 inHg, 21% RH): SAE CF = 1.176 × (99 / 101.35) × √(304.6 / 298) − 0.176 = 0.98 STD CF = (29.92 / 30.07) × √(547.67 / 537) = 1.01 Either way, the dyno barely corrects - conditions are already close to standard. The corrected and uncorrected numbers are within 1-2% of each other. Altitude car (86°F, 26.02 inHg, 20% RH): SAE CF = 1.176 × (99 / 87.27) × √(303.15 / 298) − 0.176 = 1.17 STD CF = (29.92 / 25.90) × √(545.67 / 537) = 1.17 Both standards add 17% to the uncorrected number to "estimate" sea-level power. SAE and STD essentially agree at altitude. So the dyno took the altitude car's 862.9 whp, multiplied by 1.17, and printed 1,009 whp. The math is correct. The math is also completely irrelevant to your turbo car. Why this is bullshit on a turbo car SAE J1349 was written for naturally aspirated engines. NA engines breathe atmospheric air directly. When ambient pressure drops 17% at elevation, an NA engine genuinely loses 17% of its mass per intake stroke, and 17% less air = 17% less power. The correction "refunds" that loss to estimate sea-level performance. That's legit for an NA car. Turbo engines don't lose 17% to thin air. As mentioned in my DA article, turbos were literally designed to help aircraft engines perform at altitude. The entire initial purpose was to counteract the effects of altitude. The compressor doesn't care what's coming in - it just spins faster to hit your target manifold pressure. Your wastegate is targeting a specific boost number, not a percentage of atmosphere. Pull the DME logs from both cars at peak HP and look at the actual operating data: Channel at peak HP Sea Level Car Altitude Car Cell ambient pressure 14.83 psia 12.78 psia MAP at peak 54.33 psia 53.86 psia Compressor pressure ratio 3.66 4.21 Post-IC manifold IAT 82°F (locked all pull) 93°F (locked all pull) Lambda 0.853 0.846 Timing 14.0° 14.0° Cylinder fill ~270% ~270% Manifold air density 0.2706 lb/ft³ 0.2628 lb/ft³ Same MAP within half a psi. Same lambda. Same timing. Same cylinder fill. Same fuel. The only meaningful difference: manifold air density at altitude was 2.9% lower than at sea level. That's the entire altitude tax on a properly built turbo car. Not 17%. Two point nine percent. The Turbosmart 6870's compressor walked from about 73% efficiency at sea level to about 70% at altitude (3 efficiency points off peak). It dumped maybe 60°F more heat into the charge to make the same gauge boost. But that's fine, because we have intercooled cars, so that means nothing to us. The integrated RK Autowerks A2W intercooler kept IATs steady and close to ambient. Real altitude penalty for this exact build: ~25 HP. SAE-fabricated altitude refund: 146 HP. The other 121 HP is pure fiction. OK but why does the altitude car read HIGHER uncorrected? Because every dyno on Earth has variance, and that variance is bigger than people realize. Three things contribute: 1. Cross-dyno calibration variance (±3-5%) Two Dynojets at two different shops will not read the same engine identically. Drum coast-down calibration drifts with age and so calibration recency naturally differs. Bearings wear. Some dynos sit in conditioned rooms, some sit in barns. Same model dyno, two installations, you'll see 3-5% spread on the exact same engine. That's 25-40 HP on an 850 HP car before you even start. 2. Loss of traction on the roller Age of the dyno also matters. Dynojets in particular ship with a knurling on the roller surface. Over the years and thousands and thousands of dyno runs, this wears down, which introduces more slip on one dyno vs another. I cite sources and studies on this in my other article comparing different dynos, but it's anywhere from 5-10% depending on the power level and the wear on the roller. Even if you think it's not spinning, it's still spinning. 3. Cell air, airflow, ambient conditions Engine dynos work best with cool, well-ventilated cells. Hot ambient air, recirculated exhaust, poor airflow over the intake - all of it costs measured power. A shop with a big bay door open and a fan blowing across the inlet will read different than a sealed-room dyno running off the building HVAC with dedicated flow matched inlets and oulets. And across days or seasons in the same cell, the temperature and humidity swing can dwarf this whole 5% on a turbo car. I broke that down in the DA breakdown. So the 5% variance is real but marginal vs the 25% correction lie Add up the dyno-side variables: maybe 5% spread between two installations. That's the noise floor for cross-dyno comparisons.  Compared to a 17% SAE correction at altitude, or a 22% gap between corrected numbers from two different dynos. The dyno noise is ±5%. The correction-factor lie is +25%.  And uncorrected isn't a magic answer either While in this specific case, uncorrected was much more accurate, it's not always the case. While temperature is a smaller component of the factor, it is still taken into account. And this is where the second lie comes from: when shops present uncorrected numbers in the winter and claim it actually "makes more" if a correction factor was applied. Cold temperature will always bring the correction factor down. If you're at sea level and it's 20 degrees outside vs 85, here's what it looks like: 85°F day at sea level (29.92 inHg, 50% RH → Pd ≈ 29.31 inHg / 99.3 kPa): SAE CF = 1.176 × (99 / 99.3) × √(302.6 / 298) − 0.176 = 1.01 STD CF = (29.92 / 29.31) × √(544.7 / 537) = 1.03 20°F day at sea level (29.92 inHg, 50% RH → Pd ≈ 29.87 inHg / 101.1 kPa): SAE CF = 1.176 × (99 / 101.1) × √(266.5 / 298) − 0.176 = 0.91 STD CF = (29.92 / 29.87) × √(479.7 / 537) = 0.95 Roll an 800whp uncorrected pull on that 20°F day. SAE-corrected, it's 728 whp. The cold dense air made the engine breathe better, and the correction factor brings that back to standard 77°F. The shop posting the 800 number with "imagine what this would be corrected!!" is lying. Corrected would be lower, not higher. The 800 is the inflated number, the 728 is the normalized one. Shop's up north LOVE posting pulls in the winter and claiming they're oh so honest because it's uncorrected. They're not being honest, just trying to sell you more snake oil, as always.  So uncorrected, by itself, is also misleading if you don't know the conditions. What you actually need to compare dyno numbers For a dyno number to mean ANYTHING, you need three things: Ambient conditions plainly listed. Cell temp, baro, humidity. If the run data doesn't include conditions (where the peak numbers are shown) then they're hiding something Correction factor disclosed. If your tuner or your dyno graph doesn't show what CF was applied (or if uncorrected, that's stated) then you can assume it's bullshit. A baseline on the same dyno. The most important dyno comparison is BEFORE and AFTER on the same dyno, same conditions, same tire setup, ideally the same day. So even if they're playing with correction factors, the delta gain is still a reasonable check. Without ambient conditions + correction factor + a baseline on the same dyno, all dyno comparisons are worthless. The bottom line Two basically identical Turbosmart 6870 builds, on two installations of the same model dyno, made essentially the same engine power. Within all measurement noise, they were the same car. The dyno graphs said one was 188 HP stronger. It wasn't. Next time you see "1,200whp at altitude" on Instagram, do the math. Pull 17% off for the correction factor fabrication on a turbo car. Pull another 5% for tire setup. Pull another 5% for whichever dyno is feeling generous that day. You're probably looking at a 950-1,000whp engine. Still a very fast car, but not as fast as they want you to think. The questions to ask any shop posting power numbers: What does it run uncorrected, with conditions disclosed? What was the baseline on this same dyno, same setup? What dyno? That's it. Anything else is marketing. This is what we do at Racebox. We dyno-tune cars on our Dynojet, we always show conditions, and we baseline before/after on the same dyno. No dyno queens here. If you want a real number on your B58, S58, or VR30 build, come see us for an honest BMW or Infiniti tune.

