VitR — Scraping the Bottom of the Barrel

No the title isn’t a shot at the car — I’m finding it to be a great car and the motor is turning out to be pretty solid. There’s just one thing that’s quite annoying — Honda gimped the potential of the motor with the fuel system. The D/I fuel pump is pegged out very easily once you turn the power up, which in fact leaves you literally scraping the bottom of the barrel to make a clean power curve as you’re riding a very fine line with the fuel system.

This turns out to be even more true once we put on the PRL Motorsports Intercooler on the car and found out it improves cooling efficiency to a point we were making the power we had before while running less boost. Of course we wanted to make more power and run more boost — which proved futile given the limitations we kept hitting in the fuel system.

The car is 100% OEM aside from the PRL Intercooler and we used E35 blend fuel for this test (as we wanted to see how much more we could push the motor).

Intercooler

So what did the results look like? We were able to make a clean power curve and pick up more power before the fuel system maxed out — and make roughly the same top end with less boost.

But that’s not good enough! Let’s make more… and we certainly tried. The motor wants to make power — but takes a slump after 6000 rpm. Why is this?

Well here’s why — in our MoTeC programming we’ve added code that estimates how much injection time you have left before your spark event. You need a certain amount of time left over for a clean mix in the combustion chamber before the fuel is ignited. Do not leave enough time for this to happen the engine doesn’t make power or even worse — misfires. It appears that on this motor we want at least 1.5ms of “Injection Time Remaining” to get a complete mix.

But as fuel demand on the D/I pump goes up it skips a beat and we see a large fuel pressure drop — from a target of 22.5MPa to 16.6MPa (a 6MPa or almost 900psi drop in fuel pressure). As a result of the pressure drop the injector pulse width has to increase and now we’re running out of injection time!

Really Push It

How bad does it get if you REALLY push the D/I fuel system? Well take a look at this. Wow it’s just a roller coaster ride after 5500 rpm — we were making over 390whp and would of made well over 400whp if we had the fuel…

 

You can see how tapped out the D/I fuel system was on this 29psi pull as we dropped 8MPa of fuel pressure.

Injection time remaining was super low… and the motor even misfired at 6600 rpm.

 

 

So Did the Intercooler Help?

Absolutely. Here’s some useful data from actual hard acceleration runs on the stock intercooler and with the PRL intercooler. You should note — we actually used a bit less boost on the stock intercooler runs and the stock intercooler had colder ambient temperatures versus when we ran the PRL intercooler test. However the stock intercooler still heat soaked rapidly. For comparison are runs at WOT from 1st to 3rd gear.

First on the left is the stock intercooler. Note the following — “Airbox Temperature” is the temperature sensor in the intake BEFORE the turbo and the “Inlet Air Temperature” is the temperature sensor in the intake manifold. The sensor in the intake manifold tends to heat soak at idle/low air volume conditions (not much fresh air flowing through the intake manifold).

 

On the right is the run w/ the PRL intercooler. You can note that with the stock intercooler the Inlet temperature never touches the airbox temperature and in fact starts to climb almost immediately after dropping. With the PRL intercooler the aircharge is rapidly cooled and comes down to match the temperature at the airbox — and stays steady all the way through 3rd gear and would *maybe* start to slowly climb in 4th gear.

Clearly the intercooler upgrade is worth it — especially for our road course folks.

Some Fun

After having the car for a while and getting more comfortable with the clutch — I wanted to see if it would hold a bit more torque. I brought in 23-24psi earlier in the powerband, held 24psi through the top end — and made ~390hp and just a hair shy of ~400tq.

Is there more potential there? I believe so — I think we can probably make 450wtq quite easily — if we had reliable fueling. We had a 6MPa pressure drop and the fuel pressure just did not look that great.

 

 

Some Thoughts

Yes — this is E35 fuel (we had made 360whp on 92 previously), but there will be comparable gains from this intercooler on pump gas as well. The E35 actually doesn’t bring up fuel demand *that* much over pump gas (which is E10 to begin with) — it’s about 10% more fuel. So what’s that tell us? Where the hard limit is for fuel volume with this D/I fuel system.

How much can we do on just pump gas? Maybe 400whp before we need to do secondary injection. Maybe less, maybe more. We’ll find out once we have a the PRL downpipe on the VitR!

The other option is to use a race gas that reduces fuel system demand over pump gas….

 

 

VitTuned, CivicX and MoTeC

While this blog may lead you to believe things are quiet, that would be a false assumption! The gears have been turning and I’ve spent a lot of time and had a lot of fun reverse engineering all the various systems on the new 10th Gen Honda Civic platform (“CivicX”) to implement our MoTeC M1 solution. Both on the 1.5 turbo motors and the 2.0 Type R turbo engines.

While the process to reverse engineer not only the Honda CAN (which we have plenty of experience with from previous platforms) but now in turn a whole new engine and direct injection (D/I) fuel system was involved — it was also quite fun. This involved using some advanced tools and scoping the various systems — D/I pump, D/I injectors and their relation to the cam and crank triggers on the vehicle.

So while it was a TON of work, a lot was learned in the process and we gathered a literal treasure trove of data and insight into the platform that no one else has or can provide any kind of insight on.

The Basics

So what did we need to figure out to run this motor?

