Stalling when fan comes on
 

Hey guys.
I have a simple issue that I've been unable to fix.

I have a single beefy SPAL fan that draws 17ams. Whenever it comes on, the car comes awfully close to stalling. And if the lights are on, it can easily stall. I've tried everything, it seems, and the RPM still dips dangerously close. Closed loop doesn't seem to react fast enough to the RPM drop.

Details:
- Idle is at about 1000rpm
- Fan drains 17amp (which is a lot, it seems)
- When fan comes on, RPM drops as low as 500rpm
- NB throttle body in closed-loop idle control
- Configured 5% idle-up when fan comes on - doesn't seem to have an effect o RPM
- Configured idle advance to be 10* at low load. When the fan comes on, advance is increased all the way till 20* as the load increases
- When fan comes on, voltage drops from 13V to 11.6, but comes back to 13V (if the engine doesn't stall...)

log:
http://s95367906.onlinehome.us/photo....08.52_FAN.msl

I'm by no means an expert, but in closed loop control aren't the PID "I" settings what catch the rpm's during a sudden increase in electrical load? Perhaps increasing that setting will help?

I bet you're still not using the voltage idle pwm compensation table and spark compensation table?

this is a 90-93 alt or 94-97?

Are you running the stock alternator? I thought I was running a 60amp alt when I was having a similar issue with a Spal fan and posted the issue up here. There's a load of ideas in that thread. Fortunately for me I was not running the 60amp alt, but the 45amp Miata alt and the upgrade fixed the problem. Seems that the alternator upgrade would be the simple route if it's applicable in your case. Retrofitting another larger mazda alt is a fairly simple job.

45A?!

should be 60A for manual and 65A for auto 90-93.

then 65A manual and 70A for auto 94-97.


I wouldn't be surprised if mine is only putting out 45A, however. I idle at 13.8v, it drops to 12.5v if I have the headlights on, if I turn on the a/c is constantly drains at idle, and can drop below 11v to the point my accesories start turning off (radio, LC-1).

I want to either rewind the spare alt I have to 100A, or get an 70A FC RX7 Alt.

Hmmm. Maybe it was 60a and I swapped in a 626 80a. IIRC the KL V6 alt came at 100+ amps and will bolt in too.

a v-notch as well?

stock 94-97 alt because I'm using a ribbed belt.
I was using the default idle pwm compensation table. After seeing the voltage drop, I made changed the curve to from like 1.5% to 6% @ low voltage.

hmm upgrading the alternator seems like a good idea. I don't trust mine anyways. RX7 alt bolts in as well, right?

yeah you can bolt in a FD alt. I'll post the comparison pics in a sec...grabbing lunch.

Check your AFRs when the RPMs dip. Your injector dead-time voltage compensation may be off.

Bigger alternator won't solve your problem, because the dip is probably caused by the alternator loading the engine. If the alternator is too small, it won't load the engine as much as a bigger one when the electrical load kicks in.

90-93 Alt:
https://www.miataturbo.net/attachmen...ine=1372263672

https://www.miataturbo.net/attachmen...ine=1372263672

1989 FC Turbo Alt:
https://www.miataturbo.net/attachmen...ine=1372263672

https://www.miataturbo.net/attachmen...ine=1372263672


94-97 Alt:
https://www.miataturbo.net/attachmen...ine=1372263672

https://www.miataturbo.net/attachmen...ine=1372263672


FD Alt:
https://www.miataturbo.net/attachmen...ine=1372263672

That's the only part of my tune that's dead-on thanks to the dead time tables Injector Dynamics provides.

I do think the alternator is dying, anyways - so it's due for replacement.

Chances of me finding an FD at a junkyard are slim, so looks like I'll be going for a 626/MX-6/MX-3/Probe alternator - they are 80/90A depending on engine displacement.

That assumes that the injector driver clamp voltage in the circuit they used is the same as yours.

what?
https://www.miataturbo.net/attachmen...ine=1372276613

An injector drive circuit clamps the "flyback" voltage (which all switched inductive loads show) of an injector at turn-off. The voltage to which the circuit clamps, changes the effective dead-time. The diagram you pasted, shows dead time vs. battery voltage, and fuel pressure. The flyback clamp voltage is not shown.

IIRC ID used a Motec ECU to generate their plots. If you could find out from them or Motec what said clamp voltage is, then you could compare it to the MS, which IIRC is 55 or 60V.

In this schematic it's 60V or 36 V depending on Q3 and D6 et al aer used.

https://www.miataturbo.net/attachmen...ine=1372292176

Turned out to be the battery. Swapped out the 2 year old Bosch for my massive jeep battery, and everything is great. Bosch battery has a 36 month warranty so I should be able to replace it for free. This one appears to have a bad cell - it charges fully but then discharges on its own (e.g. disconnected completely)

Oh the battery fixed it eh? Nice. When i had this problem it turned out that Travis fucked up on the injector dead times by 10x... instead of 1.5ms he put .15ms.. yeah that caused major problems especially during cranking and idling.

It still dips but a very, very small amount. The dip might be more pronounced once I go back to a miata-sized battery.