HOWTO: Injector Scaling Basics

[DanStraffa]

Yeah I honestly don’t know what happened they weren’t that bad few months ago.

I’m going to start from scratch, least I can only go up from here @Roland@pcmtec

 

I’m not sure but I think one of my dramas could be is I changed my alternator pulley to a larger one from Dynomite Performance and my battery isn’t seeing 14 bolts it’s only seeing 12.6-12.8 volts.

 

Could that be a contributing factor? 

Edited November 10, 2020 by DanStraffa

[Roland@pcmtec]

Depends if your deadtime offset has been calculated correctly for those voltages. When you scaled your deadtimes did you rescale the whole table every time, or just the 14v row?

[DanStraffa]

Nah we rescaled the whole lot, not just 14v.

[DanStraffa]

Thoughts now? After a few changes of deadtime and low slope.

[Roland@pcmtec]

Looks pretty close. Drive it for at least 80kms to dial in then see if it persists. 

[DanStraffa]

Okay will do thank you 👌

[Milanski]

On 6/20/2019 at 1:13 PM, Roland@pcmtec said:

Here is the actual equation if you want to build a slope calculator in excel. I worked this out drawing the straight lines and some year 9 algebra (yes there was a point to learning all that maths!)

IF (fuelmass >= breakpoint) y (fuelmass_lbs) = highslope_lb_sec * (x (injector_sec) - (breakpoint * (1/lowslope_lb_s - 1/highslope_lb_s) + offset_sec))
IF (fuelmass < breakpoint) y (fuelmass_lbs) = lowslope_lb_sec * (x (injector_sec) - offset_sec)

fuelmass_lbs = airmass_lbs/ commanded_AFR

 

The equation for the lines would then be

Above breakpoint

y = h * (x - (b * (l-1/h)+o))

below breakpoint

y = l * (x-o)

Where:

y = fuel_mass

x = injector_pulse_width (seconds)

l = low_slope (lb/s)

h = high_slope (lb/s)

b = breakpoint (lb)

o = offset (seconds)

 

You can then invert the equation if you want to solve for injector pulse width as well.

above breakpoint

x = (b*(h*l-1)+h*o+y) / h

below breakpoint

x = (y/l) + o

Hi Roland, just to pick this up.  Does Min Pulse Width = offset in these equations?  I think you mentioned it early on.  How does the battery offset come into equation then?

[Roland@pcmtec]

Min pulsewidth should be above the dead time. Min pulsewidth is basically saying between deadtime and now the injector will be dribbling or otherwise unreliable in its fuel delivery (misfires etc) so clip it at this pulsewidth to ensure combustion is complete (but it will be rich as a result).

Properly designed and sized injectors should never hit the min pulse width clip to ensure good fuel economy. 

Dekas are a great example of a crap injector needing a high min pulsewidth to operate without misfiring and hence have poor fuel economy and run below lambda at cruise/decel as they are hitting the min pulse width clip. 

[Milanski]

Thanks understood, maybe I will ask the questions another way,  the offset_sec  in the equations comes from the battery offset table?  

[Roland@pcmtec]

15 hours ago, Milanski said:

Thanks understood, maybe I will ask the questions another way,  the offset_sec  in the equations comes from the battery offset table?  

Yes the offset is the deadtime. They are the same thing. Minimum pulse width is simply a minimum clip value.

[john.z]

Ive just fitted 1000cc injectors, the only info ive received is a dead time of 1.59ms @ 12V and 3 Bar. running 4 bar, how should i be scaling this? what ive currently done is entered in 1.49ms @ 12V and moved the whole voltage curve the same difference. Is that the wrong approach? It's a cammed boss 290. Got it running alright, there is a slight hesitation while cranking after the first fire, but never struggles

[john.z]

@Roland@pcmteci should be able to log for individual banks yes? not sure why but this is how it came out. It's fuel flow bank 1 doing it

[JVK]

Did you try swapping the injectors from right bank to left just to rule out an injector issue ?

[john.z]

32 minutes ago, JVK said:

Did you try swapping the injectors from right bank to left just to rule out an injector issue ?

Happening with the stock and the new  injectors. 

[john.z]

noticed throughout my logs that there was no difference between the pulsewidths on either bank, it seems to be just recording bank 2 and reporting that as both. which expains the weird plot. LTFT and fuel flow show up right though. Is this just an oversight in PCMTEC datalogging?

