Calculating IQ from BSFC

Found some BSFC maps for my exact engine, for WOT that means I'll have to increase both torque max and HP max IQ, since my planned torque max is now 2300-2500, not 1900 as in the stock file.

So from my previous values of 5,6Nm/mg/stroke, which are @1900/min, my previous numbers were:

62,5mg/stroke @ 2500 (let's say 2500 for now, to save the gearbox)
56,25mg/stroke @ 4000

and now, calculating BSFC in, those numbers become:

65,625mg/stroke @ 2500 (62,5mg + ~5%, 197->207 BSFC)
67,95mg/stroke @ 4000 (56,25 + ~20%, 197->240 BSFC)

I'm not sure whether that's really going to be correct, but we shall find out :)

An alternative way to calculate necessary IQ

My first number of 5.6 Nm/mg seems to be off, as I've found in some discussions in german language forums (thanks Ulf). Namely, I'm calculating IQ directly from logged IQ at WOT at a specific RPM setpoint. However, what I didn't take into account until now is the following, which seems very convincing, the more I think about it.

To make the motor rev at 1900 rpm without any load (no "usable" torque) needs about 5mg/stroke real IQ. Meaning, the IQ of ~37,5mg/stroke at this rpm point is actually 32,5mg/stroke of "usable" IQ plus 5mg/stroke that goes into friction losses etc. just to make the motor turn over.

So, when I recalculate Torque vs IQ, the factor is now 6.46 (210 / 32.5) instead of 5.6Nm/mg, plus the 5mg/stroke for losses.

My values then become -

Torque max @ 2500
350Nm / 6.46 + 5 = 59.179 mg/stroke

HP max is more difficult, since I don't have log numbers using the stock engine...estimating from the original file, it's injecting ~32.5mg/stroke @ 4000. This has to net 90HP/66KW - so needs 157Nm @ 4k.

Judging from what happens when the quantity adjuster isn't set correctly (mechanical position), it doesn't need more than about 7mg/stroke to rev to 4k without load.

Now let's divide 157Nm by 25.5 (32.5mg - 7mg friction losses), that's 6,16Nm/mg. Pretty close to the 6.46 we had above, but less efficient. So I'll use 6.1 to be on the safe side. This is 5.57% less efficient.

HP max @ 4000
315Nm / 6.1 + 7 = 58.64 mg/stroke

Guess the true values will lie somewhere in between - but Ulf's calculation, or rather, how he arrived at it, is convincing. Also, the HP max IQ has never been above the one for torque max in any file I found, so I'm skeptical of the IQ numbers adjusted for BSFC from the diagrams I found.

Regards

Efficiency wise it does look close to mine...
0.34 lb hp hr @ 2000 = 206.8 g kw hr
0.38 lb hp hr @ 4000 = 231.1 g kw hr

350nm (167hp) = 65.0mg @ 2500 (0.35 lb hp hr)
320nm (180hp) = 64.6mg @ 4000 (0.38 lb hp hr)

Yeah, that does look close enough - from looking at the BSFC chart, I thought the value was closer to 240, or even above, maybe I just read it wrongly. Thanks again! I'll just have to figure out what the real necessary IQ is by doing log runs - will be using my values to start out with and work myself up towards yours.

In other news, almost fried my EDC15 yesterday, bought a bench flash adapter cable to make things easier on myself, then decided to use my multimeter to check whether it was contacted correctly before plugging everything in.

Good idea, that! A little paranoia can go a long way...

It wasn't an EDC15 cable, it was a ME 7.x cable, where polarity is laid out differently, K-Line is somewhere else, ...

Anyway, replacement (cable, ECU is fine ;)) will be coming in this week, then we'll see.

Regards

Good and bad news - the good news: my bench flash harness works perfectly ;)

The bad news: the remap I've been driving (paid for, not by me) is...let's say..."uninspired". I could say many bad things about it, but I won't, it's no use crying over spilt milk.

Anyway, I've decided not to even look further in the file because there is nothing new to learn in it :mrgreen:

I'll be working on a basic remap this weekend, with all relevant map axes extended to 70mg/stroke - will be a bit of work, but worth it.

Next up in hardware changes: installing a 300kPa MAP sensor and adjusting the MAF and MAP linearisation (MAF only goes up to 850 stock)

Regards

Correction: it goes up to 875mg/stroke. Still, not enough to inject anywhere near my wanted quantity and be able to regulate AFR/smoke via MAF.

So in the long run, instead of switching to a larger MAF housing and redoing air duct from filter to turbo and doing MAF recalibration and so on, I'm going to switch to MAP-based smoke regulation instead.

However, to start out with and develop a working baseline, I'll be keeping the MAF as it is for now, and save myself some money and lots of work. That means staying at ~150HP or thereabouts max for now, but that's okay, I'm not in a hurry as long as my long-term goal is reachable eventually :) That's why it's a long-term goal, after all.

