Hymee,

Thanks for the info, although I don't understand what you mean by PID and GUI. Remember, I'm a engine engineer with a mechanical engineering degree! And your profile confirms you're an IT person. But don't worry, I'll do a search on Google before you come back to this forum.

I'm very interested in your approach and project. I'd be happy to help with some info I have on Ford ECUs as it seems to be of the same type as the Mazda. It might be of some help to you in order to better understand which parameter does what.
Note for example that the MAFS readings are most probably not used as a direct input in the fuel or spark table but are used to calculate a 'LOAD' that is used as an input (vs. RPM). Note also that these MAFS measurement are not always used by the ECU to evaluate engine LOAD (e.g. in conditons of backflow although newer generations of MAFS can actually deal with backflow).
By the way, which parameter did you record to get the MAF?

Some other interesting investigations using your CAN and soft could tell us :
- where the engine works in close and open loop fuelling
- when the cat temp protection overfuelling kicks in (you might have to wait for next summer to do so).
- how the electronic throttle body respond to driver's demand, especially at low engine speeds (I have this feeling it's pretty aggressive, giving very good performance feel but perhaps low FE - if it goes WOT fuel enrichment).

I have in numerous posts on this forum already mentionned the problem of potential incompability between aftermarket intake systems and MAFS flow characterisation. The characteristic curve of the MAFS is actually measured with the complete program intent intake system upward of the MAFS !! Any change in the hardware during development leads to a re-cal (or a least a check) of this characterisation!

 

IKN,

Your points on the MAF are most valid. It may shock you to know that it was previously (probly still is) common practice to tune engines by altering the MAF sensor to trick the ECU into a "more desireable" result.

In fact the Canzoomer controller intercepts the MAF signal and sends a "modified" signal to the computer. Tuning involves finding out how much to trick it at each RPM/Load point.

The PID is a Parameter ID in ODBII speak. There are a whole bunch of them in the SAE docs I purchased. Yes - I have monitored the "Fuel System Status" to watch when it goes in and out of closed/open loop! It is easy to verify on the RX-8 as it has a WIDEBAND O2 sensor that I can also monitor. You can see when it is in closed loop, as the system does very well at maintaining a stochiometric 14.7:1 AFR (Lambda = 1.0). As soon as it goes open loop, you can see the AFR skipping around the place.

The fuel/ignition controller on my car does not intercept and modify the MAF signal. It interceps the signal to the fuel injectors, and tuning involves altering the duty cycle by a % at each RPM/Load (Throttle position) point in a table. Im not exactly sure how it alters the timing, but from studying how it connects to the wiring harness, we believe it intercepts the crank angle sensor input, and adjusts it by X degrees, via a similar table. This modified crank angle sensor signal is then fed on to the ECU. Ingenious in a way. I would rather crack the PCM cal and change it's internal operating parameters.

One drawback is the fuel controller currently only changes one of the three sets of injectors.

And the cat temp can be monitored via ODBII, so I could try and work out then the Cat Protection kicks in.

Anyway - GUI is a Graphical User Interface.

Cheers,
Hymee.

 

If it is a wide band sensor, it's a type I don't know of because it only has 4 wires (2 for the heater, 2 for the signal). Does anyone has more info about this sensor. All the wide band sensors I've ever used have 6 or 7 wires (NTK UEGO and Bosch LSU). Strange!

There is something Canzoomer should explain to me then. Considering the engine is set to LBT (optimal fuelling for torque) at WOT across the RPM range, what can you gain on a stock engine by modifying the fuelling?
Same thing with the Spark advance, it is usually set to MBT-L (Max Brake Torque - detonation Limited) :
- so the engine already run at MBT (or very close) and you won't gain anything by advancing spark.
- or the engine is detonation limited and by advancing spark at these operating points, you reduce the safety margin built into the system. Note moreover that the knock control has authority and will most probably bring the spark advance back to original level anyway.

Now, if you have modified some hardware (intake and/or exhaust systems) or, better, turbo/supercharged your engine, I can understand the potential of such a fuel/ignition controller.

 

i can field this.

right from the get go, when we had Speed Racer firing dual 2 ft. flames out of the stock exhaust, 15mi/g fuel consumption, exhaust sooty beyond belief, those not vigilant flooding their cars far too easily, and the highly publicised reduced power rating, we knew something was wrong. what exactly was wrong has been evident, also from the beginning, that the A/F ratios were completely f*cked straight from the factory. the emissions thing, y'know i'm sure.

the thing is, at the original ~250hp rating, you're right that there wouldn't be much to tune in, but since it's now really ~220hp at best, there is a whole lot of room to pick up power simply by leaning it out (back to stoich).
another thing is that all motor wankel engines don't really knock very easily, so you can go rather crazy with the ignition advance (not to mention it's measured in e-shaft rotation degrees, meaning that 3 degrees of advance is really only 1 degree of advance on the rotor) and maybe gain a little (i dunno about Maurice, but Jeff had reported that there wasn't much to get there anyways).

as an engineer, you should know better than to even mention the word "intake", in its usual context, especially this car :P.
the exhaust, i'll give ya, it could definitely use some work. but there is lots to gain in the motor itself, even before you go changing any of that stuff.

