New to this forum but not to Miatas or turbos
 

I have been looking around on this forum for a while and found some pretty useful info. I'm ready to get involved.

I am a former Garrett turbo engineer with lots of previous experience with Miatas and turbos. My wife and I have been hot lapping our 99 Miata with NASA for 5 or 6 years now and have had a blast. Miatas are great track cars right out of the box the way Mazda created them. But, after five years with R-compound tires and some racing beat sway bars on a completely stock drive train, we are ready for some more power. Time for a turbo. I am hoping to gather the wisdom of this forum to find out what works best. Hope you guys (and gals) can help.

By the way, I don't just plan to poach info without contributing. I engineered and built a complete custom turbo system for my FD Rx-7 so I have a wealth of experience to bring to the table. Let me know if I can help anyone out.

Thanks,

Dan

you're late/we've been expecting you. jay kav is around here somewhere too.

welcome to miatayourjob.net

WOW! :bowdown::bowdown::bowdown:

This Board keeps getting better and better! :)

Now, if you'd be so kind to post your location below your avatar and post some pics of your car.... or ban! :giggle:

Yeah, Dan's RX-7 is pretty badass.

But what's this? I'm the third reply to a "Hi, I'm n00b" post and nobody's flamed him yet? Weak.

he was humble and didn't ask a stupid, obvious question. what is there to flame?

oh and he's got cred.

Cool. Here's a Q.

Does the GT2860 really have a massively more efficient turbine than the GT2560? (as per the Garrett site).

Nah, put the screws to him. I insist.

JC, the peak efficiency is higher. Those turbine maps aren't very useful IMO since they don't have speedlines. Or efficiency, other than the peak. That, and it is very hard to map a turbine repeatably (hint).

But seriously, let's get to flaming Dan.

Welcome, sir.

Now post some Miata pics and add your location to your details.:)

I'd like to see the FD pics as well.

Sit down and stay awhile...

OK I'm dense, what's the hint?

Oh, Dan better answer the question.

I took it the hint was that the maps aren't set in stone as it's hard to repeat and get the same numbers. Gives you a limited idea about the turbine I'd suppose.

Oh have I got a good question for a turbo smart person- Does a little bit of "flutter" hurt anything? My GT3271 will flutter at low boost before the BOV opens. Does that matter?

Oh yeah, NEWB, read the faqs and stickies before you start asking about ebay turbo kits and how many boost you can run.

welcome sir, hope you enjoy your stay

Or ask questions like "how much boost can my built motor take" without telling us what turbo you plan to use.



Excelllent question Pat, mine does the exact same thing sometimes.



Welcome aboard DR, +1 on pics of the FD.
__________________
Best Car Insurance | Auto Protection Today | FREE Trade-In Quote

Thanks for the warm welcome (except JayKav).

Rafa, I'll get my details info updated tonight - please don't ban me until I get a chance to do something really annoying.

JC, unfortunately, I have to agree with JKav's answer to your question. The turbine maps they post on the Garrett site are not of much use. The are a rough curve fit through the peaks of the speedlines. Peak efficiencies and flows do give some point of comparison but what really matters is properly matching the turbine to the compressor. Without going through that exercise, you don't know where the turbine will really operate. The maps on the site only provide "best case" operating conditions.

For those who are interested, I will post pics of our Miata and my FD very soon.

Yes, welcome. To validate your identity please answer the following.

Q: According to Garrett, under what circumstance should a restrictor be used with a journal bearing turbocharger?

I am not aware of any restrictor requirements for a journal bearing turbo but a restrictor is recommended for ball bearing turbos if your oil pressure is above 60psi.

Pat, the flutter you are hearing is compressor surge. This is an oscillation of flow through your compressor when it is throttled as the intake closes. The durability concern associated with surge is that the flow oscillation can put a large oscillating axial load on the rotating group which must be reacted by the thrust bearing. This can be very hard on a journal bearing turbo which used an hydrodynamic thrust bearing. If the thrust load is enough to collapse the oil film, you get metal-to-metal contact which causes thrust bearing wear. This is typically not a concern for ball bearing turbos since the balls react thrust load. That said, mild surge like what you are describing is typically not problem even for journal bearing turbos. Sorry for the long winded answer - I'm a detail guy.

Thanks for the explanation. As for oil restrictors... I'm running a .070" restrictor on my GT3271. I don't have a "real" oil pressure gauge, but I'd imagine it's around 60 or so at cruise and 90ish by redline. If these numbers were correct, would a restrictor be warranted? Or even better: Would a restritor hurt anything? I've seen a lot of people run a restrictor on journal bearing turbo that had premature oil seal problems in the past that they contributed it to not using a restrictor.

pass :bigtu: (according to garrett guidelines)

I am curious like Pat regarding higher oil pressures. I can't see the harm with higher pressure in a journal bearing. It seems that the more oil the bearing gets, the better- providing it doesn't cause seal issues.

Dr. Turbo! With a name like that you can expect to get a few tech questions! I'm really looking forward to you future posts.

Welcome!

I wonder- has anyone with the Doctor's background ever evaluated one of those China chargers I keep hearing about? I mean, really took one part and evaluated the tollerances, metalurgy, design of the vanes etc. and compared it to the Garrett?

