Dyno Tuning: This is why you do it (56k beware)

[BLOZ UP]

Attached is the difference between my car tuned on Megasquirt, versus the stock ECU. All other items are the same, save for boost level and injector size.

[bradyzq]

So you're saying fueling, timing (assuming your MS is controlling spark as well), and boost are not the same.

 

Not a valid comparison, IMO.

 

If you're trying to show that the stock ECU is a weak link, fine, but if you really want to show what tuning can do, show us a graph with the boost and injectors the same as the original ECU.

[bradyzq]

From personal experience:

 

Dynapaks tend to be the most consistent dyno. Best for hardcore tuning (50+ runs)

 

* They are a bitch to set up and that will cost you, unless you can do it yourself and have friends with a dynapak lol (especially on 4wd cars)

 

* They tend to read higher than most other dynos (dynos aren't great for measuring power, they are for tuning!)

 

* I'd rather use a roller dyno most of the time just cause it's easier!

 

I have a Dynapack, and yes, I love its consistency. And I agree, on 4wd cars, it can take awhile to set them up, especially if you have to jack each corner separately.

 

But if you're looking for actually tuning, not just a baseline, that's no problem because it just adds a few minutes to the bill, and even if it is just for a baseline, that usually has a fixed (not hourly) price.

 

IME, my dyno reads slightly higher torque early on in the pulls than a Dynojet might, so your torque peak numbers might be higher if they are at low RPMs. But HP numbers are not any higher than anything else out there. Of course this all depends on how hard you load the car during the pulls. Generally, I use 2 seconds/1000RPM.

[Zmanco]

Of course this all depends on how hard you load the car during the pulls. Generally, I use 2 seconds/1000RPM.

I just looked at a datalog of a 4th gear pull I did from 5k to 6k and it took 3.6 seconds. It was at full boost already below 5k. I know that I was starting just after the torque peak so it's probably less than that from 4k to 5k.

 

I need to finish my tuning on a dyno this winter so wonder how this setting affects the measurements. Assuming that we only use measurements at full boost, my intuition says that the HP and TQ readings should be the same regardless. But I have near zero experience with dynos so am I missing something?

 

Thanks.

[BLOZ UP]

Yes it is DOS, and there is a good reason for this though it doesn't have the slick interface that it otherwise could. I have been told (but never confirmed) that the sampling rate when logging in a DOS operation is faster and more consistant than what it would be in a windows environment. Makes some sort of freakish sense to me, as you can still buy high end laboratory datalogging software that runs in DOS for some reason.

 

False.

 

It's because the company is sticking to what works, rather than having to program for a whole new OS. The benefits of a GUI in a dyno situation are worth the investment, in my opinion. It has no effect on sampling rate.

[BLOZ UP]

So you're saying fueling, timing (assuming your MS is controlling spark as well), and boost are not the same.

 

Not a valid comparison, IMO.

 

If you're trying to show that the stock ECU is a weak link, fine, but if you really want to show what tuning can do, show us a graph with the boost and injectors the same as the original ECU.

 

True, but I mainly posted it for comparison of the fuel graph. The stock ECU just bottoms out the sensor, and results in a choppy power graph.

 

"Tuning" this out, albeit at higher boost and different timing, result in a much smoother graph. The smoothing factor is "only" at 3.

 

That's an old car and old dyno. I do happen to have a different Z31 now, but it's still running the stock turbo and injectors with megasquirt, so whenever I get around to putting it on the dyno I can save all the runs. Assuming they have the ability to give me the runs on disk.

[Drax240z]

False.

 

It's because the company is sticking to what works, rather than having to program for a whole new OS. The benefits of a GUI in a dyno situation are worth the investment, in my opinion. It has no effect on sampling rate.

 

It was definitely the story from the salesman.

 

Actually thinking back on this situation now, I'll bet it was to cut costs on the hardware side, as well as keeping things the same/simple. I've used plenty of windows based data logging equipment since I wrote this, and they can sample rather quickly.

[Ben]

I've recently used a workshop with a 'Mainline' eddy-current dyno, and the software was fantastic. The vendor has an SDK and will attempt to integrate any tuning software you have into their application. So while the operator is adjusting dyno parameters, they can be tuning your car with the same keyboard - and looking at both apps on the screen at the same time. The system also has variables that you can enter in such as stall rpm to assist with that low-rpm tuning on auto's that tends to be a bitch.

 

One of the big differences between inertia and eddy current dynos is that you have to 'drive through' a load point on the inertia unit. On an eddy current, you can set it to hold the dyno at a constant speed (providing you can shed the heat generated within the dyno), and then vary the load via the throttle. It makes fully mapping a particular rpm point easy (and can really test the mechanicals of the car).

 

I've seen a few Dynapaks here in Aus, but we mainly use eddy-current types - the AWD ones are particularly handy!

[SidWell]

Do different RPM's demand a different AFR? In other words, at 3,000 RPM an engine produces best HP at 12.8:1 but at 6,000 RPM the max HP is found at 13.2:1? These are just numbers that I threw out there as an example to explain the question. Not to be taken as fact.

