Sluggish 3SGTE

[INSW20]

Can anyone on here point me in the right direction on if I'm looking at more of an ignition timing or a fueling issue (or both?) as to why my car feels soooo sluggish taking some quick drives.  It's certainly not optimized yet, but it seems like I'm less and less able to determine if it's getting better or worse with some changes I've made.

I *feel* like it wants more ignition timing, but I'm hesitant to do that at this point.  Compression ratio is around 9.3:1-9.5:1, running 93 octane.  My goal is to get it reasonably safely into boost so next time I take it to the dyno I can get some useful tuning completed.  I'm not trying to 100% street tune, but last time on the dyno the tuner had a tough time getting it into boost without severe stumbling, which seems to be mostly fixed, it's just not smooth at this point.

2.1L 3SGTE (91mm crank)

ID1050x

plug gap at .028"

1zz coils

stock cams (236/236 degrees)

 

 

[0x33]

Your timing on boost seems to be quite conservative relative to other 3SGTE's I've tuned. For sure some more power & torque to be had. Obviously with a higher compression engine you will need to be careful with increasing timing on 93 octane, be sure to have sufficient knock detection / audio equipment so you can dial it in safely. 

You have a large issue with fueling on this car. 

Firstly, the fuel map makes no sense in terms of what a typical engine's VE airflow model should look like. 

Your main fuel map uses VE numbers over 100% whilst not in boost, which if you think about it makes no sense. Theoretically you can only have more than 100% VE in a cylinder if you have pressure higher than atmopsheric pressure (aka Boost). 

So looking into your log, if you look at Fuel Pressure & Differential fuel pressure you can see that you either have a dodgy fuel pressure sensor or something really wacky is going on with your fuel pump, fuel pressure reg, block somewhere etc. Your log starts with you having 440KPA diffential fuel pressure, which is very high and not required, usually 300kpa is enough. 

Then later in the log whilst driving your diff fuel pressure drops to 3.6 bar, then proceeds to bounce back between 440kpa to 360kpa. Further to this, there are many points in the map whilst you are crusing where your fueling is very lean and could be contributing to the car feeling sluggish as the engine may not have enough fuel to pull cleanly through the rev range. 

Also you say the car is running on 93 Octane, but your flex sensor says you have 18% ethanol in the tank and someone has attempted to setup multifuel using two fuel tables that interpolate between each other based on ethanol %. However, another problem with this is the second fuel map its using to interoplate with looks like this:

This is just flatout wrong and will cause issues and leaning out. Very important that if you continue to run ethanol that you get this fuel table tuned properly. 

Also injector timing looks very wrong, basically firing fuel when the intake valve is about to close, which will not help with the car feeling sluggish, throttle response will feel better if fuel is injected as intake valve is opening or just before it does (typically somewhere like 360-450 degrees). 

So yeah quite a few things to work out, specifically what is causing your fuel pressure issues. 
 

 

 

 

[INSW20]

Thank you for the input!!

For the fuel pressure, it's an Aeromotive 340lph and a Bosch 3 bar FPR, which is rated to flow 220lph if I'm not mistaken.  So what I was expecting to happen seems to be occurring, that when the injectors open and take load off the FPR it's allowing a drop in rail pressure.  I had rescaled my injectors to 1305cc @ 65psi pressure and that helped lower my VE table numbers, but the drivability wasn't any better/worse.  Fuel pressure sensor is one of these:  https://www.haltech.com/product/ht-010920-bosch-150psi-fluid-pressure-and-temperature-sensor/

The car was previously filled with E75, and filled a couple times since with E10 (all local stations have some E content) but I believe I have the interpolation set to 100% fuel map 1 at the current ethanol content.  I'll still adjust my fuel table 2 to be less of a mess.  EDIT:  I checked and my fuel table blend is 8%.  I can adjust that to 0% for current ethanol content.

I can try advancing the injection timing as well, but on stock cams and cam degreeing, the intake valve max lift is about 250 BTDC, so I tried to aim end of injection slightly before then.  Is that still too late?

 

Again, I appreciate the help!

[Vaughan]

38 minutes ago, 0x33 said:

Your main fuel map uses VE numbers over 100% whilst not in boost, which if you think about it makes no sense. Theoretically you can only have more than 100% VE in a cylinder if you have pressure higher than atmopsheric pressure (aka Boost). 