It would be pretty neat to have a (turbocharger) engineer's perspective.

Dr. Would you be willing?

Dear Doctor Turbo,

I bought a GT25 back in the day when there were only 3 or so GT series turbos in existence--before Garrett renamed everything with a ball bearing a GT turbo. It's a GT2876R (.64 A/R turbine housing) per the naming convention change.

Why does the turbo by garrett site list it as "not recommended for general performance applications" on the info page?

I will admit it's a little large for a streetable 1.8L, but it's been great for power at low boost and isn't a bad compromise in terms of spool. Here's my dyno graph at about 9.5 psi (the RED line). (the green line is a twin turbo 1.8L miata--2x GT1548)

http://y8spec.com/dyno/twins_vs_2876.gif

+1. Great question!

LOL, funny trick question! However, you shoulda picked a harder one. Even a turboknewb dumbass like me knew how to answer that question.

If I knew it, then WHO KNOWS what kind of riffraff DrTurbo could be, posing as a Garrett engineer!

I demand more proof...something REALLY wonky and supersekrit that only a Garrett engineer would know...something NONE of us could possibly know and verify!

:giggle:

What is your favorite color!

http://www.dailyllama.com/news/2003/...idgekeeper.jpg

Heheh, how about more complete turbine maps for the GT25, 28, 30, and 32 ...

Pat,

I see no reason to use a restrictor with a journal bearing turbo. Unless you are seeing oil leakage into the compressor or turbine, you don't have a problem. If you are leaking as evidenced by oil in the compressor or turbine, make sure you check all the other potential sources of CHRA seal leakage before resorting to a restrictor. There are many potential causes of leakage including oil backup due to a improperly designed oil drain. Here are links to a couple tips on this from Garrett:

TurboByGarrett.com - FAQ's

TurboByGarrett.com - Turbo System Optimization

The last thing you want is to starve your turbo bearings of oil.

Dan

That type of competitor benchmarking is routinely performed by companies like Garrett although probably not for smaller companies turbos. Now days, a lot of these small companies are buying Garrett/Mitsu/IHI/etc turbos, disassembling them, and reverse engineering to steal their aero designs. Then, performance just comes down to how well they can measure and reproduce the originals. This usually doesn't turn out well. If they are doing their own aero designs, they surely don't have the design and testing resources to compete with the Garrett designs. Then there is the question of reliability. The materials and manufacturing processes are likely not as well controlled as for an OE type manufacturer. The ultimate test is performance and reliability in the field. If you are considering going with an off-brand turbo, I would definitely recommend doing some research first. As far as me doing any benchmarking - I must decline. While it would be interesting, it would take an enormous time commitment not to mention the money to buy the candidate turbo. I already have a day job and don't have any spare cash. Keep the questions coming though and I will answer what I can.

Dan

I think markp was having decent luck with the chinachargers last I heard. Welcome Dr.

y8s,

Not sure why the Garrett site lists that. My best guess is that it is because the sizes of compressor and turbine are not well matched. (I won't get into the details of that here.) While your power and torque numbers look good, it looks like you are not on full boost until about 4000 rpm. For a low boost setup that is not targeting big power, there are other turbos that would provide the same performance at 9.5psi with better boost response. For street use, fast, linear boost response is usually what you want. The bottom line is that the compressor wheel on that turbo is just too big. Hope that helps.

Dan

Blue, no yellow. Ahhhhhhhhh!

I said FORMER Garrett turbo engineer. Sorry, I don't have access to that stuff anymore. I'll pass along your request to my friends at Garrett though.

darn. either I run more boost or find a smaller wheel. is there such a thing as a bolt-on wheel and housing for that turbo? :)

Such as a GT2860 LOL...

y8s, the smaller bb turbos like yours have a backplate that is integral to the center housing. This means the counterbore for the comp wheel is specific to that wheel's OD.

So in order for you to swap out to a smaller comp wheel, you'd have to also swap out the center housing. You're better off just starting over with a different turbo in that case, unless you can somehow find a chra + comp housing (sans turbine housing, since you can re-use the one you have). Maybe call ATP and tell them your situation.

ATP has my original .86 turbine housing. I traded them for the .64 I have now. :)

I think I'll just stick with this for a while.

finally some pics
 

I have gotten some request to post pics of my Miata and from my turbo Rx-7 project. Funny thing is, I looked around and I don't seem to have an pics of my Miata worth posting. I guess I'll do that this weekend. Here are some pics of the 7 to tide you over.

Looks pretty stock from the outside than the CF hood.

Attachment 208425

All the good stuff is under the hood. Pretty much everything is one-of-a-kind custom fabbed except the HKS exhaust manifold.

Attachment 208426

I designed and built the 1-off turbo while I was still at Garrett. It's basically a GT3276 with a twin scroll turbine housing.

Attachment 208427
Attachment 208428

The intercooler was also a custom job using a 12x18x3 Garrett bar plate core. I use a draw through fan to prevent heat soak at idle and low speeds.

Attachment 208429

Finally, a custom parallel flow dual oil cooler setup (one on each side) with custom ducts to the front bumper inlets.

