Underboosting - Only hitting Wastegate Pressure

[Reedy]

As the title says, my car has recently decided it doesn't want to go fast anymore and is only boosting to about 11psi.

Turbo is a BNR Stage 1 that I regret buying every single day but it's what i've got. I am still using the OEM Boost Control Solenoid.

I also have a TMIC, DP and intake but that shit isn't so relevant as the car has been running without any issues since my engine rebuild.

Typical Boost is about 19-20 PSI, the boost curve is much the same as before, it just tops out at 11psi and the car can't even spin the wheels in 2nd anymore, it might as well be a prius.

I have done some poking around in the engine bay and making sure the magic pixies are flowing through the relevant electronics, I did continuity tests on the wiring harness to the EBCS and some crude PWM testing with my multimeter and the signals all seem to be working as expected without having an osmelloscope to play with.

The EBCS also actuates when I apply a 12V DC signal, and it holds pressure and vents as expected when the 12V signal is applied. This doesn't however tell us how it behaves with a PWM WGDC signal.

I have no check engine lights, and no engine codes appear using my OBD2 scanner and Dashcommand on my iPhone.

The only things I can think of is that:
A) The Boost Control Solenoid is latching closed which is basically just forcing it to run spring pressure, but the fact the solenoid appears to be working and the signals also seem ok.
B) There is an engine code present that my simple tool doesn't see that is causing the car to run in some kind of limp mode.

But something still makes me think I might be barking up the wrong tree.

I have a new EBCS that is due to arrive tomorrow that I will install as my one is 6 years old and has done some serious miles, but in the event that it does not fix this problem I don't know what else it could be.

Please school me oh internet!

[speed23]

If the vacuum line from the TIP is blocked/clogged it will do this. When I bought my car, it was running only 11ish PSI and I discovered the TIP nipple had been broken and "repaired", but the repair limited the airflow severely. New TIP fixed the issue.

[Jeff23spl]

is the boost start as low as before but peak at 11 or it makes more lag too? If it is as fast as before, you need to check your ebcs and vac tubes. If it lag more, you may check for a leak or something blocking the intake path or a bad BOV etc. You can also check engine health but despite it would take a huge damage to to the block to reduce boost that low...
Check for wastegate mecanical issue too...it would increase lag and reduce boost...

You need to datalog and check if there isn't anything wrong and a self protection lower the WG duty to avoid engine damage...

[GroceryGtr]

if you've been running that setup without issues prior then I'd say either a line came loose or is cracked OR its time to upgrade to a proper EBC and run it in interrupt mode. Its possible your wasetgate arm came loose, but unlikely. Give its a yank it shouldn't move very easily.

[Jeff23spl]

Wastegate arm may not be that easy to move, it depend on the spring rate....mine is 14psi and it is very hard to move even with the turbo in hands..

[Reedy]

Originally Posted by speed23 If the vacuum line from the TIP is blocked/clogged it will do this. When I bought my car, it was running only 11ish PSI and I discovered the TIP nipple had been broken and "repaired", but the repair limited the airflow severely. New TIP fixed the issue.

I'll be pulling the tip out to replace the boost controller anyway sow I'll make sure I check this thoroughly. Good tip thanks!

[Reedy]

Originally Posted by GroceryGtr if you've been running that setup without issues prior then I'd say either a line came loose or is cracked OR its time to upgrade to a proper EBC and run it in interrupt mode. Its possible your wasetgate arm came loose, but unlikely. Give its a yank it shouldn't move very easily.

My Wastegate arm had a problem a while ago but was fixed and re welded. That was causing massive turbo lag at the time but has been sorted out. Now the responsiveness is there, just at a lower boost level.

I used a 3 port EBC previously but capping off the turbo housing and second wastegate port was a messy solution that I didn't really trust. The 3 port made no difference to performance at the boost levels I am running so I decided to simplify.

[sharksinspace]

I, and others, run upgraded EBCS in standard two-port mode just to have a more durable and higher quality piece, FWIW.

[Reedy]

Originally Posted by sharksinspace I, and others, run upgraded EBCS in standard two-port mode just to have a more durable and higher quality piece, FWIW.

Next time I get the car tuned I'll probably do this.

[sharksinspace]

If you set it up in 2-port, you don't even need to tune for it; that's only if you run all 3-ports. I upgraded mine when I did an initial TIP install...grabbed a Grimmspeed for about $100 shipped and just blocked off the third port of the EBCS. Much more durable construction than the OEM plastic piece, and my understanding is that it has faster response time, as well. JBR or Cobb's EBCS are comparable, and CS also offers one that is a slightly different design.

