Nistune Forums

An interesting read about the Z32 (ECU) knock control system:

http://twinturbo.net/nissan/300zx/forum ... -ABCs.html

some of the new addresses of the parameters/tables that were found:

LOW OCTANE KNOCK TRESHOLD CYL 1 : $7360
LOW OCTANE KNOCK TRESHOLD CYL 2 : $7370
LOW OCTANE KNOCK TRESHOLD CYL 3 : $7380
LOW OCTANE KNOCK TRESHOLD CYL 4 : $7390
LOW OCTANE KNOCK TRESHOLD CYL 5 : $73A0
LOW OCTANE KNOCK TRESHOLD CYL 6 : $73B0

HIGH OCTANE KNOCK TRESHOLD CYL 1: $7300
HIGH OCTANE KNOCK TRESHOLD CYL 2: $7310
HIGH OCTANE KNOCK TRESHOLD CYL 3: $7320
HIGH OCTANE KNOCK TRESHOLD CYL 4: $7330
HIGH OCTANE KNOCK TRESHOLD CYL 5: $7340
HIGH OCTANE KNOCK TRESHOLD CYL 6: $7350

REGION A-B SWITCHOVER RPM: $7450
REGION A WEIGHTING: $7454
REGION B WEIGHTING: $7455
REGION C MIN TP: $7F74
REGION C MIN RPM: $7EFF

Do other Nissan ECU's (type 4 in my case, s14a UK) knock control work in the same way?

Excellent investigation and write up. Knock detection routines are very complicated and this investigation must have taken a long time to figure out.

This explains why programs like ROM Editor refered to these as High and Low octane fuel tables. During tuning we recommend to make the timing in these tables the same in order to detect when the ECU may be pulling out further timing during tuning

Also from what we have seen previously when the knock sensor is faulting (code 21) that timing is pulled out at certain parts of the map

The resistor modification mentioned is used when the factory knock control system is reporting false positivies and continuously pulling timing. This is the case when using the Z32 ECU with the RB25 (due to different knock sensor on the RB25)

In those cases yes you are running without knock control, but you tune for that factor in mind (similar to other ECUs like a powerFC which only report knock using the engine light as feedback to the driver)

False positive reporting has been seen even on factory vehicles (such as S14 and S15 SR20s) where K constant has not been adjusted and hence TP scales have not been shifted (so knock regions would not be affected) and therefore the resistor mod was required due to mistaken engine noise being detected as knock

I would say that the previous investigation would apply for other later vehicles like SR20s as JECS/Nissan implementation seems to be consistent with various ECUs. Earlier model ECUs like Z31/VLT dont have the high/low octane maps and do pull timing using a simpler form of knock detection from the investigation I've performed previously on those ECUs

Thanks for the tables Eric. I'll add these to the software and dig into the other ECUs for similar parameters