Guides – by Hussain Boxwala

Dynojet vs Mustang vs Hub: No, Hubs Don't Read Low

TL;DR In general, hub dynos read higher than rollers, not lower. Mainline themselves publish a 30-40 hp gain in their own documentation. The hub-vs-roller gap doesn't shrink on big-power cars; it amplifies. The math, the physics, and Mainline's own customer data all agree. Track data settles power claims dyno math can't. Trap speed validates raw HP. ET validates the whole package. Before you read all of this, please understand the intent. I'm not here to call anyone out or say that all shops are liars. I just want to dispel the misinformation so that you all can make better decisions when researching your future tunes, setups, and cars. Everything can be manipulated when it comes to dynos, and I talk about correction factors and DA elsewhere. This is just so you can understand what you're looking at when comparing different cars on different dynos to get an idea of where your car is or should be at. Anyway, a good example I use for people when they ask is this: a customer came in for a retune from another shop in town, complaining that the car didn't feel like it was making the power. It was originally dyno'd on a Mainline Hub with a dyno sheet reading 500whp. We strapped it on our DynoJet 224x days later and read 450whp on the old tune. Same car, same tune. Whose number is "right?" If you spend any time on tuning forums or enthusiast group chats, you've heard the universal claim: "hub dynos read low." People parrot this everywhere, especially the armchair racers who want to feel accomplished and get their ego hurt when someone with a higher dyno number comes around. It's the ultimate cop-out, "well, I would have made 10% more on a DynoJet." Even shops put it in their marketing. Much like misconceptions about DA and correction factors, people are misinformed, and shops are intentionally misleading you. It's my goal to explain to you exactly why the same car read two different numbers, why the "hub dynos read low" line is wrong, why Mainline themselves publicly contradict it, and how to read a dyno number without getting played. Three Ways to Measure Inertia roller: drum sitting on rollers, engine accelerates the drum, power is calculated from the drum's known mass and angular acceleration. The DynoJet 224x is a prime (and the most ubiquitous) example. Eddy-current absorber roller: a roller dyno with a magnetic brake (eddy current or "electromagnetic retarder") that the operator can dial up to apply load. Capable of fast inertia-style ramps AND held-load steady-state. Mustang's MD series, the Dynojet 224xLC, and Dyno Dynamics' Dynotech Series II all live here. Hub dyno: bolts directly to the wheel hubs, no roller, no tire interface. Uses an eddy or AC absorber to load the engine. Mainline's ProHub is the popular tuning-shop unit; Dynojet's Vector-X4000 and Dyno Dynamics' HUB2000 are also hub dynos. So when someone tells you "Dynojets read high vs hubs," what they're really implying (but probably don't know) is that inertia rollers read high vs hubs, which isn't true. When someone says "Mustangs read low," they usually mean eddy-current rollers operated with conservative load calibration, which is somewhat true. The brand name on the side of the dyno is mostly a distraction. Same hardware can produce different numbers depending on how the operator runs the test, and three different brands can produce nearly identical numbers if you set them up the same way. Dynos measure torque. Horsepower is calculated. A quick physics lesson before we get too far into this. Every chassis dyno on the planet measures rotational speed (RPM or angular velocity), plus either a reaction force directly from a load cell (eddy-current rollers and hub dynos) or angular acceleration from how fast the drum speeds up (inertia rollers). Both paths calculate torque from those raw measurements. Power is then always derived from torque × angular velocity. There is no dyno on Earth that measures horsepower directly. Power = Torque × Angular Velocity In SI units, kilowatts equals newton-meters times radians per second. In US imperial, the equation gets rearranged with a unit-conversion constant: HP = (Torque in lb-ft × RPM) / 5252 You might also recognize 5252 at the point at which the HP and torque curves on a dyno chart cross. It's just a unit conversion: 1 HP equals 33,000 ft-lb per minute, and there are 2π radians in one revolution, so HP = T × 2π × RPM / 33,000 = T × RPM / 5252.11. Of course, this is only true on dyno charts that read in HP and lb-ft. Each type of dyno gets this measurement and calculation differently: Inertia roller: torque is back-calculated from drum acceleration. T = I × dωdt, where I is the drum's known polar moment of inertia (this is provided by the manufacturer and part of the initial calibration process when you buy a dyno). The dyno measures how fast the drum is speeding up, multiplies by the drum's mass distribution, and gets torque. Eddy-current absorber roller: torque is measured directly from the absorber's reaction force on a strain-gauge load cell, multiplied by the load cell's known arm length. T = F × r. There is also likely some combination of accounting for the rollers inertial mass, but Dyno companies are notorious for not releasing the specifics about their calculations. Hub dyno: same idea as the eddy-current roller, but the load cell sits on the absorber housing at the wheel hub. T = F × r, just at a different location in the system. All three then multiply T × ω to get power, and convert to HP or kW for the printout. The "truth" in any dyno number depends entirely on the accuracy of those two raw measurements: the torque (whether back-calculated from acceleration or measured by a load cell) and the angular velocity. Everything downstream of that is unit conversion and atmospheric correction. SAE J1349 specifies the standard power equation in detail. How an inertia roller works (DynoJet 224x) An inertia dyno just measures from one part: the drum. The vehicle's drive wheels accelerate the drum. The dyno's controller measures roller speed and computes angular acceleration. Power is then a one-line equation: P = I × ω × dωdt Live readout from our DynoJet 224x during a S58 stock turbo 93 octane pull. Power on the left, torque on the right, RPM on the bottom. Where I is the drum's polar moment of inertia (a fixed, known number from the manufacturer), ω is angular velocity, and dωdt is angular acceleration. Multiply rotational power out and you have your "wheel power" reading.  The advantage of a pure roller dyno is that it's mechanically simple. A pull takes 5-15 seconds depending on the power and gear. Calibration drift is nearly zero because there's nothing to drift; the drum mass doesn't change. That's why DynoJet's own data correlates well to itself across years and across shops running the same model (except for botched correction factors designed for NA cars). The downside is that the dyno is measuring power going into accelerating the drum. It doesn't know how much was lost due to the engine's own rotating assembly, the flywheel, the clutch, the driveshaft, the differential, the axles, and the wheels. Now, everything but an actual engine dyno will experience the losses from the drivetrain, but hub dynos stop at the hub. They don't experience the loss from spinning up the combined weight of the wheels and tires. On a fast inertia ramp, those rotating masses are climbing through 6,000+ RPM right alongside the drum, and the dyno never sees the energy that went into spinning them up. Coast-down measures the parasitic drag of those parts at steady speed, but it doesn't measure the inertial penalty of accelerating them. A Nissan Z on our DynoJet 224x. The tire sits directly on the knurled steel drum, and that contact patch is where slip and patch losses happen. The other main issue is tire-to-roller slip. Force can't transfer between two rotating surfaces without some deformation and slippage; that's just how rubber on steel works. PEREK's technical write-up on tire slip walks through the Michelin friction model and lands on a useful number: in the acceptable measurement range, slip sits somewhere between 5% and 15% depending on the tire and roller combo. PEREK's companion piece on engine power calculation quantifies the wheel-roller contact patch as costing about 8% of transferred power experimentally, and notes that this loss is treated as part of the engine-side correction, not the wheels-side reading. How an eddy-current roller works (Mustang MD series, Dynojet 224xLC, Dyno Dynamics) An eddy-current roller looks similar from the outside but adds a magnetic absorber to the roller shaft. Apply current to the absorber, the magnetic field induces eddy currents in the rotating mass, and those eddy currents generate a torque opposing rotation. The dyno operator controls how much. Mustang's MD-AWD-1100 spec