I've read through all related threads and have not found a solution to the issue in my car, which has a streetported n/a s5 engine.

Here are my observations:

At full throttle, the power drop occurs at 6300 rpm and continues to 8000 rpm and beyond. Even in neutral, revving the engine at WOT will reproduce this; it's subtle, but perceptible. In 1st gear, it's evident at 6300 rpm, but the car will rev to 8000+ rpm. In 2nd gear, it's very noticeable at 6300 rpm, but the car will eventually rev to 8000+ rpm. In 3rd gear, it's obvious, and there is essentially no power. The car will not get past 7000 rpm. 4th gear and 5th gear -- impractical to test.

At 1/2 throttle, the observations are the same, but the drop occurs later, at approx. 6800 rpm.

At 1/4 to 1/3 throttle in 1st gear (i.e., slowly accelerate in 1st), the car will rev smoothly to 8000+ rpm, without any hesitation or power drop.

The observations above are identical with or without the silencer (restrictor) installed in the exhaust muffler.


Some history:

The car has been in my ownership since June 2007 and I've driven fewer than 2000 miles. The engine is streetported and built by Karack (Ben).

I have been in touch with Ben in the past and he said he has actually been through that engine twice. The first time, he refreshed the engine for a customer, two owners ago. The guy overheated the engine and sold the car. The next owner asked Ben for a rebuild with a mild streetport. I believe Ben used Racing Beat templates for the porting work. When I was test-driving the car, I thought the hesitation would be a simple problem to solve and, at the time, I was more interested in the handling/chassis setup.


The car has been to two good shops in the Bay Area: PR Motorsports and Rotorsport2.

Here is what has been checked. Initially, we proceeded methodically, checking possible failure points that could account for this behavior. Later, we began swapping parts in more distantly related systems.

This list is roughly classified according to purpose of the unit (i.e., it is not arranged in the order in which we tested/swapped parts). None of these had any impact on the hesitation issue. Nothing made it worse and nothing made it better.


Control Systems:

ECU

Tried ECU reset.
Tried 3 other known good ECUs of same type.

MAP sensor.
Tried 2 other known good units.

TPS
Tried 3 other known good units, including a new OEM part. Checked / adjusted each time.

Engine Wiring Harness
Swapped out with known good part. Includes wiring to the injectors. Checked grounds at that time.

VDI
Confirmed correct operation & checked actuator & solenoid
Wired valve in dynamic chamber both open and shut

6PI
Checked 5th & 6th port actuators for correct orientation -- found correct
Wired them open, then shut -- no change

Atmospheric Pressure Sensor
Tried known good unit of same type.

Air Flow Meter
Tried known good unit


Air Intake:

Upper Intake Manifold
Checked operation of flapper valve underneath.
Verified at high RPM, air is delivered to all four chambers (no restriction)

Throttle Body
Checked operation of butterfly valves.
Tried known good unit.


Fuel Delivery:

Fuel rail with pressure regulator

Pickup mesh screen in fuel tank (clean).

Fuel pressure
Well within specs -- at the high end of the scale.
Checked line fuel pressure at idle and under full load

Injectors
Cleaned, flow-tested, and balanced by RC Engineering.
Tried known good n/a injectors.
Tried known good s5 turbo injectors.

Fuel Pump
Replaced with GSS341 Walbro 255 lph High Pressure Fuel Pump
Tested for good input voltage to fuel pump under load

Fuel pump relay
Correct operation at high and low voltage verified

Fuel resistor pack
Swapped with known good.


Ignition:

CAS
Pulled CAS out -- advanced by one tooth, then tried retarding by one tooth.

Spark plugs (new)

Spark plug wires (new) & tried Magnacore

Leading ignition coil
Tried known good unit

Trailing ignition coil
Tried known good unit


Exhaust:

Entire exhaust behind the headers has been replaced with a straight-through non-restrictive exhaust system:
Replaced Bönez high-flow catalytic converter with Bönez Race Pipe
Replaced custom Magnaflow dual exhaust with A'PEXi GT Racing single exhaust

Removed exhaust to ensure no restriction


Other:

Wax pellet (replaced assembly with new)
Oxygen sensor (replaced with new)

Vacuum lines
All rubber hoses replaced with silicone vacuum lines

Disabled emissions
Disabled all emissions system vacuum sources (removed all solenoids, etc.)