If equation load source is set to MAP which it should be in the majority of vehicles then that VE number is based on the manifold pressure. For example if you have 60kPa in the manifold and 100%VE you will get 60kPa worth of air in the cylinder. More than 100% VE will mean physically more air per volume in the cylinder than in the manifold which can occur when intake lengths and resonances are just right.

[0x33]

21 minutes ago, Vaughan said:

If equation load source is set to MAP which it should be in the majority of vehicles then that VE number is based on the manifold pressure. For example if you have 60kPa in the manifold and 100%VE you will get 60kPa worth of air in the cylinder. More than 100% VE will mean physically more air per volume in the cylinder than in the manifold which can occur when intake lengths and resonances are just right.

True, can't disagree with that. However most boosted engines I map in modelled dont see 100% VE until some boost, hence the comment that something is wrong with OP's setup as high scaling and numbers don't seem right compared to the norm. Typically these sorts of VE numbers is an indication of issue with setup. 

Back to OP, 

Any reason why your base fuel pressure is so high? 4.4 bar static on such a small pump will give some problems if you plan to run any kind of serious boost on this setup. Personally I'd set base fuel pressure to 3 Bar and start from there. 

Your fuel pressure differential should be a flat line under pretty much all circumstances, which matches your base fuel pressure that you set. Any drop in differential fuel pressure means that your fuel system is bottlenecking somewhere. If its doing weird drops under cruise conditions, imagine what it could do when you go into boost and really put some load on fuel system, you might loose fuel pressure altogether if pump can't keep up. 

As for injector timing, to test this with your setup, hold the RPM at say 3000rpm. Don't change anything other than injector timing, and see where the lambda goes richest.  This will be a good indication of how to dial in injector timing for a more efficient burn, which usually makes throttle response a bit crisper. 

[INSW20]

1 hour ago, 0x33 said:

True, can't disagree with that. However most boosted engines I map in modelled dont see 100% VE until some boost, hence the comment that something is wrong with OP's setup as high scaling and numbers don't seem right compared to the norm. Typically these sorts of VE numbers is an indication of issue with setup. 

Back to OP, 

Any reason why your base fuel pressure is so high? 4.4 bar static on such a small pump will give some problems if you plan to run any kind of serious boost on this setup. Personally I'd set base fuel pressure to 3 Bar and start from there. 

Your fuel pressure differential should be a flat line under pretty much all circumstances, which matches your base fuel pressure that you set. Any drop in differential fuel pressure means that your fuel system is bottlenecking somewhere. If its doing weird drops under cruise conditions, imagine what it could do when you go into boost and really put some load on fuel system, you might loose fuel pressure altogether if pump can't keep up. 

As for injector timing, to test this with your setup, hold the RPM at say 3000rpm. Don't change anything other than injector timing, and see where the lambda goes richest.  This will be a good indication of how to dial in injector timing for a more efficient burn, which usually makes throttle response a bit crisper. 

So first, disclaimer:  I agree with everything you're saying.  I *could* update to a larger adjustable FPR, but I want to make sure I'm not missing something else before spending that money. 

My current FPR is too small for my pump, but the pump was fine last year.  Plan is to run max of 20psi boost on E85 once I get to that point, but short-term I'm not looking for over 10psi on pump gas.  Immediate term, I'm just trying to get into 5-10psi smoothly with some kind of return to how it ran last year.  Here's all what I updated since it ran last:  turbo, injectors/rail/FPR/fuel lines, coil on plug.  The internals weren't touched.  The previous FPR was a 50psi base, and the pump had no problem pumping against it.

I believe one reason my VE table is so high is that my injectors are set to 1065cc/min at the 3bar base, and my hypothesis is that the ECU does a linear extrapolation on injector flow, when it's not quite linear.  So since my pressure is ~50% higher, the ECU expects 50% more flow, when it's actually more like 25% more flow.  It then runs a smaller pulsewidth, which is offset by me adding to my VE table.  So one more question:  is there any way to set up an injector flow table?

I'll re-run the injector test.  It seemed happiest at idle at 270btdc, but since it's set up as a table, can certainly try it at 3000rpm and extrapolate from there.

[Adamw]

You need stable and predictable differential fuel pressure.  If you know the regulator is under sized then you need to fix that, you are not going to be able to get a a good tune when the fuel pressure is so poorly controlled.

