E-cool or WI?

[ckkcmx5]

Hey guys,

I've been running my 94 Coldside with e-cool and MegaSquirt with Knock sensor. With 130mm pulley I can't get rid of the knock even running 15 degree of timing at 4000-6000rpm under 10-12 psi of boost.

I need to reduce timing to around 9-10 degree to get rid of knock.

I recently added the AEM water injection kit and disconnected the e-cool injector but the result is disappointing. The intake temp seems only dropped a few degrees than e-cool but knock occurred even more than that with e-cool.

I have tried to operate both system at the same time but result is not good. I think the fuel and water being injected there will interfere with each other.

I have the water nozzle installed at the default WI location. And the intake temp sensor installed at the 1/8 NPT hole next to the WI location. And I also suspect that the sensor being placed there will not reflect the true intake air temp as it is "before" the water nozzle.

I guess one or two of you here have the same setup (MS + AEM WI) would you share your experience?

Thanks a lot~!
kevin

[Steve in VC]

I recently added the AEM water injection kit and disconnected the e-cool injector but the result is disappointing. The intake temp seems only dropped a few degrees than e-cool but knock occurred even more than that with e-cool.I just quoted you because I wanted to see this in print, twice.

Running E-cool and WI is possible, but it sounds like there is some really wrong. I would focus on that first.

Do you have an wideband A/F? What does it read?

Knock occurs when the fuel & air are hot enough (heat from combustion and the pressure from compression combined with the pressure from burned fuel / air). Retarding timing reduces the pressure from compression.

Are you pulling timing only on boost? Do you get detonation boost at all? How does the car run when you have the timing retarded? Is there any possibility the crank sensor is way off?


WI should have cleaned off any carbon deposits, minimizing the possibility of a piece of carbon acting like a glow plug. Have you pulled and looked at the plugs? What does the inside of the cylinders look like?

It could be a dirty injector, running one cylinder lean (thus hotter).

Does anyone know if you can pull each injector lead, one at a time, then rev to boost without harming the engine?

[Tom @ Fast Forward]

What brand/grade of fuel are you running? Here in Arizona, for example, Diamond Shamrock (Valero) is great and the car wont hardly run on Shell. When I was in Texas, the car would hardly run on DS/Valero but ran fine on Chevron and other brands.

All we have is 91 octane and I ran my 130mm pulley on 0* retard below 4000, 4* retard from 4000-5500 and 6* retard from 5500-7000. Normal timing under WOT on the Miata is about 24* so I was running 18-20* from 4000 to redline. That was on 91 octane in Arizona heat.

[ckkcmx5]

I just quoted you because I wanted to see this in print, twice.







haa...sorry I know it doesn't make sense but I tested 2 to 3 times alone with each system back and fore and give the same odd result.

I have a AEM WB and it ran 12:1 below 8psi and 11:1 for all above 8psi.

Crank sensor is at the right position as I calibrated it when installing the MS and tested again 2 weeks after that.

Fuel here in Hong Kong is 95 or 97 which should be good enough.

I may post the timing map here later on but even the knock sensor retard timing the car still running smooth.

[ckkcmx5]

Tom,

A side question, as I'm now using MS to handle four 460cc injector and the pulse width will be much less than orginal, will this affect the fuel amount being injected by the e-cool as the PCPRO is depends on the #1 injector signal?

[ckkcmx5]

http://www.ckkc.net/mx5/timing1.jpg

Here is my timing map used by MS. Area within the red border is the range that knock mostly occurred. When the knock sensor found knock, it will retard timing by 1 degree, most of the time it retarted 8 degree sequentially in that area.

I believe the map without boost is a bit advanced over stock but all area in boost is reduced to 15-20 degree.

P.S. those 4 injectors are 460cc and cleaned before installed.. but I also thinking if one cylinder running lean. Need to check later.

[Tom @ Fast Forward]

The 5th injector only uses the #1 injector to generate a tach signal. Pretty sure I have it set to ignore the duty cycle (load) value. You could turn it up a bit with the big pulley though. Easy way to tell if it is close is to read the A/F at 4000-5000 RPM. Then disconnect the 5th injector and run up to 4000-5000 RPM again. You should see about 1 point difference. If you had 12:1, it should go to around 13:1. Just don't go above 5000 RPM and stop if you get knock.

Someone will have to convert the fuel rating for me. I know you use the same rating system as Europe and 95 or 97 is probably like 91 over here?

We are just finishing up an install on a 96 that had #3 injector not operating correctly. Normally lean (luckily VERY lean) then would stick open if you pushed it hard. We had 12:1 A/F but it was from rich 1,2,4 and lean 3. I say luckily VERY lean as it was so lean as to not make heat. Like basically no fuel. You might want to have the injectors flow tested if you haven't already.

[FormerDatsun510Man]

One thing I notice is that according to your timing map, the timing looks awefully advanced at mid to high rpm at 101 Kpa and 84 Kpa zones. I think perhaps the knock may be occurring because you have the timing initially a bit too advanced when going into boost, even if you are bringing the timing back down while into boost. In other words, it takes time for the retarded timing in boost to take effect and once knock starts (even if at only 1 or 2psi) it won't stop until the timing is drastically retarded. Better to keep the timing more conservative from the get go. What we know works here is that the stock ECU timing coupled with only a few deg of retard (as Tom pointed out) works well with controlling knock. My suggestion is that you might get better results if in the timing map you mimick the stock ECU timing curve as close as possible in the ambient (100Kpa and less) zones and then bring it down just a few degrees in the boost zones. The stock ECU never sees more than about 24 deg at full throttle and ambient. So in the 101Kpa and 84 Kpa, try reducing the max timing to about 24 deg. Perhaps pull the 75Kpa zone down a bit to, to around 28 deg max. Below that the stock ECU also gives quite a bit of advance like your Map shows. Anyway, I think part of the problem is near ambient there is too much advance.

With WI it is important to both have the right water:fuel ratio and also the right air/fuel ratio. I would shoot for around a 15% water:fuel ratio and an a/f ratio in the 12.5-13:1 range (remember the water is taking the place of the normally rich fuel mixture for cooling purposes). On a similar parallel, I have found that with E-Cool the amount of fuel added by the 5th injector is very important for best results. Going either too high or too low can increase knock. It is all about the balance. The procedure Tom mentions for setting the E-Cool is exactly what I ended up doing and found it to yield the best results in terms of both power and stopping knock. I tried going with more than 1 point of E-Cool and also less, all while adjusting the main injectors so as to maintain 12:1 a/f ratio overall, and found that it worked like a charm set at 1 point. Tom let me learn the hard way so I learned it well :).

Bill

[FormerDatsun510Man]

My FFS line is below

[Tom @ Fast Forward]

Kevin,

I will be in HK in a few weeks. Let me know if I should bring anything. If so, it will be my furthest service trip. :) Bad part is I don't think I can drive there. :(

[Satisaii]

LOL, I might have to be out there myself. But not in HK, Nansha instead.

[Tom @ Fast Forward]

If you're driving, can I ride along? My engine is broken or I would drive. :)

[Satisaii]

Don't like the flight? You can take a boat. Long trip, though.

We would need a James Bond Lotus especially equipped for long trips in order to have a chance of making at drive.