1.5 years and $$ later, I'm back to square one. As it stands, the car is sitting at Rotorsport2 in Santa Clara. After speaking to Paul, it seems that they have exhausted virtually all possibilities. There were some things they have solved / corrected along the way, but the 6300 rpm hesitation persists. The most frustrating aspect is that they have not been able to make the problem even slightly better or worse with any of the changes above.

Suggestions?

 

I started having similar problems today but with a nice smoke at 5k+ rpms. Taking a closer look in my driveway my cats (which recently started rattling) were glowing red. I may have to accidentally hit them with a hammer to fix them.

Oh and its not the normal red, there was actually melting bits of tubing. If youve seen molten iron you can understand that when it is hot enough it starts to send off sparks. My exhaust it which I believe has made a new leak for me to fix.

 

^ that'd be the cats clogged...

 

you do know the stock computer goes stupid rich at around 6k or so. Check the rtek section and look at the graphs for the s5 na rtek 2.0. The guy in there got 40+ horsepower on the top end with fuel tuning alone.

The difference between 10.0 afr and 13.5 afr are quite profound. Especially in the case of an N/A motor.

Replacing the fuel pump with the walbro unit probably made it worse because of the increased flow.

edit: you sure your car has a fuel injector resistor pack? It shouldn't because you have an S5 89-91 correct?

 

have u modified ur igniton??? u may need better spark dont quote me im just thinking that may be ur problem u are getting fuel but is it too much or too little of a spark

 

Yea I came to the same conclusion. All i gotta do is figure out how im going to weld the exhaust back together after i cut it apart.....

!!!!!!!!!!!!!!!!!HINT!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
I dont weld!

 

^well just take it to a muffler shop

 

Sounds to me like it's too rich. You should check the mixture with a wideband and tune it to compensate (with something like an RTek 2.0 perhaps). Also consider getting an MSD box for the leading plugs (or some other sort of ignition system to boost the power). When it's really rich, it'll be harder to ignite, so upgrading the ignition can help, but it's probably just pig rich.

 

It does not; that's an irrelevant item. Sorry for the confusion and thanks for noticing the error.


Paul at rotorsport noted that the Walbro is actually putting out less fuel than the stock pump -- I think he said it was pushing 30psi. Looks like I ended up with a bad unit..... I'll go back to a Nippon-Denson pump soon.....


To re-emphasize, there are no cats at this point. Racing pipe & no real muffler.


This is something I'm considering. (The car is sitting in a different shop right now and I'm waiting for their input on this issue.) But I'm wondering if that is necessary (in general) for a conservatively ported motor. The stock ECU map may be unable to get the most from the motor, but would it actually result in a loss of power in this way? The situation I'm trying to avoid is throwing more upgrades / parts at the car, only to find out the issue persists because of something more fundamental.

 

New data from the shop:

At mid RPM, the O2 sensor voltage reads 0.65V.

At 6300 RPM, it drops to 0.04V.



(Calls into question the previous AFR measurements, etc.)

 

never trust the narrowband o2. it's too slow to respond and only works in a very narrow range.

I'd get a real o2 wideband hooked up and see what it does. chances are you are seeing a 10 or lower afr from 6300 up

 

how bout a fuel filter?

 

The smart money's on the AFR's. Get it checked with a wideband (you can probably use the shop's gauge). Without cats you can check at the tailpipe, and many dyno shops will have an attachment for doing just that.

Yes if the stock computer's causing it to go way rich it'll be down on power and wasting gas. Over a long time a good tune (and the associated hardware) might even pay for itself with better mileage, so lower fuel bills.

 

Fuel filter has been replaced (neglected to mention above). Thanks.

 

Update:

When the stock fuel pump was replaced (several months ago), it looks like I ended up swapping the old unit for a busted Walbro. Today, the shop replaced the Walbro with a new Nippon/Denso and the pressure difference is quite dramatic:


With Walbro:
At idle, 25 psi
At WOT, 30 psi


With Nippon/Denso:
At idle, ~30 psi
At WOT, ~40-42 psi


New observations:

While accelerating in 2nd gear at WOT:

At mid RPM, O2 voltage reads 0.94V (up from 0.65)
At 6300 RPM, it drops to 0.85V (up from 0.04V)

There is a pronounced stumble at 6300 RPM, the engine struggles to increase revs. After 7000 RPM, it picks up again and revs to the 8000 buzzer. In the 6300-7000 RPM range, it feels like an ignition stumble, according to the mechanic who test drove the car. So, overall there is a difference compared to the original behavior.

