turbo boost leak testing

[Metako]

I took my GT2871 turbo off the other day due to the ball bearings being shot and massive amounts of play in the shaft. I have always had problems with this turbo since I bought it off ebay (yes I know) and it never boosted like it should - far too little boost and too late in the rev range. It only started making positive boost at 3000rpm and had 10psi by 4000rpm (on a 18psi actuator spring).

I'm going to try rebuilding it with one of the ceramic ball bearing cartridge kits on ebay and might also need a new compressor wheel and comp housing, as these are probably too damaged to reuse.

But before I disassemble the turbo I wanted to boost leak test it to see if the turbo was leaking boost through the comp housing o-ring. Well I've tested it and it's leaking massive amounts of air through the inside of the comp housing through to the exhaust housing and also out the oil feed and return holes.

This doesn't seem normal and seems to explain where I was losing the boost. There's so much coming out the exhaust housing that the o-ring, if it's leaking, would be the least of my problems. Is it normal for the comp housing not to be 100% sealed against leaking? Do the seals need oil pressure from a running engine to be effective?

My test setup on the kitchen table:

[Bean]

Hi Metako!

The shot center section is likely the biggest cause of the leakage, and with a lot of play - thats plenty of gap for pressure to be lost.

If there are signs the turbines contacted the housings then both are usually shot and need replacing.

I'll be very curious to hear how the car goes with a proper GT2871 on it as it should be quite responsive when 100%.

Rabin

[Metako]

Hi Rabin, thanks for your reply. Thing is, it was leaking boost from the turbo even when there was no bearing problems and no shaft play. I tested it several times on the car and there was always a massive hissing sound coming from the turbo itself and rapid loss of pressure. The rest of the system seemed to be holding ok except for the PCV valve (so I added a check valve in there).

Also it's hard to see how the bearings themselves are providing a seal to boost - if they were journal bearings then yes I could see it but not with ball bearings - there are usually sideways gaps between the balls. Can you explain how they would provide a seal? I'm thinking the oil seal is the main way the turbo keeps boost out of the exhaust hsg but not sure.

Yeah the wheel is probably shot and I'm thinking of buying one of the Japanese GTX2867 billet copy wheels from Kinuwaga (?) in place of the 71mm wheel. Just need to do a bit more research on their reliability.

[keebs]

The oil seal between the chra and compressor back plate is all that keeps the boost in. If your chra and housings are toast its cheaper just to buy a new garrett with a warranty. Knockoff compressor wheels aren't worth it. Sometimes they're heavier to make them strong enough because of inferior alloys, and low quality control (think turbine overspeed burst).

[Metako]

Bean, I've just re-read your post and realised you probably mean that the worn bearings have created so much movement that the seal no longer works. Sorry and yes that makes sense.

Keebs, yes I've read of problems with some of the knock off wheels but read nothing but good things about kinugawa, so far at least. The other option is just a new stock wheel. Kinugawa below:

http://cgi.ebay.com.au/ws/eBayISAPI.dll?ViewItem&item=271573463576

I'd love to just buy a new turbo but over hear in Australia they're about $1200 so if I can rebuild the one I have reliably for around $400 I'd rather take that road. Of course, I don't want anything that's going to detonate my engine.

Had any experience with the ceramic ball bearing cartridges like this one below?

http://www.ebay.com.au/itm/111124857415?ssPageName=STRK:MEWAX:IT&_trksid=p3984.m1423.l2649

[Bean]

Hi Metako,

The Kinugawa / Kamack rebuild kits and components are well regarded in Volvo circles and I have bought their td04 journal bearing kits with upgrade 360 degree thrust bearings. Kit quality is top notch. I also bought an ARD Tuning compressor wheel that is likely a Kinugawa wheel and it looks beautiful - but I'll weigh it against the stock wheel when I rebuild the turbo to see how good it is.

And yes - I meant the seals in the chra as the only way boost can go through to the exhaust housing is if it was going through the chra. There should also be an o-ring seal between the compressor housing and the chra where they come together - A leak there would leak externally.

One thing about rebuilding the turbo is that the rotating assembly should really be dynamically balanced anytime a component is changed. The slightest imbalance at the RPM they spin will make quick work of the new chra. Better option might be to get a new chra with new wheels. (Unless the wheels and chra are balances individually? Journal bearing assemblies should be balanced together which is what I'm familiar with)

Rabin

[keebs]

I personally would stay away from chinese bearings on something that spins so fast. Give garrett a call and see what they can do for you. Also maybe look on www.himni-racing.com they have everything you need and may do exchanges.