 

[Roland@pcmtec]

2 hours ago, john.z said:

noticed throughout my logs that there was no difference between the pulsewidths on either bank, it seems to be just recording bank 2 and reporting that as both. which expains the weird plot. LTFT and fuel flow show up right though. Is this just an oversight in PCMTEC datalogging?

 

They are the real internal values. They may be pre fuel trim scaling though. Can you set them both to high priority and attach the log file here? I will have a look at the asm tomorrow and see if there is anything added after that value that you can log seperately. 

Can you also include the OSID of your vehicle. I will double check the addresses are different as a sanity check.

[john.z]

okay ill get that through tomorrow

 

[john.z]

aren't the first 15 already set as high priority? if so here is my first log after adding the 1000cc injectors

 

 

1000cc log1.teclog

[john.z]

also i cant log STFT for both sides?

 

[Roland@pcmtec]

5 minutes ago, john.z said:

also i cant log STFT for both sides?

Try logging F_A_RATIO[0] F_A_RATIO[1] instead as this is effectively the STFT per bank, we have never had a v8 test car so this stuff has all been developed for the 6 cylinder. The Pulsewidth calc is copied and always identical for the two banks as its actually done on a per cylinder basis and the bank reading is not really used for anything useful internally that I can see.

There are actually 8 pulsewidths you can log, one for each cylinder. This is what is really going to the injector. These are logged as FPW[0], FPW[1]...FPW[8]

[john.z]

2 hours ago, Roland@pcmtec said:

Try logging F_A_RATIO[0] F_A_RATIO[1] instead as this is effectively the STFT per bank, we have never had a v8 test car so this stuff has all been developed for the 6 cylinder. The Pulsewidth calc is copied and always identical for the two banks as its actually done on a per cylinder basis and the bank reading is not really used for anything useful internally that I can see.

There are actually 8 pulsewidths you can log, one for each cylinder. This is what is really going to the injector. These are logged as FPW[0], FPW[1]...FPW[8]

None of those show up. I’m using professional, they only on workshop?

Edited July 20, 2021 by john.z

[Roland@pcmtec]

Yes, they are not in pro afaik. 

[jpknobel]

Hi Roland,

I'm not sure if this is the right place to ask but it's relevant and may assist others in future. Please let me know if this isn't the right place.

Im using PCMtec Pro and also have HPtuners. (i usually tune diesels and have only done a small number of petrol tunes (not for paying customers).

Have a FG XR6 that has just been turbo'd (is an NA motor) and has the following relevant mods:

- Custom Alloy Welding billet centre feed intake manifold 

- Raceworks INJ-213 Injectors (1100cc at 58PSI)

- 4" turbo back exhaust

- Walbro 460 pump with a Turbosmart FPR800 reg set to 58PSI

- Innovate wideband

 

I have setup (taken a preliminary guess) some blending between tables of the NA calibration and a factory turbo calibration in terms of timing and fuelling. I've copied over most speed density, cam control and whatever else may be partially related to the current tune from a factory turbo tune.

 

The issue i am having is unstable AFRs. I believe this is due to my injector data being off (raceworks supplied) however am unsure. If i disable closed loop O2 control my idle and light cruise at 100kmhr AFRs maintain target AFRs close enough at this point (assuming some SD stuff will need tweaking due to intake manifold).

The main issue i am seeing is when closed loop boost control is enabled my AFR gauge swings between 13.5:1 and 15/16:1 and will do a full swing once every second or two.

I have attempted to log relevant data as you have previously described but it doesn't look that far out (i may be interpreting this completely wrong).

Do you have any pointers for me please?

[Roland@pcmtec]

The main issue i am seeing is when closed loop boost control is enabled my AFR gauge swings between 13.5:1 and 15/16:1 and will do a full swing once every second or two.

You mean closed loop O2 control right?

The O2 circuit purposely oscillates rich/lean to keep the cat converter working correctly. If you can show a datalog of the narrowband voltages and wideband overlaid on top this would be interesting to see.

Your plot looks pretty good to me, if you don't look at the datalog and the car drives smooth with no stalling I would call it a day. If you ever get a stock car do the same logging and you may be surprised how good yours is.

I suspect there are hardly any cell counts at the 0.95 region?

 

[jpknobel]

Sorry, yes, closed loop O2 control is what i was meant to say.

I don't currently have a way to datalog wideband against any of the pcmtec data unfortunately.

The car drives okay while its sweeping i just wasn't sure how wide of a sweep was normal. I verified the in car wideband against my portable wideband (LM2) and my dyno and they all seem to correlate (close enough anyway as dump pipe sensor vs tail pipe).

