cewrx7r1

iTrader: (2)

Kyle and I have been running numerous DATALOGIT logs tuning his single turbo, and a few on my non-seq stock turbo engine. On both of our logs, We have noticed that the MAPP (pressure row) does not match what should be used for the boost pressure which is also logged. Furthermore; the PFC and commander also show erroneous data that even contradicts the manual.

We all have noticed that the PFC displayed boost pressure is lower than what our boost gauges show and even lower than what the map sensor voltage says it should be. The PFC has another error beside this, it seems to be running the wrong P row.

The other info from the DATALOGIT appears correct and the same as the PFC. These are: air and water temps, PIM voltage, MAPN (rpm column), rpm, IGL, IGT, speed, and knock level to mention a few.

* Now to what we found using my last two runs. *

By the PFC manual for pressure per P row, they are listed in Kg/Ccm^2 absolute, and listed in multiples of 10,000 where 20,000 = 2Kg/CcmS^2. Thus 10,000 is equal to 1 Kg/Ccm^2.
Row 16 shows 16,000, thus 1.6Kg/Ccm^2 absolute, = 8.1 psi relative (boost gauge).
Row 17 shows 18,000, = 10.9 psi relative
Row 18 shows 20,000, = 13.8 psi relative

I was running 12.5 psi on my boost gauge. The DATALOGIC showed boost between 82 and 85. Now the DATALOGIT manual does not state if this is Kg/Ccm^2 or what the boys down under tend to use which is Kilopascal. But look at these conversions.
.82 to .85 Kg/Ccm^2 gives a boost range of 11.6 to 12.1 psi.
82 to 85 KP gives a boost range of 11.9 to 12.3 psi.
Both are close but lower than my boost gauge. Now if you look at the P row and the boost, you would think that the PFC should be using P rows 17 and 18 because my boost is between them. No, this dis not happen!

The DATALOGIT showed consistent P row 16 which is 8.1 psi. I ran some test with the PFC commander at 12.5 psi boost and got these results. Using the map ghost trace, the PFC showed P row 17 which is what we would expect. But when I displayed rpm and IGL, then looked at my maps, the timing being used was from P row 16.

All my further analysis show that the same problem. The PFC is fucked up in two respects. Displayed boost pressure is lower that what the boost is but not too bad, but it uses a P row even lower than what is says it should be using.

This is also included as an attachment.

mmaragos

So the PFC has been programmed (with enough fuel) to take into account that it (and/or the stock manifold pressure sensor) reads low.

Anyone know what happens with an aftermarket MAP sensor? I wonder if the voltage and display is as far off.

I did some dP testing, but I should have checked the stock sensor voltages at the same time.

One thing that is strange, is that the stock sensor/PFC agreed with my digital manometer and GReddy gauge at higher vacuum levels, then was much less than the manometer and GReddy gauge when under boost.

I was thinking that the stock MAP sensor was inaccurate. Considering that it has been in my car for almost ten years and that has not been calibrated in that period, 10% is not unreasonable at all (there is shift over time, etc.).

If I can help out by doing some pressure testing, let me know.

cewrx7r1

iTrader: (2)

Your assumption is wrong. With my original mod maps
from Ray, my AFRs were 12.2 to 11.8.

Also running one lower P row will give your more spark advance than you think you are running by looking only at the maps. Compare your 16 to 17 rows.

Kyle was running 15psi boost and was wondering why he was still in row 17, should have been in 18.

When I tested my map sensor voltages by the Mazda
manual, they were in about center range . The PFC even displayed the voltage correctly. However the PFC displayed boost pressure was about 1.3psi off at 14.2psi boost. This latest test, shos trhat it even goes lower when picking the P row.


Why don't you compare your map trace P row to what is given by looking at rpm and IGL?

mmaragos

My sensor voltages were within spec too, but that still doesn't mean that the sensor is accurate. And we don't know the complete range of the sensor, or exactly what the voltage output is for any corresponding pressure (we only have a broad range).

I haven't quite figured out why the PFC is accurate for higher vacuum. Why would the PFC display higher vacuums correctly, but then not display higher boost correctly?

You have given me some good food for thought. I will make a couple of runs this weekend and trace the cells and study the results.

......