 

1) Problem of fuelling is probably due to tha 'catalyst temperature protection strategy ' (do a search on these word for explanation) and was probably modified (improved) with later coming PCM updates (called reflash on this forum). This strategy only comes in when czlled for (high catalyst temperatures), and probably no more than 3 days a year in Vancouver...
Sooty exhaust : what else would you expect when you burn oil. My wife's diesel also has sooty exhauxt...

2) I still strongly believe (see my posts on this subject) that 'power issue' is just an issue of how ADVERTISED power is determined on different markets (US vs EU vs Japan). I posted the engineering figures as presented by Mazda engineers at the 2004 SAE meeting on this forum (->search) : power is 170kW (231 hp).

3) Stoichiometric fuelling is definitely not optimal for power. And or fuel economy neither. Stoich is just optimal for emission reduction in the catalyst (which must oxydise and reduce simultaneously).

4) Modern ignition systems are very accurate, much more accurate than just 1°CA.

5) If the rotary is detonation free, why does it feature a knock sensor? I guess it's because there is a knock control strategy integrated in the PCM. That has a cost and is therefore integrated on the Renesis for a reason. This reason is that it's detonation limited at certain operating conditions. QED.

6) I don't get the point in you last paragraph. What should I know as an engineer precisely? All I was telling Hymee is that such an auxiliary controller has little interest without significant modifying of the engine and I mentionned intake and exhaust as they seem to be the focus of most mods on this forum.

 

Hymee
I decided to see what my RX-8 Auto looks like compared to a standard transmission. As you can see the MAF readings seem to be quite different. At 5000 RPM I was getting 150gm/sec which is above the theoretical line on your chart. At higher RPM's the automatic flows less air and there is a huge drop in airflow at 5600 RPM, I'm not sure why that happens. I was using the CANScan OBD-II tool with the data logger set to 0.100sec per sample.

Sea Level, 80deg. F, 90% humidity.

C.

 

er, what?? friend, i've been around since well before the first RX-8 frame was welded together. i'm not a member of the SAE, but i know a thing or two. what's with all the down-talk?? you asked, i was just replying in a polite way.

1) duh, that's what i was saying. this is a problem. i know that the exhaust temps are high because of the low amount of metallic mass the exhaust gasses pass by on their way out of the motor, and that they can't just move the cat further back in the exhaust system because it has to cycle through that "quick light-off" feature (the reason for the double walled header) negating the presence of a pre-cat to keep warm up emissions low. so there's the conundrum. the best solution would be to tune the motor back up to where it should be (2000 degree exhaust gasses and all) and get a high-flow catalyst that will either hold up a little better in the heat, or that you choose to afford to replace.

as for the sooty exhaust tips, soot as i'm quite sure you know is hydrocarbonous waste. in this case, where it used to take mere hours of driving to turn the tips from waxed and perfect to sooty black, it was from the high level of hydrocarbons combusting in the exhaust. not even RX-7's with carbs soot up exhaust tips that badly when they're tuned well. deisel will of course soot up exhaust tips, as on start-up (accelerating from a stop) they belch black clouds of sooty black hydrocarbonous waste. and seriously, i know that if you're burning a lot of oil, it burns blue, not black.

2) i strongly believe you are very wrong. it doesn't matter what either of us thinks yet, 'cause it'll take years to find out what's really happened. but the detuning did happen, just ask Maurice (it's the reason the Canzoomer exists).

3) when did i say it was?? "back to stoich", as in toward. it was at 11:1. stoich is 14.7:1, as i'm sure you know. that's a big difference, and a whole long way to tune it. and i do know that for the best power you tune it up on barely the rich side of things, to ensure you consume as much of the oxygen as possible, wasting as little energy on the excess fuel as you can. i also know that these motors can run quite happily on the leaner side of things (Mazda was attempting to get injector spray patterns to concentrate lower levels of fuel in the Centre Deep Depression bathtub thing), which for obvious reasons is good for fuel economy.

4) yeah, that's ok, it's not news to me. i was talking about the relationship of degrees of rotation between the e-shaft and the rotor (which spins at 1/3 the speed), meaning that *30 of ignition advance is only 10* of advance in reality.