I might continue tuning with closed loop O2 off so i remove that factor and when i'm done ill just re-enable.

The only other question I had which is probably off topic is to do with the aftermarket intake manifold. What data would you expect that i need to change to account for it?

 

Thankyou for your help mate, its appreciated.

[Roland@pcmtec]

On 10/9/2021 at 7:48 AM, jpknobel said:

The only other question I had which is probably off topic is to do with the aftermarket intake manifold. What data would you expect that i need to change to account for it?

It likely affects the VE in a small way which would likely be linear to rpm/load if you plotted it.

You could most likely fudge this error in a variety of ways, tuning correction, injector slopes etc. The correct way would be to rescale the speed density offset/slope tables to suit, but that is an extremely time consuming process which requires you lock the cams at each position and datalog each cam/rpm cell in steady state on a dyno. Realistically no one would do this.

What is more likely a problem is unbalanced flow across the cylinders, the stock manifold is the best and if you've used a front facing plenum you likely will need individual cylinder air mass trims, the only way to determine this is with individual EGT and lambda probes.

edit: Do you mean the new plenum the intake is in the centre? If so its likely its reasonably balanced, but you never know.

[jpknobel]

Thanks Roland, appreciate it.

Yeah its a billet centre feed intake. I chose centre feed over front feed as i figured its inherently more likely to have a more even flow due to the centre feed design than a forward facing design.

Photo of manifold attached if you were interested.

 

[DanStraffa]

G'day men, just put some 2200cc injectors in, I got some ID2000's data off their website it seems pretty good so far I've driven for the 80ks but I'm unsure of what to change first. I'm not the smartest cookie in the jar, I've read this whole forum but yeah don't know what to change first I don't want to be stepping backwards.





@Roland@pcmtec the 2.10 beta with the DMR's can you post an updated list of them like STFT - LAM FINAL?, LTFT, is Fuel pulse width INJ PW NORM? etc
Don't know if there is another thread for this or not.

[dat111]

hey lads. currently tuning my NA+T BF MKII 

have restored 99% of stuff back to stock that had been changed for no reason at all on the tune. 
using the stock turbo injectors. an the turbo data 

Breakpoint 0.00000939999972615624
High Slope   0.00929140020161867
Low Slope   0.0135736996307969
4bar bosch fuel reg. 

LTFT's are sitting at between -15 an -16
**waiting on my tactrix, its in the mail havent flashed new tune in it yet. **
attached are both tunes. 

HACCKA5 old one V1.tec HACCKA5 - updated2.0.tec

Edited December 10, 2021 by dat111

[Puffwagon]

After looking at the old tune the other day, you can see it doesn't have stock injector data in it.

I didn't look further to see what else will be causing the trims to be out as you didn't mention it then.

[dat111]

11 hours ago, Puffwagon said:

After looking at the old tune the other day, you can see it doesn't have stock injector data in it.

I didn't look further to see what else will be causing the trims to be out as you didn't mention it then.

probs something i should have mentioned to you,
im going to wait until i flash the new tune in, an then go from there. as the tune was very average to say the least. 

[SKD]

Hi Everyone,

Just installed some Bosch 1000cc injectors and am trying to get it to run well enough to get it on the trailer to the tuner. It currently has a bad miss and afr is out (0.85 at idle), I'm praying it isn't any of the new hardware... I was hoping someone would throw some input my way as the first time I have scaled injectors! I have based changes on the information I could find within the forum but given everything is different I could be miles off. I have included my tec file if anyone wanted to have a look??

Thanks!

HAANFH4-BAE101-1000cc-Importdata.tec

[Puffwagon]

5 minutes ago, SKD said:

Bosch 1000cc injectors

 

On 4/10/2019 at 5:59 PM, Puffwagon said:

I'll just leave it here for anyone to use. Bear in mind that all cars are slightly different and you need to double check the tune with a wideband. I left the low slope slightly rich as it starts much better like that when it's freezing cold. Feel free to raise the low slope to 120 (less fuel) if you feel like it.

 

 

Bosch 980cc injector parameters.param 19.61 kB · 218 downloads

This will get you driving

[SKD]

10 minutes ago, Puffwagon said:

 

This will get you driving

I actually tried the parameters you had... Its why I'm leaning to a possible issue elsewhere. It would idle though again afr was quite a bit out? I guess I could be expecting too much from the changes too and need to wait my tune?? 