sheet lists the absorber as an "Air-cooled eddy current power absorber (model MDK-250)" with a strain-gauge load cell measuring reaction torque, and describes their load control as "Vehicle Simulation technology where the dyno automatically controls the loading to simulate actual on-road conditions." SAE Technical Paper 980406 covers the use of eddy current absorbers for transient simulation in chassis dynamometry. Dyno Dynamics, the Australian manufacturer that's been building chassis dynos since 1982, calls their absorber an "electromagnetic retarder" but it's the same eddy-current principle. Their current Dynotech Series II 2WD chassis dyno uses a 450 kW / 600 HP electromagnetic retarder per shaft, upgradeable with a twin retarder to 1,200 HP. They also sell the DynoDynamics HUB2000 hub dyno. Dyno Dynamics is the source of the "Shootout Mode" standardized test that became infamous on tuning forums for reading conservative, and that mode is still listed as a feature on their current site. What an eddy-current roller gets you that the inertia-only dyno can't: held-load steady state, road-load simulation, controlled-rate sweeps from any RPM to any RPM, and the ability to load the engine at low RPM for tuning part-throttle areas of the map. For diagnostic and ECU-mapping work that's exactly what you want. It's especially useful in calibrating track cars that don't just see WOT all the time, and for dialing in daily driving loads if you've extensively modified the vehicle to the point where OEM calibrations in those areas are no longer useful. But, to be fair, you can also do these with real world logs in different situations, it just requires more time and customer input. The downside: load calibration is subjective and chosen by the manufacturer, and this shows in the spread of readings across different brands. The eddy current load curve gets baked into the software defaults. Dyno Dynamics' Shootout Mode was famously conservative and that reputation followed the brand for decades, even though their newer Dynotech firmware can be operated with different calibration. A roller dyno with eddy-current load can also be operated as a quick-ramp inertia-style pull (turn the absorber off, run a short sweep), and at that point it behaves much closer to a DynoJet 224x. So the same physical machine can produce different numbers depending on whether the operator runs a long loaded sweep, a held-load Shootout-style test, or a fast unloaded ramp.  Roller dynos aren't all built the same Calling something a "roller dyno" lumps together two structurally different machines that produce different numbers for the same car.  Single big drum (DynoJet 224x and the older 248). One large-diameter drum spans the full wheelbase. Both drive wheels sit side-by-side on the same drum surface. The 224x's drum is 24 inches in diameter, 81 inches wide (60.96 cm × 205.74 cm per Dynojet's pre-installation guide). The original Dynojet automotive model, the 248, used a 48-inch drum, literally twice the diameter, with much more rotating mass, and is preferred by a lot of tuners who can still get their hands on them. A bigger drum has more rotating inertia, which means more of the engine's energy goes into accelerating the drum during a pull and the test "feels" the drum more. It also means the dyno is more stable and less twitchy run-to-run. Twin-roller cradles (Mustang MD series, Dyno Dynamics Dynotech). Two smaller-diameter rolls per axle. Each tire sits in a cradle between the two rolls, contacting both. This adds a roller-roller interaction on top of the tire-roller interaction, but lets the manufacturer build with less rotating mass per dyno. Mustang publishes their rotating-inertia spec in their 2025 catalog: it ranges from 630 lbs (MD-150-SE, 8.575-inch rolls) to 2,530 lbs (MD-1750-SE, 50-inch rolls), with popular tuning-shop models like the MD-1100 sitting at 1,500 lbs. Dyno Dynamics doesn't publish a rotating-inertia number on their current product pages. An empty Mustang twin-roller bay. Each tire sits in the cradle between two smaller-diameter rolls instead of on a single big drum. Why does this matter? The lower the rotating inertia, the more the eddy-current absorber controls the measurement and the less the drum's own physics dominates. A Mustang MD-150-SE with 630 lbs of rotating mass is an "absorber-driven" dyno: turn the eddy off and there's almost nothing for the engine to push against. A DynoJet 224x is the opposite, with its single big drum having substantial inertia, and a 224x without the eddy upgrade can run a full pull on drum acceleration alone. Same calculation framework (T × ω = P), but very different physical setups, different sources of measurement noise, and different tradeoffs between repeatability and absorber-control flexibility. To be fair, DynoJet doesn't publish a rotating-mass-equivalent inertia spec for the 224x in their public manuals, only total dyno weight (3,500 lb / 1,588 kg), which includes the frame and electronics, not just the drum. Dyno Dynamics is the same. Only Mustang publishes per-model inertia numbers. So a clean apples-to-apples cross-brand inertia comparison from public data isn't possible. You can compare drum geometry and total weight, and Mustang model-to-model, but a complete "DynoJet vs Mustang vs Dyno Dynamics drum inertia" table doesn't exist publicly. Even then, the logic stands. You can't compare roller dyno numbers to each other, even within the same brand. How a hub dyno works (Mainline ProHub, Dynojet Vector-X4000, DynoDynamics HUB2000) A hub dyno cuts out the rollers and the tires. The car's wheels come off, hub adapters bolt directly to the studs, and each adapter spins a shaft connected to an absorber. The absorber is typically eddy-current or AC, and the load measurement is the reaction torque on the absorber housing measured through a load cell. You're closer to "real engine power" technically, since you've eliminated inertial losses from spinning up 60lbs of wheel+tire on each corner. Mainline ProHub bolted directly to the hubs of an R35 GTR. No tires, no rollers, no slip — just the engine spinning the absorber shafts through the OE drive flange. The cleanest source for what a hub dyno measures and why it reads differently than a roller dyno comes from Mainline themselves. Their Hub Dyno vs Roller Dyno information sheet exists specifically to explain the gap, and it's signed by the same engineers who design the ProHub. Their own words: "First and foremost, wheels and tyres of a motor vehicle are an 'inertial mass', and as such, consume horsepower during acceleration. The 'acceleration' we are interested in is the acceleration of the wheels and tyres during a 'Ramp' or 'Graph' test on a typical roller type dynamometer." And then the part the "hubs read low" community never quotes: "To make it easy to understand for all technical levels of readers of this article, the average amount of horsepower consumed to accelerate a set of street car wheels and tyres, is around 30-40 horsepower and this number increases as the acceleration rate increases. So at a bare minimum, typically you will gain at least 30 to 40 horsepower just by taking your wheels off and bolting your car up to, and running on, a hub type dyno."via Mainline Automotive Equipment, Hub Dyno vs Roller Dyno information sheet That's Mainline directly stating their hub reads at least 30-40 hp HIGHER than a comparable roller dyno on a typical street car, before tire slip is even in the picture. The same document goes on to give two real customer examples: an XR6 Turbo Falcon that picked up 100 RWKW (134 hp) when moved from roller to ProHub, and a heavily modified R35 GTR that went from 1,300 RWHP on a roller to 1,500+ AHP on a PH4000 hub.  