Autocross turbo setup - mitigating lag
 

My FM2 kit with GT2560r on non vvt 1.8 is not cutting it on response time. It is awesomely responsive and reasonably powerful, but I want more. I actually want faster response more than I want more power.

The goal:

I want substantial boost pressure within 8 ft coming off the slowest corner I'll ever encounter. That means about 200 ms (0.2 second) from trailing throttle to substantial/full boost at 3000 rpm in second. I figure I need response down to about 10 ms by 5000 rpm, which seems far less challenging. The present setup almost does that, with a poorly repaired FM DP. I don't want to give up any more top end power than I already do with the GT2560r.

To encourage spool with valve timing:

1.8 with vvt. Stock high compression (not building the engine yet).


To encourage spool on the exhaust side:

I'm using a port matched, short, small diameter runner manifold with semi-tangent collector. I'm doing a DP with taper from the turbine outlet diameter up to 3" into a 5" 100 cpi cat, through a 3" perf'd core straight through muffler.

To encourage spool on the intake side:

I'm doing a 9" barrel AEM dryflow with a 6" velocity stack into a 3" pipe, with a long taper transition into the compressor inlet. I'm planning on 2" OD pressure side intake plumbing, small water to air intercooler, most direct route possible, and maybe water injection too.

I've looked at various types of two turbo setups to get good response and good efficiency over a broader rpm range. Obviously this will be a lot of money and a control system nightmare in the making, but it is still a possibility.

I'm looking at anti-lag with a small-ish frame, large-ish ar turbo. If I had anti-lag setup so it engages partial misfire with retarded timing and throttle bypass solenoid in a narrow rpm range and low but not zero throttle position, I could have the system only beat up on the engine/turbo when I intentionally pre-position the throttle just before going WOT on corner exit. I'm still concerned that this is going to murder my exhaust valves and turbo.

Should I go smaller on the pressure side intake piping? Think the GT2560r will get me where I want to be with just a little anti-lag? A little anti-lag is a lot less bad than a lot of anti-lag, right?

Thoughts?

the vvt will help a good bit
the higher comp should a bit too
the least restrictive hotside and exhaust setup will help

So yeah, your plan sounds legit.
Not sure about antilag, never used it on a miata, but should help too.

You could also switch to a tubular fm replacement log like Tim makes, that's a significant improvement over the cast log.

Switching to 2" ic piping and the smallest ic to achieve your goals would also help.

Efr6258?

ha I was going to say that too, but he's wanting to get it spooling faster, and iirc the efr spools about the same just provides a ton more power

I assume you are using EBC? Since you are talking about setting up anti-lag I assume you are already using EBC, but just want to make sure.

Since when is ebc>mbc when discussing spool? I thought they were equal

I ask because I see in multiple threads lately people telling other people to switch to ebc for better spool, and I just don't get it: ball/spring mbc is closed til desired pressure is reached, and afaik they would have the turbo spool just as quick as an ebc.

The only thing I can add is that I was under the impression "anti-lag" likes to eat turbos.

18psi, The adjustable bleed MBC is probably what they are talking about as opposed to the ball and spring MBC. I still think an EBC will be more effective because it doesn't open as gradually as the ball and spring does.

Very short and small diameter manifold tubing sounds good because of the reduced volume. Likewise, reducing the intake volume, shorter runners, including intercooler and piping may help. But shorter runners can effect low RPM torque.

Lighter weight rotating and reciprocating parts are a big plus for responsiveness. If you are serious, you already have looked at the lightweight dual plate clutch, an aluminum driveshaft, a tubular front subframe, lightweight brake rotors, shim under bucket lifters with lighter springs, titanium spring retainers, etc. Forged rods are much heavier weight than stock, so be careful when you upgrade.

If you want more torque and responsiveness down low, have you advanced your cams yet to shift your torque curve down the RPM range? That should help the lower RPM spool as well.

I understand that retarding your ignition timing in the spool ramp up cells can aid in spool by causing expansion of exhaust gasses to continue to occur in the exhaust stroke and into the manifold, but doesn't that also impede acceleration? I'd like to see the differences on the dyno.

I see. I had no idea people still used bleed-type mbc's lol.

I don't think anti-lag can even really be tested on a dyno unless you do hard shifts to simulate real world conditions and log pressure recovery time (as OP kinda talked about).