[Puffwagon]

Hardware changes can throw out a tune to the point of it being not drivable.

If you only want to get it on the trailer, just add 10 to each of the slopes and try again.

[Roland@pcmtec]

If you haven't seen it already check this out

 

[Andre34]

Hey I’m not sure if this is a wrong question but I’ve just got some new injectors in the car and I want to have a go at getting the idle stft right. At the moment I see it go as high as 23%. 
 

My question: 

if the stft is negative, does that indicate it is pulling fuel which means the minimum slope is too high? and if it is reading positive, does that mean it is adding fuel meaning the minimum pulsewidth is too low?

[Puffwagon]

25 minutes ago, Andre34 said:

new injectors

What injectors are they? Someone might have used them and can get you in the ballpark.

[Andre34]

25 minutes ago, Puffwagon said:

What injectors are they? Someone might have used them and can get you in the ballpark.

They are Bosch 627cc/min EV14 Injectors 0 280 158 123

ive used some dekka 650cc injector data to help me get started because I couldn’t find anything online at all for mine. 
 

they’re on an na bf xr6. If someone has data for them that would be awesome.

 

I’ve got screenshots of the data I used for them attached here

Edited October 13, 2023 by Andre34

[Puffwagon]

Don't touch the minimum PW or breakpoint at this stage. They will work for now.

It looks like your previous Deka data was based on a 4 bar fuel reg. At least the high slope is anyway. An na car runs a 2.7 bar fuel pressure iirc.

Try 0.0125625 for the high slope and 0.020925 for the low slope.

I based the high slope number on the injector flow at 2.7 bar and I based the low slope from your data extrapolated from 4 bar down to 2.7 bar. Long story short I multiplied them by 0.675.

It's gonna add a bunch of fuel and should see the trims swing to -10 or so. Better to start rich and lean it out gradually, rather than have it lean af under load. The trims only work at idle, cruise and mild acceleration, full load does what you tell it to do which in this case is probably very lean. Doesn't matter on an na car, it won't hurt the engine but not ideal and loses power.

Haha see what I mean about needing a wideband now?! Heh mebbe we can get an admin to make a new thread ay?

[Andre34]

9 minutes ago, Puffwagon said:

Don't touch the minimum PW or breakpoint at this stage. They will work for now.

It looks like your previous Deka data was based on a 4 bar fuel reg. At least the high slope is anyway. An na car runs a 2.7 bar fuel pressure iirc.

Try 0.0125625 for the high slope and 0.020925 for the low slope.

I based the high slope number on the injector flow at 2.7 bar and I based the low slope from your data extrapolated from 4 bar down to 2.7 bar. Long story short I multiplied them by 0.675.

It's gonna add a bunch of fuel and should see the trims swing to -10 or so. Better to start rich and lean it out gradually, rather than have it lean af under load. The trims only work at idle, cruise and mild acceleration, full load does what you tell it to do which in this case is probably very lean. Doesn't matter on an na car, it won't hurt the engine but not ideal and loses power.

Haha see what I mean about needing a wideband now?! Heh mebbe we can get an admin to make a new thread ay?

Hahah yes life would be much easier with a wideband!! 
 

im doing my best to get fuel trims as close as possible for my own knowledge and to also save on hours once I take it to the dyno. Once I get the injectors at a good spot, I’ll put a 255lph pump and take it for a proper tune with the dyno. 
 

I’ll try those numbers tomorrow and check out the fuel trims. Once they’re at ~-11, do I just slightly increase the low slope until it’s closer to 0 (around -5)?

i don’t think I’ll do anything to the high slope since I don’t have afr readings. I’ll have it rich until the dyno.

once I get to the dyno, should I adjust the high slope until it gets a decent afr reading and then fine tune using the fuel tables, or should I keep my high slope as is and just adjust using the fuel table?

 

 

also, should I leave the dead times how they are? 
 

this is the only info I was given about these injectors and I couldn’t see anything from my understanding that would help

 

Edited October 13, 2023 by Andre34

[Puffwagon]

You can adjust the low slope to get the idle trim closer to 0. Leave it at a negative number rather than a positive number, otherwise you'll probably have poor cold starts and possible stalling at intersections.

There is a method for dialling in both slopes in the driveway but I ain't about to type it out on my phone. I'll see if I've posted it and copy paste it for you. You need a wideband to do it tho so get onto it.

On the dyno the wideband should read the same as the base table. You use injector data and speed density tables to make this work. Start with high slope adjustments, this is likely all you'll have to do with a fairly stock na car.

