Hey all,

I currently own a 1.8L NA8C that came with a refreshed engine, running 10:1 pistons but no VVT.

I am ordering a turbo kit for it but I can't decide between the GBC series (20-300 or 22-350) vs the Pulsar 2860 Gen 2 GTX clone.

All my research shows that a GBC is much better value for money (roughly 1/3 of the price) compared to the GTX2860 GEN 2 *but* this assumes both are original Garrett.

The Pulsar 2860 GTX Gen2 clone is the same price as a GBC20-300 however which really throws a wrench into the gears of the decision process.

Power goals would be approx 280whp peak with safe torque but a built motor can be part of the car's future since the budget allows it if needed.

The GBC series can do up to 300hp or 350hp depending on variant at the flywheel and the 2860 can do ~475hp, so a lot more room for any future idea.

Personally, even without experiencing it, I doubt I'd ever go over 350 at the wheels.

I saw some comparisons of other journal bearing vs ball bearing but I am not sure they apply to these models since the GBC seems to be using GTX parts that provide better spool.
I also saw people mentioning that journal bearing are easier to service/rebuild for the DIY enthusiast..

Thoughts?

 

Might be difficult to get 280whp from a 10 to 1 engine on pump gas without a somewhat compromised spark map. Be sure to use det cans when tuning spark and always leave yourself some cushion.

A retarded map will increase EGTs, which is tough on turbos.

 

Well 280whp isn't set in stone, I could settle with 10-15 less for now I suppose. I was also planning to add water injection

 

I'm interested to see what people say as well, because these are the exact options I'm considering for the future as well. I'm also looking at the ~350 WHP range.

I'm not sure how much it matters, but the GBC line is only oil cooled, where the GTX is water cooled. Unfortunately there are only a few threads I've found on the GBC line.
Pulsar does have a GBC20-300 clone, which is the PSR 3942G, although it's only ~$200 cheaper ($750 vs $950) than the actual Garrett unit. I haven't found a GBC22-350 clone from them yet.

 

I am in Europe so the difference between a GBC20-300 and a Pulsar 2860 Gen 2 is ~100$ so .. practically same price

 

I was comparing the Garrett vs Pulsar versions of the GBC 20-300. Personally I'd shell out the $200 to get a genuine Garrett over the clone.
I noticed you didn't mention the Pulsar 3942 so I figured I'd put it in the mix, although it doesn't seem that attractive unless you're on a really strict budget.

 

I can tell you the GBC line is very efficient and will boost creep with a full 3" exhaust. My power goals were based on a stock bottom end, and I'm currently struggling to keep the GBC22-350 under control! I have a cat-less exhaust, so if you're running a cat you should be able to keep the spool in check a little easier. My plan is to add a cat, #1 so it doesn't smell like a boat, #2 so I can keep this thing at or under 10PSI at redline.

 

Tangential to the main thread, but I'm thinking about adding a cat to my exhaust as well so it doesn't stink so much. I guess that's how you learn though. It's not horrible, but I definitely get whiffs when I roll up to stop signs.

 

Well I didn't mention it because wherever I found it for sale it's the same price or even sometimes a bit more with tax than the original GBC /shrug

 

I am going for the 3" full Kraken exhaust that includes a 200cel cat

 

Not on a Miata so no input on sizing there but I run a GTX3584 clone. I pieced it together with bare center section, turbine, Pulsar compressor, bearing kit. They are as easy to work on as old T series journal bearing Garrett's except pulling the bearings out pulls the cartridge apart.

 

Surprised nobody's asked yet, is the car going to see track use? Some have done it, but I can't imagine putting a non-water cooled turbo on track duty.

 

More like 80% spirited street / 20% track (the car is not my daily driver so no compromise required)

 

Power or response, what are you after?
3942 vs 4647...
2560R could also (still) be an option, it's just such a well placed turbo for the BP

 

DV, good to hear from you!

He brings up a good point^^ Regardless, I'd want my turbo water-cooled if it's going to see any track time at all.

 

I am not aware of the 4647 but I am guessing it's similar to the 2860 ?

