It's that time of the month again. After another race where some slow car tries to hang, and I spend the whole time modulating between 70 and 80% throttle to keep the boost under the 18 psi boost cut.... I decided to get back in there.

The FM holder.. leaves something to be desired. Admitedly, I'd lowered the can (it used to be at THIS angle


if you install it how they ship it). That always seemed a bit funky to me, so I'd lowered the wastegate so it fed straight, but probably pulled it back another 2mm of preload. This was the straw rebreaking the old camel's back war wound.

Finally I had a dig in a friend's box of turbo parts (Thanks TeamPlur!) and found an adjustable wastegate with a long throw. It's much lower pressure. The FM provided one really starts to open at ~9 psi, and doesn't fully open till 15. This one starts around 4 and is full open at 10-11. Much better.

Anyway, the shaft was too short, so I welded a piece of stock on the end. LIFE SAVER. I left another hole in the middle in case this ever sees use somewhere where that would help.

5 psi is JUST where it starts to move. I welded the can onto the bracket (I didn't need to, but I wanted to. :-) You have to watch you don't burn up whatever soft bits there are in there, I weld a bit, spray with water, weld a bit more....




Anyway, here's the end result:



This is the wastegate closed, the middle position is the FM setup, and the rightmost is the new one. I've got a decent amount of preload on there but I can add more. I'll need to drive it and report back, but I wouldn't be surprised if I can run 12 psi again. My poor stock rods are begging for it. Anyway, you can see I about doubled the travel.

Lastly:

That's my preload (not really in force, but whatever), and there's plenty of adjustment in both directions. You can see everything lines up straight. If someone wanted to produce this, they'd just need a piece of pipe, tap it, and hammer the end. Done deal.

 

Looks like I'm between 8 and 9 psi, steady state. I didn't touch the MBC, so I have no idea where it was set.

What's odd: I seemed to have ridiculous overshoot?! I only tested it once, and I was more looking for cops, but the time I watched the overshoot closely it was like 12-14 psi or some such. But then it holds nice and steady well into redline. I was thinking perhaps the dead volume in the can is higher but I don't think so, it's smaller. Perhaps the 'leak' in the MBC, which is calibrated for the bigger can, is somehow too big for this one? Seems doubtful.

This morning I got the feeling that maybe I'm building boost more slowly, but I'll check. Time to use the logs. Anyway, the car feels way more refined at 8 psi steady than at whatever overheated, overspun, stressed state it normally is. Now I could set it for 15 PSI and tune it.

 

UHG! So, it seems 8 psi is the highest I can boost?? I get a couple psi of overshoot, but... I dunno!

My MBC was a little dirty inside, so I cleaned it out. On one assembly, it would absolutely hold pressure. On another, if would leak very fast to the set pressure, then leak very slowly after that.

I guess the smart thing is to see if, with the wastegate signal tubing unhooked, if it holds pressure. If not, the can is too small (i.e. 5 psi over 1.5" diameter's area is much less force that 12 psi over 2.5" diameter's area - over 6 times) and the pressure on the flapper is pulling it open. To a small degree, more preload will help here, but not enough. My remaining option is to add an assist spring, which always seems like a good idea, but rarely works.


Edit: Also, it seems like I'm building boost more slowly. I'll need to look at some logs for this.

 

Yep, even with the control line unhooked (wastegate exposed to atmosphere) I get the same spike to 14 psi, level out at 8-8.5.

The one on the car was too hot to test, but the old wastegate was ~23 lbs to open halfway, another candidate replacement I measured was around 19.5. I have a 9 lb helper spring I might put on if I need it, but I don't think I will.

Maybe the wastegate just isn't closing all the way? Will get a look tonight.

 

You need a larger diameter wastegate can to prevent the flapper blowing open.

 

I think that's going to be the answer. Well, it's just how strong the pull is. And that means an assist spring should be just as good, only it'll raise the pressure it opens at somewhat.

 

Send me your turbo, I will port it for you @ n/c.
Stephanie

 

Thanks Steph, I really appreciate that!

As it turns out, though, I've already done a bit of a job on it... And don't plan on having the turbo off any time soon.

Because:
It works *perfect* now! I haven't seen how low a boost I can hit, I was running ~12 psi at the time.


What I did:
I added a helper spring as I discussed. I took a scale to the wastegates, and found the OEM unit to be ~24 lbs to open (once I measured 33 to get it all the way open) - that's lbs absolute, not psi.

The other wastgate I had was ~19 lbs. The one on the car was ~12, so I think I know why it was coming open. I added a helper spring of ~ 9 lbs at a moderate stretch, and it was mostly ok, back to the stock "FM" set up - i.e. I could JUST stay out of boost cut in low gears by easing into the throttle, and keeping the pedal off the floor. Still, it seemed to hold 14-15 psi pretty well in 4th, 5th, most of the time.

I couldn't shorten the spring or there'd be no preload (hence no effect). If I lengthened it, there'd be even more force. But I attached the spring to the halfway point (or so) on the wastegate flapper arm, which should halve it's effect, PLUS halve it's effective spring constant (as a function of actuator travel: i.e. for an inch of travel of the can, the spring is only stretched one half inch.) This means the force becomes more additive/constant, and doesn't increase as the gate opens.

Result: I get minimal overshoot, since the gate actually opens when I want. I hit a NICE steady 12.5 psi very quickly, with reasonable spool. It's not much of a tune (stock rods and laziness), but I get 9 psi by 3,000 RPM, and spike to 13.5-14 psi by about 3700 rpm. There's plenty more in the turbo.