Even if you think I am full of shit and trying to promote the dyno numbers from my shop (I'm not, we have enough Dragy data to back those up), you can't argue with the manufacturers. But let's say you don't want to take Mainline's word for it. Maybe we can find a peer-reviewed academic study. Oh, just kidding, no one in the aftermarket automotive world has done such a study, because at the end of the day the majority of shops in this industry are run by people who don't even know how a dyno works.The best I could find in academia was a 2022 paper on emissions dynos published in Energies by Giechaskiel et al. in which they aimed to test the feasibility of hub dynos for emissions testing in place of currently accepted roller dynos. "Why are you talking about emissions?" Stay with me. The part we care about is where they state that in calibrating the dynos against real emissions from a road test, the hub dyno required two extra inputs (dynamic wheel radius assumed at 650 mm, tire longitudinal slip estimate) to translate hub-shaft rotation into equivalent vehicle speed/distance, since there's no tire to give a rolling radius. If you're smart, you realize this is the mirror argument to what I've cited above from Mainline. Why the same car reads three different numbers 1. Drivetrain inertia consumed on fast accel ramps (the biggest one) An inertia dyno only measures power going into accelerating the drum. It has no way to see the power that's simultaneously being used to accelerate the engine's flywheel, the transmission's input shaft, the driveshaft, and the wheels themselves. PEREK's write-up calls this out directly: "In measurements other than steady state, [engine elements inertia] absorbs part of generated power. It can be roughly estimated from engine displacement. If engine inertia value is correct, engine power measured on different gears should be the same." That last sentence is the clue. If a chassis dyno can't account for engine and drivetrain inertia, you'll see different power numbers in 4th gear vs 5th gear vs 6th gear on the same car. Slower gear ratio means slower angular acceleration of the engine, less inertia tax, higher reported number. That's exactly what every tuner sees when they sweep different gears on a DynoJet. It's also why some shops will dangerously, in our opinion, run a car in 6th gear for a true "1:1" despite the vehicle reaching speeds upwards of 180mph. It shows a higher number than if they ran it in 5th, and their marketing team can yap up the fact that it was run in the "true 1:1" gear. Mainline's hub-vs-roller paper puts a number on the wheels-and-tires share of that inertia tax: 30-40 hp on a typical street car, scaling with acceleration rate. A hub dyno doesn't see the wheels and tires at all (they're not on the car), so that 30-40 hp recovers automatically. The remaining drivetrain inertia (driveshaft, axles, gearbox shafts) is similar between hub and roller, so the difference is the wheel and tire mass. Roller dyno companies don't bake corrections for tire/wheel inertia or contact patch slip into their power formulas. They can't, honestly. There are infinite combinations of wheel size, tire compound, pressure, and strap-down out there, and any baked-in number would just make the dyno wrong by a different amount on every car. The DynoJet Power Core User Guide doesn't use the word "drivetrain" once, and Mainline's hub-vs-roller paper says it explicitly: "the vehicle's drivetrain's inertia is an unknown fact... the inertia of the vehicle's drivetrain should not be assumed." Estimating it would just make the number dishonest in a different direction. 2. Tire-to-roller slip on roller dynos You've seen the videos of people loading up in the trunk or backseat of a car strapped to a Dynojet. No matter what you do, some cars just spin on the rollers when going WOT. It's why you generally switch to hub dynos when tuning above 1000whp. Mainline's Craig made the case in a 2017 Street Machine article on roller dynos as bluntly as anyone has: "Tyres are the biggest killer from a drivetrain loss perspective… As tyres heat up they absorb energy that isn't being transferred to the roller. Tyres distort, the circumference of the tyre changes and the contact patch is reduced, resulting in traction loss. All tyres slip a little on a roller dyno, but as long as the tyre slip is consistent and repeatable, it's fine from a tuning perspective. But the more horsepower you're dealing with, the greater tyre loss becomes." And the punchline: "You can add retarders to a dyno but it won't help traction, and with anything over 1000hp, you're generally wasting your time on a roller dyno." Mainline's hub-vs-roller paper backs this with a concrete analogy: a drag car that runs over 300 kph at the strip but is "speed-limited to 240 kph" on a roller dyno isn't really speed-limited; the rest of that delta is significant tire slip. A peer-reviewed 2022 paper in Mechanics Based Design of Structures and Machines built a twin-roller chassis dynamometer model that explicitly included tire-roller slip interactions and showed that ignoring slip produces large modeling errors that you only see correctly when you account for it. The PEREK piece estimates the wheel-roller patch alone costs about 8% of transferred power experimentally, and slip itself eats more on top depending on the strap-down, the tire compound, and the roller surface. A drag radial at 20 psi being asked to put down 800whp through a single drum is going to give back more than a sticky street tire at 35 psi making 350whp. That's why you can't apples-to-apples a 305 drag radial pull against a 245 street tire pull on the same DynoJet, let alone against a hub dyno that doesn't see any tire at all. The bottom line? The power gap between a roller and a hub compounds as you accelerate faster, both through the power required to spin up the wheel and tire, and the slip between tire and roller on torquey cars. 3. Load curve and sweep speed Eddy-current rollers and hub dynos both let the operator decide how much load to apply during the pull. Run a fast unloaded sweep (close to inertia-mode) and the dyno behaves like a roller. Run a slow loaded sweep with the absorber holding the engine at a controlled accel rate and you maximize the percentage of engine power that goes to the dyno load, which inflates the reading. Held-load testing is the entire reason eddy and hub dynos exist. The number on the page reflects how the test was run, and the same operator can produce two different "peak power" numbers from the same car within the same hour by changing the sweep parameters. Forum-grade comparisons that ignore this end up being meaningless. A Dyno Dynamics Shootout pull and a fast inertia-mode pull on the same machine are not the same test, and the printout says so. The "hubs read low" myth, debunked The universal forum and shop marketing claim is "hub dynos read low." It's just misconception and old tropes that have carried over in the tuning world, kind of like how old gearheads will tell you ethanol will hurt your car. The most damning evidence against the myth comes from the manufacturer everyone points to as the canonical "hub reads low" example. Mainline's own published documentation says the opposite. Their hub-vs-roller paper explicitly states their hub picks up at least 30-40 hp over a roller on a typical street car, and gives a real customer pickup of 100 RWKW on a Falcon and a 200+ RWHP pickup on a GTR. Anyone marketing "hubs read low" while running a Mainline ProHub is contradicting Mainline's own engineering team in print. From Mainline's own Hub Dyno vs Roller Dyno information sheet. The bottom paragraph is the smoking gun: a heavily modified R35 GTR went from 1,300 RWHP on a roller to 1,500+ AHP on a Mainline PH4000 hub. Their words, not ours. Sweep rate, briefly Sweep rate is how a dyno operator decides how much load the absorber applies during a pull. It's expressed in RPM/sec or MPH/sec, and it sets how long the pull takes from starting RPM to ending RPM. Pick a fast sweep and the absorber doesn't load the car enough, the boost doesn't fully build, and the dyno reads low. Pick a slower sweep and the car loads correctly, or overloads, and the dyno reads higher. The 30-40 hp figure Mainline quotes is on a sweep rate that's properly calibrated for the specific car. Mainline, Dynapack, and Dyno Dynamics all publish guidance on sweep rate selection, but in practice it's the operator's call, and most operators end up bowing to customer pressure on what the numbers should look like. So, can a hub dyno read lower than a roller? Sure, if the sweep is way off, the boost is wrong, and the load is wrong. Can it read higher than a roller? Sure. Can it read the same? Sure. Which brings us back to the main point. The numbers don't matter. Baseline, gain, and street performance are what matter. I'd rather see hub dyno operators take pride in the fact that their dynos eliminate the wheel and tire losses, let the numbers read high, explain it to their customers, and use the dynos the way the manufacturers intended. Back to our customer data: 500 vs 450 Customer came in wanting a retune, so we decided to get a baseline on his old tune. Luckily he had the old dyno graph for us to view. Both pulls performed with the same fuel, same tune, same tires. Mainline ProHub: 500whp peak, slow loaded sweep, absorber doing the work. Racebox DynoJet 224x: 450whp peak, fast inertia sweep. That's a 50whp delta. Now that you've read through the article, you know that all of it is some combination of: Inertia tax on the DynoJet (the stock turbo's narrow peak window means the engine and drivetrain are accelerating fast through the measurement window, and that energy doesn't show up). Mainline pegs the wheels-and-tires share at 30-40 hp on a typical street car; that alone covers most of the gap. Tire and patch losses on the DynoJet that the hub doesn't see at all. PEREK's experimental ~8% wheel-roller patch loss accounts for the rest. Sweep methodology difference (loaded sweep on the hub vs. fast inertia sweep on the DynoJet). If you ran the hub with a fast inertia-mode sweep instead of a slow loaded sweep, the hub number drops noticeably. If you somehow ran the DynoJet with held-load (you can with the 224xLC eddy upgrade, but the 224x can't), the DynoJet number rises noticeably. The cars don't lie. The dynos don't lie. The methodology gets the credit and the blame. Dragy and track data, not dyno math, settles big-power claims Drivetrain inertia tax on a roller scales with power. Ptax = Idrivetrain × ω × dωdt. A 1,500whp big-single accelerates the same drum and drivetrain harder than a 500whp stock-turbo, so dωdt is higher at peak, and the absolute hp lost to drivetrain inertia is bigger on the more powerful car. As a percentage of engine power, the tax stays roughly fixed. It's set by wheel/tire mass, drivetrain mass, gear ratio, and the dyno's drum inertia, none of which depend on engine output. Patch loss is the same story. PEREK quantifies it as roughly 8% of transferred power experimentally. 