installed 4 bar map sensor--one question

[big-phil]

on my ve and spark table the kpa used to read:

 

250

221

199

180

161

140

120

101

89

70

50

20

 

now it reads:

 

408

360

327

295

263

230

198

165

149

117

84

36

 

now I don't plan on using all 4 bar of the sensor, maybe 30lbs at most, so can I adjust the ve and spark table to get more resolution up to say 30psi then clump the rest up to the max of 44psi? kinda like this, this is an example I'll have to figure out what 30psi is in kpa...

 

408

295

230

198

170

150

120

101

90

70

50

20

 

you get the idea, my thinking is I should keep the drive ability more spread out??? But now that I think about it in boost would be more important so maybe the bottom part could get more clumped and the boost up to 30psi more spread out??

 

anyone got the 4 bar sensor??

[TimZ]

30psi of boost is about 308kPa (assuming ambient pressure at 14.7psi=1atm).

 

It is a good idea to compress your scale a bit to give better control over the range that you will be using.

 

Just keep in mind that no matter what you do you will be limited by the range of the sensor and the range of the A/D converter. For instance if your A/D is 8 bits, then the smallest kPa increment that you can work with is 1LSB = 408/255 = 1.6 kPa. You should keep your breakpoints far enough apart that there will be at least a few LSBs in each range.

 

I tend to leave myself more breakpoints in the "driveability" part of the map, but only by a breakpoint or two. If you think about it, there are a wide range of conditions where you will be driving normally, but when you are in boost, you will almost always follow the same boost curve. For instance if your boost controller is set to 15 psi, then you will almost never be running at 13 psi except when you are passing through it on your way to 15psi.

 

Also, does the MS allow you to set the rpm breakpoints? If it does, then you might consider setting up the rpm breakpoints so they are evenly spaced on a logarithmic scale instead of linearly. You don't see many people thinking of it that way, but RPM response is more akin to frequency response than anything else - RPM is a measure of frequency after all. This will end up looking like it's bunching up the breakpoints at low rpm and spreading them out at high rpm.

[big-phil]

I'm trying to find what my processor is, I found this: pre-programmed 68HC908GP32 processor . I'm not sure how manny bit that is? Yes I can adjust the RPM rang as well........

[big-phil]

ok I've matched the old kpa, with the new kpa in my new map. I need to ge for a drive and will report back.

[big-phil]

what I have done is started to tune the car with its KPA settings. Drove it last night, boosted it, its running great. I'm liking it. I've got to get some bigger injectors so I can try 30 psi.

 

p.s. I got some 100% methenol last night!!! so I'm going from windshield washer fluid, about 8% meth? to 100% meth:mparty:

[Careless]

someone stop this guy before he starts putting 100% meth in his coffee.

 

LOL j/k phil.

 

30psi? I don't remember any links to any forged bottom end goodies... you withholding information?

[big-phil]

someone stop this guy before he starts putting 100% meth in his coffee.

 

LOL j/k phil.

 

30psi? I don't remember any links to any forged bottom end goodies... you withholding information?

Thats going to be my downfall, the motor has a fresh rebuild but is stock (2mm head gasket) I'm trying to tune it so she wont detonate.

 

My longterm goal is to get forged pistons, if I can work out the tunning bugs and run 30 psi on stock pistons, then I'm getting forged. If it detonates and breaks ring lands (again) then I'll get forged pistons and run it to the point of detonation. I'm running 19lbs now with windshield washer fluid:-)

 

so either way I'm getting forged pistons.

[X64v]

The highest KPA entry in your fuel map doesn't have to be the max that the MAP sensor will read. The highest entry I have in my N/A fuel map is 100kpa, though the sensor goes much much higher. Right now I have one that reads:

 

100

90

80

70

60

50

40

30

20

10

0

0

 

(The 0s are at the bottom because I really don't need any more than 10 pressure entries in my 12x12 table, though the 12 rpm tables help a lot)

[TimZ]

The highest KPA entry in your fuel map doesn't have to be the max that the MAP sensor will read.

That's true, but it's a good idea to understand what the controller will do if you go outside the map - does it flatline (probably), or does it use the last known slope and extrapolate?

 

(The 0s are at the bottom because I really don't need any more than 10 pressure entries in my 12x12 table, though the 12 rpm tables help a lot)

 

I guess I don't understand the logic of not using all of the resolution that you have available - if you have 0-100kPa range and 12 cells, why not use 8kPa steps and all of the cells? Or better yet, since you probably aren't going to actually be doing anything besides closed throttle decel in the 0 - 20 kPa range (possibly higher depending on your cam: 20kPa equates to ~24in of vacuum), why not go from 20-100kPa in 7 kPa steps?

 

Also, if you are using big injectors (like Phil is), it's a good idea to take a look at the point where the commanded injector pulsewidth will be less than the min injector on-time. For instance, if 40kPa MAP is going to result in you being under the min on-time, then having cells at 20 and 30 doesn't make much sense.

[TimZ]

I'm trying to find what my processor is, I found this: pre-programmed 68HC908GP32 processor . I'm not sure how manny bit that is? Yes I can adjust the RPM rang as well........

 

Looks like 8-bit

[z-ya]

Why do you guys care if the processor is 8 or 16 bit? MSI uses and 8 bit processor, and MSII uses a 16 bit one. Extra code gives you a 12x12 table, and from my experience the 8 bit processor has enough cycles to handle it. It is when you start running boost control, water injection, etc. that you might start running out of processor cycles.

 

That's true, but it's a good idea to understand what the controller will do if you go outside the map - does it flatline (probably), or does it use the last known slope and extrapolate?

