K04 TURBO REPORT FROM NEUSPEED

[enganear]

The information below, gleaned from a B5 site, offers some insight into the limitations of the K04 turbo that I have not seen mentioned before. High temperatures and softening of the shaft could also be the root cause of some of the reported turbo problems and also may explain why Mazda intentionally pulls the plug on boost at 6k rpms. Is anyone measuring exhaust temp in the turbine housing?
-enganear

K04 TURBOCHARGER REPORT FROM NEUSPEED

For those readers unfamiliar with the background of the 3K-Warner K04 turbocharger here is a brief summary of past events. The K04 turbocharger (manufactured by 3K-Warner GmbH) is sold as an upgrade for 1.8T engines by several international tuners, including NEUSPEED. In the United States, NEUSPEED has sold the largest number of K04 turbochargers in the aftermarket for use on 1.8T engines. Over the past two years, we documented 8 K04 turbocharger failures and reported this information to 3K-Warner�s United States importer, Dura Products Corporation in Bradley, Illinois.
After validating the statistical significance of the failure rate, NEUSPEED suspended its sales of the K04 turbocharger, and with the assistance of Dura Products Corporation performed extensive testing to determine the cause of the failures. Based on our joint findings, we prepared this report to brief current K04 users, and potential users, on the proper installation, operation and maintenance of the K04 turbocharger to maximize its service life.

1. INSTALLATION
Each re-seller of the K04 turbocharger is responsible for providing installation instructions. The following information is intended to supplement those instructions, not replace them.

A. Always drain engine oil and remove old oil filter prior to removing existing turbo from the vehicle.

B. Always inspect factory oil feed line to determine if line has become coked with oil residue. If the oil feed line has dark, hardened coating inside of it, discard the line and replace it with a new Audi or Volkswagen genuine part.

C. After the K04 turbocharger has been bolted onto the manifold and oil feed and return lines have been reconnected, install new oil filter and use synthetic engine oil only. Mobil 1 and Redline are the recommended brands. 20/50w is recommend for summer. 10/40w is recommended for winter.

D. Before restarting the engine for the first time, the engine, oil lines and turbocharger must be dry-primed with oil. To perform this procedure, first locate the Engine Control Computer (ECU) and disconnect the multipin connector from the ECU. Next turn the ignition to the RUN position and crank the starter motor in several 4-5 second bursts. (Your oil pump will circulate oil, however no spark or fuel will be introduced into the cylinder.) Next turn the ignition key to the OFF/LOCK position and reconnect the multipin connector to the ECU. Next turn the ignition key to the ON/POWER position, but do not start the vehicle. Leave this key in this position for at least 30 seconds. (This sends power to the ECU, but the engine will not be running.) Finally turn the key to the START position to start the engine. Allow the vehicle to idle for several minutes, then shut down the engine and check the oil level. Top off if necessary.

2. OPERATION

The K04 is designed to provide reliable, long-term service as long as its performance parameters are not exceeded. Driving the car at engine speeds above 5800 rpm with 15spi or greater boost significantly increases the exhaust temperature measured inside the K04 turbine housing. This signals that the maximum efficiency of the turbocharger has been exceeded, and that the energy produced by the turbine housing is now converting to heat rather than accelerating the compressor wheel faster. Unless a careful and extended idle-down is performed at the end of every hard driving session, the rapid heat build-up from high boost/high rpm operation causes the turbine shaft to soften, and allows the inconel turbine head to droop. The result is an imbalance that ultimately leads to a shaft failure.



A. NEUSPEED has revised its ECU programming to reduce K04 boost pressure above 5800rpm. Current K04 users are encouraged to contact their ECU software supplier and request this change. Existing NEUSPEED K04 customers will be offered this software upgrade at no charge. Peak horsepower, measured at 5700 rpm, is unaffected by this change. The reduction in high-rpm boost pressure lowers the peak turbine exhaust temperature to approximately 875�C. As long as peak turbine exhaust temperatures do not exceed 875�C, the K04 turbocharger should continue to operate reliably.



B. We encourage K04 users to install an exhaust gas temperature probe directly in the turbocharger housing to monitor peak temperature during operation. Do not mount the probe in the exhaust manifold. Do not mount the probe in the exhaust downpipe. The measurements obtained in these two locations cannot be compared to the 875�C critical temperature.