 

27 minutes ago, INSW20 said:

I believe one reason my VE table is so high is that my injectors are set to 1065cc/min at the 3bar base, and my hypothesis is that the ECU does a linear extrapolation on injector flow, when it's not quite linear.  So since my pressure is ~50% higher, the ECU expects 50% more flow, when it's actually more like 25% more flow.  It then runs a smaller pulsewidth, which is offset by me adding to my VE table.  So one more question:  is there any way to set up an injector flow table?

 Injector flow is correctly modelled using Bernoulli's equation.  Flow change is roughly proportional to the sqaure root of the pressure change.  

 

Injector timing isnt really relevant at this stage, but just FYI, at idle and cruise with normal injectors close to the valves, the ideal EOI will usually be around 360-400BTDC.  You want the fuel to sit on the back of the hot intake valve for as long as possible to be best vapor in the port.

[koracing]

I assume you went with the Racer X rail and their bosch regulator option -- I always advise my customers against that regulator option and just stick with a proven adjustable regulator like the Aeromotive 13109.  Yeah they cost more and you have to buy fittings and lines, etc.  They *always* work. 

A lot of your adjustments and trims are a bit... different... than I'm used to seeing and I can only guess the fuel pressure being so high is a big part of it.  Your dead times didn't match the information I have exactly from Injector dynamics website, but it wasn't far enough off to make that large of a difference.  Is your engine a Gen2?  I woudl start as a safe base of about 14 degrees of timing at 14-15psi boost (100kpa boost).  If it's 3rd gen or higher I would start a couple degrees less as a base.  

[INSW20]

2 hours ago, Adamw said:

You need stable and predictable differential fuel pressure.  If you know the regulator is under sized then you need to fix that, you are not going to be able to get a a good tune when the fuel pressure is so poorly controlled.

 

 Injector flow is correctly modelled using Bernoulli's equation.  Flow change is roughly proportional to the sqaure root of the pressure change.  

 

Injector timing isnt really relevant at this stage, but just FYI, at idle and cruise with normal injectors close to the valves, the ideal EOI will usually be around 360-400BTDC.  You want the fuel to sit on the back of the hot intake valve for as long as possible to be best vapor in the port.

I didn't mean it like that.  =) As always, thanks for the help and clarification on all of this!

 

33 minutes ago, koracing said:

I assume you went with the Racer X rail and their bosch regulator option -- I always advise my customers against that regulator option and just stick with a proven adjustable regulator like the Aeromotive 13109.  Yeah they cost more and you have to buy fittings and lines, etc.  They *always* work. 

A lot of your adjustments and trims are a bit... different... than I'm used to seeing and I can only guess the fuel pressure being so high is a big part of it.  Your dead times didn't match the information I have exactly from Injector dynamics website, but it wasn't far enough off to make that large of a difference.  Is your engine a Gen2?  I woudl start as a safe base of about 14 degrees of timing at 14-15psi boost (100kpa boost).  If it's 3rd gen or higher I would start a couple degrees less as a base.  

Correct, mostly went with that regulator for ease of fitment.  Actually about to order a regulator from you to get my pressure issue fixed.  Was going to get the Radium, but if you think the Aeromotive is better quality, I'll go with that.

A couple of the dead times I adjusted to get stable AFR's trying to load the electrical system (turning lights and fans off/on).

It's a gen2 stroker with around about 9.3:1 compression.  It has 9:1 CP pistons, TRD head gasket, and the head was milled.

[Rustam]

11 hours ago, Adamw said:

Injector timing isnt really relevant at this stage, but just FYI, at idle and cruise with normal injectors close to the valves, the ideal EOI will usually be around 360-400BTDC.  You want the fuel to sit on the back of the hot intake valve for as long as possible to be best vapor in the port.

If at idle this value equals 400 BTDC, thus what you suppose would be the most relevant value at 8000rpm?

[INSW20]

1 hour ago, Rustam said:

If at idle this value equals 400 BTDC, thus what you suppose would be the most relevant value at 8000rpm?

It's been my understanding that it has most impact at low rpm and loads when the injector pulsewidth is short, and can be completed within a small window of crank rotation.  At higher rpm and pulsewidths, the injectors are open for a much larger portion of the cycle.

[koracing]

11 hours ago, INSW20 said:

I didn't mean it like that.  =) As always, thanks for the help and clarification on all of this!