[GroceryGtr]

Originally Posted by Reedy My Wastegate arm had a problem a while ago but was fixed and re welded. That was causing massive turbo lag at the time but has been sorted out. Now the responsiveness is there, just at a lower boost level. I used a 3 port EBC previously but capping off the turbo housing and second wastegate port was a messy solution that I didn't really trust. The 3 port made no difference to performance at the boost levels I am running so I decided to simplify.

3 port mode will give you slightly quicker spool times(it holds the wg shut as opposed to bleeding/cracking it open), but you are correct that at lower/stock boost levels it doesn't really make a noticeable difference. Between not having to retune and the negligible performance gains I would agree that at stockish boost levels 2-port is the better choice. As other have pointed out an aftermarket EBC is still a good idea as they generally have a higher volume chamber allowing for slightly better control even in bleed mode.

[Jeff23spl]

3 ports interrup mode unload all the back pressure on the wastegate actuator to give full spring pressure resistance to exaust flow instead of cracking just a part of it to trigger the boost against spring resistance...But this won't make the k04 flow more at high rpm...

When you talk about work on your wastegate arm, what was it exactly ? are you running IWG ?

[sheston]

I've found this ^^ to help my K04 hold more boost. The stocker may have been acting up as well. I picked up around 4 psi at redline with the 3 port upgrade.

[Fstrnyou]

Originally Posted by sheston I've found this ^^ to help my K04 hold more boost. The stocker may have been acting up as well. I picked up around 4 psi at redline with the 3 port upgrade.

Hmm...good to know. My boost tapers off after 5-5.5k...not by choice like some tunes do. 100% WGDC with OE 2-port EBCS just doesn't cut it. Stock intake/exhaust isn't helping either.

OP, to check limp mode, log requested load vs actual load.

[Reedy]

Ok some developments:

I replaced the boost controller with a new one from Mazda.

Also checked return lines and TIP for blockages and everything is fine.

Aaaaannndd...... The problem still persists.

I will see what kind of load logging I can do with Dashcommand, since I got rid of my AP and went back to the ECUTEK tune I haven't been able to do real datalogging which is a bigtime bummer. I was going to buy the ECUTEK cable and software anyway so I might grab that asap so I can go back to real datalogging capability.

Next option will be to check with a Workshop spec OBD scan tool like a Snap-On one. I have booked the car in for a dyno health check and touch up tune but would like to nail down this problem before we need to go and spend valuable dyno time chasing gremlins.

Originally Posted by Jeff23spl When you talk about work on your wastegate arm, what was it exactly ? are you running IWG ?

The weld had failed on the little pin that the wastegate arm attaches to, and was flopping around. This meant there was no preload on the arm and the WG never closed fully. End result was extremely bad turbo lag.

Last time we had the turbo out of the car we had it welded back up and fixed.

[Jeff23spl]

Originally Posted by sheston I've found this ^^ to help my K04 hold more boost. The stocker may have been acting up as well. I picked up around 4 psi at redline with the 3 port upgrade.

interresting, what was the before and after numbers psi and g/s? This would mean that even with a 100% WG duty and 2port valve, the wastegate can still open under exaust pressure...someone may just play with restrictor pills to get the same with 2ports
Will it last for long pushed that far? i was on the impression it was already maxed out on turbine speed creating higher BATS...

[Reedy]

Originally Posted by Fstrnyou Hmm...good to know. My boost tapers off after 5-5.5k...not by choice like some tunes do. 100% WGDC with OE 2-port EBCS just doesn't cut it. Stock intake/exhaust isn't helping either. OP, to check limp mode, log requested load vs actual load.

Disregard last post, I have spent some serious time looking at datalogs now. I picked up the ECUTEK cable and software so I am able to log everything really nicely now.

[Reedy]

Latest developments,

OK, Looking at the ECU with a snap on scan tool reveals some codes that other readers can't see.


We are able to recreate these codes by unplugging and shorting the EBCS respectively, clearing the codes does not fix the underboosting issue that I am having.

Next, for some ECUTEK datalogging, to start with I re-flashed my existing tune back into the ECU to see if this would have an effect, this did not change anything.


ECUTEK software is very nice for viewing log data. The main downside is you need a laptop in the car to do datalogging with.



I have overlaid these logs with some old, healthy logs that I did a while back when I was still running my Accessport and you can see the WGDC and Engine Load values are acting very strange.





So where to from here? Well here is some information on the history of my car that I haven't mentioned so far.