Another error I've realised - the stock file from the T4 2.5 that RPM Freak recommended and uploaded is not designed for a 10mm pump. I read the number wrongly, it's actually an 11mm pump! So that means the pump voltage map can be used pretty much 1:1, it's just larger (17x16 vs. 15x16) than the one on my ECU dump, so I'll have to decide which columns to keep and which to remove. However, realising this error makes my work a whole big lot easier - getting the pump voltage map as right as the OEM did for the T4 would have taken an enormous amount of time and effort!

So, now that the pump voltage map is figured out, diagnostic limits are adjusted, the IQ limit needs to be adjusted up from 51mg/stroke, as well. I've already found all the 8-bit values that need changing (I think), but haven't managed to find the associated 16-bit diagnostic scale factor yet. So right now the reported IQ via diagnostics would be way off but real quantity correct, will have to research more.

Modified maps will be uploaded tomorrow for some review.

Regards

It's been a while since I had time to do anything, been sick and had lots of work to catch up on.

In the meantime, I've ordered an ARL front cooler packet including braces, bumper support and intercooler and fans, since the stock side mount intercooler is really not enough to be safe at WOT for the power I'm planning. The only thing that has to be changed is the air ducting from turbo to intercooler and back to the engine inlet, will see whether I can use stock parts there (I like my car to look OEM/original), the rest should be a plug-and-play change.

So, on to some maps!

Here's my revised pump voltage map -



I've rescaled it to 65mg/stroke, which should suffice for now and be within safe limits.

The map is taken from the T4 bin RPM freak posted earlier, pump voltage is extrapolated > 50mg/stroke. Changes shown are in regards to the original map for the 10mm ALH pump using .184 stock injectors.

The only part of the map I didn't raise quite as far as calculations said is the bottom right field for 65mg/stroke @ 5100 - since that would go over 4.7V (values shown are in real pump voltage, with the correct factor applied).

Next is my revised smoke limiter - axis values extended from 853 to 960mg/stroke (linearisation max value @ 4.6V is 960, I had the axis off by two bytes, not 875)



Not so sure about it yet, but I'll see what happens. Calculated 17:1 AFR for (hopefully) smoke-free operation.

Now for the torque limiter - also not quite sure whether it's "rounded" enough yet, but still not overtaxing the gearbox below 2k rpm. We will see :)



Next up will be Turbo related maps, will have to guesstimate what sort of airmass it can push at what pressure and compare with log values.

The torque limiter is already scaled to 360Nm@2500 falling to 320@4000 for about 185 HP@4k.

I'll do two sets of maps, one "full power" and one @ ~150HP. This is the "full power" variant.

Regards

Hi,

I've noticed another big error in my calculations - studied what Bosch wrote in the function description some more and noticed that the linearisation for MAF is really not in mg/stroke but in kg/hour.

Stupid of me not to notice...and I was wondering why actual MAF goes so far above the linearisation values for a long time - now I know.

The actual formula is calculated from the raw value via the linearisation table like this:

actual air mass [mg/stroke] = linearised value [kg/hour!] / rpm / 60 / 2 * 1000000 [kg->mg]

Using this formula, I get very close to my actual observed log values @ 4000 rpm, where the MAF is limited due to the housing size/air flow speed.

All of a sudden, things become much clearer :) Will have to revise my smoke limiter some...let's scale that sucker up!

Regards

You can run the car on MAP limiter and trow away the MAF.

Yeah,

that's what I'm planning in the long run, anyway - just found the switch bits in my file yesterday night, I think (51AF0/51AF1, 71AF0/71AF1), as well as N75 upper/lower limiter to be able to open the vanes at high IQ a bit further.

I found a file from a 07 Fabia 130 PD which has a MAP-based smoke limiter, which will help in checking my modifications. To switch to MAP-based smoke limiter (will leave MAF installed for now), AFAIK I'll need:

• 2 bytes changed from "00 00" to "01 01" per coding to switch from MAF to MAP based smoke limiter
• populate the "IQ by MAP limiter" map with meaningful values
• adjust the default values in the "boost pressure by temperature limiter" map to match the ones from the Fabia instead of no correction as in my MAF-based original file right now (MAF sensor does this on its own, it's temp-corrected)
• adjust requested boost to match modified IQ
• adjust N75 upper limit (to open vanes further at high IQ, meaning @torque max as well as at high rpm) and N75 map in general to take higher IQ/boost into account and limit boost spikes

Will use the IQ by MAP limiter map from the Fabia as a starting-off point, probably will have to adjust for my upgrade turbo, GT1752VB vs. GT1749VA. My turbo should actually give more airmass at the same boost pressure/temp, so that should be fine to start off with, less smoke anyway.

Did I forget anything except doing lots of log runs to validate everything? ;)

Right now, I'd probably be okay with MAF-based smoke limiting, although at high RPM IQ would have to be reduced somewhat, at least > 4k RPM - so that's why I'm definitely going to switch, but maybe not right away.

Regards