8% of 500whp is 40 hp. 8% of 1,500whp is 120 hp. Bigger absolute hp lost, same percentage. Tire slip also gets worse at higher torque, not better. Every mechanism we've covered either holds steady in percentage or grows in absolute hp as power scales up. Mainline's own documents back this up. Their hub-vs-roller paper gives two real numbers: an XR6 Turbo Falcon picked up 100 RWKW (134 hp, ~17% gap) moving from a roller to a Mainline ProHub, and a heavily modified R35 GTR went from 1,300 RWHP on a roller to 1,500+ AHP on a PH4000 hub. That's 200+ hp / ~15% gap. Percentages similar, absolute hp gap bigger on the bigger-power car. So why does the dyno-vs-dyno argument feel less important on a 2,000whp big-single than on a 500whp street car? Because the 2,000whp car has track data. The 500whp street car usually doesn't. Trap speed validates raw engine power directly. A purpose-built track car running an 8-second 1/4-mile pass with a 175 mph trap is making the engine power its dyno number suggests. Trap speed can't lie because at the strip the air, the road, and gravity don't care which brand was on the dyno. A 500whp stock-turbo street car with no strip data has nothing independent to validate against. The dyno number is the only data point.  ET tells you something different and arguably more important: how the whole package performs. Engine power is one input. Weight, traction, gearing, transmission tune, launch, and driver execution are all inputs too. Look at Qzilla, our VR30 world record car: 825 whp / 674 wtq on a Dynojet 224x, with twin Garrett G25-550s, full E85, and our VR30 tune. Plenty of built-motor VR30s make more power than that. None of them run quicker. Qzilla owns the platform record outright at 8.89 @ 148.40 in the 1/4 mile and 4.60 in the 60-130. And the cleanest data point we've got to show it's not all in the power is Qzilla going from a 5.06 60-130 to 4.60 just by adding our custom RE7R TCM tune. Same engine, same turbos, same dyno number. Half a second came off the ET. Qzilla on a Dynojet 224x: 825 whp peak, 674 wtq peak, on stock engine and stock trans. So when people flex dyno numbers online or at a car meet, just ask them what they run. If they don't have Dragy or track data to back it up, it doesn't matter. How to actually read a dyno number We tell people this all the time. Dyno's are just tools. They are only as good as the operator, and as the data provided. Every experiment in science needs a control to establish context, and changes. You need the same with a dyno. Always get a baseline. Yes, we have approximate power ranges for what certain cars on certain setups should make, but you really should only be looking at how much power you gained. Don't compare dyno vs dyno. A 500whp ProHub number and a 500whp DynoJet number on different cars are not the same. Even on the same car they wouldn't be the same. Treat absolute peak HP as a marketing number unless you have track data backing it up. A good shop will be happy to back it up with customer Dragy and track data. Look at correction factors. On Dynojet, ask for the run conditions and look for yourself. On Mustangs, Mainlines, and others, ask them to display atmospheric corrections and the total correction factor "TCF." This is the one that's the most telling. And more importantly, make sure the correction factor is the same across all runs. Flip side: if the strip backs up your dyno number, your dyno was legit. If it doesn't, the dyno was lying regardless of what brand was on it. That's the test you can't game. So which Dyno is "right?" They all are. That's the point! Every dyno is right in its own regard. As long as you have a baseline, run conditions, and correction factors displayed, you have the information you need to make a solid conclusion. The facts are that most dynos have a custom correction factor that operators can apply to skew the results. So even if a Mustang Dyno out of the box reads low, the dyno operator can add a TCF and bring all the readings up. A Dynojet operator doesn't have a custom TCF to play with, but they can always stick the ambient sensor somewhere really warm just to bump up the numbers. Then there's always the sweep rate on load absorber based dynos that will result in different numbers as well. Most shops aren't out there to deceive you, but occasionally you'll spot the TCF and wild numbers if you're looking, and it's important to know that so you don't fall for sleazy marketing. Operators can make the dynos intentionally read lower, too. There's a reason SCCA horsepower limited classes require runs on a Dynojet - it's harder to fake or lower the number. At the end of the day, the actual numbers themselves don't mean as much as the power gained. If your Q50 dyno'd at 450 on E85 but runs a solid 7.5s 60-130, then you know the dyno read low. If it made 550 and runs a 9s 60-130, you know it read high.  Final Thoughts Look, I know the majority of you reading this care a lot about dyno numbers. Hell, we advertise tunes based on dyno numbers. But we also have the times to back it up. When we throw dyno numbers at customers, we tell them approximate 60-130 times. We encourage them to get out there and run. We only advertise the dyno numbers because we know what they mean. And it is my honest hope that most shops operate the same way. There are definitely some that hide behind big dyno numbers, and a lot try to underplay their hub readings to sell tunes. At the end of the day, it doesn't matter as long as they're putting out the results. The problem is, a lot of them don't. And I'm not saying dyno numbers are useless. We know you guys like to see them. We even post things like "highest HP records" and shit. But we don't caveat everything with "hey but its hot, it would have made more up north!" and, more importantly, our claims are backed up by record setting times. The next time someone on Facebook tells you "well my car actually would have made MORE if it were on a DynoJet," call them stupid and move on. And if a shop is telling you that? Steer clear, they're just trying to market with lies. Sources Dynojet 2WD Automotive Chassis Dynamometer Model 224x product page (inertia-only base unit; eddy-current upgrade sold separately) Dynojet 224xLC product page (224x with eddy current absorber bundled) Dynojet Vector-X4000 Hub Dynamometer product page Mustang Dynamometer MD-AWD-1100 product page and specifications (eddy current absorber MDK-250, Vehicle Simulation load control) Dyno Dynamics company history and product line (founded 1982, Australia) Dyno Dynamics Dynotech Series II 2WD chassis dyno product page (electromagnetic retarder spec) Dyno Dynamics HUB2000 hub dyno product page Mainline Automotive Equipment, "Hub Dyno's vs Roller Dyno's" information sheet (manufacturer's own primary documentation; 30-40 hp hub-vs-roller delta, real customer examples) Street Machine, "Tech Torque - Roller Dynos" (March 2017) with Craig from Mainline DynoLog quoted on tire loss Mainline DynoLog Premium Range Dynamometer Information Guide (PDF, 51 pages, Oct 2013) PEREK technical documentation, "Tire slip on roller chassis dynamometer" (5-15% slip threshold for valid measurement) PEREK, "Engine power calculation on roller chassis dynamometer" (~8% wheel-roller patch loss experimental, engine inertia absorbs power in non-steady-state) Lourenço, M.A.M., Eckert, J.J., Silva, F.L., Santiciolli, F.M., Silva, L.C.A. (2022). "Vehicle and twin-roller chassis dynamometer model considering slip tire interactions." Mechanics Based Design of Structures and Machines. DOI: 10.1080/15397734.2022.2038199 Giechaskiel, B., Forloni, F., Otura, M., Engström, C., et al. (2022). "Experimental Comparison of Hub- and Roller-Type Chassis Dynamometers for Vehicle Exhaust Emissions." Energies, 15(7), 2402. DOI: 10.3390/en15072402 SAE Technical Paper 980406, "Eddy Current Dynamometers Suitable for Transient Simulation" SAE J1349 standard, "Engine Power Test Code"

