Wargasm

IMPORTANT IMPORTANT

READ THE SECOND POST FIRST

IMPORTANT IMPORTANT
----------------------------------------------------------------------------------------

Hello,

I wasn't sure if anyone had made a sheet like this...

This spreadsheet should help you determine the settings for Scale and Offset for any typical linear 5V MAP sensor.

The idea is that you measure your new MAP using:
1) a 12 Volt or 5 Volt source (for powering the MAP - I think most use 12V?)
2) a pressure/vacuum tester (like a MityVac or similar)
3) a multimeter (for reading output).

You put the info you gather into this sheet and then play with Scale and Offset until it matches your experimental data.

Don't bother downloading this unless you have all the items listed above in 1, 2, and 3.

I think this sheet works, but I haven't actually calibrated a new MAP this way since I still run the stock one.....

Comments?

Brian

Wargasm

IMPORTANT:

I'd like some folks who have actually done a calibration to check my work! I'm seeing a lot of posts like "use 41800 and 0 for a GM 3 Bar"... which either means my spreadsheet is wrong, or the GM sensor is reading 0-1 volt (rather than the stock 0.5-5V), or something...

Please don't blindly rely on my as-of-yet unproven sheet!

books

iTrader: (8)

curious...
did you use this formula or any relationship derived from it?

P-Row Value = (Sensor Voltage x Scale x 9828.8 / 65536) - Offset

a version of this formula was on the Datalogit site

when I was intalling my 3 bar, I did some comparisons between the stock sensor and the 3 bar using this formula

Wargasm

books,

Nope, didn't do anything with that formula. I wish I had a GM 3 bar to test with...

books

iTrader: (8)

Here is a log which was recorded with pressue applied from an outside source.

The sensor used was the GM 3 bar with a scale of 4400 and an offset of 357.

Just open it in Excel and sort either by boost or pimv.

I believe I have a similar log of the stock sensor too.

books

iTrader: (8)

Here's the stock sensor logged. It may be messed up.

Wargasm

I have compared my spreadsheet to various references.

books' data
internet data for the GM 3 bar on other car forums
etc.

Seems like the correct ballpark settings on a GM 3 Bar is around 6500 Scale and 0 Offset. Your precise values may vary depending on your exact sensor, but even changing the Offset by +/- 250 at 6500 Scale results in almost no change. Even scale values from 6250-6750 only cause variations of a few PSI at most.

Anyhow, I think my spreadsheet works. It should work for any linear sensor as long as you test it out and put the data in there.

cewrx7r1

iTrader: (2)

Brian,

I just am not understanding your instructions!

The stock map sensor values as comes on the PFC, are 35086/3947 not 5363/3947.

After entering my ABS/PIM values into the light blue matrix (which were taken with my above stock values), I played with the orange scale and offset to get my yellow line directly on top of the light blue line. This gave values of 5400/3900.
After entering this into my PFC, all PIM ABS values for vacuum and boost come out "0". My light blue line was almost exactly like your experimental line.

Then I adjusted the yellow line to be linear from lower left corner to upper left corner. This gave 6775/100 which was better at 0 psi but at 14PSI was 4420 as compared to the old 19357.

Please explain.

Chuck

twokrx7

iTrader: (1)

Arbitrary IMO unless you want your displayed value on the PFC to be dead on accurate. I used the popular 0 and whatever, then calibrated each of my p-rows to correspond to a known boost psi, set my initial timing according to boost, then tuned the map from top to bottom.

Wargasm

I looked at about 15 *.dat files that I could find, and they all had Scale=5263 and Offset=3947 except one map which did have a 35086/3947 set up (called timimgandstevekinj.dat). So, I'm not sure about which numbers are the true defaults from Apexi/Mazda, but if majority is an indicator of this, then it seems like the true base settings are 5263/3947.

To enter data into the light blue area you need a power source for your MAP (most run on 12V), a multimeter that can read from 0-5V (most run in this range), and a vacuum/pressure pump like a MityVac or similar.

Did you do this with your sensor? If so, what numbers did you end up with in the light blue area?

After you do that... then as a final step, you move down to the yellow/orange area and see if you can make the yellow line lay on top of the light blue one.