We recommend using the HKS EGT gauge and thermocouple, it reads from 500�C-1200�C. It is easily purchased from any HKS distributor and incorporates a Peak Hold Warning feature. To install the HKS thermocouple, you will need to remove your turbocharger and drill a 5/16� hole into the turbine side housing. Use a 1/8� pipe tap and cutting oil to tap threads into the hole. (We used Sears Craftman tap #9-54531) Thoroughly clean the housing and hole of all cutting debris, coat the thermocouple threads with anti-seize compound, and install the thermocouple.



C. K04 users should always bring their vehicle to a complete stop and allow the engine to idle prior to shutting down the engine. Recommended idle times vary from 1 minute after mild driving, to 5 minutes after aggressive driving. The idle-down procedure circulates fresh oil and coolant through the turbocharger, allowing it to cool gradually, prior to shut-down.



D. NEUSPEED also now offers a 5 bar fuel pressure regulator to be used in conjunction with the factory 235 CC fuel per minute injectors. The 5 bar regulator does not materially affect the turbine housing temperature. However, our testing showed that it does provides a more consistant Air/Fuel ratio during high rpm operation.

3. MAINTENANCE

A. Engine oil and filter should be replaced every 3000 miles. Do not change engine oil without installing a new oil filter. Check your engine oil level frequently. The turbocharger shaft spins over 10 times faster than your engine�s crankshaft, so an adequate oil supply is critical.

B. Periodically inspect the turbocharger to determine if the wastegate rod and hardware has been bent or damaged. The K04 wastegate rod comes from the 3K- Warner factory with a locking clip over the adjustment nuts. DO NOT attempt to recalibrate the wastegate by moving the adjustment nuts. Any tampering with the wastegate can dramatically alter the boost characteric of the turbocharger and lead to serious engine damage.
C. In the event that any air intake hoses are removed or disconnected during service, thoroughly inspect every hose and remove any foreign objects or debris that may have fallen or collect inside the hose(s). Any loose objects
inside the air intake tract may be sucked into the turbocharger upon start up, seriously damaging the compressor blades.

[Mr._Pickles]

There's several different K04s. The K04-001, K04-021, K04-022 and I think there was a 023. I don't know how this factors in.

[NYpest]

WOW great work !!!

[Haltech]

Good work Enganear!

[enganear]

I was watching my catalyst temp earlier today and it was routinely exceeding 820C. I am not sure how this relates to temperature in the exhaust turbine housing. The exhaust entering the catalyst should be cooler than the exhaust in the housing, but there is an exothermic reaction that occurs in the converter and I am not sure where the temperature is measured. If it was not such a pain in the ass, I would install a thermocouple in the housing as suggested. If I ever remove the turbo, I will definitely do this before re-installing.
-enganear

[Haltech]

Originally Posted by enganear I was watching my catalyst temp earlier today and it was routinely exceeding 820C. I am not sure how this relates to temperature in the exhaust turbine housing. The exhaust entering the catalyst should be cooler than the exhaust in the housing, but there is an exothermic reaction that occurs in the converter and I am not sure where the temperature is measured. If it was not such a pain in the ass, I would install a thermocouple in the housing as suggested. If I ever remove the turbo, I will definitely do this before re-installing. -enganear

i thought the purpose of the catalyst was to heat up the unburned fuel, so the temp in that system is going to be higher since its a ceramic material?

[enganear]

Originally Posted by Haltech i thought the purpose of the catalyst was to heat up the unburned fuel, so the temp in that system is going to be higher since its a ceramic material?

Yes, the converter is its own heat source and I would guess the sensor is somewhere near the center or end of the converter body. The temp at this point should be unrelated to the temp in the turbine housing. I was just surprised to see the temp so high under light loading.
-enganear

[Haltech]

Are you using the DH to measure this temp? If so, i would say that PID is off.

[enganear]

Originally Posted by Haltech Are you using the DH to measure this temp? If so, i would say that PID is off.

