Rocwandrer

I bought the car turbo'd and it is mostly the FM2 original generation kit. It has a recirculating valve with a little tiny K&N filter on it instead of recirc.

I tested it with a plastic bag in front of it to detect flow.

It leaks air in under vacuum (a lot, almost lost the bag into the valve)

It leaks air out under very low boost pressure (a lot)

I think it leaks air out under moderate boost pressure (a little tiny bit?)

I don't think it leaks above maybe 5 psi, but it is hard to tell because rev limit comes on quick in neutral

My spool time from coasting to full boost (13 psi) with a GT2560 is:

About 3 seconds at 3400 RPM (which is the full boost threshold in an off idle pull)

About 1 second at 3800 RPM

About 2/3rds second at anything above 4000 RPM.

I'm expecting the GT2560 to hit hard and fast at anything over 4k. I've got a somewhat constricted (dented) 2.5" DP, and this recirc valve as potential culprits for the slow spooling of a near new BB turbo. Anybody comment on what I should do (on a low budget build) to fix the problem? Do all recirc valves leak a little? Is there a hose clamp on fitment BOV that is any good and not 1/4 what I paid for the car?

bmxfuel007

iTrader: (21)

I'm using a Forge RS bypass valve that I have vented to atmosphere. I love how it sounds. It's adjustable and seems to be holding my boost up until I let off on the gas and it does its job admirably. got mine off ebay

should I ask how much you paid for the car?

codrus

I'm told that all BOVs leak a little bit, that's why recirculating is so important in a car with a MAF system instead of MAP.

I think exhaust backpressure is a more likely candidate for slow spool than intake tract leaks. Have you checked the cat? What kind of midpipe & muffler are on the car?

--Ian

triple88a

iTrader: (2)

The bov will leak at idle since there is a lot of vacuum applied to it compressing the spring. Unless its a 2 valve design is normal to leak. When you hit boost where there was vacuum, now there's pressure so it will hold it closed and it wont leak any more. Disconnect the vacuum line that goes to the intake manifold. If it still leaks then it will leak on WOT. Revving in neutral is a horrible way to test a bov since the turbo doesnt really spool up. Also it's sometimes normal for the bov to leak at about 60-80kpa during cruise as the turbo is starting to push and there is still vacuum.

Braineack

iTrader: (62)

should not leak under boost.

skidude

iTrader: (3)

The valve should never leak when the throttle is fully open. If it does, something is wrong.

BOV

There are a bunch of BOVs here. I have one of theirs, and it works very well. My only complaint is that it whistles at certain RPM/vac points.

matthewdesigns

iTrader: (8)

Slow spool is more a function of exhaust backpressure, leaking pre-turbo or a wastegate that is not staying closed than intake track leaks. I recently found my wastegate was loose and gained over 1Krpm in spool.

Braineack

iTrader: (62)

replace 2.5" exhaust with 3", straight through muffler, high=lfow cat, and ensure wastegate is tight and holding closed until boost threshold is met and you'll be singing a different tune.

Rocwandrer

Did you have one of those when I drove your car last fall? Is it the same one now? Did you reduce the spring preload or stiffness between the two times I've driven it? I ask because last fall your car seemed to have a subtle point in th throttle travel where slightly less throttle blew all the boost instantly out the BOV, and slightly more and pressure would build until I needed to lift again to keep the car going where I wanted. With such long sweepers, it was really annoying/challenging to drive around. I could feel none of that this year.

Rocwandrer

Probably I think I nolonger have $100 worth of problem to deal with, and I wonder if people don't just not realize their cheap ATM vented valves are leaking unfiltered air in under part throttle vacuum. Have you tested your Forge valve?

there certainly is some obstruction in the form of dents in the downpipe. There is some soot on the manifold to turbo flange, so there might some high boost leakage there. It seems like that leak is too small to have a dramatic effect on spool....


Sounds like an excellent test to do, and I will. I'm not certain the valve isn't seating well sometimes and not others, but if it leaks with the manifold reference line disconnected, it is going in the trash. I have a gt2560, and it spools enough to blast hot air out for several seconds after reving it hard in neutral. I am pretty certain it is generating positive gauge pressure reving in neutral.

I think the leak at 60-80 kPa is the reason the car came with a breather filter on the recirc valve. If decent low end BOV's still leak in that regime, I guess I'm going to plumb the recirc to actually recirculate.

Rocwandrer

I edited the last reply instead of making 5 posts in a row....

Those that pointed the finger at the wastegate were right on. The tune our Link ECU came with wasn't the worst it could be, because it didn't blow the engine. Other than that....

I noticed the higher the RPM I rolled on to WOT, the faster it built to 5 psi. After that, boost built to setpoint slower than it built to 5 psi, more so the higher the rpm. After I started this thread, but before I event noticed the weird spooling speed cusp at 5 psi, the wrench in this project did some wastegate tuning. What he did made a huge improvement. I think he got it basically as good as possible in 4th and 5th gear. He followed the Flyin' Miata tuning procedure explicitly. The problem is, their tuning approach is to use linear control adjustments to make a smooth boost curve that never undershoots or overshoots. SCREW THAT!!! I want boost, and I want it RIGHT NOW whenever I go WOT.