Either way, they're all very closely priced, making a decision difficult

 

Here’s my take:

If you think you might build the engine and you want 300 or more whp on pump gas, go with a G25-550 clone from pulsar. It seems to spool just as fast as the 2860r GTX Gen2 from back when I comparedy data to others’ data (it spools faster than a 2867r GTX Gen2). Additionally, the bigger/more free flowing turbo will likely help you run more ignition timing on your high compression stock engine.

edit: you’ll want the 7psi actuator on the G25-550. At 18psi it made like 290whp. At 14psi, I’ve not tested it but I think it would be right at the limits of the stock block at peak torque.

If you want to keep the stock engine and not have to build it go with the GBC-350. If you are planning on tracking your car, you will need to have an oil cooler regardless so the non water cooled turbo shouldn’t be too much of an issue. If these things needed to be water cooled, Garrett would’ve done so.

Also keep in mind the advertised crank horsepower that these turbos can achieve are not very realistic. They’re often at very low pressure ratios with perfectly ideal conditions, and if you’re on pump gas you’ll have to play it safe with ignition timing which will further reduce their maximum output. For perspective looking at the G25-550s compressor chart I’m likely near 3:1 pressure ratio and at least 45lbs/min of airflow, which means I’m likely within 25-50lbs/min from maxing out the compressor of that turbo, yet I’m only making 375whp on 100 octane gas (pump 93 in the US + VP racing Octanium). So basically, I’m thinking with the GBC 22-350 you’ll likely end up right around 280-300whp without absolutely wringing the turbo out and risking over spinning it.

If you’re worried about boost control, or lack there of, on the GBC 22-350, port the wastegate. That and a 7psi actuator, especially with a catalytic converter, should definitely let you keep the boost at a safe level. Lastly, I wouldn’t target more than 220 lbft on a stock motor. Sure 240 lbft is technically doable, but just one cool morning or small overboost can be the end of that engine. 220 lbft @ 6000 rpm is 250whp, achieve that with like 10psi or less at peak torque and 16psi or so at redline for example (not actual numbers, I just pulled these out of my *** for an example, testing will be required)

If E85 is available near you, add 10-20% to the power estimates of these turbos.

 

Thank you for taking the time to write all this information I live in EU with no access to E85 but I have access to 100 octane pump gas.

The pricing however for the G25 is not the same so we're drifting a bit away from the original question - Here are the prices I've access to, converted to USD, ignore the actual prices if you compare to the US, just look at price difference between models

Garrett GBC20-300 = 600$ + tax
Garrett GBC22-350 = 700$ + tax

Pulsar GTX2860R GenII (stamping actuator) = 615$ + tax
Pulsar GTX2860R GenII (billet actuator) = 650$ + tax

Pulsar G25-550 clone 4849G (billet actuator) = 900$ + tax


Regarding building the engine, since it's freshly rebuilt with less than 1000km on it since, it feels foolish to start over *just* to get more power. I'd keep power to a level where it doesn't blow up and if it does or if I really feel the need for more, I'll just drop it and do rods and possibly drop piston compression along with whatever is the usual route for building the engine "while it's out" within budget.
The purpose would be to introduce more reliability however, since I feel that I shouldn't be going for over 350 at the wheels in any case. I am guesstimating a 300-320whp in this car would be already too scary.

"Saving" the rods was one reason I was looking into the 2860 Gen2, assuming it would have less torque below 4000rpm then tune it so i get a flat torque curve. In theory, if the dyno below was used as a reference, if my math is correct, managing boost so the torque curve is flat from ~5000 up to 6750 (where it seems to shoot up), you can keep 315nm (232 lbft) at the crank. meaning you can peak at approx 300hp (again at the flywheel).

I am uploading a picture I made by scaling and merging 2 dynos to compare the curve - forgive the color issues, I am not an expert
Both cars are running stock 1.8 engines with stock compression. The 2860 is actually a Gen 1 Pulsar GTX.

There are numbers at the crank. One dyno comes from a local car and the GBC comes from SPS Motorsport as seen in one of their own videos about the GBC.