I was using my Auterra Dash Dyno. What are people seeing for catalyst temps with the DH?
-enganear

[Haltech]

Originally Posted by enganear I was using my Auterra Dash Dyno. What are people seeing for catalyst temps with the DH? -enganear

Oh yes, i forgot you had one of those. Temps vary, from 600s to 900's. But the second you shut the car down, let it sit for a second, than turn it back on, the temp was higher than when it was running.

[digitaljedi]

Great work, this is useful information, I will review it in depth tonight with my mechanic (roomate, lol). Thanks. I really am enjoying all the usefull information on this forum, I have learned a great deal about our car.

[TurboWagon]

The K04 is designed to provide reliable, long-term service as long as its performance parameters are not exceeded. Driving the car at engine speeds above 5800 rpm with 15spi or greater boost significantly increases the exhaust temperature measured inside the K04 turbine housing. This signals that the maximum efficiency of the turbocharger has been exceeded, and that the energy produced by the turbine housing is now converting to heat rather than accelerating the compressor wheel faster. Unless a careful and extended idle-down is performed at the end of every hard driving session, the rapid heat build-up from high boost/high rpm operation causes the turbine shaft to soften, and allows the inconel turbine head to droop. The result is an imbalance that ultimately leads to a shaft failure. K04 users should always bring their vehicle to a complete stop and allow the engine to idle prior to shutting down the engine. Recommended idle times vary from 1 minute after mild driving, to 5 minutes after aggressive driving. The idle-down procedure circulates fresh oil and coolant through the turbocharger, allowing it to cool gradually, prior to shut-down.

I am usually pretty good about letting the car idle, even after regular driving. This is something everyone should practice! Those 2 sections are "the money". I would assume using synthetic also wouldn't be a bad idea.

Thanks for the info! I had a good amount of Neuspeed products on my GTI and all were great.

[Crossbow]

I'd have to disagree with the extended idling. It's a water cooled turbo. Even with the engine shut off, coolant is still going to circulate through the turbo...the whole water cooled turbo system was developed to prevent turbo turds, or baking of the oil in the turbocharger.

I'm not saying its bad to let the car idle for a bit...but excessively is a bit silly, unless you just finished a 25 minute track session. I think most of us are smart enough to not shut the car off in the middle of a 6,000 rpm 5th gear run.

TurboByGarrett.com - Turbo Tech101

The intake and exhaust plumbing often receives the focus leaving the oil and water plumbing neglected. Garrett ball bearing turbochargers require less oil than journal bearing turbos. Therefore an oil inlet restrictor is recommended if you have oil pressure over about 60 psig. The oil outlet should be plumbed to the oil pan above the oil level (for wet sump systems). Since the oil drain is gravity fed, it is important that the oil outlet points downward, and that the drain tube does not become horizontal or go “uphill” at any point. Following a hot shutdown of a turbocharger, heat soak begins. This means that the heat in the head, exhaust manifold, and turbine housing finds it way to the turbo’s center housing, raising its temperature. These extreme temperatures in the center housing can result in oil coking. To minimize the effects of heat soak-back, water-cooled center housings were introduced. These use coolant from the engine to act as a heat sink after engine shutdown, preventing the oil from coking. The water lines utilize a thermal siphon effect to reduce the peak heat soak-back temperature after key-off. The layout of the pipes should minimize peaks and troughs with the (cool) water inlet on the low side. To help this along, it is advantageous to tilt the turbocharger about 25° about the axis of shaft rotation.

Oh and here's a compressor map to add to the discussion.

MS6 K04-2280 compressor plot

[gsrtype1]

good write up... I am glad to say I obsessivly ALWAYS let my car idle a while before i shut her off, though its prob not needed unless your on the track or a long highway run. The car manuel sais let the car idle for a while after a long highway trip or going up a long hill.

[AutoXRacer]

Originally Posted by enganear The information below, gleaned from a B5 site, offers some insight into the limitations of the K04 turbo that I have not seen mentioned before. High temperatures and softening of the shaft could also be the root cause of some of the reported turbo problems and also may explain why Mazda intentionally pulls the plug on boost at 6k rpms. Is anyone measuring exhaust temp in the turbine housing? -enganear K04 TURBOCHARGER REPORT FROM NEUSPEED Unless a careful and extended idle-down is performed at the end of every hard driving session, the rapid heat build-up from high boost/high rpm operation causes the turbine shaft to soften, and allows the inconel turbine head to droop. The result is an imbalance that ultimately leads to a shaft failure.

