jetlude

iTrader: (9)

Well, I am going to hit the dyno for the first time and guess who the tuner will be? Yes, its yours truly. Never done it on a dyno before, just on the street. Been learnig to tune on my own cars, for the past 5 years . First started on my turbo charged Honda H22 and now for the past 3 years on the rotary. I have had the rotary reliably street tuned to just about 24psi on an EMS Stinger. I have lately switched to a Haltech E8 and now tuning for boost. I am confident that I can successfuly and safely extract 500whp from the Large street ported 13B. There is a mobile dynojet available and I will be booking some time very soon.

Setup:
FD running large street port
Garrett GT4094R 106 a/r hotside
Dual Bosch 044 fuel pumps
1000cc/1600cc fuel injectors
HKS twin power
Homemade preturbo water injection 500cc
Tial 44mm w/gate
3" exhaust pipe
FMIC
dual egt setup (not logged)
wideband (cabin display and logged)


Just wanted to know if you guys are willing to share the methods of tuning you have seen or done on the dyno like loading. Also any tips you have will be welcome. The car will be already street tuned to about 12psi before hitting the dyno and raising the boost.

Please remember I have 0 experience with dynos, however, the operator will be there.

Thanks

arghx

iTrader: (3)

Well, does this dyno have loading capability then?

-- There are two buttons on a dynojet. Red button is for the brake, green button is to begin logging. If he will let you, play around with the buttons a little bit and get comfortable with them. Normally you hit the red button as you let off the gas for a pull, but you don't have to do that immediately. You can coast down for a second. Make sure you hit the red button again to release the brake. If you don't, the car will stall. Don't hit the actual brake pedal in the car though.

-- There's no law that says you have to immediately put it in 4th gear and slam the gas like you see in dyno videos. If you want to, you can be easy on it and try to lay in it in 2nd at first or something like that. See how the car behaves on street-like conditions. For example, you can adjust your throttle pump/accelerator enrichment. The big advantage is that you aren't going to wreck it or get a ticket.

-- Driving on a dyno feels weird. It just feels--weird, like you're driving on nothing, especially when you first let the clutch out. Driving with the loading on is a little bit different--you can set a % load or you can hold a particular rpm. When you hold the rpm, your EGT's will rise pretty quickly and the engine bay can get hot. It's good for cruise tuning or getting random lean spots adjusted, but don't do the hold rpm thing for a long periods of time--maybe a minute at most and then let things cool off.

-- It's highly likely that the first numbers you see you won't like. Don't worry about it. Chasing the internet dynosheet rainbow can tempt you to do stupid things, and it can make you think there's something wrong with your car even if everything is running fine. Every dyno is different, even dynos of the same brand.

-- Don't be surprised if the AFR's end up being a little bit different when you get back on the street. A nice clean map with good interpolation between cells will help mitigate this.

-- Watch your air temp compensation. Sitting there with a big fan and the hood popped is not a good representation of street conditions. You will have to keep tweaking this when you get back on the street.

-- Reading boost and AFR on the dyno is helpful but they are not going to read the same as your on-board instruments. Boost will probably read a little lower and AFR will probably read a leaner, sometimes a full point leaner. When in doubt I usually trust my on-board readings. Try to log in your engine management system at the same time you log on the dynojet (hitting the green button)

-- Download the WinPep7 software off the dynojet website and become familiar with it before you go in there. Bring a flash drive and ask the dyno operator to give you the run files when you leave. Play around with setting different scales and chart configurations so that you know what you're looking at. Graphs can be very deceptive. Note the differences in the shape of the curves as you change the scaling on each axis. Note the difference in horsepower readings when different correction factors are selected.

-- with the water injection notice the before-and-after effects on your power output. You will probably drop a few hp when you switch it on but make no other changes.

-- if possible, try to dyno without a crowd around. It's just a distraction, and honestly it can subconsciously put pressure on you to impress people. Think of it as a testing session only.