They use it in wrc and other types of racing with great results, I'm sure there's merit to it.

Another completely random idea is to eliminate the bov altogether. If you're gonna be hard on the turbo I don't think that's any less damaging to it than anti-lag.
This way you get minimum pressure loss in the intake side during shifts or throttle transitions.

Just a thought

PS: oh and definitely do not shorten the IM runners. That is for sure going to hurt your lowend

Ditch intercooler and go with meth/water injection so you can run the pipe straight from the turbo to the intake?

(I don't know if this is a good idea or not)

I actually have no idea what the difference between the 2 types of MBCs are, despite being on this site for 7 months now. Because nobody talks about them here anymore.

In my defense though, OP didn't specify any boost control, so it could have been wastegate only (though I doubt it).

Ditch intercooler and replace with air/water cooler... fill with ice water before each run :D

supercharger.

:eek: Bring on the hate! :firedevil

I've played with both. One is stupid, one is not.

Bleed types: Only pro for these guys is you can adjust it from the cockpit with some models. But the line is constantly open to the wastegate, you're just bleeding off some of the pressure, so it opens a little later. You'd achieve the same effect by cutting holes in the line going to the wastegate. Spools sucks cause technically the wastegate starts to open at .000...00001 psi.

Ball/spring types: Can't adjust from the cockpit unless you run long lines into the cockpit, but that seems stupid. Spring/ball design keeps the line shut until close to target pressure, where your boost is strong enough to push the ball out of the way and open the wastegate. HUGE pro to these guys is that you spool the turbo with the wastegate clamped shut.

even the ball/spring type is inferior to EBC because once the pressure is enough to overcome the spring, the ball/spring MBC turns into a bleed type until target boost is reached.

EBC holds the wastegate shut until the very last moment.

How about left foot braking and keeping your foot on the gas a bit to load the engine and keep turbo spooling? A-la brake boosting

when i read this .. i tought of that ...i will use a photo because my engilsh is bad bot the idea is that the trothel body move at the same time.I really dont know if it doable just an idea
https://www.miataturbo.net/attachmen...ine=1363999263

You recall wrong. It spools faster, responds SIGNIFICANTLY faster, and provides far more power.

The OP wants an EFR6258.

Incorrect. An 8psi wastegate spring begins to bleed pressure at ~5psi, which slows spool. An electronic valve that provides no pressure to that wastegate until ~7.9psi is going to spool faster for obvious reasons. This has been proven in dyno pulls (IIRC Soviet has some pulls in his build thread that illustrate this very clearly).

This is correct.

I came in here to give some advice, but sixshooter beat me to all the points I would have suggested.

You may consider some cylinder head work. It wakes up the BP everywhere.

Anything that improves naturally aspirated torque at 3000 RPM will help spoolup at that RPM.

For example:

- more displacement
- VVT
- compression ratio

First step IMHO would be a APEXI AVCR EBC. Sav is correct about the benefits of an EBC vs MBC. What most don't know is gain is set to avoid spiking in top gears. In lower gears you would want more gain for better boost response. The AVCR is the only boost controller (that I am aware of) that allows you to adjust boost gain by gear. Just think of how a turbo is loaded differently in each gear. That's why being able to adjust boost control by gear is important to someone like you.

On the other hand if you don't have the cash to buy a AVCR or want to deal with it try using a Greddy Profec B. Crank up the gain for the autocross runs and set it back down to normal for street driving.

Ebc/mbc: I currently run ebc with the boost threshold set above the actual boost threshold to just barely not overboost in second with the current spool characteristics. I have the other parameters also set for second gear. Since I'm never bogging 3rd on course, this works out ok. I just lower the boost threshold low enough so it is laggy but safe in 4th and 5th.

My plan is to run a ball and spring mbc with the bleed blocked and a check valve'd back bleed passage referenced to manifold pressure. This way, I get full wastegate actuation as soon as the ball leaves the seat. I'm having trouble finding a little check valve with a low enough cracking pressure, but I'll make it work.

bov:

I've thought about this a bunch. I'm less concerned with boost recovery during a shift (just because it isn't bad now at higher rpm) than I am with going from off throttle while cornering to wot through the apex. I'm not sure if I'd even see a gain in that scenario with no recirc valve.

left foot brake pre-spooling:

This does not work well with tight corners, rwd, lsd. It ends up altering the line enough to be a net loss loss compared with living with some lag. If I could coordinate ebrake with throttle well enough to avoid the understeer that comes from the front brakes engaging, maybe. I'm not interested in going down that road rather than fixing the car, though...

meth/water injection:

Rules don't allow meth. I plan on a small water to air system supplemented with water injection.

intake piping:

Anybody think can get away with smaller than 2" od charge piping (sweeping mandrel bends and so on)?

efr turbo:

That's a lot of money compared to a turbo i already own. Just how much faster "Just above boost threshold" spool are we talking?