It's past my bedtime now so I'll get back to you later.

[Andre34]

2 hours ago, Puffwagon said:

You can adjust the low slope to get the idle trim closer to 0. Leave it at a negative number rather than a positive number, otherwise you'll probably have poor cold starts and possible stalling at intersections.

There is a method for dialling in both slopes in the driveway but I ain't about to type it out on my phone. I'll see if I've posted it and copy paste it for you. You need a wideband to do it tho so get onto it.

On the dyno the wideband should read the same as the base table. You use injector data and speed density tables to make this work. Start with high slope adjustments, this is likely all you'll have to do with a fairly stock na car.

It's past my bedtime now so I'll get back to you later.

I ended up increasing my low slope to 0.020925. Here is the data log.

 

I noticed that the stft goes very rich on decel. Does this mean i need to increase my deadpoint? Also, i noticed alot of fluctiuation in my rpms and stft on idle. Is that normal and if not, does that mean my minimum pw is too low? 

 

I also noticed that my stft was reading lower before i took the car around the block a few times. It looks like the car didnt go into open loop at all. Im not sure if that's normal or not.

 

Cheers

14-10-2023 02-29-40 AM Log new injector data.teclog

[Puffwagon]

7 hours ago, Andre34 said:

I noticed that the stft goes very rich on decel.

No it doesn't, rather the opposite is true. Minor decel events are lean and the major decel event is also lean. Likely this is caused by the high slope being lean.

You are asking questions that could be answered if you understood how a few things work. As I mentioned, you need to read up on the difference between a wide band and a narrow band O2 sensor. Specifically read up on how a narrow band sensor operates and what output you should expect from it when you read the log.

You need to go back into the log and check it a lot more carefully, the car did go into open loop when it was supposed to. Read up on the difference between open and closed loop O2 control. After you understand this and how a nbO2 sensor works you will have answered your questions.

Also don't floor the car until you have a wide band installed. If it was a turbo car you would have likely kicked a rod out by now. As I said you won't blow up an na car but it's not worth trying to find out.

The trims look pretty good overall, I wouldn't bother touching the low slope any more, it's more or less where it needs to be. Change the high slope to where I said because I know you're gonna floor it again. When you install a wideband and dlp8 you can check it.

Do the reading and if you need specific things answered about them I will help. Nothing you've asked so far can't be answered by doing your own research.

Honestly, get that high slope to where it needs to be and it will be good enough to drive.

[LRiordon]

When logging STFT is it normal to read for 30seconds then go back to 1.0? 
Reading matches the afr gauge when it’s reading however when it goes to 1.0 it’s super rich.

[BeerTurbo]

On 10/13/2023 at 7:33 PM, Andre34 said:

They are Bosch 627cc/min EV14 Injectors 0 280 158 123

ive used some dekka 650cc injector data to help me get started because I couldn’t find anything online at all for mine. 
 

they’re on an na bf xr6. If someone has data for them that would be awesome.

Bosch 627cc/min EV14 Injectors 0 280 158 123*

i used these on my bf/turbo r31. i found they were near impossible to dial in with the 2.7bar factory regulator. i found some flow information whitch indicated they dident like opening at 2.5bar and i figured at idle maybe that's why i was having issues.

so with a  4 bar regulator, this is what i was last using.

[Andre34]

2 hours ago, BeerTurbo said:

Bosch 627cc/min EV14 Injectors 0 280 158 123*

i used these on my bf/turbo r31. i found they were near impossible to dial in with the 2.7bar factory regulator. i found some flow information whitch indicated they dident like opening at 2.5bar and i figured at idle maybe that's why i was having issues.

so with a  4 bar regulator, this is what i was last using.

Awesome! Thanks for this. I’ll go get myself a 4bar as well in that case

[Andre34]

2 hours ago, BeerTurbo said:

Bosch 627cc/min EV14 Injectors 0 280 158 123*

i used these on my bf/turbo r31. i found they were near impossible to dial in with the 2.7bar factory regulator. i found some flow information whitch indicated they dident like opening at 2.5bar and i figured at idle maybe that's why i was having issues.

so with a  4 bar regulator, this is what i was last using.

Just ordered one. Hopefully the na likes the same scaling as the turbo. I’m assuming it should work with the same scaling

[Roland@pcmtec]

The scaling should be identical however if you've gone from NA to turbo you'll need to tweak the speed density due to increased back pressure.