-- be wary of dyno operator tuning advice, unless you are with a rotary tuner or someone who has significant familiarity with your engine management system. If he freaks out because his wideband is reading lean, you may want to ignore him if your on-board instruments are good. Try to be confident but cautious.

-- Wear ear protection! Earmuff style is better than earplugs because you can take them on and off quickly. There's no point in not wearing ear protection. It accomplishes nothing. Assume the dyno operator does not have ear protection available for you. Bring your own.

-- try to get a good tach signal. If the tach signal cuts out your torque reading will be off. Hook the inductive clamp to the front trailing plug wire and have it set up like a 4 cylinder.


I have made some dynojet run files are available here: https://docs.google.com/leaf?id=0Bzo...kMzVk&hl=en_US . rename to a .zip file and extract. there are some runs from 2nd gen Rx-7 turbo, an FD, and an STi. The 2nd gen is having tach signal problems so note how the rpm signal drops off and the torque reading is all wrong.

Indian

iTrader: (1)

awesome info arghx

jetlude

iTrader: (9)

^^^^^^Damn, very informative.

Few questions,

Is the brake only used to slow the engine after a pull, or can it be used to load up the engine?

indio84

iTrader: (7)

good luck, let us know how she does

arghx

iTrader: (3)

It depends what model of Dynojet you are using. A Dynojet 224x, which most shops have, does not have loading. Check this out http://www.dynojet.com/automotive_dy...o/default.aspx as it explains the differences in the models available




Loading is a more expensive feature, and Dynojets are significantly more expensive than competing Dynos so they're not likely to have it. Usually the higher end shops have loading, shops that deal with nicer cars. Your typical Billy Bob's Performance will just have a basic 224x with one large roller. The 224xLC has load control. The 424x has two rollers, one for each axle (for AWD), and the 424xLC^2 has four individual rollers for AWD. Most of my experience is with the 224x and the 424xLC^2.

Getting back to the topic of tach signal pickups, here is an example of a tach signal dropping out:



The run just stops at about 6500rpm. But if you change the axis to vehicle speed (speed of the roller) you will see a crazy torque spike:

j9fd3s

iTrader: (3)

good advise ray!

i find all the noise and the fumes basically turns my brain into mush, so i actually like to make a checklist of things to do BEFORE you go to the dyno.

it really helps to just have a roadmap to follow so you don't have to think.

part B is that you probably want to start at the low boost and work your way up. the idea being that low boost = more margin for error. this also builds a bigger broader map, so in case something happens, and you end up in a weird map cell, the ECU is still tuned for it

jetlude

iTrader: (9)

Thanks for the advice guys. Low boost is already tuned on the street, will be concerned with tuning high boost on the dyno.

jetlude

iTrader: (9)

What about the ignition timing proceedure. I was told to use moderate to low timing until the AFR's are established good. Afterwards, I could begin doing doing pulls raising the timing a degree or two at a time until no more noticable power gains.Then take out about two degrees.

Is this the correct proceedure.

arghx

iTrader: (3)

That's the basic idea, yes. As for the whole part about "no more noticeable power gains"

Since you do not have a working/effective knock control spark retard system (almost no modded Rx-7 does), you need to focus on how much horsepower you are getting each time you make a timing adjustment. It's not like tuning a naturally aspirated engine, where the risk of knock is more limited. Are you willing to add two more degrees to get only 2 more horsepower? That's up to you. It depends on your tolerance for risking your engine and it depends on the application.