Guides – by Hussain Boxwala

Stop Crying About Altitude: Heat & Humidity Hurt Turbos More

What density altitude actually is, how it's calculated, and why a 68mm-turbo build doesn't lose nearly as much as the altitude shops claim. With real Houston and Denver weather data and a PTE 68mm compressor map walkthrough.

Guides – by Hussain Boxwala

BMW B58TU Tuning Guide: Supra, M240i, M340i, M440i & More

The Racebox B58TU Tuning Guide If you own a B58TU powered BMW - whether that's an M240i, M340i, M440i, 540i, X3/X4 M40i, or a Toyota Supra (or an X5!) - you're sitting on one of the best tuning platforms BMW has ever made. The B58TU replaced the already-excellent Gen 1 B58, and it responds to modifications like nothing else in its class. This guide is built from years of hands-on tuning and wrenching experience at Racebox, and it's going to walk you through exactly how to make real, reliable power with your car. We're not going to throw inflated dyno numbers at you. Everything in this guide is based on what we see on our dyno, with real cars, real fuel, and real results. That matters, and we'll explain why. A Word on Dyno Numbers Before we get into the builds, we need to talk about dyno numbers because this is where the tuning industry loves to mislead people. Not all dynos read the same. Some dynos read higher than others. Some have adjustable correction factors that allow you to make the dyno read high OR low (Mustang Dynos and most Hub Dynos). Even two Dynojets can read wildly different numbers based on the ambient conditions and elevation, as the SAE and STD correction factors were developed for NA cars, not modern turbocharged vehicles. The numbers you see from different tuners are often measured on different equipment, which makes direct comparisons meaningless unless you're looking at the delta from baseline. That's the only number that matters: how much did the car gain over stock on the same dyno, same day, same conditions? We've seen competitors advertise "550whp on stock turbos" for the B58TU. Sounds impressive until you realize their bone stock baseline is showing 400+ whp. That's not a 550whp car, it's just a generous dyno with a 150whp gain, which is exactly the same as a car that makes 500whp with a 350whp baseline. On our dyno, a stock B58TU puts down 320–370whp depending on the model, and our tune numbers are measured against that honest baseline. When we tell you a number, it's real. So when you're comparing tuners, don't compare peak numbers across different shops. Compare the gain over stock. That's where the truth lives. Stock Baseline: Know What You're Starting With The B58TU doesn't make the same power across every model. The biggest factor is whether you have a 2-port or 6-port head, as the DME calibration for the 6-ports is more aggressive from factory. Configuration Models Stock WHP (Racebox Dyno) 2-port X5 40i, X7 40i, 2-port Supra  320–340 whp 6-port M240i, M340i, M440i, 540i, X3 M40i, X4 M40i, 6-port Supra  350–370 whp In practice, we've seen that both attain the same results when tuned with similar mods. Differences in flow only become apparent and restrictive when we exceed 850whp with 6-ports making more power with less boost on larger setups up top.  These are your starting numbers. Everything we talk about below is measured as gains on top of these baselines, on our Dynojet. The basics: Downpipe + Tune (Where Everyone Should Start) This is the single best modification you can do to a B58TU. A quality catless or catted downpipe paired with a custom tune from Racebox is where the B58TU really wakes up. The factory downpipe is restrictive and is the single biggest bottleneck in the exhaust system, and replacing it with a higher flowing aftermarket setup lets the turbo spool faster and breathe significantly better. Most BMW cat-backs are around 2.75 in from the factory and thus a full aftermarket exhaust is not necessary for power on stock turbos. 93 Octane Downpipe + Tune On 93 octane pump gas with a downpipe and our custom tune, expect: 420–450 WHP on 93 Octane That's a gain of roughly 60-100whp over stock depending on your model and baseline. The 2-port cars see the bigger percentage gain since they're starting lower, but the 6-port cars hit higher peak numbers. The midrange pull is where you'll notice it most. The turbo comes on harder, stays on longer, and the car just doesn't run out of breath the way it does stock. This is the stage where most people new to modifying their cars should live. It's reliable, it's repeatable, and it makes the car genuinely fast without stressing anything beyond what the factory hardware can handle. The stock turbo, stock fuel system, and stock intercooler are all perfectly happy at this power level on 93. New to all this? Read our Car Tuning 101 guide first for a primer on what a tune actually does. E40 Blend Downpipe + Tune If you have access to ethanol and are running an E40 blend (roughly 40% ethanol, 60% gasoline), the B58TU responds extremely well. Ethanol has a higher octane rating and better charge cooling properties, which means we can run more timing and more boost safely. 480–510 WHP on E40 Custom Tune This is still on the stock turbo, and stock fuel system with just two mods: a downpipe and ethanol content analyzer (ECA), paired with our custom calibration. The spread in numbers comes down to the health of the motor, the condition of the turbos, intake air temperatures, and how consistent your ethanol blend is. A car with fresh plugs, clean injectors, and a healthy turbo on a cool day will hit the top of that range. A car with 80,000 miles on the original turbos in Houston summer heat might sit at the lower end. The majority of the B58s out there are on this setup and there is a reason why. It’s a simple recipe for an enjoyable daily driver. If you're going to run ethanol, you need to be monitoring your ethanol content. We like the MHD Flex Fuel Kit as it works with all tuning platforms (EcuTek and MHD) and it is a very easy install across all Gen 1 and B58TU chassis. The flex fuel kit also allows our calibration to change on the fly depending on what exact ethanol blend is in the tank without having to change maps. In addition, we can put in safeties baked in the tune in case you overshoot your ethanol content. Don't guess. Remember, ethanol dilutes your oil faster, so drop your oil change interval to 3,000 miles max on any ethanol blend. Read our B58 Maintenance Guide for the full breakdown on that. Max Effort Stock Turbo: Full Bolt-Ons (FBO) Once you've done the downpipe and tune, the next round of mods is what gets you to true FBO status. This is where you add the supporting hardware that lets the engine and turbo work more efficiently. FBO is not necessarily for massive peak power gains, but for a broader, more consistent powerband and better heat management. What FBO Means on a B58TU Charge pipe upgrade - the OEM plastic charge pipe holds up at stock boost on the B58TU, but once you're tuned and pushing higher boost targets, swap it for an aluminum unit with T-bolt clamps. Don't wait for it to crack. Also, if you're tossing on an upgraded intake mani or PI anyway, you'll have to change the charge pipe to fit. Upgraded charge cooling - the stock water-to-air system is decent, but it heat soaks on hot days and repeated pulls. Upgrading the stock intake manifold to an aftermarket unit such as the RK, DO88, CSF, etc is the best cooling mod you can do once you start to add power. An upgraded heat exchanger keeps charge temps consistent, which means consistent power. Intake - the stock intake is fine for performance. If you want sound and aesthetics, go for it, but the gains over a quality drop-in filter are minimal. Exhaust - while the stock exhaust is plenty for the stock turbos, you can still gain a little bit on the top end. Similar to intakes, do this for more noise per your preference rather than performance. Port injection - this is the final step to gain every last bit of performance out of the stock turbo. Adding PI via a PI rail or with the upgraded intake mani will allow you to