1. My Turbo to DP gasket failed at the racetrack a few months ago. This vented hot gases right onto my O2 sensor leads, which melted the plugs and caused a dead short between cores. The car still ran for a while and we fixed this by splicing the wiring harness and soldering new plugs in. One theory we have is that the prolonged short might have scrambled the ECU's brains a little.

2. I have had a P0134 (HO2S-11) no activity detected fault code appear intermittently since we repaired the O2 sensors. AFR and O2 sensors seem to be working fine 99% of the time and this code has only appeared twice in the last few months.

We also spent a lot of time doing continuity tests on all earths and all related circuits and everything seems to be working just fine. Voltages are correct and I have no concerns about earths and continuity in the wiring harness.

Everything We are seeing points towards the ECU needing to be replaced, but it's a lot of money to throw at a problem that might not be a 100% guaranteed fix.

What do you think wise old Internet? Should I pull the trigger on a new ECU or is there something obvious that I am still overlooking?

[speed23]

Now that you have codes, you could pull up the shop manual thats somewhere on here and follow the guide to ensure you're not missing anything before replacing the ECU?

[Reedy]

Originally Posted by speed23 Now that you have codes, you could pull up the shop manual thats somewhere on here and follow the guide to ensure you're not missing anything before replacing the ECU?

Done that, that's what we used to step through all of the continuity testing on the wiring harness and earthing checks. The codes also don't reoccur, they only do when we disconnect or short circuit the EBCS deliberately.

[sheston]

I'm assuming you reloaded your map to see if that helped?

I think I'd mess around with the wastegate (maybe there is something wrong still with the actuator) before spending big bucks on a new ECU

[Reedy]

Originally Posted by sheston I'm assuming you reloaded your map to see if that helped? I think I'd mess around with the wastegate (maybe there is something wrong still with the actuator) before spending big bucks on a new ECU

Reflashing the tune did not help

I've done as much poking around with the wga as I can without removing the turbo, pressure tests to actuate it etc & it's working just fine.

There's a problem with the pwm signal.

[Jeff23spl]

with a Boost solenoid CEL and turbo not boosting, you should investigate there.

Check Boost solenoid wiring from the ECM to the EBCS. You can start by checking coil resistance at the solenoid itself and check if you get the same resistance at ecm wiring harness disconnected.

There is also just 2 connectors in this car that are identical and in the same region, one for the EBCS and 1 to be a filtering capacitor. If you reverse them, you get the problem you have now. The capacitor is just 1 wire but the EBCS should be 2 wires...IT is probably not that but i made that mistake once....Assuming all the oem connector were not compatible with others...

[Reedy]

Done all of that. The pwm signal drops out when the ebc is connected.

[Jeff23spl]

is the valve resistance in the specs ? Can you operate it by putting a constant 12v on it and check for the spool and open the port ?

Did you check if it is the positive side that vanish or if it is the negative ? Positive is from a constant 12v and the negative is pulsed from the PCM. You may have a resistance somewhere that take all the voltage when you put load on it...

[Reedy]

Well we pulled the trigger on a PCM replacement and what should have been a 10 minute job at Mazda on a Saturday turned into the immobilizer fucking everything up and needing to leave the car at the dealer until Monday when the guy who knows how to use the software gets in. It just doesn't end.

If only I had found this thread the day before, but silly me thought the Mazda guys actually knew how to do an ECU replacement.

http://www.mazdaspeedforum.org/forum/foru...acement-85575/

Car alarms and engine immobilizers have caused more pain and suffering than they will ever prevent. That much I am sure.

[Reedy]

Ok back on track.

The ECU replacement finally got sorted when Mazda technical got involved - As it turns out their workshop procedure in the software is incorrect for smart key cars so keep an eye out for that in the future. So next time you go in for a 10 minute job, expect them to take 5 days.

In addition to the immobilizer finally being sorted - The ECU replacement seems to have fixed the problem. It's boosting to 17psi now so everything should be back to normal once we get the correct tune loaded back in, which I will sort out tomorrow.

Thank fucking christ for that. a new ECU costs an arm and a dick so I woulda been pissed if the fix didn't take.

[speed23]

Glad you're back up and running!

[Jeff23spl]

how much was the pcm?
Have you try to open the old one ?

[Reedy]

Over $2k Australian. Don't forget I have to pay the Australia tax on parts, so when I say it cost an arm and a dick I am not kidding.

The old ecu is gonna be used as a diagnostic spare so we're probably not gonna open it up.