Also:
The idle is slightly higher, and surges less.
Engine doesn't stall as much (used to stall at every stop light/sign).

 

Yesterday I visited the shop and drove the car myself. So, I'll have to correct the mechanic's driving impressions slightly.

First, power is up throughout the entire RPM range. The increase in torque is noticeable and, for the first time, I feel the car accelerates adequately at mid RPMs at WOT, before hitting 6300 RPM. Above that, the power drops, with the car jerking back and forth. However, it "powers through" a bit faster, and reaches 8000+ RPM more easily in 1st and 2nd. In 3rd, it's just as sluggish and I don't think I was able to get it much higher than 7000 RPM. At 1/4th to 1/3rd throttle, it will accelerate smoothly, just as before.

Out of gear, the engine spins up much faster at WOT.

Undoubtedly, these improvements are due the increased (i.e., normal) fuel pressure.

So now I have the same hesitation and power loss, albeit at higher fuel pressure and torque throughout the RPM range.

Next steps:
The shop is going to measure the pulse pattern / width for all four injector outputs on the ECU.

 

and why have you not gotten it tested on a dyno with a wideband? that's the only way you are gonna get to the bottom. some crazy *** injector pulse width testing ain't gonna tell you nothing. put it on a dyno. run it with a wideband and then tell us what happens.

 

Solareon -- I plan to get some dyno runs eventually. But for the time being, I'm letting the shop do things their own way.

Right now, the car is in a new shop (i.e., I have never taken any cars there until now). On some days, I've seen a dozen rx-7s in their parking lot.

I've owned four s5's and an FD during the past two years; that is essentially the extent of my rotary experience. The owner has been working on rx-7's for 25+ years. I'm feeling a little outclassed.

I will let them tool around some more and, who knows, maybe they'll come to the same conclusion and take it to a dyno.....

 

The shop called me today with bad news. They have "tried everything" and there has been no change in the power drop @ 6300 rpm.

I took some notes on what they've done:

Hooked up lab scope
Monitored pulse width of primary & secondary injectors.
Primary pulse -- starts at 2.6 ms, reaches high of 8.4ms @ 6300 and drops down to 5 ms, staying there all the way to redline.
Secondary pulse -- @ 3500, starts at 5ms, slowly climbs to 6ms, jumps to 11ms and @ 6300 drops to 5ms and then to 0 by 7000.

Swapped ECU -- same exact numbers.

Some conclusions:
not a computer problem (tried known good N352, N351 ECUs)
not the injectors
not the fuel supply

** Inputs to the computer are giving an erroneous reading @ 6300 rpm and causing the injector pulse width to drop down. **

Shop said:
at WOT, the s5 ECU does not use MAF signal (only at low and midrange)
at WOT, the s5 ECU uses the built-in fuel map and TPS.

2009-01-22
Checked crank angle sensor inputs -- ok.
Checked coolant temp sensor input -- ok.

** Bottom line:
At 6300+ rpm, there is a lean condition due to incorrect injector pulse width, cause unknown.


If anyone in the SF Bay has insight into this and has time to kill, PM me and I'll bring the car and demonstrate the problem.

 

If you believe that the pulse width is incorrect then you should check everything associated with it.

Start with the injector plugs and work your way back (I assume you have had the injectors checked)

Check for damaged wires including stretched wires. then check the plugs that go into the controller (on the harness side) could one of the leads be maing poor contact.

Next (because you've eliminated the ecu) check ALL the sensor input wires on the plug as well. Anything that may be loose or making poor connection could be a cause.

And just for ***** and giggles check how the wiring is laid out. Remember that other wires can and will induce currents and could consume current due to flux. Avoid signal (sensor) wires being parallel to driven (outputs like the injectors). if they must cross try to do so as close to 90 degrees as possible.