So I did a test. I rolled on at 3400-3500 rpm to WOT and when it hit setpoint, I noted the RPM. 4700-4800. Then I did the same thing from 3k. 4700-4800. Then I retuned using roughly their procedure, but in second gear. Now it builds to full boost from 3500 RPM roll on to WOT in about 500 rpm, overshoots slightly, then settles at setpoint. In first it is fine. In 3rd it hits the overshoot safety limit, undershoots a little, and then either stabilizes or goes nuts in a narrow window around setpoint. I did not try 4th.

I'd say from a 2nd gear corner exit of 25 mph, I've got easily 50% more average power up to 35 mph, then the same power above that until the shift, if needed, or end of straight section.

I want a wastegate controller that either has a PID controller with tunable integration, constant, and derivative control, or I want it to work like this: Any time boost pressure is less than 5% below set point, KEEP THE DAMNED WASTEGATE CLOSED!!!. Then I want the initial wastegate duty cycle a mathmatical function of RPM. I want the controller trying to close it a little on the low side with change proportionate to pressure error. I want it to open more when on the high side of setpoint pressure, change proportionate with RPM times error.

I'd rather have the pressure hunt a little, loosing some pressure after it has spooled up fully, than live with slower spool up. On average, WOT straights at autocross are frequent and short. Quicker spool is everything on the power side.

I've tried getting into the throttle earlier to get it spooling. Enough throttle to make a difference puts me off my ideal line. I've tried reving with the clutch in, in the middle of a sweeper. That is easier said than done, sounds like a missed shift, works great when timmed perfectly, but doesn't apply to all situations.

So i think I still have issues with the BOV and exhaust restrictions/leaks, but the primary problem turned out to be wastegate control.

matthewdesigns

iTrader: (8)

You may have answered this already but does your turbo have an internal or external gate? If internal, check that the arm movement is not obstructed, and that the flapper is not jammed or blocked in any way. Mine is internal, and the actuator mounts on the inlet side of the comp housing, so I was able to shim it out a bit to tighten it up. If yours is on the chra side of the housing you could probably bend the mounting flange a bit to get a little more tension.

Rocwandrer

Internal. We have checked for obstruction. I don't think it needs tightening since it seems to function very well when we tell the controller to hold the wastegate closed. Our wastegate problems seem to be some combination of limitations in the LinkECU boost control, or limitations in the setup scheme in the FM Link ECU manual.

triple88a

iTrader: (2)

Well the difference is the turbo is spooled enough to where youre leaking air from the turbo outwards so theres no unfiltered air going into the engine. Most of us are running a MAP sensor so that leak really doesnt matter. It actually lowers lag as the turbo is spinning a bit faster in case you punch it.

As far as recirculating your setup or leaving it the way it is.. it really makes no difference. In one case it will leak through the little filter on the bov, in the other case it will leak out of the big filter. Are you running a MAP sensor or the stock maf? If map, find out of the bov is leaking under boost. If it's not leaking, leave it the way it is. If running maf, get rid of ur bandaids and go with a standalone.

Rocwandrer

Wow, stupid mistake on my part. Clearly it is only when the turbo is making so little boost that the intake upstream from the manifold still has vacuum, like throttle tip in, that I am worried about leakage in of unfiltered air.

I think the little filter represents a much larger restriction that the primary air filter. I also think most of the flow that affects spool would, in a recirc, just go right into the turbo, not back out the intake filter. In this case, all the air out the BOV has to pass through that tiny little restrictive filter.

triple88a

iTrader: (2)

Probably idle and up to about 1500 rpm <40kpa is when you'll get air through the bov in the engine but you have a filter on there so it doesnt matter. For the air to be effectively helping the turbo the air will need to be directed to hit the blade head on. That's why the borgs work great however with a random bov a mile away it wont benefit anything but add complexity.

The only vacuum in your system will be in the intake manifold and few inches in front of the compressor impeller, all the other piping such as the intercooler piping will be above atmo.

skidude

iTrader: (3)

Why yes, I actually did tighten it up quite a bit between those two events. Same valve, but before it only had one spring, now it has two and is recirculating instead of VTA. There's a tension screw, but that doesn't matter. It's tighter now. The one I'm using is the greddy type S imitation, I think.

Rocwandrer

I wasn't talking about recirc helping the turbo spool because of the air hitting the input side of the turbo. I was talking about how the breather filter is restrictive enough to slow the pressure relief on the output side of the turbo, making it spool down quicker during shifts.

Rocwandrer

Does that mean increasing spring tension made it more progressive by still cracking on lift throttle, but not opening as far, and thus not letting the pressure vent down past equalizing the pressure ahead and behind the throttle plate, where the last setup was venting to 0 psig whenever the throttle closed more than a certain amount?

skidude

iTrader: (3)

I don't really know why it does what it does, but increasing the spring tension definitely holds the valve closed better. I also added a boost-signal line to the second hose barb on the valve, so it sees manifold vac/pressure on the top of the diaphram, and charge pipe pressure on the other side. Helps it open a little quicker so I can get away with the higher spring pressure than when it just saw atmo on the bottom of the diaphram.