Another thing to do is look for dips in the torque curve. This can be tricky because it depends how noisy the horsepower curve is. Even with maximum smoothing ("5" on a Dynojet) it can still be noisy. If you notice a dip in the torque curve you can attempt to add a degree or two to that rpm and load range. Also, adding timing at high rpm sometimes will widen the power/torque band a little bit.

jetlude

iTrader: (9)

Generally about how much how much HP gain should I look for with about 1 or 2 degrees advance?

arghx

iTrader: (3)

I don't have a simple answer for that. It depends on too many factors. I will say this though--the "closer" you are to MBT, Minimum Spark Advance for Best Torque, the less power you will incrementally gain--diminishing returns. But without knock control I personally believe it is risky to get too close to MBT on pump fuel because you have insufficient safety margin. The same idea goes for AFR. I have posted this graph before:



It is the relationship between the excess air factor, known as lambda, and torque output on a single cylinder research lab engine. Along the X axis Lambda = 1 is 14.7:1 AFR on E0 (100% gasoline). A factor less than one is richer, and a factor greater than one is leaner. Typically on a modified boosted engine you are going to be between .72 and .80 lambda or about ~10.5 and ~11.76 AFR as indicated on a typical wideband display. While the actual chart isn't for a rotary engine, the basic relationship (shape of the curve) between AFR and engine output remains.

I am posting that chart to illustrate the concept of diminishing returns. Look right along the X axis and note how the slope (rate of change) for the torque output changes as the mixture becomes leaner (we are assuming negligible knock). The safety margin aspect also comes into play with AFR. On a modern engine you have a self-learning system called fuel trims. These fuel trims will use feedback from the factory wideband sensor to calculate compensation in order to keep the AFR more consistent as conditions change. An Rx-7 doesn't have fuel trims, and it very rarely has effective knock control. There's nothing to bail you out, so if longevity is a goal you need to seriously consider building in a wide safety margin. A wide safety margin means sacrificing engine output, at least to some extent.


Finalizing the wide open throttle AFR and spark is the part of performance tuning that is often closer to art than science. It is one of the areas where a good professional performance tuner should be earning his pay. Wouldn't it be cool to have a nice chart that tells us the relationship between spark timing and horsepower output or knock activity? If you were a Mazda engineer you could hook up all sorts sweet equipment to your engine and calculate all sorts of cool stuff to figure out the spark timing. In their labs they have completely different knock detection systems, gas analyzers to tell how the flame is progressing inside the chamber, and thermocouples to read temperature along the rotor wall and the housing wall.

But you are making do with a chassis dyno in a performance shop. You just have to do what you can and put one foot in front of the other. If this were say, a Subaru STi, you have the option of advancing timing until the knock sensor starts acting up. But that's often not a reliable method when tuning a modified turbo rotary. It's a big reason why I was emphasizing the psychological aspect here. In most cases you shouldn't try to make a spectacle out of an initial dyno testing session. It subtly messes with your head and could influence your decision making. Trying to put on a show for people (intentionally or not), trying to have internet bragging rights--there's a place for that, and everyone has the right do what they want with their own engine. But for most sessions it should not be standard procedure.

smg944

iTrader: (34)

are you tuning on pump with the water or on race fuel?

the only time you can raise timing on a rotary until it stops making power is on race gas as the chance for knock is very low. on pump fuel you want to stick with the basic numbers adding timing doesnt add a ton of power if the motor flows well. plus as arghx said its not worth the risk for 2 whp or so.

arghx

iTrader: (3)

Yes, but I was also pointing out that it does depend what your starting map looks like. Just like the benefit from leaning out the engine depends on how rich it was in the first place.

jetlude

iTrader: (9)

Will be tuning on pump fuel.

jetlude

iTrader: (9)

Hey Arghx, is it possible to get the tach signal from the Rear Trailing plug? The front trailing will be a bitch to get to.

arghx

iTrader: (3)

I always use the front but I don't foresee any problems with the rear, unless somebody else wants to chime in

scotty305

iTrader: (2)

Make sure the vehicle is in race-ready condition... a dyno can put more strain on the engine, cooling system, and drivetrain than the racetrack. Although the building should have fans, most dyno setups won't supply as much cooling air as driving the car at speed on a real racetrack. If you're using a roller-type dyno, make sure the wheels and tires are in good enough shape to spend a lot of time above 100MPH.