[Bean]

I'm a little more lenient when it comes to Chinese products. While they are well known for cheap knock off parts that are sub quality, they still have OEM quality capability since many OEM's and their suppliers have production facilities in China.

I would definitely look at the chra's much closer and do a ton of research on quality before spending the money and risking them in my motor, but if they check out they should be fine.

Be very sure though since the turbo shares the engine's oil system.

Rabin

[Metako]

Done a bit more research and found this thread on Nissansilvia.com which gives the brand a thumbs up:

http://www.nissansilvia.com/forums/index.php?showtopic=502715&hl=%20oil%20%20restrictor&st=0

I think I read somewhere that Kinugawa actually make Mitsubishi turbos for them although this could be a myth.

Yes it's probably too late to do a valid boost test on the turbo now that the bearings are shot so I'll make sure I test the rebuilt turbo before I install it this time. I'll check that the comp o-ring is in place when I disassemble the old turbo in a day or so.

It's strange but even though the turbo was leaking boost on the car it never smoked which seems to suggest the oil seal was ok, so maybe it was the comp o-ring after all. Will know soon I hope.

Btw from what I've read from a turbo builder the wheels and shaft are balanced seperately and then balanced again as an assembly, even for ball bearing turbos. However Garrett use a fairly wide tolerance in their balancing so you could get a good one or a bad one. Sounds like it's worth doing a final balance on any rebuilt turbo because of this.

[keebs]

How many miles are on this turbo. I'm just curious as to the conditions for its failure. You know lack of oil, too much oil, ups dropped it (a lot), or maybe a noticeable defect that slipped past QC. If you find anything interesting please post a picture.

[Metako]

Hi Keebs, the turbo was bought from ebay as "brand new" from a private seller. When I recieved it in the post it didn't look new to me so I took it to the local Garrett shop who confirmed that the covers and core certainly were not new, although the compressor wheel looked new. Without taking it apart they could not confirm it's condition apart from no significant shaft play and good overall condition.

I contacted the seller who said "well it's been rebuilt by Garrett and it's as good as new". I told him he had misrepresented the item and I wanted a discount. In hindsight I should have demanded a refund and sent it back but I had caught "big turbo fever" and accepted a $50 refund on the original $550 price I paid. I figured $500 was a good price to pay for a fully rebuilt turbo. In short I was naive and took a big gamble. Expensive lesson learned!

I installed it myself together with a Nistune USB chip which came with a map for the GT2871 and my other mods. I didn't use an additional restrictor in the oil lines after the Garrett shop said they had their own SR20DET with a GT2871 and it works fine without one, and that GT turbos come with a restrictor already installed in the core. I never had blue smoke out the exhaust but had some oil in the intake pipe, more than a tiny bit. So maybe I did need that extra restrictor after all to reduce the oil pressure which I've read can go up to 85 psi on the SR20DET. Garrett recommend no more than 60psi and 30psi is ideal.

The turbo was always very slow to boost and only ever made 10psi. I disconnected the actuator, and made sure I had no pipes leaking but I could always hear air escaping from the turbo itself when leak testing. It now sounds like the leak was probably from a bad oil seal. Maybe it also had too much oil pressure, or I overboosted (I left the actuator disconnected and used the car whilst trying to track down the leak). In fact I first heard the strange grinding sound after I overtook another car at around 4000rpm. It could have already had rubbish bearings when I bought it (although no shaft play yet).

Here is a pic of the compressor cover. It's not as bad as I thought with slight scoring only in one place and the rest of the surface is still ok. Not sure this can be salvaged so will take it to the Garrett shop soon. Any opinions?

Also attached a pic of the comp blade which is ground on one edge of each blade-pretty sure this is unuseable.

The last pic is the oil feed orifice in the core where the inbuilt restrictor is meant to be. All I can see is a 3mm hole and no allens bolt which is supposed to hold the ball bearing cartridge inside the core. Anyone know if this 3mm is a removeable plate/restrictor covering the allens bolt/pin? It looks like maybe I could grab it with some long nose pliers but I don't want to damage it if it's not removeable. I tried sticking a 3mm allens wrench through the hole but it's not fitting into anything.

[Bean]

I'd agree that the cover is likely fine, and the wheel is done. I can't comment on the restrictor at all as I've got 0 experience, but googling it should confirm unless Keebs can confirm.

As for the leaking - I doubt you would hear air leaking internally on the turbo, but leaking from the compressor housing would be something you would hear as it would escape into the engine bay. Not knowing how many connections between the turbo and intake manifold - it really could be anywhere between those points.