Regards,
Brian

cewrx7r1

iTrader: (2)

(1) I looked at about 15 *.dat files that I could find, and they all had Scale=5263 and Offset=3947 except one map which did have a 35086/3947 set up.
CEW1: All the maps I ever seen and have on my PC; have my values for normal.

(2) To enter data into the light blue area you need a power source for your MAP (most run on 12V), a multimeter that can read from 0-5V (most run in this range), and a vacuum/pressure pump like a MityVac or similar.
CEW2: I use what the DL shows for PIM voltage. When I verified this against my digital volt meter on the sensor, both were the same.

(3) Did you do this with your sensor? If so, what numbers did you end up with in the light blue area?
CEW3: Yes with the stock sensor, I did vacuum from 20"hg to 0 in steps of 5, and boost 0 to 18psi in steps of 2. Different than yours but close, here are two examples: 20"hg relative = 4.8 ABS at 1.36V and 0 psi rel = 14.7 ABS at 2.62v.

(4) After you do that... then as a final step, you move down to the yellow/orange area and see if you can make the yellow line lay on top of the light blue one.
CEW4; did that and came up with 5400/3900. Changed my PFC with it and retested a few vacuum and pressure. All PIM ABS then came out 0??????

I'll try again tommorrow and also try your values.

Chuck



Regards,
Brian[/QUOTE]

Wargasm

Chuck,

I just discovered something a bit disturbing. Follow me here, it gets confusing.

When I wrote this spreadsheet a few days ago, I used the 2.1.0.17 version of the "FC-Edit Universal" program.

When I made my little "how-to-tune" your car two years ago, I used the classic FD-Only FC-Edit program -- version 1.101.

Based on your comments, I went back and re-evaluated my work... looking for a way to resolve the apparent conflict you and I had regarding which are the stock values.

The answer? We're both right - sort of.

I looked at my screenshots using version 1.101 of fc-edit, and YOU WERE RIGHT. My own maps even show a "30-something-thousand" value (hard to see) for the stock scale.

Then I looked at the SAME *.dat file using the new universal 2.1.0.17 version... guess what? The same *.dat file now says I have a 5263 scale.

Sigh...

I'll sort this out.

By the way, the only map I found with the 30-thousand value was not viewed using FC Edit. I viewed that map using FC Tune... and I guess it agrees with the older FC Edit versions.

Wargasm

Here's what's up.

The newer version of FC Edit (2.1.0.17) that is "universal" correctly puts the Scale in terms of ( (kg/square meter) / Volt ). For a stock MAP this works out to around 5263 for Scale.

The older version of FC Edit (the FD3S-only one) simply identifies Scale in arbitrary units. Are they...
kg/sq.meter/Volt?
Kips/sq.inch/Volt?
Torr/sq.inch/Volt?
Foot-long Hotdogs/Pound of Solar Masses?

Who the hell knows.


The spreadsheet works if:

1) You use the new version of FC Edit (the universal one)

OR

2) You use the old version of FC Edit, put in your data, calculate your Scale and THEN MULTIPLY BY 6.66 to put it back into your old version of FC Edit.

My personal theory is that there was a bug in the older FC Edit versions, or possibly, it was not reverse-engineered quite right, resulting in the odd units for Scale. It didn't hurt anything because everyone used these odd "whatever they are" units.

Anyway, I'm rambling now, but from now on, the answer to "what scale should I use with X sensor?" will have to be given the reply, "That depends on what version of FC Edit you are using".

Chuck, try out my theory... let me know if it makes more sense now.

Thanks! (and sorry if I confused people who are reading along - feel free to PM me or reply here)

** EDIT : By the way, I think the newer version of FC Edit is actually more buggy, so I like the old one **

Wargasm

A little screenshot showing the differences....

https://www.rx7club.com/attachment.p...d=214482&stc=1

cewrx7r1

iTrader: (2)

You got it dead on correct now Brian !!!!!!!!!!

From my previous results of 5400/3900 and using the 6.66 multipler;
I got 35964/3900. Doing a 0psi test, the boost showed slightly vacuum.
I then changed it to 35964/3850 which gave a very close 0 at 0 boost.

Reran my test of 20"Hg vacuum down to 0psi in 5"Hg steps, and 0 - 18 psi boost in 2 psi steps.