5) i didn't say "free of detonation", i said "easily". the knock sensor is there to prevent knock; suppose someone gets a terrible tank of gas, orrrrr the plug wires got screwed with and are getting old, and you've got voltage peaking in all four when they shouldn't be... i dunno. but again, i didn't say they don't knock, nor that they didnt' need a knock sensor and ignition retard safety routine. i did say that you can advance ignition a lot when you're counting degrees on the e-shaft, although it doesn't appear there was much to gain there anyway.

6) the last paragraph was a tongue-in-cheek joke (hence the smilie), as every intake mod is pretty much a bullshit mod. popular doesn't mean good, as is evident with so very many things in our advertising saturated consumer culture . ricers don't understand how to make cars fast, which is why they're slow.
btw, 20-30 rear wheel horsepower (say, 45-60 flywheel horsepower) for $600 dollars is a whole lot better than the most expensive intake or exhaust system is ever going to get you "without significantly modifying the motor".

 

That is a good graph CygnusA.

You will find that the dip is when the S-DAIS is changing over the port runners. on the 6PI like mine, there are in fact two dips. And they match the two dips in the power curve (which I didn't plot). The dips on the plot I posted are not noticible, as I only plotted MAF at 500 RPM increments, so I have lost some of the resolution. See any of my posted Chassis Dyno plots, or acellerometer plots and the dips are apparent. It makes it quite apparent that the amount of power being made is related to the amount of air (intake charge) taken by the motor.

Cheers,
Hymee.

 

I'm not a SAE member (nor your friend as a matter of fact). I don't see any down- talk in my post. Can you? The fact you politely replied does not force me to agree with all you said. You don't need to be around from the very first day on this forum to be aware of the history of the car. Indeed, all the info is archived on this site.
If you look below, we in fact agree on most things.

1) We basically agree on the problem. And there is one otherwise the subsequent PCM updates wouldn't have fixed it. And they apparently did for most. I just doubt the Canzoomer and other auxiliary ECUs can correct the fuelling in a reliable or safe way, especially since the updates came in. I can explain in details if needed (not in this thread) or ask the question straight to Maurice in his vendor thread.
I'll get my first update (Std Power 5MT) next Wednesday. I'll see if the sooty exhaust remains.
And anyway, I prefer to sacrifice a few hp for better durability, especially for costly parts like a catalyst.

2) Compared to the initial figure of 250 hp (marketing ****-up?), I might be wrong indeed. But I'm convinced this engine delivers 231 hp on the engine dyno and that it's advertised at 240 hp in US for the reasons mentionned in my other post.

3) Apparently "back to stoich" confused me.

4) We agree, great.

5) So basically, you cannot 'go crazy' with the ignition, can you?

6) Thanks for the clarification. Sorry but I don't understand that particular smilie. We agree on the intake BS. Especially when you read all the problems some people have had after installing them .

 

Time to jump in.

The issue with the catalytic converters is really not as big of a deal as Mazda would have you believe. Yes it needs to heat up fast but with modern technology, there is no reason why they couldn't program the ecu to run a different timing/fuel mixture to get a hotter exhaust which would heat up the cat quickly. After the car senses operating temperature the map changes. It would be simple. The old 13B rotaries had hotter exhaust gasses and their cats could be made to last. The last nonturbo rotaries had 3 cats in series. 2 small cats for early lightoff and then a large main one. The small ones were very close to the engine yet these could last to 100,000 miles. The cat would not be an issue if Mazda would tune the damn thing properly. From the maps I've seen, they aren't even close to how I'd do it. Yes, we all agree that the cat lightoff situation is a current issue. It shouldn't be though.

The power issue goes back to the whole improperly tuned thing. The RX-8 doesn't make 247 hp and probably not 238 hp in it's current state of tune. The engine is easily capable of it though as has been varified by Paul Yaw and others. This isn't an engine issue. It is a tuning issue.

Stoich is not optimal for power but it is the best for emissions. Stoich is 14.7:1 on pump gas. Nitromethane is 2.5:1 !!!!! Rotaries make best power near 13:1.

Spark advance is measured in degrees of eccentric shaft rotation not in degrees of rotor rotation. Wakeech's 1/3 less for rotor rotation is correct though.

The knock sensor is obviously there to prevent detonation. Detonation isn't an issue on a naturally aspirated rotary. This assumes of course that it is tuned properly. The way I tune n/a rotaries, I can't get detonation. It is difficult. Why does Mazda have a problem with it? Once again it is lousy tuning. This is the running theme here and the source of all of our problems. I've seen their maps. They suck. They are overly complex and make little to no sense. Tuning should be kept simple. Mazda isn't doing it this way.

Basically there is only one issue that is really the cause of the problems that have been going back and forth here. Tuning.