Rabin

[keebs]

Now don't take this as gospel since I dont play with BBs very often, but I believe that "oil restrictor" piece is what holds in the outer bearing race. To remove I think you thread a bolt into it and pretty much slide hammer it out. Remember to check runout on the turbine shaft after any housing contact. A guess on tolerance would be around .0005 inch or .0127 metric.

[Metako]

Thanks Keebs, that sounds like it might do the trick. After doing some more searching I think the pin looks like the pic below and the actual restrictor is on the side of the pin not the bigger top hole. Now to find my slide hammer...

The leak was definitely coming from the turbo itself Rabin as I had blocked off both the intake and outlet of the turbo and eliminated the piping from the equation. But I don't know if it was an external to atmo leak or a leak from compressor to exhaust housing.

[keebs]

Yep that looks like the one. If its a comp housing leak to atmosphere break out the copper rtv and see if it works.

[Bean]

If you could hear it my money would be on compressor housing to atm... I doubt you would be able to hear it if it was leaking internally via the chra...

Air will take the path of least resistance, so if the chra is leaking more now then it would be harder to make the housing leak.

Regardless - I think rebuilding it as you are planning should be the way forward. Very curious to hear how it turns out once done as far as when it starts building and how fast it reaches peak boost.

Rabin

[keebs]

When rebuilt it should be about the same as my t3 s60. Full boost at or before 3000rpm.

[Metako]

You're probably right about the leak Rabin. Will also have to find my hi temp RTV sealant. And full boost by 3000 would be very nice! Will let you know what happens.

[Metako]

Well I finally got the pin out with an M4 x 0.70 bolt, which threaded into the pin, and a homemade visegrip slide hammer. The bearing cartridge then just slid out after I removed the endplate, comp wheel and shaft.

On closer inspection it doesn't look promissing - the comp wheel is quite ground down on one blade, the shaft looks scored and even the exhaust wheel is has grind marks where it touched the exh housing. The ball bearing cartridge seems to be missing the rear bearing carrier which holds the balls in place! Wonder what happened to it. The shaft was only supported by the front carrier/race so no wonder it went pear shaped.

I'll take the box of parts to a Garrett shop and see if anything can be reused. If not then I think I'll just rebuild a journal bearing T28 I have and make do with that.

[keebs]

Man that sucks, sorry about your luck.

[Metako]

yeah thanks keebs. I went to the turbo shop today and only thing I can reuse are the comp housing and exhaust housing. The comp wheel and shaft are too far gone he said, and the missing bearing carrier is probably sitting in the sump! He said the oil supply line ALWAYS needs to be replaced with a new one whenever you fit a new turbo, whereas I reused the old one. The tiny restrictor hole in the side of the pin apparently can easily get clogged up and when it does it's goodby ball bearings.

I'm a bit over the whole ball bearing thing as they seem too susceptible to oil starvation - the plain bearing turbos use larger holes in their restrictors and are cheap to rebuild. My next step is to disassemble the T28 I have and have the parts inspected and hopefully rebuild it with a kit.

[keebs]

You're right, BB turbos need better oiling tech than those tiny holes to be reliable. Get an inline turbo oil filter while you're at it. Are you able to pull the sump with the engine in.

[Bean]

If you can get the 360 degree thrust bearing upgrade for the T28 in the rebuild kit you should be set.

With a fresh T28 you should still be worlds better than the knackerd BB. If they make a billet compressor wheel for it that works better than stock you may not even need or want to replace it.

Rabin

[Metako]

You're right, BB turbos need better oiling tech than those tiny holes to be reliable. Get an inline turbo oil filter while you're at it. Are you able to pull the sump with the engine in.

the filter's a good idea. I'll have to have a closer look at the sump but from memory I think I can take it off with the engine in the car.

If you can get the 360 degree thrust bearing upgrade for the T28 in the rebuild kit you should be set.

With a fresh T28 you should still be worlds better than the knackerd BB. If they make a billet compressor wheel for it that works better than stock you may not even need or want to replace it.

Rabin

yes will try to get the 360 degree thrust bearing-any idea if Garrett make these as I would prefer a Garrett repair kit? Yes it would be very nice to add a billet wheel on the T28, maybe something like this one:

http://www.ebay.com.au/itm/Billet-Turbo-Compressor-Wheel-Garrett-GTX2863R-47-1-63-4-Trim-56-11-0-/291240375983?pt=AU_Car_Parts_Accessories&hash=item43cf4776af

[Bean]

No idea if Garrett makes a kit ot not, but the Kamak/Kingawa kits seem top notch.

Billet wheel would have the be spec'd out to the turbo and comptessor housing of course, but if they made them they'd be a great upgrade - I'd just want to back up that mod with some reliable dyno proof hopefully from other SR20DET owners...

Rabin