You make lots of points, all of which are valid, and many of which are ignored by most of the marketplace.

So, out of curiosity, what kind of run to run repeatability can you get with all the bells and wistles you have at your disposal? Much less than 1% of peak HP? And what kind of measurement standard deviations do you see? I am very curious because I come from a measurement background, but a very different one than you.

Anyway, thanks for yet another dose of reality...

Matt

 

Huh? No such thing as 'the truth' on a dyno. Too many variables, and especially given the intercooler/airflow, the only 'true' info is what happens on the road.
A tuner that uses inflated numbers will be discovered and outed, usually in an expeditious manner. But hey, people still buy '20 HP modules' and the like - it's a crazy world out there.

 

Perhaps the two dynos relate with a common percentage of difference?

The dyno pulls should be repeatable on the same type dyno under similar conditions. Perhaps a repeatable percentage of difference can be had with experience on the different dynos.

I firmly believe a dyno soley for max power is a waste of time so I try not to get lost in the numbers. Dyno for max performance through active tuning is different. Whatever the parts installed require an "ideal" tune to work in partnership. What counts most to me is real driving performance. Show me the part consistantly out-performs in real road conditions, and you'll have an interested buyer. A dyno queen is just that, pretty.

I believe there is no good answer to your dilema. In your position, the higher number is usually going to be the better seller. Ethics don't sell well (glass have full or half empty?). But like I previously stated, show me it consistantly out performs in the real world.

In Peter's case, the Nederlands(?) DNRT (link) series seems quite competitive. I don't speak spanish (ron burgandy), but his team of four has three placed in the top six in points. To me, that appears consistant.

 

Hey Scott, I have an idea for valid dyno testing of MINI performance parts. It's someone in the MINI community with access to testing facilities, and...

Thank you for the great read. It's nice to get facts from someone with direct professional knowledge.

 

The dyno was made as a diagnosis tool & has become a bench mark tool. Unfortunately, as many of the posts on this thread point out, the data is not the same from dyno to dyno, even if the brand is the same. There is 1 piece of data that can be constant ( within reason ) if all the test conditions are similar and you have the baseline data, percentage can be used.

90STX, another thanks, nice read & to the point.

 

a single final dyno number is useless alone. in my business, we use the term 'provenience' - the definition = "A source or origin". Basically....the reference point with which to judge meaning....something not in situ is virtually worthless.

The change, or delta, is what's relevant...not the actual number itself. Any company that throws out a topline number and not the actual dyno proven delta is suspect in my opinion. I find it a bit odd in the MINI world that dyno plots are scarce...in the Subaru world, they are much more common. It's understood that some chassis dyno's are optimistic and others are heartbreakers - someone bragging about numbers without disclosing baseline reading, the dyno type, manufacturer and even the operator is shunned. As it should be.

 

The delta is relevant, however, as I have written in many threads: unless you control for all variables, e.g., wheel slippage, IAT, fan cfm, the data is invalid.

 

Dinan has moved from dynos that use the tires to one that bolts up to the flanges...... No slippage but a change in rotational mass. But Steve was saying it's easier to get repeatable numbers....

FWIW....

Matt

 

i rather see 216. I can sleep better at night

 

of course! a run at 50F ambient isn't comparable to one at 90F ambient. apples to apples, not apples to groundhogs

 

Dinan uses a DynoPack (I have access to them as well). Since you bolt up the hubs it controls wheel slippage.

 

"Why you might ask...Because if you are true to your self (read ethical) and to your customers, you would always give them the actual non fictional numbers not part+part+part= x amount.

A lot of numbers are thrown around here on NAM, and it might been the number that the Dyno chart showed at the moment but how true are they?

If you would own a company like M7, which one would you use?

202 whp or 216whp that is the question..."


I say post all results, indicate as much about baseline, measuring conditions, before and after mods, etc. and let the buyer beware.

but to look at the posted example:

where is the baseline?
what were the dyno conditions?

I don't see fiction as the issue, more a confusion over applicability.
on the other hand, this may be a landmark: M7 actually posting dyno numbers, however disguised they may be, and they represent an M7 car reaching 200whp (I think it is not drive-train corrected )

 

A dyno is a great tool to be able to use. If today my car is 180 at the wheels, great. However tomorow, the weather has changed, 172 at the wheels. Both days the data is correct & valid, but for those 2 days only.

If I were selling a product that promissed big HP gains, I could easily wait for the best air day & cool the snots out of everything, use 0 weight oil, get the good pull & hey look at this!
Almost always under similar conditions, the percentage 10%, 15%, whatever, increase remains about the same.

IE: 180hp today - 10% XYZ part added = 198hp. Next day 170hp -10% XYZ part added =187hp. On the other hand, if I said, 20hp gain with XYZ part, well, you get the picture.

After having many engines on different dynos, I have seen this more often than not. A percentage is more repeatable even with, sometimes, awfull conditions.

 

I prefer seeing deltas on the same dyno vs simple peak readings on various dynos. Inertia dynos seem to have the greatest variance from unit to unit. Load cell dynos, be they engine or chassis (bolt to hub) tend to produce the most consistent numbers and are not affected by wheel/tire inertia and traction issues. They also have the benefit of slowly ramping the engine powerband or holding it at certain levels for more 'scrutiny'. Every dyno installation requires correlating numerous data inputs, and each of these measurements (temp, baro, etc) needs to be done with instruments that may not always be fully calibrated, affecting the results.