My first statement was meant to be a question. If the PFC thinks it has lower boost than what is actual, then it would have be programmed pretty darn rich to compensate, correct?

machinehead

Well, that is kind of scary. I will check mine as well. So much for my sentiments on the quality of Apexi products...

adax

iTrader: (2)

This is a pretty big deal. What version of PFC and commander do you have Chuck? So you are saying that the PFC pulls timing and fuel data from the wrong P-row based on it's own pressure reading (kg/cm^2) according to the Datalogit row and the commander's indication of timing row (or was that from Datalogit too)? What I'm asking is, is there a chance that the problem is with the Datalogit data stream or can it be shown that it's using the wrong row based only on commander data? I thought we (you) once established that the commander displayed timing was correct according to the map?

I guess I still don't completely understand your post from about 6 months ago (and therefore this one) about the PFC's inacurate calculation of boost based on PIM voltage. It looks like you are using your boost gauge as the gold standard, how do you know it's that accurate? Did you have it T'd into the MAP sensor when you were comparing pressure to output voltage and find that gauge pressure was within voltage spec? Doesn't this introduce an error as you are comparing absolute to gauge pressure? I've plotted PIM voltage against PFC calculated pressure and it's perfectly linear.

My gauge (SPI) reads LOWER than the PFC (160 ft above SL). I've never really worried about the difference since I assumed the PFC was using the appropriate data based on whatever pressure it was reading, but now this changes everything. Why do you bother with all the back and forth PSI conversion?

So how did you effectively wide-band/dyno tune your car since you were tuning the wrong P-row?

I'll look at some logs this weekend to see if I have the same problem.

As always, thanks for the info.

Alan

cewrx7r1

iTrader: (2)

Alan,

PFC 5.07, Commander 3.32.

Reread my post about row 16 and it's absolute/relative pressure value as given by the manual, and what I read with the commander. All facts are there to answer all your questions about the Datalogit and PFC.

Now about wideband tuning. First I used the PIM voltage map for the top three voltages to make across the board changes to get the AFRs in a safe range. Then I made equal column changes for rows 16 through 20. I did this at 12 psi boost then at 14 psi boost basing my belief that 12 psi used 17 and 18 and that 14 psi used 18 and 19. If I added 3% for N10 for 17 and 18 then I did the same for 16 and 19. I have 20 maxed out as a safety buffer. Now it looks like I could nevr even get to 19.


MMARAGOS,

Even if the PFC gives a different boost level than my Autometer boost gauge, there is still the problem of the used P row not matching the boost the PFC listed!

From an old post of mine:
When my AutoMeter show 12 psi, the PFC showed 11.2psi in equivalent Kg/Ccm^2. Thust when I was running 12.5 psi boost, why would the pfc when listing over 11.2 psi boost use row 16 which by the manual is 8.1 psi, row 17 is 10.9 psi.

I can accept the pressure diff but not the wrong row unless someone can explain it!!!!!!!!!!!!!

machinehead

Chuck, I reread your post and I have some questions. Before we jump to conclusions I would like to know if the timing is actually off or if it is just being displayed incorrectly. The commander may say 15 deg or whatever but it may actually be at 10 or 12 or whatever its supposed to be at. Also it would be nice to know if the injector pulse width correlates to the proper row or if it is in the same boat with the timing.
Either way it is a little upsetting.
Lastly, how do you find this stuff?

cewrx7r1

iTrader: (2)

That is the problem. If the datalogit and PFC show the same error, how do you know if it is an error or false reading. It would take a lot of money and lab equipment to verify that.

Take timing. You would need someway to actually measure timing based upon the actual spark produced by the coils and compare it's timing to the e-shaft position. At minimuim, you need a degreed timing wheel attached to the e-shaft, an accurate timing light , an accurate rev counter, and a boost and load controlled brake or eddy engine dyno.

Which of you can afford this?

I am just presenting facts as seen by a few people who
really get into trying to understand their machines.

Maybe that is why the serious racers use other computer systems. Apexi might be only trying to fill the void between them and reprogrammed ecu.

I now bow out of this discussion.

GoodfellaFD3S

iTrader: (213)

Well, this is very unsettling. So should we all turn down the boost at a minimum, and sell our Power FC's as a final resort?

Can we get this thread to Apexi somehow, so that their engineers can tell us what the deal is (hopefully not in engrish)?

This is not good....

mmaragos

Just make sure that your AFR and EGT are OK. If the boost and/or timing is not correct in the PFC, but your AFR and EGT are OK, then you are fine.

adax

iTrader: (2)

Too bad Chuck bowed out, I've got some more data.

I talked a friend into doing some 3rd and 4th gear 3.5k to 7.5K RPM pulls about midnight last nignt. Not hard to talk friends with slower cars into doing this, even at that hour.