So how this affect PG's reworked turbo...
They replace the turbine shaft with a light-weight one...though I don't know what material its made from... Are there any concerns there...

So no matter if its the stocker or reworked, you'll still be limited to 6K RPM range with this turbo...

I wonder how the reworked compares to the stocker in terms of failure, reliability, and efficiency...

[ispypsi]

same argument i have with people when i call turbo timers 'rice mods'....water cooled turbochargers aren't as fickle as people make them out to be. i agree with crossbow.

but nice to see some info on it. if some of it is true about the shaft and such, that kinda sucks. looks like alot of people will be forced to upgrade sooner or later.

[ptperformance]

Originally Posted by enganear Yes, the converter is its own heat source and I would guess the sensor is somewhere near the center or end of the converter body. The temp at this point should be unrelated to the temp in the turbine housing. I was just surprised to see the temp so high under light loading. -enganear

Coverter temp are high both before and after the converter. You will get readings that are higher 6 inches in front of the converter and 8 inches after. The pre inlet temps are elevated from radiant heat reflecting back up the pipe. I did a whole class on this crap and if you get under a car at idle with a good temp gun you can see what I am getting at.

You are also getting calculated temps, not actual temps. There is no temp sensor in our cats, its a guess at best.

[ptperformance]

Originally Posted by AutoXRacer So how this affect PG's reworked turbo... They replace the turbine shaft with a light-weight one...though I don't know what material its made from... Are there any concerns there... So no matter if its the stocker or reworked, you'll still be limited to 6K RPM range with this turbo... I wonder how the reworked compares to the stocker in terms of failure, reliability, and efficiency...

If the did there home work, the modded stockers will work better.

No reason for the LW shafts because of the clipping taking away so much material (if they clip theres).

That is still TBD, no one has been able to run the turbo out past that due to the throttle closing.

Compairison from us will be coming soon. Keep an eye out in our section.

[ptperformance]

Originally Posted by Mr._Pickles There's several different K04s. The K04-001, K04-021, K04-022 and I think there was a 023. I don't know how this factors in.

Wondering the same thing. Hell there have been KO-4 ran clear back into the 80's. Why all the sudden is there an issue? Or is it just our car that has these issues?

[Ziggo]

I wish I could find a definitive turbo map for our turbo. I don't think that one that everyone likes to post up is right. We would be getting surge at 3000RPM if it was right, or I am reading it wrong.

[ptperformance]

Originally Posted by Ziggo I wish I could find a definitive turbo map for our turbo. I don't think that one that everyone likes to post up is right. We would be getting surge at 3000RPM if it was right, or I am reading it wrong.

It will never happen unless someone wants to spend the money to make it happen. Shoot me $10k and I will get it for you. Thats how much they wanted (turbo shop in Cali) for getting compressor maps layed out for a turbo.

What someone needs to do (I have but not releasing it) is get before and after turbine housing pressures. This will give you an idea for pressure ratios from turbine to compressor. Then you can calcualte off the inducer and exducer of larger compressor wheels to find the correct clip and relative (calculated) pressures. This is where the homework needs to be done to find the failure points in our turbo. I would do it but I am having enough issues with finding cores for the big wheel units that I don't want to buy new units to do more testing. The dealer has them on back order last time I checked anyways.

[Ziggo]

Thats ridiculous. I could probably build everything I needed to create a good test setup for less than 5k.

The choke point isn't too terribly difficult to figure out using theory and I wouldn't expect it to be too far off. The Surge really has to be tested though.

I have never heard of a turbo running out of steam at the top end like these are claimed to due to efficiency problems. Reaching the choke point sure, but never hitting a wall like this due to efficiency.

[ptperformance]

Originally Posted by Ziggo Thats ridiculous. I could probably build everything I needed to create a good test setup for less than 5k. The choke point isn't too terribly difficult to figure out using theory and I wouldn't expect it to be too far off. The Surge really has to be tested though. I have never heard of a turbo running out of steam at the top end like these are claimed to due to efficiency problems. Reaching the choke point sure, but never hitting a wall like this due to efficiency.