run full E85, unlocking another 10-15whp with the stock turbo. With FBO and our custom tune, you're looking at similar peak numbers to DP/tune. The gains from charge pipe, intake manifold, and intake are more about consistency and reliability than peak horsepower. The car holds power better through multiple pulls, on hot days, and in higher gears where heat soak kills stock setups. Without the timing corrections from the increased IATs, you can hold onto your power especially as you increase power which increases heat. Port Injection: Why PI Over DI, Always If you're thinking about upgrading your fuel system on a B58TU, we need to talk about port injection versus direct injection upgrades. Our position is simple: there is no reason to run upgraded DI when fully integrated, closed-loop port injection is available. Every single customer we've had who started with upgraded DI injectors eventually sold them and switched to PI. Every. Single. One. PI is cheaper, supports full ethanol at any power level, and integrates fully with the DME through MHD+ or EcuTek, meaning the DME sees everything the port injectors are doing, maintains all safety features, and keeps fuel trims accurate. With CAN-bus integrated port injection on MHD+ or EcuTek, the DME monitors everything. If a port injector fails, the DME catches it immediately. On a non-integrated setup, a failed injector means a melted piston and a $15,000+ problem. This is why we always recommend MHD+ or EcuTek to integrate Reflex+ port injection seamlessly into the calibration. Full Send: Big (Top mount) Turbo This is where the B58TU becomes a completely different car. The stock turbo is efficient but it's physically maxed around 550whp on ethanol. Once you bolt on a bigger turbo, the ceiling disappears. Turbo Kit Options The RK Autowerks Top Mount kit and the Black Market Race (BMR) Lightning kit are what we recommend the most. The BMR Lightning uses a Turbosmart 6870 compact frame turbo with dual 45mm wastegates. It is the easiest top mount kit to install and makes excellent power with great spool characteristics. Doc Race, ETS, and KLM also make solid kits depending on your goals and budget. Supporting Mods for Big Turbo As you can imagine, you cannot just bolt on a bigger turbo and call it a day. You need all of the supporting mods from the FBO section above, but we'll list them again along with some others. Port injection — ID1050X injectors are the go-to for most builds. Go ID1300X or ID1700X if you're targeting 900+whp. You'll also need a fuel rail if you don't have an upgraded intake manifold, but to be honest, if you come to us with a PI rail and stock mani, we are going to recommend upgrading it anyway. Upgraded intake manifold — RK Autoworks makes the best performing B58 intake manifold we've ever tuned. CSF and Do88 are also solid options. Motiv Reflex+ controller with the Spool PnP harness — seamlessly handles boost control, PI, and aux fuel pump control together with the DME to provide perfect, closed loop integration. This is the standard on every big turbo B58 we build. Upgraded LPFP — Visconti or Precision Raceworks for the low-pressure side. The stock pump can't keep up with the fuel demands of a big turbo on E85. Built transmission — the Spool Stage 2 Transmission DIY Kit is what we run in our builds. The stock ZF 8-speed will slip once you push past 700whp/650wtq. On stock trans, we keep torque low, the goal is to make 700+ horsepower while keeping torque as low as possible to keep that stock trans happy. xHP Stage 3 TCM tune — faster shifts, custom launch control, and torque limits per gear for quarter-mile consistency. Crankcase management — at 30+ PSI boost, you need a proper catch can setup. The BMR Gen 2 B58 Catch Can Kit with PCV block-off pins keeps crankcase pressure in check and protects your PCV system. Intake — on a big turbo setup, you will absolutely feel the restriction of an intake. Run a turbo guard or otherwise approved intake based on your turbo kit. Exhaust — the factory exhaust is great until you start pushing big power. We recommend going with a larger diameter exhaust at this point. Big Turbo Power Numbers On our dyno, big turbo B58TU builds on pump E85 with proper supporting mods: 700–900+ WHP on E85 4.2–5.0 60–130 MPH Where you land depends on the turbo, supporting mods, and how aggressive the tune is. Some specific examples from our shop: Alex Ye's 805whp Supra Turbo: Spool Performance top mount kit, PTE 6466 Gen 2 Trans: Pure Stage 1 built Intake mani: Alibaba billet Boost/PI control: Motiv Reflex+ with Spool Performance Reflex PnP harness Injectors: ID1050X PI LPFP: Visconti Stage 1 Catch can: BMR Tune: EcuTek Motor: Stock Power: 805 whp Time: 4.5s 60–130 (50,000 miles, flawless) This is our shop reference build. 805 whp on a stock motor, 50,000 miles over 5 years on this exact setup, completely flawless. The PTE 6466 with the 0.84 A/R housing is the sweet spot for stock-motor builds that actually get driven, not just dyno queens. BMR catch can keeps the PCV system honest at boost. Ryan's 921whp Supra Turbo: KLM 2-port top mount kit, PTE 6870 Gen 2 Trans: Pure Stage 1 with high-stall converter Intake mani: Aftermarket billet Boost/PI control: Motiv Reflex+ Injectors: ID1050X PI LPFP: PFS drop-in with Walbro 525 secondary pump Cooling: CSF heat exchanger, aux radiators, trans cooler Tune: EcuTek Motor: Stock Power: 921 whp Time: 4.3s 60–130, 9.0s 1/4 mile (slip) Ryan's Supra is what happens when you actually take a build to the track. He can't get the car to dead hook on the street, but when he does, it's quite a bit quicker than a 4.3. Ryan's deep in the street scene, so he doesn't want us putting all his numbers out there. Here's the 9.0 1/4 mile slip he's willing to share, set on a stock motor at 921 whp. The car has gone significantly quicker on both the 60–130 and the quarter mile, but if you don't want to take our word for it, you can always line it up with him. Alejandro's 911whp M240i Turbo: ETS top mount kit, PTE 6870 Gen 2 Trans: Pure Stage 2 with high-stall converter Intake mani: CSF Boost/PI control: Motiv Reflex+ Injectors: ID1750X PI LPFP: Custom split-feed system (PI/DI split, regulator, return line, twin Walbro auxiliary pumps) Tune: MHD+ Motor: Stock Power: 911 whp Time: 4.20s 60–130 Alejandro's M240i is the quickest 60–130 of this group. 4.20 seconds on the street at 911 whp, stock bottom end. xDrive makes the difference here. Being able to put 900+ whp down without spinning the tires is what closes the gap on rolling acceleration that RWD cars give up. The custom split-feed LPFP setup with twin Walbro auxiliaries, plus a PI/DI split with regulator, is the kind of fueling that lets you run ID1750X PIs reliably at full E85 boost without DI starvation. The 60–130 is a more convenient measure of real-world power than a quarter mile. It normalizes across cars because you don't need a perfect 60-foot or a hookup-friendly launch to get a meaningful number. That said, with high-power RWD cars like Ryan's Supra, the 60–130 can still get dragged down by the car spinning through the rolling pull. For really high-power builds, you ideally want to see a 60–130 paired with a 100–150 to filter out the traction noise entirely. For ease of understanding here, we're sticking to quarter mile times and 60–130s, which is what most readers care about anyway. All of these are stock engine builds. The B58TU bottom end can handle 900+ whp reliably. But if you're going past that, or putting the car through sustained track abuse, a built motor is cheap insurance. Get Started Ready to wake up your B58TU? Start with the basics: B58TU Custom Tune — Remote tuning via MHD or EcuTek B58 Stainless Steel Downpipe MHD CAN Flex Fuel Kit — Essential for E40+ builds MHD Super License — For logging and flash tuning Spool Performance Top Mount Turbo Kit — For big turbo builds For maintenance on your B58TU, read our BMW B58 & S58 Maintenance Guide — especially if you're running ethanol. 3,000 mile oil change intervals, no exceptions. Questions? Hit us up. We respond within 24–48 hours, guaranteed.