I have the latest PFC and commanders (5.08 and 5.0 IIRC). I controll boost using the PFC, set for 0.8 kg/cm^2 and 1.0 kg/cm^2. Intake temps stayed around 50 and water around 87. This should put me squarely in row 17 at 0.80 and row 18 at 1.0 with no interpolation between rows.

Observations:

The commander map trace and datalogit map trace are identical and correct according to indicated and logged boost and RPM. Datalogit P and N values are incorrect at all RPM and boost levels. They always read one cell too low for boost and for RPM. It's easy to observe this by watching the map tracer and the N and P values in real time while logging (or even idling).

At 1.0 boost, logged leading and trailing timing values were 2 degrees retarded from what they should have been based on row 18. Row 19 is pretty close to being 2 degrees retarded from 18 so it's hard to tell if it's reading one row too high or it's being retarded for some other reason. I suspect the latter. Regardless, I did not find that it was reading a lower row as Chuck did. The error seems to be on the safe side for me.

At 0.8 boost, logged leading and trailing timing values were 1 degree retarded from what they shouls have been. this did not match with any nearby cell very well so I suspect timing is being retarded across the board for some other reason.

Air and water temps were not high enough to cause retard according to my datalogit settings.

So, with my car at least, tune based on the map tracer, or remember to add 1 to the N and P logs.

Alan

Tim Benton

iTrader: (1)

So does that mean the DATAlogit is incorrect in the data it pulls for the logged runs, but the map trace for both the commander and datalogit are showing the same cells? It's a little confusing but I think the Apexi unit is indeed one of the top ecu's over here and in Japan. I have 7 volumes of the japanese RX-7 Magazine and in their articles, they are shown using it to tune normal cars, to single turbo, 1.0 to 1.2 boosted cars. And that is in different issues as well. I doubt that the home market would continue to use it if there were so glaring a problem really existed.

One other thing is that some people's cars have different pressure values in the P column than other. In that P might be 18000 for some, then 20000 for other in the same row. As far as being so off in that in Chuck's case, he was making 12.5 psi on the boost gauge, but only in row P16, is that on the map trace from the commander or the datalogit? Also bear in mind that you might be making 12 psi, but since there is room to fall inbetween teh P rows vaules, it then goes to the nearest row to what your making, meaning you need to tune the rows, on top of and underneath for the fuel in that map area.

Tim

spyfish007

Ok there has been talk of this before. We all should know that the data from the PowerFC comes in groups and not all is from the same sample time! I have never seen the P&N number come in correctly once I start logging a couple of groups. Also be sure no other issues were coming into play when items like timing are concerned ... what I mean is that there are several parameters that will affect the final ignition timing ... like too high air temp .. too high coolant temp ... etc, etc.

adax

iTrader: (2)

Tim,

At least in my case, the PFC appears to be using the correct rows/columns. The datalogit N and P values are incorrect. Boost values assigned to rows is entirely user determined (RPM too!) using the datalogit software. I'm considering adding more high boost rows and fewer low to increase the resolution where I spend most of my time. Trying to figure this out using the PSI you read off your gauge just adds another layer of confusion in my opinion. Set the boost (as read off the commander) equal to a row value, then tune that row and you won't have to worry about between row stuff.

Spyfish,
As I mentiond, air and water temps were below that which should cause retard. But I agree, ther are probably other parameters not accessible via datalogit that affect it. Probably day of the week just to fvck with us.

Alan

cewrx7r1

iTrader: (2)

Alan,

Only using the commander:
(1) observe the ghost map trace and read the P row that lights up for a range of rpm.
(2) go and display only rpm and IGl and freeze it for that same rpm range under the same run conditions.
(3) compare your IGL per rpm to the previously displayed area of the ghost trace P row for the rpm area in the map.

ie. your map trace show P17 for 4000 5000 rpm at 13psi.
Make the same run but show IGL and rpm and then freeze the reading when the revs are between 4000 to 5000 rpm. Go to the N columns for 4000-5000 rpm and find the match on the IGL. Is it P17 ot not?

Tim Benton

iTrader: (1)

I'll check mine out like Chuck just stated, once I get the turbos back on. Damn 99 spec turbos and the fact that I tried to do it in my garage. Man 2, freakin' bolts took 2 days for me to get out....from the manifold to the turbo bolts. Plenty of cussing and swearing was heard by the neighbors as they watched in horror as a B-movie but with real blood ws seen streaming out from underneath my car.


