Engine Update - 1997 Mazda MX-5 Miata Long-Term Road Test

1997 Mazda MX-5 Miata Long-Term Road Test

  • Full Review
  • Pricing & Specs
  • Road Tests
  • Comparison
  • Long-Term (10)

1997 Mazda MX-5 Miata: Engine Update

October 14, 2013

1997 Mazda MX-5 Miata

Look familiar? The engine shown in the back of our long-term Silverado belongs to Project Miata. After a hiatus, activity on Project Miata is chugging back to life. Rejoice! The Keegan Engineering-built long-rod engine has lain relatively dormant during this period, patiently awaiting several odds and ends and little fabrication jobs while I traveled the globe, worked on countless other projects and, hey, somebody's got to drink all this beer. In the meantime, the engine sat. And sat.

Hey, better slow progress than none at all. It could be worse. Just look at the House of Representatives.

Anyway, recently I managed to bring the longblock over to Specialty Cars so that head honcho John could tackle items that require actual skill. You see, I can turn wrenches, but am hopeless with fabrication tasks involving any kind of finesse, like welding or bodywork. It boils down to a patience thing. I have none.

John's working on a brace to support the Garrett GTX2863R turbo at its Tial turbine housing. This brace will lighten the load that the manifold has to deal with, true, but there's more to it than that.

Typically, the thinking is to support a turbo's weight by something other than the manifold, and to do it in a way that allows the manifold to expand unconstrained when heated. The idea is that this approach reduces the loads and stress in the manifold and, thus, staves off cracking of the manifold. This concept usually manifests as a turnbuckle/clevis/heimjoint crane-looking hanger thing from which the turbo dangles. Problem is, Miata engines are thrashy and have a reputation for shaking the bejeezus out of anything attached to them. When you mount a turbo to a manifold, the turbo's mass lowers the manifold's resonant frequency. This resonant frequency is excited by the engine in operation, and the manifold lives a short life as a result. The problem with the hanger approach is that it doesn't really increase the natural frequency of the manifold/turbo system at all, and so the manifold (while relieved of turbo weight) is still subjected to vibration-induced pain. Don't underestimate the violence of a BP at full song. It has shaken many a manifold (and other ancillary items) to death, hanger or not.

So, the goal instead is to stiffen the manifold-turbo system up. Raise its natural frequency so that it evades excitation by the engine. Project Miata's brace will pick up the 321 stainless boss that was welded previously onto the turbine housing (as close to the turbo-plus-downpipe's center of mass as reasonably feasible). At the other end the brace will tie to the engine block's unused bosses just aft of where the A/C compressor lives. It turns out that many OEMs have adopted a similar approach on turbo four-cylinder engines. Look for a rigid brace that ties the manifold or turbine housing directly to the block.

The new Mercedes-Benz CLA45 AMG has probably the most elaborate turbo support I've ever seen, complete with springs to accommodate expansion. Project Miata's brace will be simpler, more like what's found on factory Mazdaspeed Miatas and Mitsubishi Evos.

John will also be sorting out some of the turbo's coolant plumbing. A turbo wants the coolant path to be as "uphill" as possible, so source coolant from the lowest point, and route this continually uphill to the turbo's center housing. Same idea applies to the coolant port exiting the turbo — no low spots — so that the turbo's "thermal siphon" action can work effectively when the engine is shut off. More turbo-geek reading on this subject on Garrett's website.

In our case we will source coolant from the port on the driver's side of the engine with a -6 AN banjo from Goodridge (Love Goodridge. Very high quality fittings, available in virtually every conceivable configuration. Peruse their catalog and see for yourself.) since a straight fitting will foul the downpipe. On the return side, we're thinking of welding a -6 fitting onto the heater core bypass tube that runs right by the turbo. This is the same location used by the factory in the Mazdaspeed Miata.

We'll see what we come up with. And by "we" I mean John.

Jason Kavanagh, Engineering Editor


  • duck87 duck87 Posts:

    Somewhat interesting (not) is that the bracket for your Dart multi-air is a pretty beefy piece, namely because for whatever reason the 1.4L turbo has the turbo facing UP and sitting on the manifold. I noticed this in the Abarth and it was kind of odd. As for coolant pathways, keep in mind that if you have the turbo as the highest point, you need to be very careful with bleeding the system. Many OEMs are turning towards coolant afterrun pumps and/or continuous bleed through a hot-bottle recovery system because of the obvious risk of local boiling in the turbocharger upon shutdown and when running the vehicle very hard.

  • Really appreciate this update. Still a bit more time before the swap will happen I see.

  • noburgers noburgers Posts:

    I thought you had more patience than you let on. I am really excited to see something happening with this project. Updates have been so long overdue, I wondered if you were moving forward with it or not.

  • bassrockerx bassrockerx Posts:

    im not an expert but i would imagine you can modify an engine bracket that has a some kind of bushing material to suport the turbo. allow a little flex so that the turbo is not strained. the spring idea is brilliant because you can get a progressive spring that is harder to compress the further it travels so it not only supports the weight but provides like shock absorption

  • mechengftw mechengftw Posts:

    Any news??

Leave a Comment

Past Long-Term Road Tests

Have a question? We're here to help!
Chat online with us
Email us at help@edmunds.com
*Available daily 8AM-5PM Pacific
Call us at 855-782-4711
Text us at ED411