[burninator]

Most people shoot for a pretty flat AFR curve, but OEMs and people who are tuning for mileage may tune for different AFRs in different areas. I believe common wisdom is that NA engines produce best numbers at 13.2 and FI engines a bit richer (something like 11.5 or 12 depending on who you ask). You may be able to make more power (especially on an FI engine) by running leaner, but it's not safe for your engine to run too much leaner. And in my experience a few points doesn't make a huge difference in power.

[SidWell]

Burniator... I agree with what you said. In fact, at freeway cruising speeds I have leaned out my AFR to over 15:1. I need less than 25hp to move the car at 70 so there is little load on the engine. My question revolved around wide open throddle conditions: Does the engine produce best horsepower at a given AFR at one RPM and at a different RPM does it require a different AFR?

[hoov100]

One thing I would like to know, is if a dyno that uses a large singular roller corrects the power, as when a large singular mass is spinning at a fixed point, it will take less and less force to move it, as the inertia builds. (like how pushing a car thats already moving takes less effort, then pushing a car from a dead stop)

[Drax240z]

One thing I would like to know, is if a dyno that uses a large singular roller corrects the power, as when a large singular mass is spinning at a fixed point, it will take less and less force to move it, as the inertia builds. (like how pushing a car thats already moving takes less effort, then pushing a car from a dead stop)

 

Is there a question in there somewhere?

[X64v]

Hoov, that's incorrect. Just as F = MA, , where Tau is torque, I is the mass moment of inertia, and alpha is angular acceleration.

 

In short, the force it takes to accelerate the drum does not depend on the drum's speed.

[hoov100]

wouldn't a dyno with a single large roller, read out more power and show a more vertical power band, because a large rotating drum requires less force to rotate the faster it goes?

[X64v]

because a large rotating drum requires less force to rotate the faster it goes?

 

It does not require less force. It takes the same amount of force to add, say, 50rpm to the drum in 1 second, whether the drum was already spinning 2rpm or 200,000rpm.

Edited March 8, 2010 by X64v

[Leon]

... My question revolved around wide open throttle conditions: Does the engine produce best horsepower at a given AFR at one RPM and at a different RPM does it require a different AFR?

 

I know this is a bit old but I wanted to add my two cents...

 

The answer is yes (kind of), and it's because of the engine's volumetric efficiency (VE). VE is the measurement of the engine's ability to completely fill the volume of each cylinder. Having 50% volumetric efficiency would mean that at an arbitrary operating point, the cylinder is being filled with air (and fuel) at half of its capacity.

 

The volumetric efficiency curve will essentially resemble the torque curve that you see in dyno plots. What this comes down to is, around peak torque, your VE will be maximized meaning that the cylinders are being filled as much as they can be with air.

 

Thus, at peak torque, you will likely tend to have your leanest AFR given that the engine is fueled linearly with respect to MAP (manifold air pressure). Of course, with standalone/tunable ECUs you can tune your AFR to whatever you like and if you want to run it rich to be on the safe side, you're probably not losing much power (although you are using more fuel and dumping more crap out the tailpipe).

 

So I guess this only half-answers your question, as I'm not sure what would be the best AFR at each RPM point given the ability to tune the AFR at every point. It would be something to test on the dyno, by loading it up and playing with the ECU fueling curve to see what gives you best torque. So maybe I didn't really answer your question at all, but hopefully it will give some insight as to what is involved in getting an answer.

[zredbaron]

i'd be very interested to know how the AFR's relationship with torque changes over the rpm band.

 

if i had an ECU, i'd do dyno runs with flat AFRs at 12.0, 12.5, 13.0 and 13.5. then i'd superimpose the torque graphs over each other and see where they differ. then i'd pick the best graph (for each rpm range, in case ideal AFR actually varies) and do two more pulls with +/- .25 on the AFR and superimpose them over the best torque graph from the previous round.

 

i think that experiment would produce the graph we want to see. perhaps i'll do that one day when i go to ITBs.

 

i imagine that as VE varies, so does the limiting factor in complete combustion (ie fuel vs air). i doubt the ideal AFR graph per rpm would vary by much. no more than 0.5 i would guess, but hey, i'm just making this up as i go.

 

the question is, would this target AFR graph be universal for all 4-stroke engines, or all NA/turbo, or all L6, or completely unique for each application?

 

speaking of turbo, i bet CR changes target AFR, too. damn variables! hence we all have to go to a dyno to see what our engines really want...

Edited May 20, 2010 by zredbaron

[Guest ttoude]

Thank you for the article. We have 3 dynos in town now I have an idea of what to look for as well as what directions to follow with parts.

[duragg]

Can I resurrect an old Pinned thread with a new question?

 

 

How does a dyno graph change with a different rear end ratio?

 

Same car, same dyno, same time, just magically switch final drive.

Is the HP and TQ plot identical?

 

 

Thx