Guides – by Moeed Qureshi

Car Tuning 101: What is it & How it Works

Car Tuning 101: What It Is, How It Works & How Racebox Helps You Unlock Your Vehicle's Full Potential What Is Car Tuning? Car tuning refers to the process of modifying a vehicle's engine control unit (ECU/DME) parameters to improve performance, throttle response, and drivability. In simple terms, we are overwriting the manufacturer's calibration with our own to optimize for your modifications, fuel quality, and environment. What Does Tuning a Car Do? Tuning changes how your car's engine behaves by adjusting variables like: Air-to-fuel ratio (AFR) Ignition timing Forced Induction boost control Throttle mapping Rev limiters and speed governors Variable valve timing Many more The result? Increased horsepower and torque, smoother shifting, improved throttle sensitivity, and overall a more enjoyable experience with your car. How Does Car Tuning Work? Modern vehicles rely on an ECU or DME—a small computer that controls how the engine runs. By accessing and modifying the ECU's software, we can change how the engine behaves under different conditions. Depending on the platform we use a variety of software suites such as EcuTek, MHD, Bootmod3, or HP Tuners. There are two popular options for tuning: Preloaded/Off-The-Shelf (OTS) Tunes – Maps intended for a broad mod set and to work on a wide array of cars with the ability to adapt to different fuels and environments. These are a one and done flash, no data logging or revisions required. Custom Tunes – Personalized adjustments tailored to your specific vehicle to optimize the tune based on your modifications, fuel, and environment. The custom tune will result in a more powerful tune because we can optimize for your specific use case and vehicle. We also have the ability to adjust the drivability to your exact liking whether that be a more sensitive throttle, torquier powerband, or another specific preference. How to Tune a Car There are a few different ways to tune a car depending on your goals and limitations. We are located in Houston, TX but we have dealers and customers all over the globe. OTS Tunes – Flashed via EcuTek or MHD. Custom Dyno Tune – Your car is tuned on a dyno at our Houston location or one of our partner locations. Remote Custom Tune – We send comprehensive data logging/flashing instructions once tune is purchased. Over 75% of our tuning customers are remote. What Is Dyno Tuning? Dyno tuning is performed on a dynamometer—a machine that measures power output while simulating real-world load conditions. Think of it like a treadmill for your car. It gives tuners a safe, controlled environment to adjust the ECU in real time and adds another layer of data - a power and torque graph. Dyno tunes are the most effective for high horsepower cars where we need to limit horsepower or torque for engine safety, or when learning a new platform. It also allows you to see the exact power output of a car, although dynos can read differently from dyno to dyno. It's best to run a baseline in stock form to see the power gains rather than relying on the exact power figure spit out by the dynamometer. What Is Remote Tuning? Remote tuning is performed on the street by the customer. Once the customer flashes the tune file to the ECU, they will need to complete data logs so the tuner can dial in the tune. These data logs often require wide open throttle (WOT) pulls through various gears, but can also include capturing idling and normal driving depending on what the tuner needs to dial in. Remote tunes are best for dialing in with real world conditions and are more cost effective than a dyno tune. Remote Tuning vs Dyno Tuning Dyno tuning would be advantageous if you do not want to do the data logging on your own or do not have access to a private road for the data logging process. Remote tuning is by far our most popular option as you can data log at your own convenience. We will take about 1-2 business days to review your data log and get back to you. For our tuning process, we do not require you to respond right away but you do have 90 days to finish the tuning process with us from the first date the initial tuning file was sent. OTS files from Racebox do not require data logging. However, if you have a specific concern regarding your car, you are welcome to send in a data log for review by our tuners. Keep in mind, certain tunes such as flex fuel may take several revisions as we start off with pump gasoline then move onto your car's specific target ethanol mixture. I heard Dyno Tuning is better than Remote Tuning This depends on your tuner's experience with your platform. For many vehicle setups, we have our tuning strategies honed to a tee and it's better to dial in fueling, boost, and timing in real world conditions (remote tune) as loads vary greatly between a dyno and an actual road. If we are unfamiliar with a particular turbo, fuel component, etc, we may ask to do a dyno/remote-dyno as we can keep a close eye on critical parameters during the tuning process. Plus, we can make rapid changes and they will be within a reasonable constant environment (dyno load, temps, etc). The other time dyno tuning is beneficial is when power or torque needs to be limited to protect weaker components such as stock engine or a stock transmission. Is Tuning Safe? When done correctly by an experienced tuner, tuning can be very safe. We base all custom calibrations on: Your vehicle's exact modifications Your environment, such as elevation and climate Manufacturer safety limits determined by our extensive knowledge of testing the drivetrain's limits Extensive logging and data review Does Tuning Void Your Warranty? Tuning can void your warranty, especially if the dealership spends the time and effort to determine if the ECU has been altered. However, many tuning platforms offer ways to revert to stock before dealer visits. It's always best to weigh the risk vs. reward and check with your manufacturer or dealer before tuning. In addition, we would advise you to speak with your local dealership to feel them out for what modifications will void warranties. Dealerships do have some discretion and will vary from location to location. Why is my friend's car with the same mods faster? "I got the same modifications as my buddy and he's faster!" Well, let's break it down. Engine Variance: Not every engine is built the same and if your engine has wear and tear, that can also impact engine output even with the same modifications down to the brand. We've put hundreds of cars on our dyno with identical modifications and power can vary as much as 40 + WHP/WTQ. When customers come to us and ask us for a power figure their car will make with a specific mod set, we always give a broad range with this information in mind. Weight: This goes without saying, but less weight will equal a faster car. Removing static weight such as removing interior components (seats, panels, etc.), upgrading components to lightweight items (2 piece rotors, carbon fiber hood/doors, etc.) can shed hundreds of pounds given the extent they go to. DA (Density Altitude): Depending on where you live, this can play a big difference in your times. DA is a reading composed of pressure, temperature, and humidity. We are in Houston, TX which has a sub tropical climate and about 50 ft above sea level. In recent years, we've had extreme summers (100 + F/90% + humidity) and some cooler weather than usual (25 + F) which can impact performance. I have two solid examples where we take one particular car with the same tune, same modifications, and run them in two drastically different temperatures. This is a BMW M4 (G82) with SPOOL hybrid turbos, upgraded fuel system, built transmission and tuned on MHD +. This car is making upper 800'sWHP and has 0 weight reduction. In the summer months, it's running a low 5 second 60-130 MPH in 3K DA. In cooler months with absolutely no change to the car, we see that 60-130 MPH metric dip down to 4.69s. Driver mod: Yes, even automatic cars require driver mod. Sometimes, your friend just knows their car better than you, can handle it better, or has better traction. How Much Does It Cost to Tune a Car? There are three parts to tuning costs: hardware, licensing, and the tune itself. The hardware needed to tune a car is an OBD II module and they typically run about $80 - $250 but can vary depending on the software used. Certain software such as EcuTek require their own specific module to interface with their app. Licensing costs are generally around $600 but vary depending on software used and vehicle tuned. Tuning costs vary based on modifications, vehicle being tuned, and software. OTS tunes are the cheapest offerings and are around $300–$600. Custom remote tunes for simple bolt on modifications are around $500-$700. Modifications like turbos, or flex fuel can bring the cost around $1500-$2000. Dyno tuning generally charges a flat rate or by the hour depending on the dyno shop, the additional cost for dyno tuning is around $300-$600. In total, for a car that has never been tuned before you can expect to spend anywhere from $300 (if you already have the hardware and licensing from a previous tuner) to $2000 including hardware, licensing, and tuning costs. Where Can I Get My Car Tuned in Person? You can visit our Houston location or one of our approved vendors below: Texas Racebox 2121 Brittmoore, Suite 3200, Houston, TX, 77043 New York Driven Motorsports 18 Neil Ct, Oceanside, NY 11572 Florida Temis Motorsport 4270 NW 19th Ave F, Deerfield Beach, FL 33064 1320 Motorsports 13470 SW 128th St, Miami, FL 33186 Maryland Kinetic Autoworks 8224 Veterans Hwy Unit 1, Millersville, MD 21108 SB Automotive, LLC 8041 Queenair Dr #8, Gaithersburg, MD 20879 Why Choose Racebox for Remote Tuning? At Racebox, we've tuned thousands of cars globally across multiple platforms. Our success is driven by our precise, reliable, and powerful tunes paired with our industry leading response time of 1-2 business days. As our customer base and demand continues to grow, we have grown our team to ensure we can always offer the best support and maintain our quick turnaround times. Here are some of the platforms we support: BMW (B58, S58, S55, N63, S63) Infiniti/Nissan (VR30DDTT, VR38DDTT, VQ37VHR, VQ35DE, VQ35VHR, VK56D) We have tuned the most VR30s in the world Toyota Supra A90 & GR 86 Audi & VW (4.0T, DAZA, MQB) Our Remote Tuning Process: How does remote tuning work? Remote tuning allows us to calibrate your vehicle's ECU from anywhere in the world, without requiring an in-person visit. Here's how it works: Purchasing: Select your platform (e.g., Infiniti, BMW, Nissan, etc) from the Racemode drop-down menu on our website and choose the appropriate options. Once purchased we will ship out your hardware and email you instructions to begin the tuning process. Preparation: Ensure all maintenance items (e.g., oil changes, spark plugs, filters, tires,etc..) are up to date. Install any modifications you've communicated to us (e.g., larger intakes, high-pressure fuel pumps, turbocharger) before flashing your car with the base calibration file. Process: You'll receive a base calibration file and detailed flashing/data logging instructions. A fully charged battery or a battery tender is highly recommended during the flashing process. Custom tuning customers will log data for review, while OTS tuning doesn't require data logging. Data logging: Depending on the platform, we will ask you to perform wide open throttle pulls in a specific gear from a certain RPM to redline while data logging. Data logging will need a specific set of parameters selected and will be detailed in our tuning instructions. You will need a safe stretch of road to perform the data logs. If this is not an option for you, you are welcome to find a local dyno and coordinate a remote dyno tuning session with us. Map Selection: When getting a custom tune with us, the base file will have only 2 maps (Daily & Racemode) which are meant to be low/high power maps. After completing the data logging and revision process, we'll finalize the tune and offer additional map options.

Guides – by Hussain Boxwala

Infiniti Q50, Q60, and Nissan Z VR30 Maintenance Guide

Proper maintenance is crucial for the longevity and performance of your Infiniti Q50, Q60, or Nissan Z equipped with the VR30 engine. Regularly monitor and replace fluids, filters, and spark plugs to ensure optimal operation. Adhering to a consistent maintenance schedule not only enhances reliability but also maximizes the benefits of your vehicle's tuning. For detailed guidelines and recommendations, refer to Racebox's comprehensive maintenance guide.