Attachment 208430
Attachment 208431

Engine management is provided by an Electromotive TEC3.

Now for the interesting stuff. This is the closest I've gotten so far to a turbo Miata setup.

http://i68.photobucket.com/albums/i1...7/IMG_1713.jpg

Yep, that's right - I am a member of JKav's Lemons team and had a hand in building that monstrosity. Part of what inspired me to turbo my own Miata. Now the real fun begins.

that rx7 is awesome.

are you the homeless person doing all the welding for that build?

I wish. No welding skills yet. We have another guy on the team with those. I do have access to the welder though and have been thinking about taking a class.

I showed your car's pics to a friend (he also owns an RX-7 like yours) who's rebuilding his engine and he wants to know at what RPMs do you see boost with that T3276.

FWIW, he drags his car regularly but also uses it as a dd.

Thanks

I'm currently running an 8psi WG spring with no boost control. Haven't had time to tune for higher boost. My setup makes full boost (8psi) by 3000 rpm. Keep in mind that this is a custom 1-off I designed while I was at Garrett. Unfortunately, this combination of parts in not generally available. Its a ball bearing turbo and has a divided turbine housing which really helps with boost response, especially on a rotary when combined with my HKS divided exhaust manifold. This turbo will also support more than 18psi at high efficiency. I designed my system for road course duty so spool up from low rpm wasn't a big concern for me. Still, full boost at 3000 rpm is very streetable and only ~500 rpm later than the stock twins. Plus, I'm making 50 more rwhp than stock at 4psi less boost. My backpressures and EGTs are way down which makes for a much more reliable system. That was my main goal for a track setup that regularly sees 20-40 minute sessions at full load.

So is my GT3271 a good size turbo for my 1.8? From what I can tell, it's efficient to 23 PSI. I bought it because I wanted to go to 300whp from the begining and after building a motor sit around 350whp. I'm around 270 right now. It's not the fastest spooling turbo, but the power is nice.

That turbo setup you posted is badass. What paint is that on the charge pipes? Makes them look cast.

EDIT: And why is my turbo called a GT3271 instead of a GT3278? It's a .78 hotside divided turbine.

What makes a "good size" turbo is somewhat dependent on your goals. My 1.3L rotary breathes like a 2.6L piston engine so it is much bigger than your 1.8. Turbo selection is often a compromise between top end power (max flow capacity) and boost response. You should choose your turbo size based on what is the most important to you. There is also the important issue of properly matching the compressor to the turbine. This is very important in achieving the maximum possible efficiency from both wheels. From there, it gets pretty technical so I won't get into it. My guess is that you could have made your power targets with a 3071 and potentially achieved better spool up but it is impossible to say for sure without doing the design calcs. Anyway, it sounds like you are doing pretty well with what you have.

With regards to your GT turbo nomenclature question, the reason your turbo is called a GT3271 is because it has a 71mm compressor wheel. The GT32 identifies the turbine size and the 71 identifies the compressor size. The wheel trims and housing ARs are not indicated in the designation.

There is no paint on any of my aluminum hardware. It is all bead-blasted. That is what give it the dull mat finish look. You have to be really careful when handling it though. If you get oil or dirt on the mat surface, its really hard to clean.

Thanks to everyone for the warm welcome. I've been scouring the forum for the past 2 days straight gathering info and trying to learn as much as I can from what people have been doing. I think I'm ready to start making decisions and gathering parts from my own turbo build. I have lots of questions and want to get as much advice/input as I can as I go. My last question for this de-noobing thread is - where is the best place to put my pre-planning/build thread. Here is a little background:

I'm starting with a 99 NB with a completely stock drive train. The car has already seen extensive track time (~10k miles). It's a fantastic track car just as Mazda built it but, as everyone who has a Miata knows, its HP challenged. The goal is ~200rwhp with a focus on durability and thermal management for road course track duty. The car will regularly see 40 minute sessions at full load on the high desert tracks of SoCal. I already have a brand new Garrett GT2560R that has been sitting in a box in my garage for 7 years waiting for this moment.

My suggestion would be in the "Builds" Sub-section of the "Media" Section but you could also do it in the "DIY Turbo" Section.

Dear Dr. Turbo,

Can you please point me to SAE papers or other technical publications on CAC sizing and the relative merits of bar-plate vs. tube-fin CACs? I personally think the latter is better for a street-driven or auto-x vehicle where low lag is important. I have searched the SAE database and found none. Does garrett not publish at least some of their findings?

Thanks.

Unfortunately, I don't have much for you. I used a bar-plate Garrett cooler in my Rx7 because it was readily available. I think the Garrett cores are all bare-plate but I am not sure. I used the internal CAC design code to select the fin sets and do the heat transfer calcs. As far as pro-cons go, the tube-fin is definitely lighter which has its obvious advantages. A heavy front mount gives a big polar moment penalty. A bar-plate core has higher thermal mass and would take longer to heat soak giving some advantage for short duration performance like autocross or drag runs.

Started a build plan thread
 

For anyone who is interested, I started a thread on my build plan here:

https://www.miataturbo.net/forum/t31289/