 

Correct, mostly went with that regulator for ease of fitment.  Actually about to order a regulator from you to get my pressure issue fixed.  Was going to get the Radium, but if you think the Aeromotive is better quality, I'll go with that.

A couple of the dead times I adjusted to get stable AFR's trying to load the electrical system (turning lights and fans off/on).

It's a gen2 stroker with around about 9.3:1 compression.  It has 9:1 CP pistons, TRD head gasket, and the head was milled.

The Radium is also a great choice and I think it comes with it's own fittings.  It is as good or better quality than the Aeromotive.

[Adamw]

8 hours ago, Rustam said:

If at idle this value equals 400 BTDC, thus what you suppose would be the most relevant value at 8000rpm?

Injector timing really only has a noticable effect (even then sometimes barely measurable) at injector duty cycles less than about 40%.  So for most users with a single injecor per cylinder, injector timing becomes irrelevant above medium load/RPM conditions.  In a high output NA engine where you are trying to find every last bit you typically use staged injection to keep DC below 40% at redline.  The injector timing will generally trend earlier with increasing RPM (bigger value).

[INSW20]

First of all, thank you to all who have provided input and guidance!  I upgraded the FPR to an adjustable unit and set base pressure to 43.5psi (3bar).  I sent my injectors off for cleaning and flowbenching, and they did confirm that I must not have gotten one of my fuel supply hoses all the way cleaned out of rubber dust after cutting....  anyway, they're clean now and flowed as follows:

 

 

So I updated my injector setup to 1045cc, and re-cleaned all of my supply hoses and rail and confirmed there's nothing coming out except for fuel now.  What I'm seeing after my first drive is that my VE table is skewed a lot higher than I was expecting.  Is there something else I missed during setup?  Log file and cal file attached.

The car feels MUCH smoother now, and my differential fuel pressure is very stable.  Please excuse my very rough VE map, I'll be smoothing and filling it in soon, but don't want to spend a lot more time if there's something else in setup that I need to adjust/fix.  My fuel maps match, fuel pressure is well-controlled, injectors are verified clean and flowbenched, but I'm still needing to ramp my VE table up to 85%+ at idle and up from there.  I also updated my injection timing and will fine tune that once my VE map is in better shape.

Thank you all!

EDIT: I need to update my short pulsewidth adder for the correct fuel pressure, and double-check my deadtimes.  The deadtimes are pretty much spot-on, but would the SPWA be causing some of my issue?  Injectors are ID1050x's.

 

[Adamw]

If VE is that high at idle but looks closer to normal at higher loads then I would be most suspicious of deadtimes.  A quick test is to hold it at constant RPM about say 2000RPM and adjust the target lambda by 10%, if the deadtimes are correct then the measured lambda will also change by 10%, if not then scale the whole deadtime table up or down until measured lambda follows a target change closely.

[INSW20]

Sounds good!  I'll double check deadtimes and SPWA, and then test using that method.

To confirm, there's not anything necessarily "wrong" if my VE table has for example 110% while I'm still in vacuum and 125%+ while in low boost?

[INSW20]

Confirmed my fuel pressure sensor is accurate using a mechanical gauge.  Verified deadtimes and SPWA are exactly from ID's website for the ID1050x's.  Does this look like a deadtime issue to you then?

Aside from that, is it normal for me to see VE table values as high as they are at higher RPMs and loads?

 

[Adamw]

Did you do the test I suggested to confirm deadtimes?  Injector dynamics use old Motec ECU's to derive their deadtimes so they dont always match up with other ECU's.

[INSW20]

33 minutes ago, Adamw said:

Did you do the test I suggested to confirm deadtimes?  Injector dynamics use old Motec ECU's to derive their deadtimes so they dont always match up with other ECU's.

Yes, and that confirmed my deadtimes are incorrect. I’ll get them fixed so my actual lambda changes more closely with my target changes and report back.  I think my log captures the discrepancy as-is.  Would you expect that change to also drop the rest of my VE map?  Or should boost regions still be well over 100%?

[0x33]

Fuel pressures and other mechanical bits seem to be behaving now. 

Deadtimes & short pulse width look correct, I've used same values previously with no issue on multiple setups. 

This table should be set to 0 whilst you are setting up your pump fuel map. 
Currently, whilst you are sitting at 15%, the following variables will interpolated between your first fuel and second fuel settings depending on ratio amount. Hence why your main VE table makes no sense and are very high. 