 

wideband verifies it is running lean, at WOT it is 11.5 up to 4k RPMs, 12.5 up to 5k 13.5 up to 6k and 14.5 at 6k+.

i've checked:

wide range TPS
narrow range TPS
IAT
coolant thermosensor
swapped MAFs with an S5 N/A and S5 T2 for kicks
checked voltage at injectors and patched in a wire directly off the battery to the main relay power feed
bypassed the fuel pump relay to bump the fuel pump voltage up from 6.7volts to 10.6 to keep the idle from being so assy
ohm tested the secondary injectors since it only happens after the transition
removed the wide range TPS wire to trick the ECU
added a ground to the pressure sensor to get a better signal from the ECT, IAT and pressure sensor
cleaned and tightened the battery terminals, negative was pretty loose aux negative wire was corroded

the engine now has 2 or more vacuum leaks from the various shops its been to, which aren't doing any favors for the idle and lean condition since it is tricking the MAF to think the engine is needing less fuel since less air is going past the MAF and some past the vacuum leaks under the UIM. unless these leaks in conjunction with several other points are met, i still don't think this would be the cause of the power loss issue.

the main thing i have not checked is the driver's side engine harness which runs both the coil packs and the CAS wiring, the engine harness has been said to have been replaced and from my tests nothing appears wrong with it, the driver's side harness can be getting some noise induction from the high amp alternator or a short/bad connection in the coil wiring harness. this hopefully hasn't been checked for yet, the last possible cause aside from this is one of the unknowns of the ECU misinterpreting a sensor and keeping it out of its high load high RPM open loop state, which sensor it is, i guess i will have to test one by one since the FSM does not give a chart of exactly how the ECU operates at all times. since they all test OK, how the ECU will act when they are removed and if it inputs a default value or goes into limp mode, i guess i will find out.

 

Did you try cleaning the gas take and removing any rust? I noticed it wasn't on your list and that was something that we ran into at the race track.

We changed only the fuel sock initially because of the side to side hesitation. Later in the day the car began hesitating at high rpms, so we opened up the tank and found that the sock was gunked again and noticed that there was rust at the bottom of the tank. We cleaned the tank but also replaced the sock and inline fuel filter because they were victims again of rust and never had the problem again.

Just a thought, hope it helps!

 

This has already been checked, probably more than once. This is not the source of the problem.

As noted above, the ECU is cutting voltage to the injectors. Karack verified the AFR going very lean at high RPM with a wideband sensor early last week.

 

I'm gonna resurrect this because we saw pretty much the exact same behavior out of my FC at the track last week. Car is a '91 NA. Stock motor. Header and race exhaust. Cone filter on the intake. All emissions removed. MSD box on the leading coils.

This was the first track event for the season. Car lost power above about 6k. We tried a little of everything. Bypassed the MSD box. Adjusted the VDI (which is wired in one position). When we pulled the plugs we noticed that they looked a little lean (tannish color on both leading and trailing, running 9's all around).

This sounds just like what you have been seeing.

Two of us were double stinting the car, so on the fourth session out (my second) the engine just magically started working again. The interesting thing is that we changed nothing between that session and previous one. The only thing we did was to add about 6 gal of fresh fuel (w/premix) to the tank. This got me wondering if the whole incident wasn't a case of some bad gas that sat in the tank over the winter. I topped off the car on the way to the track, which at the time I figured should have diluted the old gas enough to run on, but maybe it just wasn't enough.

But the lean condition shown on the plugs still bothers me. It would seem that the my ECU is doing similar things as yours.

So why would the ECU cut fuel at the higher RPMs?

Longshot - a stuck BAC vale?

Hmm, it maybe time for a new tunable ECU.... >:-)

Any thoughts?

-b

 

As a little followup, I was looking around at the Rtek7 ECU upgrades and I noticed that Chris Ludwig posted some stock dyno runs for an S5NA that may appear to have the same affliction.

If you go to

http://www.digitaltuning.com/index.p...age=2&ecu=S5NA

and look at the stock dyno charts (click "show dyno charts"), you see a big drop in hp and torque right around 6300 rpm. Now the funny thing is that there are no blips in the stock AFR chart, showing things running pig rich across the board. He actually has to lean out the fuel mix to get the extra power.

Just a data point. Perhaps we should ask Chris is he has seen anything like this?

-b