AEM does it too.

If you are you going to convert over to water/air intercooler, heat soak may be a problem. I'm running frozenboost's type 20 core and their small sized heat exchanger, and I don't think it's enough for auto-x. It starts to get soaked by the time my runs are over. On the street or in the mountains it's never been a problem though. Either a larger reservoir or heat exchanger would probably fix this. I also would advise against smaller than 2" piping. It will choke up top. That said, I love my water/air system. My charge piping is all of 2' long, and response time is absurd, but I'm also using a teeny turbo.

You want the EFR, barring that, you want a 2 liter stroker and a turbo just big enough to hit your power goal. The only turbo miata I drove was a 2 liter stroker with some sort of turbo setup from FM, no idea on the size of the turbo supposedly it was making 300 wheel. There was zero turbo lag. Some of that could be contributed to the fact that he has that super rare (and horrible) adjustable australian helical diff and a rear sway bar, so I spun the inside rear coming off of every corner.

Or if you're super money bags. Beyond the Dyno: Rocket Anti-lag Teaser I know how to make this work and it'll cost you between 2 grand and 8 grand depending on how many prototypes it takes to get the aerodynamics correct in the rocket.

You don't have a turbo'd Miata?

There will be no getting into the engine internals this year. There will be no substantial displacement changes. There will be absolutely no FM parts on the car (nor anti-roll bars).

I also know how to make that work, and I don't think it would cost anywhere near that much. It is just some basic engineering math to figure out the right geometry. The basic principle is that a nozzle has a high pressure, low velocity region and a high velocity low pressure region, and by doing the combustion just after the peak velocity region, you don't get backfeeding of your flame front. It would require a special bypass valve actuation routine, two exhaust gas capable 0 cracking pressure check valves (eccentric butterfly in a rectangular section of pipe will do), and "rocket" fuel injection that is a function of engine rpm, turbo rpm, and tps. The tricky part is mapping the likely turbo speed for a given wot condition, and then using closed loop control to achieve that speed any time als is engaged and the tps is low.


I'm hoping to make (hopefully minor) compromises to total area under the curve of interest and avoid needing that sort of ALS, and also avoid needing multiple turbos.

The electronic control takes a bit of thought but it shouldnt be that difficult an MS3 should be able to do it but would have to be separate from the engine control. Or you could do it with a pectel SQ6. There shouldnt be any check valves in the design since its basically a gas turbine that someone left the shaft out of. But the expense of prototyping is that the combuster needs to be made out of inconel for it to have any sort of longevity.

why not focus on things that actually improve spool/low-end?

Run variable pulsed N20 off of a 4d map that uses TPS, RPM, and MAP as inputs?

N2O is also not legal in his class.

why not focus on things that actually improve spool/low-end?

/\ Alzheimer's?

Repetition or rehearsal is necessary for information to be stored in short-term memory or retained in long-term memory. Especially when additional unrelated information is presented before repetition or rehearsal is complete, it interferes with the brain’s ability to transfer the original information to memory.

One of the check valves is needed to reduce the effective exhaust volume upstream from the turbo when the system is not in operation (this also prevents flow reversion in transient pressure situations where no fuel is being injected). The other is needed as a safety against engine fire. Inconel isn't necessary for my application (less than a half second of operation perhaps 4 times in 50 seconds, then minimum 5 minutes of rest time.) I've got other ideas for how to implement such a system as well, but at the end of the day, I'm trying to start simple.

The compressor efficiency island is above the max pressure ratio I plan to use. Still think it is the best choice?

I'm considering the EFR, but also thinking of a lightened billet turbine wheel upgrade for the gt2560r (claims to slightly improve boost threshold, dramatically improve spool, and slightly improve efficiency at high flow rates.)