 

Yea for Hymee. I hate the term "volumetric efficency". I don't even know what it is.
Do you remember when your professor took a big pyrex bell and hooked a pump through a manometer? You know a u shaped tube with a given amount of mercury in it? When he pumped the air out the pressure would go down. And the opposite.
Did the volume change under the bell???

Of course not, it can't. The bell is the same size so the volume is the same. Now is there more orr less air in there?? yes there is, but this is mass.

So no matter how well or poorly your engine pumps air it alwways pumps the same volume.

Therefore, in order to fuel the engine properly the ECM must know the mass. If that were not the case this would be a lot easier.

I havn't read this whole tread and I;m going to do so right now. I just wanted to put that in since it is one of my pet peeves.

To answer one other question, of course mass flows over 100% is possable even on slightly modified engines. Without knowing for sure I wouldn't be supprised to see the Honda hi po engines dong this stock.
My ski boat engine is fairly high performence modified but not a race engine and it showed 109% on the dyno. And that was carb restricted. I could have gone to a bigger carb and lost some intake vacuume. More properly, increased the intake pressure, But the set up as it was gave better control when pulling up skiers.

Back to the thread. It is very telling that the 8 is pulling down very close to 100% stock. This means Mazda has done it's job.

Ok I'm going back and read this thing and will speak up later.

RAP

 

I thought we had reached an agreement on this thread. I stopped after Hymee and I agreed on that stuff we were talking about.

 

yeah, that it's measured in e-shaft degrees means that seeing an insane number like 15* or 24* of advance can be "normal" for a rotary as it's only 5 or 8 degrees (respectively, of course) of ignition advance.. which is what i meant by "advance like crazy".

and thanks for the backup. sheesh.

 

Bis repetita placent!

What do you technically mean by 'their maps suck'? Can you be more specific (on both fuel and ignition) as you're one of the happy few who saw them?

It's true that the OEM strategies are very complex and would not make sense to an engine tuner whose only focus is on performances. But it does for an OEM engine management system calibrator who has (like engine harware designers) to find the best compromise between the following (sometimes antagonist) aspects :
- durability
- performance
- driveability
- fuel economy
- exhaust emission legislations
- customer requirements
- drive-by noise
- etc.
and this for a global market (= different requirments) that includes Rovianemi (Finland) and Lake Havasu (Arizona).


Hymee, I can't wait for you to post more interesting graphs with other parameters, like lambda, spark advance, mid-cat temp...

I can't find a CAN OBD tool so far in Europe. I keep searching. Which one did you buy?

 

Sorry to bring the dead back to life, but I found the thread somwhat interesting given the subject. And I've also been bugging some people about this for the past month or so.

I logged data in a similar manner, using Harris R&D CANScan tool with a PDA. I logged RPM & MAF input through out a 3rd Gear WOT run. Having said that, I only plotted Volumetric Efficiency. And I assumed that the Renesis would have similar characteristics to a 2.6 4 stroke engine.



Any comments are welcome!

 

Nice plot,

At a quick glance, and doing some correlation between the different way we plotted our axis, it seems very similar to what I measured.

Your assumption on the theoretical displacement of the 13B is correct. It displaces 2.6 litres of air every 2 revolutions - the same as a 2.6 litre 4-stroke piston engine.

Cheers,
Hymee.

 

i know i am jumping into this thread, but i just wanted to put my .02$ on volumetric efficancy. as to moving more air than the theoretical volumetric efficiency, it is quite possible, even without forced induction or ram air. its called harmonics. look at the integra gsr, with long and short intake runners, the rx-8 with all the different runners, lexus and their high output motors, all with variable length intakes.

reason behind the different lenght runners, as the air moves down the intake tract, it resonates, and the resonation is directly related to engine rpm. at certain points in the rpm, the resonation (sound waves) actually benefit the intake charge by pushing a little more air into the motor than it would have otherwise, so there are different length runners that are tuned to different frequencies. thats why the aem v2 intakes make more power than a standard cai.

again, just my 2 cents

 

i'd like to see someone do this with a revi... for some reason i think it would show to be pretty worthless under 8ooo rpms, but you never know -shrug-

 

[*]Hello Hymee, I am thinking about to add a centrifugal supercharger to push the near-ambient intake pressure to 1.3 - 1.5Bar. To do that I am about to pair the flow diagram of the chagrer to a natually aspirated RX-8, but did not find any excel sheets about RPM-air mass flow- except what I found here as you mentioned it, but I am quite new here and did not find the attachment or list if there is any. May I ask you to send it to me? Thanks, Janos [/list]

 

It's an 18-year-old thread and the member was last online 11 years ago. You won't get a reply.

You can start your own thread or if not able to, there is a newbie thread where you can ask for help.