They're in HP / NM rather than HP / LBFT but you get the idea.
The GBC seems better up to 3500 to 3750rpm then the 2860 overtakes (granted there's 1.5psi difference)

Below is the same 2860 dyno overlayed with a GBC22-350 dyno I found on this forum from a VVT motor (the thread was about overboosting )

 

I understand. If I were in your position with a stock engine, I'd go for the GBC 22-350 in a heartbeat because of one reason: it is the smallest turbo that can meet your power goals.

The GBC will almost definitely spool faster by a significant amount. It's smaller on the compressor and exhaust wheel by a considerable amount. Both turbos use the GTX Gen2 compressor tech. The only downside to the GBC, other than the already mentioned oil cooling, is that it is a journal bearing. I can't say for sure but I have heard other people say that new journal bearings spool better than older ones. I can't attest to that but it is what I have heard. Even then, the size difference should more than offset the spool difference due to the journal bearing.

Both turbos will be able to make enough power to blow up a stock engine in almost the entire rpm range. With that being considered, the smaller turbo is almost always going to be a better decision if spool is most important, in my opinion. The only caveats are that the slightly bigger turbo might allow for an extra degree of ignition timing at most, but that shouldn't be a concern since you're not going to be able to handle any more power than the GBC can make on the stock engine, and that the 2860r is larger which means it will likely create less heat for the same power when used on the track where heat management is important. But then again, with the GBC you get a genuine Garrett turbo compared to a Chinese copy (admittedly a good copy).

I've got two things I don't like about those dynos you provided:
1) obviously, the GBC 22-300 is not the same as the GBC 22-350, but I know you understand this. The GBC 22-300 is likely completely maxed out at 0.9 bar / 240whp. The GBC 22-350 should be capable of right at 280whp considering it can flow 16.6% more lbs/min based on the compressor chart and ignoring the turbine. Which would make it perfect for your goals. EDIT: I didn't even notice that the second graph was of the GBC 22-350, given if it is the same one I linked below it is only at 12psi.

2) You can't really compare dyno graphs from two different dynos. Depending on the brands of dynos and settings there can be large amounts of variation. Also, comparing the boost threshold on a dyno without very high load/controlled settings is somewhat meaningless. If they started at different rpms or with slightly different load settings, it basically makes the spool incomparable. The best way to compare the boost threshold is going to be a tall gear pull starting at the same rpm/gear/differential in each test.

The only reason I mention the G25-550 is that if you plan on sticking with the stock internals, the GBC is an easy pick over the PSR 2860R Gen2, in my opinion. Once you even start to consider building the engine, the PSR clone of the G25-550 is the only budget turbo I'd consider for over 300whp. Just skip the 2860R and 2867R entirely. Lastly, our 93 octane is equivalent to your 100 octane due to different scales. My 100 octane is likely equivalent to 108 octane for you. Just different methods of calculating the rating.

In conclusion, while I know which turbo I would pick, neither turbo is really a bad decision. I also like trying new turbos that not many other have before just for the simple reason that it is more unique. I feel like I was one of the first people to provide decent data on the G25-550 and that was the biggest reason as to why I chose this turbo. I like data and sharing it with others. If you want to go a more tried and true route the 2860r might be better. I do think it would possibly be a better track turbo because it is bigger and water cooled, but I'd also prioritize spool time over everything else, personally. Also, I believe that if the oil cooling wasn't sufficient, Garrett would've included water cooling. The lack of having to deal with water lines might be a positive factor to some, and once you get either turbo out on the track, an oil cooler will be necessary regardless. Either way, I think you will be very happy with your choice. There really isn't a wrong choice here.

 

Also, I think I found the same graph you're referencing for the GBC 22-350 here. The fact that he is making 200lbft of torque at 2800 rpms and 220whp at 12psi seals the deal for me. Let me just tell you now that the GTX Gen2 2860R will not make that torque down low. Compare that to AndyFloyd's post on the 2860R Gen2 where he doesn't make 200lbft until 3300 rpms or so. Then again, you can't directly compare two different dynos, but that is a considerable difference.

Edit again: I didn't realize the graph was estimated crank hp so I changed the numbers to make it more correct.