If the 6258 efficiency island is above where you're running you need a much smaller turbo than the one you have now.

That's where I started. I almost bought a brand new eaton blower for $250. I was designing brackets. Then I started doing the numbers.

I couldn't find a supercharger that would make 10 psi at 3000 rpm and also make 10 psi at 7000 rpm. I didn't look at anything crazy expensive, but the usual suspects have poor volumetric efficiency/effective displacement ratios at low rpm. An eaton blower will only make about 6-8 psi (depending on which one) at 3000 rpm if setup to make 10 psi at 7000 rpm. A rotrex needs a calibrated restrictor plate in order to make boost at 3k without blowing the engine at 7k, and the adjustment factor isn't linear, so you can't have constant boost without using a wastegate on the charge air. The response time with with a belt driven blower and a compact air to air intercooler is on the order of 300 ms to full boost at 3000 rpm, and only slightly faster at higher rpm. The thermal efficiency is lousy on the external compression types, and the price is lousy on the internal compression types. None of them are going to make as much hp in the mid-range as a decent turbo setup, given the same safety factor against engine failure.

None of that sounds appealing when compared with a more advanced turbo setup.

I dont think you were looking at the right blowers, the M62 on my last car (2.2 liter engine) would make 10-12 psi from pretty much idle till 7k. Of course because the cams werent right, the tune was as good as you could get with alpha-n, and the port flow sucked, it only made 226 hp and 210 ftlbs. There was no discernible delay between the the throttle and the power.

Ever think of picking up a cheap 2554 from the Nissan world? You can generally find them for $100-200 and would swap right over. Might offer you the better response that you want.

My plan would be to have 2 maps, switchable. I plan to have a race map where, at minimum TPS, the ignition timing is normal. At the point above minimum TPS where power delivery equals engine braking, between 2500 and 4500 rpm or so (I'll tune all this once it is up and running) I'll run maximum retard and a throttle bypass solenoid. I plan to run the retard and bypass flow as high as possible without the engine putting out enough power to accelerate. Once I've found that point, I'll tune backwards to make it so that the retard drops off slowly, the bypass solenoid cuts out, and power delivery increases somewhat smoothly as TPS further increases.

The bottom line is I'll decrease the retard to normal advance values long, long before WOT. The effect should be that I can get pre-spooling on demand in the rpm range where it is needed by pre-positioning the throttle into the low power output spool region before I want power. Then as I roll on the throttle past the apex, the timing advances to increase power, the bypass closes, and boost rises rapidly if I get on it hard.

why not focus on things that actually improve spool/low-end?

Look at the VE graph here:
CAPA : Eaton Superchargers

My powerband would run from ~4700 to 13000 rpm on the MP62. I calculate that based on 90% VE at 13000, and overdriving the SC to get 10 psi even though the mechanical displacement times drive ratio implies 11 psi. Then, at 4700 rpm, the VE is about 60%. 0.6x11=6.6 --> less than 7 psi at less than 3000 engine rpm. All this normalized for temperature.

As stated in OP, because he's already underwhelmed by the topend on the 2560 and doesn't wanna give up any more.

Otherwise that would make perfect sense.



oh and OP: if you're gonna stick fancy billet wheels onto the 2560 just get the efr now and don't waste money/time.

Can you explain your thoughts? Here are mine:

The center of the island for the GT2560r is basically at 10 psig at sea level and about 200 hp.

The bottom of the island for the EFR6258 is at about 15 psi at sea level and about 260 hp or so. The middle of the island is at about 22 psig.


The EFR may well spool better, but it seems far from ideal, given it will have lower compressor efficiency pretty much everywhere.

Why do you care so much about compressor efficiency? Yeah its nice to wave your dick around but if you only want to make 200 hp, your turbo is too big. For an AX car you should be aiming to have turbo that you ride right along the surge line, then the top, the the choke line of the compressor map. So what if thats in the 65% efficiency range. You'll still make the power you want to make, without blowing up the engine and the smaller turbo will spool faster and have a lower boost threshold. And 200hp is not enough for a SSM car to be competitive never mine an XP car. You should be shooting for at least 400 in XP just to keep up.

Okay, are we talking about lag, or are we talking about spool. These are different things.

