shlammed

iTrader: (8)

Hey, joining the AEM crew shortly... just wondered what the MIATA guys have been having issues with that have went unspoken on here.


feedback on the setup besides the first impressions i have read here would be great.


im on the AEM forums too i just want to compare what i read there to what you guys are having issues with.


thanks

-Matt

jtothawhat

iTrader: (7)

I have one, I have had no problems...

Full_Tilt_Boogie

iTrader: (1)

Its a great ECU with great software, its just exensive

IcantDo55

iTrader: (11)

Did not work with my ford dizzy in my mustang forcing me to go to LS-1 coils......oh wait we are talking Miata not Mustangs. Carry on.

JasonC SBB

1) Very random misfire out of nowhere, like every 15 mins. Slightly annoying when it happens. Cause unknown. I've seen this in 3 ECUs and 2 cars, inluding a miata ECU and a Honda ECU.

2) No 'D' term (the D in PID) in the boost controller. Therfore difficult to tune out overshoot when running closed loop boost.

3) No 'D' term in VVT.

4) No raised idle speed *target* for when a/c is on. (It does have a rough idle up function which simply adds duty cycle to the IAC.

5) Can't have 2 separate temperature setpoints for the 2 fans

Of the above the lack of 'D' in boost control is the most irritating.
The misfire is the only one with no workaround.

Other than that it is very good. It does a lot of things few other ECUs do, and some which not all ECUs have. Just off the top of my head:

1) fast datalogging of ANY variables, internal ones included
2) individual cylinder spark and injection trim
3) enrichment on high EGTs
4) copy tables to/from Excel
5) inj duty cycles can be multiplied by MAP so that it looks flat vs. MAP - this makes tuning/reviewing fuel much easier
6) can use MAF input

shlammed

iTrader: (8)

Great!

I doubt many people are using the V2 system yet, but if there is, it would be nice to see if they are having the issue with the missfire.

TurboTim

iTrader: (9)

I got this issue as well. Very bad with the 1.6 coils/external ignitor (every 15 min bad misfire), better with 1.8 coils (every hour or so), with the COPs/sequential ignition it has done it maybe 2 times ever. The tach still drops often, but no actual misfire.

I don't get your lower #5. Are you talking about the 3 different ways of viewing the fuel map (raw bit data, duty cycle, and pulse width in msec)?

EDIT: I like the fact that you can configure the check engine light as an indicator of basically anything, as wired. I've used to indicate when it senses knock, when my antilag is on, right now it's used to show when the radiator fans turn on, to see if my setpoints were correct with the coolant reroute. It's good to know if your fans are running when they shouldn't be.

And I like my Serial Gauge, which can be configured to display a bunch of different parameters in a 2" gauge.

JasonC SBB

#5 is aka "boost comp" method. The Electromotive TEC was the first to do this, and they called it "linear thermodynamics" or some such.

If you set it up as a straight, then your fuel map looks flat. Instead of actual on times, it roughly displays voumetric efficiency, which doesn't change much with MAP.

e.g. at 4000 rpm, instead of

25 kPa - 1mS
50 kPa - 2mS
100 kPa - 4mS
150 kPa - 6mS

you get

25 kPa - 4mS
50 kPa - 4mS
100 kPa - 4mS
150 kPa - 4mS

.. adn the AEM automatically multiplies the numbers by MAP/100.

Makes viewing understanding and editing the fuel table much easier.

TurboTim

iTrader: (9)

Oh ok, I know what you mean then. From what I recollect from the AEM class I went to, there is a 2-D fuel trim table (map & fuel trim) that you use for this. You set your fuel map to all the same values per rpm, and the trim table adds fuel per MAP. I forget what the value was, but on all their cars and all sorts of setups they found a value of xxx more fuel per PSI worked pretty close.

Or I could be way off on this, it's been many years.

I would find that more confusing, as the fuel map is flat like you said. I like seeing the fuel map in graph form, where a higher peak is more fuel.

JasonC SBB

But you WILL still get more fuel with higher numbers.
The resulting graph will be more like the motor's VE curve.
i.e. with increasing RPM you get a variation in fuel
e.g.

3000 3
4000 3.5
5000 4
6000 4
7000 3.5

TurboTim

iTrader: (9)

That's what I figured, as my fuel map isn't a constant slope per RPM column. I think Mitch actually said they use closed loop WBO2 fueling to trim the fuel beyond what the boost comp table. I never was able to figure out how to get it to O2 trim quick enough w/o overshooting, etc so I only tried it for an hour and decide it wasn't worth the hassle setting up. I actually run open loop everywhere but idle, that's how lazy I am. To each his own.

dustinb

I love my EMS. It took a while to figure out, simply because there are a million options, but it is super powerful. Only complaint I have is on earlier motors it takes more cranks to get it started (EMS waits for more revolutions to see a crank signal).

Also for the vehicle hick-up. My engine will just randomly drop in RPM for a split second and come up back, but only when cruising in 4th or 5th. I figured it was a problem with my old ECU, but problem exists now with two stock 1.6 ecu's, two motors, and on the AEM EMS. I tried fixing it with cops, but that didn't work either. I've just given up and accept it.

TurboTim

iTrader: (9)

I was just looking through AEM Pro and the EGT feedback portions. It does not look like you can do fuel trims based on EGT and O2 sensor at the same time. ????

JKav

iTrader: (1)

This is actually really good info. You guys with EMSs should voice these wants/needs to the forum on AEM's website.

AEM's engineer peeps have been taking user feedback pretty seriously for the Series 2 development, and then working them into future releases of the software. Okay, this doesn't necessarily help the Series 1 guys but would be a big big plus for Series 2 users.

JasonC SBB

For posterity, which I posted in another thread

The AEM EGT inputs are NOT compatible with standard thermocouples. If you want to use standard thermocouples, you will need an interface box (I explained this in a thread which minime started). You will also need to modify the AEM circuit board. BTDT.

The AEM knock sensor input is NOT compatible with the 99+ factory knock sensor. It is compatible with GM "tuned" knock sensors (e.g. Saturn), but those knock sensors are NOT tuned to a frequency which the miata motor knocks at. To connect the 99+ factory sensor to the AEM, I posted an interface circuit, which is tuned to the 13 kHz which my motor knocks at, in the ECU section. (built motors may knock at a slightly different frequency)
__________________

JasonC SBB

JKAv,

Years ago I posted this stuff on their "wish list" thread. If they are smart they will check in there.

Rafa

iTrader: (22)

Jason; what about NAs?

If I have a Saturn knock sensor; what do I have to do to tune it to the Miata motor's frequency?

Thanks,

JasonC SBB

All the same story.
There is no Saturn knock sensor that is centered at the miata's knock frequency - they are tuned too low, 1.8 miatas knock higher - I imagine the 1.6's even higher. If you use it it will be half assed compared to a more properly tuned circuit.

shlammed

iTrader: (8)

Thanks for the info guys!
I decided to go through with getting the EMS from Emilio @ 949racing finally.

Slightly off topic, but would anyone be willing to send their tune to me to look over for a 1.6 setup to compare to the base maps that AEM should include?
My actual setup will be 2860RS 0.86 with RC 440's and GM 3 Bar map/IAT.


Another question is widebands. What widebands are people using, what are the most accurate and what hooks up easiest to the EMS? Im on dialup at the parents house over the break, or i would research a little more.

dustinb

I am using the PLX wideband. It's really easy to install, and it's pre-configured in the AEM EMS.