·Stoich Ratio

·Injector Flow

·Fuel Density, Fuel Density Temp. Coeff., Fuel Charge Cooling Coeff.

·AFR/Lambda Target
 

Do yourself a favour, and do the following cause you are going about this a bit backwards. 

- Drain tank and put only pump fuel in. 

- Change fueling equation mode to "modelled" rather than "modelled- multifuel"

- Set up a proper AFR target table that isnt pig rich, and setup a charge temperature table similar to whats shown in help file to get you started. 

- Tune car fully on pump fuel, get everything dialed in and working as intended. Save file, and make copy. Turn the newly copied file into your Flex Tune file. 

- Open new file and change fueling equation mode to "modelled - multifuel"

- Now tune car using ethanol and fill out all required maps etc for flex tune. 

- Enjoy 

 

[INSW20]

So should I still have it set to 0 if all pump gas in my area is E10?  Does that extra 4% make a big difference in where I’m currently at?

Everything you’re saying makes sense. I’ll turn off multifuel and keep at it. Then work up to multifuel. Thanks again!

[0x33]

When you are on your primary fuel you want the ratio table to read 0%. That way the map is scaled completely for just that fuel. 

If your pump fuel happens to have E10 in it consistently, then you will still want that ratio table to read 0 as you need your first fuel map to correspond directly with that fuel and nothing else. 

Once you go to E15 or above by adding in ethanol, then you can start blending in your secondary fuel settings / table slowly as the ethanol content increases.

Bonus points: A pump fuel that has E10 will have slightly different Fuel Density, Fuel Density Temperature Coefficient & Stoich Ratio settings compared to regular pump fuel with no ethanol. Try and find out if what each of these variables are for your given fuel if you want to try and make your modelled fuel equation as "accurate" as possible. 

[INSW20]

Okay, that makes perfect sense.  So this table I have (recently within the past few days) zeroed out as such:

So I'll continue working on getting this dialed in and then plan on draining the tank and filling with E85 (hoping I don't have to order a container of it) and go from there.  Since I have multifuel disabled, will the ECU still use the fuel blend data between gasoline and ethanol?

@Adamw, here's what I ended up with after adjusting my deadtimes table and checking different target lambda settings at constant RPM and load:

Does this seem close enough I hope?  I'm still ending up with some very high VE table values, though.  Could this now be somewhat a function of running E14 with multifuel turned off and having my primary fuel set to gasoline only?  I guess next time I run it I can enable multifuel just to test if it changes my VE map values.

Then finally ran another injector timing test:

The runtime marker is at the richest spot, which was at 330BTDC @ 2000rpm.  Then I ran at 3000rpm and it read richest targeting end of injection at about 280BTDC.  Does that seem reasonable?

Thanks guys!

[INSW20]

Hello all!  I've been making some progress on the car, and it's definitely improving in terms of drivability and power.  I did most of the work with multifuel turned off (with corrected fuel properties for E14).  I zeroed out the table blends so it's 100% on Table 1 for fuel and ignition and everything else I could find.  When I turned multifuel back on (and reset appropriate fuel properties for petrol and ethanol) it seems to run exactly the same lambda values as when MF was turned off.  It still definitely needs more mapping and logging, but I wanted to see if there was anything else glaring that stuck out to everyone/anyone.

It seems much happier with some more timing in it, so I think that's where most of my sluggishness is coming from currently.  I have some detcans (mechanical) that I'm using and I'm not hearing any knock yet, but I'm trying to very carefully ease into it, and likely won't go much farther advanced until I'm on E85.

I'm still needing 90% VE to get my idle portion of the map reasonably close to target.  If there's nothing else out of place, I can scale up the "engine size" to bring it down, but also at this point I'm not maxing out my VE table (yet).  The lean bump during idle at the end of the "drive to home" log looks to me like when I turned off my IC fan and power inverter and load dropped into the -9.4psi cell (which I have since added fuel to).

Attached logs and current cal.  I'll be doing a little more work on petrol and then planning on starting the E85 tune/blend.  That will give me some extra comfort when adjusting the ignition map, and I'll make sure I'm 100% on fuel/ignition table 2 when doing that work.

As always, thanks to everyone for the input!

 

 

[INSW20]

Fixed it!  I'll work on smoothing out the timing table, which I think will get rid of the waviness.  Otherwise it's running well now.  Thanks to everyone for all the help!