The dual throttle setup is a very cool idea, that will do wonders to minimize lag (which is why it was used in F1 way back when), but it wont make the turbo spool any faster.
Of course I question how much true lag you would even be experiencing, since that turbo shouldn't have any trouble getting going at high rpm when you suddenly give it the throttle.

Do you have something to contribute here, or just showing off your people skills? :D

Better exhaust manifold.
Smaller IC.
Larger TB.
Better IM.
Better Head.
VVT.
Better exhaust.
EBC.


When you can easily build a 2560 to make 250rwtq at 3500RPM at 14psi with simple bolt ons...


If you went to a 2554, you might gain 20-30rwtq below 3K, but above that, youll still peak at 3500 but much less overall torque than the 2560 will provide in that range, then it will run out of steam at 5.5K.

compare the two here:

https://www.miataturbo.net/attachmen...ine=1351171600
140tq @ 2.5K
190tq @ 3K
225tq @ 3.5K


https://www.miataturbo.net/attachmen...ine=1351171600
140tq @ 2.5K
180tq @ 3K
225tq @ 3.5K


You can see the output below 3.5K is very similar, but you can see by the rounded curve of the 2554 that it spools faster and more linear, the problem is, the actual output is on par with the 2560 spooling slightly slower and then increasing expotentially.

The gain in spool is outweighed by the nil extra output. and then youre left with a turbo with less top-end, where your 2560 can potentially hit ~300rwhp.

Below 2800 or so doesn't exist for purposes of my build, because if I get down that low, I'll always shift.

2800-3500 having a large chunk of the extra power of forced induction is important. 6500+, having a large chunk of the extra power of forced induction is important.

Goal: As much power as possible between 3500 and 6000 rpm, without sacrificing having a chunk of power on tap within 200 ms of going WOT between 2800 and 3500.

Those graphs pretty much match my impression of the gt2560 vs gt2554 in terms of power on tap. The gt2560 wins at every single point in the critical region of 3500-6000 rpm. On top of that, my old gt2560 setup was still pretty close in terms of responsiveness. It was still the fastest car at our biggest event last year. If a turbine upgrade on the GT2560r will give it faster spool, and less exhaust side pressure ratio above the wastegate cracking pressure, that plus the rest of this might just be enough. I'm not completely convinced that the EFR6258 will put down more power in the 3500-6000 rpm range than a worked over 2560 at 10 psi on a 1.8l engine. Can I get a probably from someone with some experience running an efr6258 at low boost? :)

why do you love to stroke off 10psi so much? if youre goal is: As much power as possible between 3500 and 6000 rpm, then why are you limiting yourself from being able to make as much power as possible between 3500 and 6000 rpm by keeping the boost at low donkey dick level?

18psi on is super amazing. I mean the GT2560R would love being run at 18psi, heck it would take 20 like a champ.

Thanks
I know.

:brain:

lololol

Better exhaust manifold - I think I've got a decent compromise between cost, flow, and volume right now. I'm doing some minor porting on it, and it seems the runner are slightly smaller than the exhaust ports, with semi tangent convergence to the t25 flange.


Smaller intercooler - well, I'm starting from no intercooler, and working up. It will probably be a small inline water to air. No FM retardedness on this build.

Larger TB - Stock is 2.125 ID. I think I have a ~2.75" bore throttle body around somewhere... With ~2" OD charge piping I don't think going any bigger will help....?

Better IM, check.

Better head..... I'm leaving that alone

VVT, check.

Better exhaust - I'm making huge improvements here, with 3" piping coming off of the transition from the turbo exit flange, 5" 100 cpi cat, and straight through muffler. I haven't decided on the wastegate plumbing, but thinking I'll do a 9.4" circumference bellmouth with a slow transition to 3" round. The other option (presuming I stick with a garret turbo) is a 2.25" to 3" tapered 90 bend off the turbine exit with a divorced DP rejoining the flow about 10" downstream.

EBC - I'll do something that holds the wastegate completely closed until just before the boost target is achieved.

I'd like the stock internals to stay internal, and I'm not opening the engine up this year.

you can push more than 10psi. i like the exhaust plans.

If he runs a non-fail intercooler, Brainy? :giggle:

What's with the FM hate? They make a solid proven product that a lot of users here have had great success and reliability with for years.

I don't have a horse in this race, I'm just curious.

Generally speaking, people skills are not required in totalitarian regime ;)

They dont make anything superior in terms of turbo parts.