Be sure the cooling system is bled, the oil is fresh and at the correct level, the fuel filter is fresh, fuel system doesn't leak, fuel in the tank isn't too old. If you've got time, a compression test is probably not a bad idea.


It's not a bad idea to bring spare spark plugs, extra oil, extra coolant, laptop charger, flash drive (to get files from the dyno computer). If you have a local gearhead friend you trust enough to help work on your car or troubleshoot in case things go wrong, it's nice to have another set of eyes around to watch things during runs. If you've got a fuel pressure gauge installed in the engine bay, give him a camera and have him record the fuel pressure durig runs.


Keep an eye on air temps and coolant temps, and try to make sure the engine hasn't gotten too warm or too cold between runs. Personally I like to save lots of copies of the ECU calibration, and write detailed notes for each dyno run in the calibration notes and dyno run notes (for instance: "run 11: removed 2 deg ign from 3000-6000 RPM" or " run 15: 0.6mm plug gap, no audible misfires"). If possible, it's very nice to have boost and AFR logged on the dyno. Try to decrease the 'smoothing' factor on the displays, lots of inconsistency or jagged lines in the power curve can mean the engine is not happy.


Ask to visit the dyno you plan to use and watch someone else tune a car, but don't be surprised if the shop or tuner says no. Try to stay out of the way, but observe what they do and learn what you can. Many 'tuning' sessions get interrupted by vehicle problems and become troubleshooting sessions, don't panic if something unexpected happens. Finding a problem on the dyno is better than finding a problem on the racetrack: the car is stationary, there isn't any traffic to avoid, and it's easier to stop and re-schedule another dyno session than wait for the next race. Think logically, try to fix the problem if possible but don't risk your safety or the vehicle's safety just because you want to finish the session. As others have said, don't chase internet numbers... there are all sorts of ways to manipulate the numbers shown on that chart. The goal is to get the car into good enough shape to perform well at the racetrack, not to pretend the dyno IS the racetrack.

jetlude

iTrader: (9)

Well, today was the day. Got on the roller for the first time in my life. I felt quite confident since I knew what to do and an expect, thanks to you guys for the tips given in this thread.
The dyno operator did all the controls for the dyno, all Ivdid was tell him when I was ready for a pull. All pulls were done in 3rd gear.

Now for the results.

First pull i made 325 rwhp with 5 hundred and something torque. I was like WTF, only 325? I expected my first pull to be much higher. And the torque, i knew something was wrong there. The operator had two inductive clamps hooked up to my plug leads. One on L1 and the other on T2. I figured he knew what he was doing so i said nothing at first. I rememberd Arhx posting to use T2 rpm.I suggested to the operator to remove the L1 clamp, and he did. This solved the incorrect torque problem, however, power remained the same.

My initial power was 325rwhp, much lower than I expected. After three hours of tuning and 22 dyno pulls, the car made 423rwhp on about 18 psi.

Had to end my session prematurely due to my battery dying and causing leaner AFR's.

I first was a bit dissapointed with the output since I was not able to reach closer to my goal, 500rwhp. I had to remind myself tihat this was my first shot at dyno tuning, and i am much closer since I was able to successfully add 100rwhp.

Thanks again for the tips and info.

jaxmacintosh

i am looking into buy a rx7 that has dyno tune and on 1300cc fuel injectors. i am unfamiliar with tuning in general, but i know that the car comes with an apexi commander pro. can i reset the dyno tune with it and just go back to stock apexi tune? and will 1300cc injectors be okay to run with it?

fendamonky

iTrader: (6)

It's been 11 years since this thread was started or posted in, not exactly the most beneficial thread to post in.

If you have no idea what you're doing then DO NOT make any changes to the car!!! I'll say it again for clarity, DON'T change the tune yourself, you're more likely to blow the engine than improve anything.

arghx

iTrader: (3)

If you reset the tune, the car may not run, or it may run poorly, or it may poorly and then blow up.

Monika

iTrader: (1)

This is a very bad idea if its running ok now.