 

Yes, it takes the last VE table entries and extrapolates up. The nice thing about the MS is that you can customize your load and RPM range scale to fit the application. The highest load range bin should be the max boost you plan on running plus a few kPA for some headroom. For NA, it should be 100. The lowest load range should be the minimum pressure the engine operates within minus a couple kPA, not your idle pressure. Many engines will make more vacuum under other operating conditions besides idle. Especially if you are running a non stock cam. For a stock turbo cam, the idle pressure is more than likely the lowest.

[TimZ]

Why do you guys care if the processor is 8 or 16 bit?

 

I was referring to the A/D converter resolution, not the processor itself.

 

Actually, the 1.6kPa per LSB resolution assumes a full 5V span for the sensor. I have seen 4 bar sensors that don't use the full 5V. So, if your sensor has a 4 volt span, the min resolution is 2kPa per LSB

[cygnusx1]

I have been running the 4-bar map sensor since day one. Phil, did you config Megatune and tell it that you are now running the 4-bar? You need to do that in Configurator.

 

You might want to invest in some ARP rod bolts pretty soon.

 

 

SEEE ATTACHED

[z-ya]

I was referring to the A/D converter resolution, not the processor itself.

 

You are correct, the ADCs on the 8 bit processor are 8bit in resolution. This not all that important to the end user because the voltage range from the MAP sensor is scaled to match the input range of the ADC. Based on what MAP sensor you tell MS you are using, the software will do the appropriate scaling.

[TimZ]

This not all that important to the end user because the voltage range from the MAP sensor is scaled to match the input range of the ADC. Based on what MAP sensor you tell MS you are using, the software will do the appropriate scaling.

 

Yes it's scaled, but as you go to higher range sensors, you still only have 255 steps of resolution, as I mentioned above. This is important to understand.

 

For instance, for a "normal" engine that typically sees 20 to 100kPa when off boost, if you are using an MPXH6400 4 bar sensor (as cygnusx1 is above) with a 4.6 volt span and 20 - 400kPa range (1.6kPa per LSB), only 50 steps of the total 255 will be available for "drivability" tuning. That's pretty coarse.

 

If Phil decides to put the big cam back in, he's more likely to only see 50-100kPa when off boost, so with that same sensor he'll only have 31 of the total 255 steps available for drivability.

 

Also, note that this does not take into account the error band of the sensor, which is spec'ed at +/- 5.5 kPa.

[cygnusx1]

Yes it is noted on the MS website that the 4Bar MAP will reduce your resolution. On the other hand, mine runs, idles, transitions, fine with the stock cam up to 15psi. When I expand my table to higher boost, that will change....or maybe not.

[X64v]

That's true, but it's a good idea to understand what the controller will do if you go outside the map - does it flatline (probably), or does it use the last known slope and extrapolate?

 

 

It flatlines, I've seen it on my own datalog of some unexpected boost creep. It will use the highest VE value in the current rpm column. I'm not saying that Phil should put it right at 310kpa if he expects to see 308kpa manifold pressures, but simply that his highest number does not have to be 408kpa. For example, when I expected to see a max of 150kpa, I'd write my fuel map up to 170kpa, and set my overboost cut right there at 170kpa.

 

 

I guess I don't understand the logic of not using all of the resolution that you have available - if you have 0-100kPa range and 12 cells, why not use 8kPa steps and all of the cells? Or better yet, since you probably aren't going to actually be doing anything besides closed throttle decel in the 0 - 20 kPa range (possibly higher depending on your cam: 20kPa equates to ~24in of vacuum), why not go from 20-100kPa in 7 kPa steps?

 

 

I could, but it's a matter of time vs. results. Since Megasquirt extrapolates between values, and I already have six kpa values between cruise and W.O.T. (50-100kpa), I've already got all the control I need over the map. Going between 20-100kpa in 7kpa steps would give me eight kpa values between cruse and W.O.T. instead of six - not enough to make a difference in driveability in my opinion, especially since Megasquirt would have extrapolated my numbers to those anyways. Those two extra values, however, would give me 24 more entries to edit when I change the curve of the fuel map any more than just some fine tuning.

 

In short, if I can get the AFRs I want everywhere with six even kpa values, why use eight oddly spaced ones?

[big-phil]

I have been running the 4-bar map sensor since day one. Phil, did you config Megatune and tell it that you are now running the 4-bar? You need to do that in Configurator.

 

You might want to invest in some ARP rod bolts pretty soon.

 

 

SEEE ATTACHED

Yes I did.

[big-phil]

It will be nice to go from a max of 408kpa to 320kps or so? Thanks so much for the input guys!!!

[z-ya]

Yes it's scaled, but as you go to higher range sensors, you still only have 255 steps of resolution, as I mentioned above. This is important to understand.

 

I totally agree.

 

There are 256 "voltage" steps (0-255) from 0V to full scale on the input of the ADC. The MAP sensor converts pressure to voltage. The circuit in the MS translates the MAP sensor output voltage range to match the ADC input voltage range (or at least it should) to maximize dynamic range. The firmware then uses a lookup table of some sort to convert voltage levels to actual pressure which the interpolation algorithm uses. This is how I think it should work (then I did a little testing).

 

The firmware I'm running (Extra code 029v) appears to limit the max kPA value to 255 no matter which sensor I have selected under "Engine Constants". In my mind this should be a bug as the maximum value you can enter in the VE or spark table in the load band column should equal the max boost your sensor supports.

 

If this is truly the correct behavior, then you would need to think of it as an 8 bit number like TimZ said, and do the scaling in your head. Again, this does not seem right. Why not have the processor do this math for you (unless MS1 with extra code and a 12x12 table does not have the cycles to do this computation in real time).

 

Make sense?