my former Greddy setup

my former Greddy setup

Before you continue on this page, I recommend this page on Greddys:

http://www.greddyturbofaqs.com/

I used to have a 1.6 Miata in Manila. The Manila Miatas are the Euro/Asian version later ('94-'97)1.6's, which are rated 95hp and use the same cams and pistons as the US 1.6's with automatic transmissions. Cams have 10 deg less duration, and pistons are 9.0 compression ratio.

My Greddy setup was 6 psi, with a small intercooler, a BEGI AFPR, an MSD, and a puny exhaust with straight-through mufflers.

Note: Dyno plots on this page are done in 2nd gear. Dynojet runs in 4th gear, as is common practice, show around 8 hp and 10 ft-lbs more. See comparison. These plots also do not use the SAE correction of around +7% from the perennial 95F temperatures and 95% humidity in Manila. To get dynojet equivalents, add about 15%.

When I first got my 2nd hand Greddy turbo, the wastegate was set wrong. It's set like this from the factory for some bizaare reason. It was open even with no boost pressure. It should be closed, and will open only when the set pressure (5 psi for a stock Greddy kit) is reached. Here is the result of shortening the rod until it's just closed. Notice the huge improvement at mid-revs.

Next I moved the wastegate sense signal from the compressor outlet to just before the throttle body, after the intercooler. This way, if there is a drop in pressure across the

intercooler, the wastegate will make up for it.

It was losing more than a pound of boost across that tiny intercooler. A bigger one would have been good.

Just to make sure that 14 deg base timing is not overadvanced, I check the dyno results of 11 deg timing:

11 deg loses power everywhere, so 14 is better, and I don't get detonation.

Next I try to up the boost by using fishtank restrictors bleeders. It now says 7 psi peak but droops to 6 psi near redline:

That's +10hp for $0.40! (Addendum - I took it out when I sold the car because I didn't want the new owner to diddle with it) (Bottom trace is stock)

I wonder why the improvement begins at 4.7k rpm. Is my wastegate cracking open before it reaches 5 or 6 psi? So I try to do a quick dyno run with the wastegate

disconnected, until the boost gauge hits 8 psi, then I let off. The result is no additional power until 5k rpm - that is, the wastegate isn't beginning to open until 5k rpm, by which I

have at least 5 psi.

I took off the fishtank thingy and just shorten the wastegate rod as far as it will go. The results are quite close. I also tried my BEGI AFPR, which after several rebuild tries

(without the rebuild kit) doesn't work. The car is obviously rich below full boost.

The Greddy AFPR is pretty lousy. See AFPR saga. I tried reversing it, and it was very different, but not necessarily better. Instead of being too rich everywhere except full

boost it was lean at low boost (<3psi).

I finally installed my O2 sensor (see install story) and find a huge lean spot around peak torque.

So I tried building a special circuit to solve the lean-out and that helped big time. See report.

Anyway here's the circuit's effect on power:

and torque:

Wow, nearly +12 ft-lbs!

Black is with the fuelling circuit. Red is with the Greddy AFPR as I got it, Blue is with Greddy AFPR reversed.

Wow, it must have been really lean. That's +10 hp at 4.7k rpm! I lost at tiny bit at the top probably because it likes to be a bit less rich there.

Then I find the dreaded Greddy exhaust manifold crack, so I have it fixed. While I'm at it, I had the exhaust manifold "ported". Actually I had it port matched, "dingle berries"

cleaned out, and corners (that the gases have to turn) radiused. Result is +6 ft-lbs and +6hp: Quite a lot for such little porting effort.

I strongly recommend anyone buying a Greddy kit to do the mod described and to port it before installation.

So here is my % increase now compared to how it was stock, with K&N filter and 16 degrees timing:

Wow, that's a 60% increase at peak torque and >50% increase in peak power! IMHO that's a substantial increase, all for <$1,500. In contrast, an 8 psi non-intercooled Sebring gets a flat +32% from 1.5k rpm up. (very nice low-end though)

As a quick experiment, I put in fishtank bleeders to get 9 psi, and retarded timing to 11 deg (from 14) to stop detonation. Indeed, there was no detonation. Air fuel ratio monitor

shows proper richness. Here's the result:

That's + 17 hp for 141hp peak. (With +15% correction it's closer to 162 hp)

And about 5 ft-lbs per additional psi above 5k rpm. That's diminishing returns, it should be 7 ft-lbs per increase psi of boost. Note that it's

slightly weaker below 4.5k rpm, probably from the reduced timing, and the gain is only present above 4.5k rpm. This is a bit annoying that the gain is over a narrow band of 5k

rpm up. Sure it will help in a drag race or in a full power pass, but if you're dawdling along at 3.5k rpm and mash it, hmmm. You can tell I'm not very impressed.

The tach needle sweeps so rapidly now above 4.5krpm that the fuelling circuit is now playing catch-up. Thus the kinks in the curve above 4.5k. I may have to do dyno runs in

3rd gear ( - my dyno strip isn't long enough to do it safely though - )

I have to get more boost at low rpm. +10 ft-lbs in the 3k to 4k region would be nice. Hmm, bigger intercooler, custom bigger downpipe, bigger exhaust. The Greddy downpipe

looks restrictive, the wastegate outlet is < 3/4th inch into what's practically a wall. If I could put in an exhaust cam that opens earlier - like the manual transmission cams but

advanced by one tooth. Maybe I should measure exhaust backpressure. A 6 inch length of 3/16th copper tube brazed into the exhaust pipe would do it.

Comparing my 9psi run it to Soren's car (US Miata, 8psi BEGI IV) http://www.turbomx5.com/

Here I put in some estimated correction for temperature and humidity, +8% for my run and +4% for Soren's run. Soren's dyno run was done in LA on a cool day, whereas

Manila is always hot and humid. was extracted using my program too. Unfortunately at the time this run was taken he was running a bit too rich and it was hurting power a bit.

BTW I forgot to mention that during this run the boost brielfy spiked to 10 psi, and that's what gives the high torque peak at 4.7k rpm.

Note that Soren's is still a bit more powerful up top - probably because of the cams and his porting, in spite of 1 lb lower boost and richness. Note my car is torquier below

2.5k rpm - that's the effect of the cams. And, the BEGI spools up quicker - 10 ft-lbs more in the 3.5k to 4k RPM range. This is because I have a puny exhaust.

Or looking at it another way, he gets the torque 500

rpm earlier in that region. Soren had a 4.3 differential while I have a 4.1, so that should be a disadvantage for Soren's turbo spoolup in the dyno plots. In spite of this he has

more midrange torque.

16 Feb '00

I rebuild my BEGI AFPR to put in the new shim that Corky Bell sent me. Flat spot between 1 and 3 psi disappeared, but the end dyno results are the same. Air fuel ratio

monitor shows a momentary leaning out near peak torque.

I did a 2nd and a 3rd gear run at 7psi. Here is my 2nd gear run at 7psi vs . my previous run at 6 psi

Here's a 2nd gear vs. 3rd gear run. 3rd gear data is valid above 3k rpm.

Wow, that's a lot of torque. Adding 15% correction is around 162 ft-lbs. Power is above 143 hp from 4500 rpm up. In a drag simulator, this yields a mid 14 sec quarter mile, from a 6 psi turbo setup costing $1,500.

I'm within 5 hp of peak power from 4600 to 6700 RPM, a spread of 1.45x - that's a very wide power band. I really think if I put a bigger intercooler and exhaust the torque will drop less at hi rpm yielding more topend, and the bigger exhaust might improve spoolup, improving 2nd gear torque from 3000 to 4000

rpm.