Some of you know about my miata build with a K24z3 engine ect. I started this build a few years ago before the billet manifold was available. There was a part available that was v-band flange welded to a four bolt flange. I figured that if it's welded it's just going to crack and break so why not just weld up something myself? (that's going to crack and break) So I did.


This lasted about a year before the first cracks formed at a very inconvenient time. I think I posted about it somewhere? That weld lasted a while longer and was totally destroyed at a track day from the extreme heat. So then I mig welded the crap out of it (manifold as pictured). Then the engine blew up. (couldn't mig weld it)

I really liked the control I got from the external wastegate.

So here we have the latest creation from KPower. The part I would have bought if it was a thing back then.



Bonus, it works with lock nuts. Just clears them. The problem I started with is having to buy a whole new turbo with internal waste gate. Then I had to fab up a new down pipe.



There was even less room to turn down the pipe and I had to buy a super tight radius elbow. 4 5/8" outside radius. It was still too tight and I cut it back about 3/8" which makes the fitment in the flange oval.





The wastegate is really in the way here. You don't get to choose the placement like an external gate. I used a 5/8" extended drain with a slight bend in it. You can find them on ebay/amazon. The bend wasn't enough and I had to hammer bend it some more. Also the clearance to the frame is tight. If you want an upgraded waste gate, it might take some notching.

Now the biggest problem with this setup isn't the manifold. I think it's a good part and it places the turbo well. The space around the turbo is tight but it can be overcome. The problem is
BOOST CREEP!
This sucks. I want to be able to control down to a few psi for control in 1st and 2nd gear. My external setup would handle 5-7 psi low boost all the way to red line. The part number on the actuator is 480009-9 which should be 0.5 bar.



I can't get it to run less than 190kpa absolute which is pretty much the most boost I want to run at. (13psi gauge absolute) I disconnected the actuator and let the flap stay fully open. It just boost creeps all the way to 160kpa. So that's my baseline I guess. The actuator doesn't seem to have the full stroke to open it all the way either. It takes 20psi of compressed air to move it all the way. Such bs.
So now I need to figure out how to get it to open all the way at low boost and stay fully closed without leak by. My controller is the megasquirt and it will operate a 3 port mac valve. I can totally block pressure to it until it needs to open.

Has anyone used any aftermarket wastegate actuators? I would think this situation requires a top and bottom port actuator with a low boost pressure spring. I really didn't want to spend $300 on a turbosmart actuator but that's looking like the route to go. (que next post about buying turbosmart actuator)

I've looked at some hacks about adding external springs to help open actuators at lower pressure. The problem is the arm isn't straight and I'm afraid it will cause leak by. I was thinking about testing with a tension spring mounted to the chassis though.

Has anyone ported the wastegate in a G25 turbo before? I want more flow.

I was thinking about putting an air compressor in the trunk to have a 20psi pressure source to operate it. The computer can run its closed loop boost control to do it's thing. But that's complex and adds weight. Complex is fun though.

 

Yeah something's up with your setup. I didn't have issues with boost creep even running 4-5PSI on the same motor/manifold/turbo setup(I didn't see the exact turbo mentioned, but looks like a G25. I ran a G25-660). Porting the internal gate is 100% not needed.
Needing 20PSI to move your gate... well ding ding ding, that's your issue. Either the arm is jamming or the internal diaphragm is FUBAR. The turbosmart gates are very nice, and yes get the dual port so you can run your 5PSI spring and still dial it up to +20PSI at will. I think you need to trim a bit more of that chassis support to fit it as its a bit girthier than that can you have on.

 

It is a G25-660. Even with the arm disconnected and the flap all the way open it still boost creeps. The arm seems to move freely from about 7psi to 20psi. I set the preload to 2mm. It's not leaking by the diaphragm. So idk.

 

I have the same actuator on my g25-550 with the big hot side. Pretty much the same manifold. Mine is custom designed to be low mount but I’d expect the flow to be very similar. I have no issue running 7-10lbs all the way to redline. No more boost with that without commanding it. Do you have a crap load of preload on the actuator?

 

I have 2mm of preload. It creeps even with the arm disconnected.
I am running e85. Are you?

 

yeah pump e85. I don’t have a lot of final photos of my setup. Here is one in process.

 

How to port a G25 Turbo: Hell if I know, hold my beer.



These turbos are already ported really well to direct the exhaust flow. But I see that ledge could be rounded off and get a little bigger on the diameter.






Rounded and enlarged. All the work had to be done through the inlet. I used a small right angle grinder and a good 3/8" round end bit. I only enlarged it about 2-3mm.



The flap hits so I clearanced it. The actuator doesn't have enough stroke but I'm keeping my options open. pun intended.

 

25 Kpa reduced with the porting. This is with the wastegate disconnected on both runs. So this is best case low boost. I'll reconnect it and see where it's at now.

 

Sometimes I do stupid **** with the stuff I have laying around. Heigh pressure air control anyone?

 

Any idea how often you'll have to fill up that tank or are you still thinking you'll get a setup onboard to refill it?

Any testing on the custom firmware to run it yet?

 

I have the air horn compressor. I just need to install it when I know this works. I tested today. There are bugs to work out and tuning to do. But it follows the commanded target really well. Steady state isn't working as well yet.

 

That looks really good for a test run. I'm sure it'll be great once you've got it dialed in.

 

I think I may have some kind of mental disorder. I've somehow written a few hundred lines of code to create a boost controller instead of doing the sensible thing. Why am I the way I am?

 

Can I introduce you to our lord and savior the Link G4x?

 

But then how am I going to do mad scientist stuff in the off season?

 

New thread to cover the boost control stuff. https://www.miataturbo.net/diy-turbo...st#post1674297

So far, no issues with the turbo setup.