Using the PFC formular of RELATIVE BOOST = ((PIM ABS / 10000) * 14.22) - 14.7;
I checked all the PIM ABS values.

20"Hg to 0 psi, this set was a little low: 20"Hg was 1.2% low; at 0 psi it is less than 0 but was so close it is not even worth measuring.

From 0 to 16PSI, no more than +_.1 psi off. At 18 psi where the map sensor is maxed, it was .5psi low.
* * * * * * * * * * * *
SPECIAL NOTE:

The boost value given by the DL in the format of NN is really
equal to ((NN / 100) * 14.5) which is BAR even though the PFC uses Kg/Ccm^2.
I verified this at a few points. EXAMBLE: at 14 psi my Blitz which reads BAR showed .97 bar which calculates to 14.065 psi, the DL showed boost as 98!!!!

Wargasm

OK I spent some time and wrote up a new version of my spreadsheet.

In this version, you first pick what version of FC-Edit you use... then go from there.

It's big, it's colorful, it's Excel-from-hell....

Chuck, can you test this out for me?

Brian

cewrx7r1

iTrader: (2)

Brian,

I'll check it out, may not be until Monday since I am suppose to be tuning a car tomorrow.

PM me your e-mail and I'll send you a gift!

Chuck

Wargasm

Heh heh, no need for a gift, but let me know if the sheet looks good.

If so, I'll see if we can get a mod to delete my first attachment since it's not as good.

moconnor

iTrader: (14)

Where does this number come from? The only pressure conversion I can find that is close to this number is:

1 atm = 14.69595psi = 6.66597kg/sq.in

or

1 bar = 14.7 psi = 6.6678kg/sq.in

But who measures pressure in kg/sq.in?

moconnor

iTrader: (14)

Interestingly, I can now calculate the PIM fairly closely from PIMV, scale and offset using this magic number:

PIM = (scale/6.66) * PIMV - offset

The calculated PIM numbers are very close the the Datalogit PIM value. In fact, it looks like the Datalogit PIM value in a row lags PIMV because for identical PIMVs I have seen slightly different Datalogit PIM values. My guess is that the Datalogit is not freezing the data when it does all its calculations for a row.

But what are the units of scale (and offset)? And PIM?

moconnor

iTrader: (14)

Boost is then (PIM - 10332)/100 perhaps, where 10332 is standard atmospheric pressure in kg/sq.m. Or is there some conversion number?

I am also wondering if boost is actually (PIM - atmospheric-pressure-when-car-is-switched-on)/100.

This would be a better approximation of gauge pressure. Obviously, if pressure changes during a run the number would not be fully accurate. However, given the PFC's lack of a reference for atmospheric pressure, it has no way to account for it varying.

Again there is a lag between Datalogit displayed boost and PIMV for a row. I need to do some more tests with a MityVac.

Wargasm

I have no clue what the 6.66 means. I'm sure it's some crazy combo of conversions. I prefer the kg/sq meter/volt that the newer FC Edit uses

Matt Hey

I adjusted my PFC scale and offset some time ago so that the PFC boost reading matches my VDO mechanical boost guage. My scale=5325 with offset=3327 under the new FC-Edit and scale=35500 with offset=3377 under the old FC-Edit. Elevation is ~1000 feet here and my 0 boost P-row was set to 10332. With these changes, be warned that the Boost settings under Boost Control are NOT working correctly for me and could cause an overboost situation. For example, .80 boost was equal to about 12 psi and lowering it to .70 boost made it go to 14 psi. I was going to do more testing but winter happened. Every thing else about changing the scale and offset seems to work great except for the boost control settings. Maybe it's just my 7? Here's my PFC map and more info...

http://home.earthlink.net/~hprx7/pfc.html

I'd like to hear any comments.

cewrx7r1

iTrader: (2)

That's what you get for using the very simple PFC boost control. Better to have a seperate good EBC.

I definitely love my Blitz DSBC R with it's four boost settings, boost scramble, and over boost protection. It keeps boost constant no matter if it is summer or winter.

Matt Hey

Yea, that's always an option but if I overboost and blow my engine I'll go single and get an EBC for sure .

P.S. typo in my last post: offset=3377 should be for both old and new FC-Edit