Its the enviromentally controlled room that costs a lot. The compressors/turbos have to be tested in a control room. That rooms costs $1000 a day to keep that way, no matter what the outside temps are, baro, ect ect that room is always the same no matter what.

Take a look at the pictures in my section on the turbo. The exhaust wheel is damn near the same size as the compressor wheel. Surge is going to be something you figure out on the exhaust side, clip to much and you have lag, not enough and you end up with surge. Its a balance that needs to be found and the only way... pressure testing both before and after turbine wheel. It sucks but someone has to do it. 3 units so far junked for our testing, number 4 is on the way.

[cld12pk2go]

Very good info.

No doubt that removing exhaust restrictions (before, in, and after the turbo) will help bring these temps down.

Also, I hadn't considered that it is possible we are hitting choke flow limits in the turbine; however, I wouldn't be surprised given the noted small turbine/housing.

[86AmishMs3]

Originally Posted by Crossbow I'd have to disagree with the extended idling. It's a water cooled turbo. Even with the engine shut off, coolant is still going to circulate through the turbo...the whole water cooled turbo system was developed to prevent turbo turds, or baking of the oil in the turbocharger. I'm not saying its bad to let the car idle for a bit...but excessively is a bit silly, unless you just finished a 25 minute track session. I think most of us are smart enough to not shut the car off in the middle of a 6,000 rpm 5th gear run. TurboByGarrett.com - Turbo Tech101 Oh and here's a compressor map to add to the discussion. MS6 K04-2280 compressor plot

Yeah true but what neuspeed was saying that the shaft droops when the shaft is not rotating and when it cools it causes the shaft to deform. If you keep your car running the shaft will still rotate and will not droop and will harden in a non deformed state.

But anyways all this means is that our turbocharger is pretty much at and exceeding the design limits, and the range of the ms3 turbocharger probably exceeds maximum efficieny. And neuspeed inspected these cases primarily on the 1.8t engine. Given our engine 2.3L and DI, (flow is important not pressure) the efficiency is even more surpassed at 15psi and 5800 since our engine should flow more at that boost. But we already knew that the ms3 turbo is undersized.

[Funky]

John, its good to see that you were around to see this while designing your reworked turbo setup. I was wondering if and how you went about solving these issues with your design.

Thanks,

[ms3077]

I'm happy with the turbo they chose for this car. Plenty of grunt down low / midrange where it needs to be for daily driving, I love the effortless passing at only 3k rpms.

[08MazdaspeedSleeper]

good post, thanks for the info

[skubidoobie]

If they went with a bigger turbo, blown motors would more than likely been a bigger issue than Mazda would have been able to deal with.

[Grim]

Originally Posted by skubidoobie If they went with a bigger turbo, blown motors would more than likely been a bigger issue than Mazda would have been able to deal with.

not really since we don't have a defiant answer why the engines blow.

[memin]

great job

[rremente]

If the K04 is undersized for MS3s. Can anyone recommend a more appropriate upgrade that can handle 350hp?

[kore2000]

GT2871 should do fine.

[holeinmypocket]

good info

[cageyvet]

i like the stock k04

[phong3992]

really good info

[flyrevs2]

yea nice job, enganeer

yea nice job, enganeer

[webster13]

Cool, I always thought it was good for any turbo to have a "cool" down period before shutting down...anywayz, i was looking around the internet and the forums trying to find the specs of the the k04 that we have in our cars..anyone know?

[TRex]

i dont know if this has been posted before but i couldnt find it via the search....i believe this is our compressor map

[Mrmatt3465]

Originally Posted by Crossbow I'd have to disagree with the extended idling. It's a water cooled turbo. Even with the engine shut off, coolant is still going to circulate through the turbo...the whole water cooled turbo system was developed to prevent turbo turds, or baking of the oil in the turbocharger. I'm not saying its bad to let the car idle for a bit...but excessively is a bit silly, unless you just finished a 25 minute track session. I think most of us are smart enough to not shut the car off in the middle of a 6,000 rpm 5th gear run.

This was an excellent read. Just to put more perspective on the idling, Mazda DOES state in the owner's manual that after "long periods of driving or after hard driving" you should allow the car to idle for at least 30 seconds prior to shutdown. An extensive period like 5 minutes seems like a bit much but 30 seconds seems about right.