 

very true . i can say that aside from the look i bought an alta header and cat back because the ad said between x and x hp and it wasn't an out ragious claim . i've come to realize the big claims are usually bull so a company that keeps it real will get my business . (like when webb says he sells throttle bodys but doesn't see a real gain to them) you gotta respect that.

 

Getting back to the original question.

Both dyno runs are what they are. One optimistic? One not so much? The dynos don't produce anything but numbers & they can be interpeted as you like. Who's to say if I take the same car & dyno it, result 218whp, or 200whp? If you only use the numbers as a tool to find performance, they can be a great R&D solution, for benchmarking braging rights that too. I think the Dinan full boat car claims 208hp as apposed to the JCW 214hp. I've driven both & there is a difference, a suprise too.
At this stage of M7 development, as a company, IMHO you should do what you're doing, marketing the inovation, engineering & design you already have. Why muddy it up?

 

what's with that torque curve ?

 

Wheel slipage?

What factor if any foes tire pressure play in dyno runs?

 

20 hp modules?

 

Plasma Booster!

 

Fast moving thread!

Dr. O,
To reply to post #26, yes we can hold to less than 1% variation in multiple runs on the same engine and hardware. Our ability to control various parameters varies on the measurement. Holding coolant, air, fuel and oil temperatures within a 5degF range is fairly easy.

When we run torque curves, we typically start at the top of the desired rpm range and stabalize conditions. Just as an example:

Coolant out (going from the engine to the radiator) temp at 202F (slightly above the allowed tolerance on thermostats to all be fully open)
Fuel temp at 110F
Charge Air Cooler outlet temp (air temp going into the intake manifold) at 115F
Inlet restriction at X inches of water
Exhaust restriction at Y inches of mercury


We then typically hold those conditions while a weighed amount of fuel is consumed. Torque is averaged for that duration. Repeat two more times, then drop 100rpm and do it all over.

Temperature and pressure measurements are largely limited by the precision of the thermocouples or transducers being used. Selection depends on the required degree of precision, as does cost. Dynos and transducers are regularly re-calibrated.

We can vary any of the above conditions and many more depending on the test. The test cell controls also let us make the engine "jump through hoops" during cyclic tests and transient maneuvers. We can simulate any running conditions that an engine might see in road or off-road use, or make those conditions far more abusive to accelerate wear or test for failure points. Of course we can also edit any calibration parameter such as fueling and timing, plus "lock" those components such as VG turbos and EGR valves which are controlled by the ECM.

Motoring dynos can even drive an engine to simulate the effect of compression braking going down a long grade. Tilt stands allow testing at various angles up to 45deg in any direction.

My group rarely uses it, but we have test cells that will very accurately duplicate driving a vehicle across the country on specific routes (such as Baker Grade in CA, or through Indianapolis city traffic). The engine experiences exactly what it does on a cross-country trip (shift frequency, highway speeds, loads, even different driver styles).

Yes, we'll occasionally find the limits of parts or even engines (that gets messy though). Our job is to find the potential failures so our customers don't once the engine is in production. It's a pretty cool job for an engineer and "car guy".

Dynos are wonderful tools. You just have to understand what you are looking at or what kind of variation to expect due to uncontrolled (or uncontrollable) conditions. Reporting increases of a couple hp, or peak hp values to .5 hp, is essentially meaningless. If a dyno operator really works at controlling all the conditions that they can, such as monitoring coolant temperature, humidty and IAT, and can make several runs to check on repeatability, then consitantly seeing a 5hp increase should indicate that the engine is putting out more power than the baseline. Not that any driver will really be able to notice 5hp in cars as heavy as ours.

FYI, the torque numbers that I'm used to dealing with are a full order of magnitude greater than what the MCS will put out, even modded pretty heavily.

Scott
90SM

 

Dang, another novel. Hope no one hurts themselves when they fall asleep wading through it and plant their face in the keyboard.

Scott
90SM

 

I brought this type of thing up awhile back

https://www.northamericanmotoring.co...t=Dyno+Testing

(Sorry... the link feature won't work tonight for me)

I agree there are just too many variable to know for sure what a cars mods have really done, other than showing an increase of some kind... at least outside of some margin of error. When I see what vendors claim, I know enough to take it with a grain of salt.... just like folks will take dyno results from your average member that's sharing what his new exhaust system did.

I guess all I can really offer here is that I think doing the dyno runs before/after mod install on the same day in the same conditions is about as close as you can try and get... naturally on the same dyno.

 

No worries it's a great read

It's fantastic to finally have a Subject Matter Expert (SME) post on this -- Johan

 

another point: posting the graphs brings another level of confidence, rather than just quoting numbers. for example, that torque bump would be discounted upon seeing it in a graph. If you rev the engine to 7500, you will get a higher hp number from revs alone (for these cars) compared to revving ghe same car to only 6800. If you were really trying to show accurate data, you would have several dyno runs and quote an average, rather than the highest readings.
Finally, the shape of the torque curve says a lot more about the performance than peak numbers alone. Of course, I am assuming the car owner wants useful comparative data, not just bragging rights (fat chance, eh?).