  • Engine Trigger pattern. Scoped and submitted to be added to the system.
  • D/I Fuel pump ECU control strategy.
  • D/I cam lobe delivery angle data.
  • D/I Injector ECU control strategy.
  • Bypass valve control strategy (the “BOV” on this car isn’t just a BOV — the ECU tells it when to vent and when not to vent).
  • Turbocharger wastegate control strategy.

Obviously there’s a bit more to it in the tuning, but these are the critical systems we had to look at to fire up and run and then tune the engine.

Injectors

The injectors were actually fairly straightforward — I grabbed some data using our Rigol oscilloscope and from the wave forms you can extrapolate the data needed to run the injectors precisely.

Minimal amount of snafus along the way.

 

 

D/I Fuel Pump

Now this was more involved — we started with scope data as well, but there is a lot involved in running the D/I pump correctly. You have to determine WHEN to actuate the solenoid in the pump relative to crank position as well as determine how long to pulse it, at what angle to pulse it and what angle to hold it open to.

Get something wrong — and you don’t make any fuel pressure, or you get sporadic fuel pressure.

 

Boost Servo & “BOV”

Traditional boost control is done using a boost solenoid — a simple valve. Honda implemented an internal wastegate in their turbochargers using a servo to open and close the wastegate. This is actually superior to using a solenoid as you can have PRECISE control of your wastegate, which allows for much more predictable boost control — with the correct software implementation. What Honda did in the stock ecu is actually pretty lame, and obviously it was done so with their own goals in mind.

But we don’t care about any of that — we implemented our own strategy for controlling the wastegate (“Boost Servo”) that literally allows you to snap to your boost target as fast as the turbo can do it and just SIT on that target without any fuss.

The BOV is also run by the ECU — this is so that on overboost situations the ECU can “vent” (recirculate) boost to get back to target. We implement an advanced strategy for running the bypass valve as well — with the ability to not only vent on overboost (if the situation arise) but during rapid throttle changes as well as complete lift off.

Honda CAN

There’s all kinds of useful goodies here!

 

 

 

Traction Control

Of course everyone’s biggest complaint is how intrusive and difficult to disable the factory VSA/TC is. This is no longer a problem — this system does not interfere with driver input and we have our own traction control strategy that is much more advanced and fully tune-able in place.  You don’t even have to push the VSA/TC button anymore. It just works.

You can chose to adjust ignition timing, boost, throttle position, cyclic ignition cut or cyclic fuel cut in any fashion you chose. And as you can see, it works quite well!

Knock System

Protecting the engine is key. The knock system we have in place is based on a strategy we use on other vehicles such as the 370Z — not only can the ECU do “live” corrections for detected knock, the ECU will also “learn” and adapt to make broad timing corrections as you chose in case you run into poor fuel or just an unexpected situation.

But how does the ECU detect knock? We’re using the OEM Honda knock sensor, and like any knock sensor — it is essentially a microphone. The DSP logic in the ECU is then programmed to listen for “noise” at a specific frequency during a specific time. The ECU is then “tuned” with an acceptable noise threshold and anything over this threshold is “knock” that the ECU must then make a decision on whether to react to it or not. Honda did this by implementing a moving target called “Knock Control”. Some so called “pros” have made comments about this system and how it functions — demonstrating their absolute ignorance. We’ve dug through the stock ECU in depth and seen how this “Knock Control” value moves — in several cases it will move without actually seeing any knock.

Why? Because protecting the engine is always key. It’s always better to err on the side of caution and reduce power than to ruin a motor. I don’t disagree with this kind of strategy from the OEM’s perspective — their goal is to produce a reliable car that will run for hundreds of thousands of miles.

But, let’s show you some examples of what “might” look like knock. Here’s a couple screen shots of datalogs recorded using knock detection on not just ONE frequency, but FOUR frequencies. You

can see the noise levels and the threshold I’ve tuned for each cylinder.

You can spy that in the screenshots there are HUGE spikes past the threshold. Man this thing’s about to eat itself alive.

Wrong.

In the first situation the “knock” actually happened during anti-lag when a spark cut was active. Just so happened the pops and bangs produced the right frequency at the right time through the engine/engine bay to trigger the knock sensor (“microphone”).

In the second situation… well see for yourself. The car was in first gear and very actively on the traction control — which caught a bit of wheel hop, and you guessed it, just so happened to make the right amount of noise at the right time.

I also have knock plots of REAL audible knock. They actually look nothing like these examples, I’ll be keeping them to myself for now.

But to summarize — the ECU has to make decisions… so, once again, it’s better to err on the side of caution. But just because someone stuck “pro” to their screen-name, doesn’t mean they know a damn thing or understand what the ECU is doing in the background in it’s glorious totality.

Data Data Data

In this case a picture is worth a thousand words, for sure. The treasure trove of data we’ve recorded is priceless. At any time the engine is running, we’re recording over 300 channels of data, with even more available if we need anything on those other systems.

This system also helps facilitate in rapid development and testing that is unparalleled. As we develop  and test parts — we can put them on and have them tuned with every data point imaginable filed away for future reference. Because we’re working on the same car/engine, the things we learn translate over to tuning we do for the platform using the stock ECU — and our recommendation is getting yourself a KTuner unit if you cannot afford our MoTeC programming. They’ve been awesome about listening to input and providing rapid development for features and products we’ve been developing and testing!

Type R

Is next. Stay tuned for another update.