Tim

twokrx7

iTrader: (1)

As Chuck mentioned (I am the "Kyle" that Chuck is referring to), our PFC testing indicates that the P row and N rows do not correspond to the map reference values. On thing to note, when the datalogit says its in P#N#, the timing is not spot on but it is close. When it is not spot on it is difficult to discern why.

Fuel is another question mark. I'm trying to pay closer attention to the map cell influence on the AFR. There appears to be a delay but it is not intuitive. If anyone has this figured out please educate us.

I have to agree with what has been said on this thread ... if you must tune your PFC, then don't get hung up on the N and P row discrepancies. You have to accept it and learn how to adjust the fuel and timing to get what you want. Makes it difficult to compare to the other popular ECU maps, very difficult. You cannot simply look at your PFC and say you run x degrees of timing at an rpm and boost. The n cells are used about 200+ rpm later than indicated by the map ref table and the p cells are all over the place. Example p row: P16 should be 16000 boost level, it really is 16800 to 19000 or there abouts. Example n row: did a run to 83xx rpm and only got to N19, N19 should start around 7600 rpm.

Kyle

machinehead

Won't changing the rev limit affect N rows? If you raise the rev limit to 8500 then N20 should be 8500 and on down the line.
I think the discrepancies are in the resolution. 20X20 is really covering allot of ground and it leaves allot of room between cells.

spyfish007

Rev limit is seperate from N20 value ... that needs to be done with the datalogit on one of the settings pages.

es

Alan (or whoever else wants to do some logs & tests)

Do you think the ignition retard that you are seeing could be caused by the "IGN vs TPS" or the "IGN vs RPM" settings? The datalogit manual is quite vague on these settings.

Wargasm

Some comments... this is an OLD thread so these issues may have been brought up elsewhere.

1) Map Numbers
---------------------
The Datalogit log files ALWAYS read 1 lower than the actual P or N being used and this is normal. This is because the Datalogit uses the counting system of 0-19 just like the PFC internals do. Therefore it identifies what humans see as "P1" as "P0" in the log.

So for example, look at what cell you reach your max boost in on a run according to the "map trace". Let's say it's P18. If you now look in your Datalogit logs, the highest P you will see is P17. This is normal because the Datalogit is only logging what the PFC is using... and the PFC is using P17 on its internal counting system of 0-19, which matches up with P18 on the humans-counting-system of 1-20.

2) Absolute pressure values
------------------------------------
I've seen a lot of talk saying that the PFC is "off" by 1-2 PSI on the boost values compared to what an aftermarket gauge shows. First, see my note on P and N above, that could explain some of it. Sencondly, who cares? If the whole deal is off by 2 PSI, just tune the car like that and now it's magically right on the money. Do you really think the car cares if P10 or P11 is the zero point? Just tune the car with a wideband and it will all work out.

IMO, the important thing to look at is not where 0 is, but rather where your max boost is. If you are reaching max boost at P15 and your boost gauge shows 8 PSI and you think that should be 7 PSI by your calculations... who cares? Just tune the car.. that's not hurting you. On the other hand, if you were tuning for 15 PSI (which should be within the range of the stock MAP sensor), but your datalogs were going up to P20 and staying parked there... you might actually be going outside the range of the program and be at risk of damaging something.

Anyhow, that's my two cents

Regards,
Brian

es

The PIM offset can correct the boost value that the Power FC Commander reads, & also which cells the PFC uses...

I have set mine to 3600, & this corrected the boost reading when the engine was off at sea level. Different altitudes will affect the ambient pressure.

tfhuth

Brian,

That's one of the best replies I have ever seen!

Best Regards, Tom

Boostn7

OK guys...here's my findings based on my own logs....

Like Wargasm mentioned, P values will go from "0" for P01 to "18" for P19 and same follows for the N value. So many will think that the timing is always advanced since it's logging a lower "P" row.....*but* when comparing the map reference numbers for both boost and rpm they match nicely......so we need to simply disregard the idea of logging the P and N position.

When logging boost...PFC will always show lower boost numbers vs what is actual......anywhere from 1 to 2 psi.

You could change the PIM scale offset (lower #'s) and you will narrow that gap.
I personally left it alone....since it allows me to run 17-17.5 psi on gauge before MAP sensor maxes out at 1.15 kg/cm2 (16.33psi).....as you could see barely over 1psi difference.

Under Map reference I've changed the complete boost (P) values to add more pressure rows.....10000 now is on P08 and 23000 for P20........adds more resolution for higher boost #s'.

Hope this helps

JD