Now second gear instead of pulsing at 1 hz due to the overboost protection, it just holds dead solid all the time. No thinking, just drive the car like you want to. It's pretty amazing, I'd forgotten how much fun the miata could be.

Net pull on the new set up: 22.5 lbs, about. I didn't check what PSI the can opens full at. Pics:





I still might take you up on more porting, though I thought I'd done a reasonable job I'm sure you guys have a magic touch. :-) Not till I get the new motor ready.

 

I wound up with a very similar setup - small 8-9psi can with a helper spring. BUT I didn't straight my rod out like you did and eventually it failed in the can. I never got a chance to open up the can (garage crushed) but it looked like the rod end was flared and then "trapped" into whatever disc was on the WG side of the diaphragm. Yeah, chinese cheap. And that lateral load from the rod angle just wore it out. As ghetto as the helper spring looks, it IS a totally viable solution. No different than swapping the spring out in the can (if you could).

 

There is one caveat: A well designed can, as Jason suggested, really IS the best answer. A large spring can be overcome by lower pressure with a big diaphram. Adding spring only raises the the base pressure it takes to move the wastegate, and the force increase per unit movement.

I'm still not totally happy with the set up, but I also don't feel like spending $50-75 every time I want to "try" something new.

Someday, a external wastegate is in my future. But that means a new manifold. And that means good time for a new turbo. And then it becomes a whole new car with piles of money shoveled into it, where as (even counting buying the welder) this was a $100 fix. :-)

Curiously, I turned up the MBC adjuster last night, what I assumed was only ~ 1 psi. Boost was ~13 (so I don't know if it made a difference or not) but I did hit overboost cut once each hard flooring in high RPM second gear. I'll try turning it back down, or perhaps it's a daytime thing or...?

I can't imagine MORE porting is even terribly possible. If anyone thinks they could tell what I did from my terrible pics, take a look here:

AbeFM : photos : Fix for ridiculous Boost Creep on FM IC kit III - Porting
-Abe.

 

There are internal WG actuators with replaceable diaphragms and springs - same stuff used in the external gates. Also the same cost... if not more than an external gate.

This seems like a good time for you to jump into dual port actuators. I'd just started working with one when the tree fell on everything.

 

they also machine a channel out of the housing so exhaust can flow more easily into the wastegate, not just port the opening. It's an offer i should really take them up on, but i hate pull the turbine housing.

 

I just hate to be turboless for so long! A friend wants to borrow it as he's developing some 8v volvo manifolds so he keeps trying to get me to pull it.

I did SOME channeling, but I'm not sure how much is enough. That was the point of the straightedge in the picture, to show how much I removed.

Stephanie, can you comment on how much I took out compared to what you do? Trade secret? :-)


FYI, backed off on the MBC a bit, and stopped hitting the cut, still at the same 12.5 psi. Somewhat strange, but SO much more confidence in turns not waiting for the lurch from overboost cut that I'm a lot faster. Not that I would disobey the posted speed limit.

I'd enlarged my bleed hole in the MBC at some point, perhaps I'll shrink it down again. That should help with the spiking.

 

I love fixing creep.

I've always been privy to it since I went with a 3" exhaust. My solution was running an EBC at my peak boost (around 12psi). I was able to tune it out pretty flat, however the MS EBC solution is anything but great.

After chatting with Abe for a bit, I decided to make a few changes. First off was going swapping to a MBC, no more temperature dependent boost levels.

Then I lengthened my wastegate arm a touch. That helped immensely with the creep, allowing the wastegate to open just a little bit further helped flatten out the boost up top. The only thing was it killed my spool, but for the most part I was happy with at a constant boost level morning or night.



After this I moved my boost source to after my [new] IC (yet before the TB). Walla, the planets have aligned. Moving the boost source after the IC helped the turbo ramp up faster and I'm left with solid boost to redline.

 

wow, that's how it's SUPPOSED to be. (and walla --> voila)

amazing too that it made your logging colors brighter.

 

That so shouldn't have worked. :-) Still, I'm glad it did.


Slight update on my set up - I put in a bigger vavle, I did some testing and found that the FM supplied MBC valve doesn't flow that well. So I took a much bigger one and found... that it doesn't flow a lot better. :-) But it is somewhat better, and pretty cheap. I soldered on hose barbs I'd bought at home depot and cut in half (actually, I lathed nice ends on it so it would solder well, mainly to save weight threading on adapters).

This valve is interesting, the poppet (is that what it's called?) is not a ball. It's a disc with an o-ring, so the seal is very good (the FM one - really just another mcmaster one, I'll call it the "ball style" for now - leaks as you blow into it. Also, my sealing surface had got dirty over time). There's a long set of fins to keep it centered in the hole. Like a very cheap, o-ringed engine/head valve. The adjustability it awesome. I'm not sure how great it is over temps, etc.

Since I'd put barbs on it, I drilled a hole (increasing sizes) in each step of the barb, so by putting the hose on further I could exponentially decrease the air leak which allows the wastegate to close on throttle closure. In theory this should effect boost building after a lift/shift.

I go out to test it and get 6 psi. Sigh, my spring assist broke.

So I went and got a great spring (from my collection of springs), and I attached it to a timing belt cover bolt to keep it away from the turbo and it's heat. I put a turn buckle on it, and now I can set the tension and protect the spring from the heat at the same time, as well as select a length.

So far it's working very well. The car feels weird with predictable boost, stepping on the gas almost provides an understandable relationship between throttle and power.

Pics to follow.