| Author | Topic |
|---|
Location:
Lost in the K hole
Registered:
May 2002
|
|
knock sensor location ??
|
Wed, 02 October 2002 11:39
|
 |
ok...
i have a 7mgte block that was designed to use twin knock sensors over cyls # 2 and 5. now im going to use this block to build up a 7mge which only runs a single knock sensor, usually located over cyl # 3.
now is it ok to run that single 7mge sensor in one of the 7mgte locations (2 or 5, and which one?) instead of over its original # 3 ???
or, is it possible to rig up twin sensors to run on the 7mge ecu, even though the ecu has provision for only one?
cheers
ed
|
|
|
Location:
North Sydney
Registered:
May 2002
|
|
Re: knock sensor location ??
|
Thu, 03 October 2002 01:48
|
|
Just wire both of them into the same input.
That will work or you can make a bracket and bolt it onto the block anywhere within reason as long as you use a solid metal bracket for the ultrasonics to move easily through.
Just take the first option and use twin knock sensors. 
|
|
|
I supported Toymods
Location:
I renounced punctuation
Registered:
May 2002
|
|
Re: knock sensor location ??
|
Thu, 03 October 2002 02:03
|
|
Mmm...be wary.....the 2JZGE/GTE have 2 knock sensors and each go into separate inputs on the ECU - what happens inside ECU I can't say but suggests each signal is processed separately.
The summation of the combined piezoelectric signals potentially may result in additive phenomena.....excess signal when no knock resulting in unnecessary retard, or cancellation of signals resulting in no retard during detonation. Clearly the latter is no good.
Food for thought...... people's thoughts?
|
|
|
Location:
North Sydney
Registered:
May 2002
|
|
Re: knock sensor location ??
|
Thu, 03 October 2002 05:11
|
|
They would tend to be over reactive and detonation would never be a chance if it was the case. Remember that Toyota use the sensors when in closed loop operation only. So when you've got your foot heavily into it the ignition maps are pre-determined.
If on the 7MGTE the inputs are separate and dont split before going into the ecu, does this suggest that the ECU will retard timing on an individual coil pack where the detonation is occuring and leaving the timing advance on the other 4 cylinders alone? A tough question, anyone?
If 1 coil pack is causing detonation because its timing is too advanced then the other might be fine but what about the third coil pack?
Now i've just gone and opened up a really frigin big can of worms!
One advantage of aftermarket, you know exactly what its going to do. Downfall is the lack of features before anyone jumps in.
|
|
|
I supported Toymods
Location:
I renounced punctuation
Registered:
May 2002
|
|
Re: knock sensor location ??
|
Thu, 03 October 2002 05:47
|
|
Paul I'd imagine it processes each signal separately but then the output is universal...ie to all 3 coil packs.
No idea.....merely speculating.
|
|
|
Location:
Lost in the K hole
Registered:
May 2002
|
|
Re: knock sensor location ??
|
Thu, 03 October 2002 06:07
|
|
well, for what its worth, and GT we were talking bout this last night, ill probably just run the sensor on the block at around #5... as the back of the block does get hotter, and is more likely to incur detonation back there.
ill then run a second sensor up the front mount from which ill run my own black box with an LED on the dash letting me know exactly whats going on in the fron 3 cyls at least.
the problem i see with temporal summation of 2 sensors (if i were going to do that) is that the ecu would never get a chance to back off the timing retard... thus not getting the most out of the engine... it'd be wrapped in cotton wool so to speak, overly cautious timing.
out of curiosity..,. when does the cu run in open and closed loop modes? i thought it was open only during warm up, then once she hit operating temp, bang into closed loop??
QUESTION: is there any way to get the ecu to run in open loop mode full time? i reckon that would be the way to go. screw standardised emmsions and ignition curves, i want to see the engine run with full performance feedback.
cheers
ed
[Updated on: Thu, 03 October 2002 08:40]
|
|
|
Location:
Brisbane
Registered:
May 2002
|
|
Re: knock sensor location ??
|
Thu, 03 October 2002 08:20
|
|
If the 7M is anything like earlier Toyota engines, it goes into open loop as soon as you hit more than X% throttle opening, where X is usually above 75%. I guess if you fooled the ECU into thinking you're at WOT all the time, it would stay in open loop permanently - but that may have other consequences, eg ignition timing?
|
|
|
Location:
Lost in the K hole
Registered:
May 2002
|
|
Re: knock sensor location ??
|
Thu, 03 October 2002 08:40
|
|
ok, back the truck up...
closed loop is when its using the sensors in a real-time feedback way to adjust engine parameters?
open loop its just using lookup tables and ignores various sensors?
argh 
cheers
de
|
|
|
I supported Toymods
Location:
I renounced punctuation
Registered:
May 2002
|
|
Re: knock sensor location ??
|
Thu, 03 October 2002 08:45
|
|
|
Just as an aside, have you guys seen what you have done to the Mk2 mailing lists? I don't think they're ever going to be the same. Tsk, tsk.
|
|
|
Location:
Lost in the K hole
Registered:
May 2002
|
|
Re: knock sensor location ??
|
Thu, 03 October 2002 08:54
|
|
they deserve full inboxes 
|
|
|
Location:
North Sydney
Registered:
May 2002
|
|
Re: knock sensor location ??
|
Thu, 03 October 2002 10:04
|
|
The ecu goes open loop until operating temp is reached then closed loop whilst "cruising". IE 0% to 75% as Norbie says then anything else it goes to open loop lookup tables. But there are also other times when closed loop activates. IE if you accelerate quickly(faster than a preset milisecond value, not sure what it is on a standard 7mgte ecu) it's commonly refered to as "PUMP". It gives the engine extra timing and added fuel so the response feels crisp in response to your foot.
Ever see a Sti WRX blow a heap of black smoke when he's dragging someone off? This is exactly whats happening, well similar actually cause the added fuel actually helps the density in the exhaust gases be heavier and this will turn the exhuast turbine faster thus minimising turbo lag.
Closed loop, also reads in the O2 sensor to get the stoichiometric ratio of 14.7 by triming fuel mixtures.
|
|
|
I supported Toymods
Location:
Perth
Registered:
May 2002
|
|
Re: knock sensor location ??
|
Thu, 03 October 2002 11:02
|
|
|
Is this how cars are getting better fuel economy on the freeway also as they get better at this? They go into closed loop on steady throttle and lean out the mixture and adjust the timing?
|
|
|
Location:
North Sydney
Registered:
May 2002
|
|
Re: knock sensor location ??
|
Fri, 04 October 2002 13:41
|
|
Yep precisely, they crank the timing constantly to the point of detonation and retard it a few degrees atleast 10 times a second this gives the most power from the engine at that point in time. The a/f ratios are leaned off alot, instead of a setpoint of 14.7 for perfect combustion and minimum emissions they aim leaner like 16:1 and some like Puegot's are getting really economical in closed loop like close to 18:1.
Toyota Diagnotic V2 in the US which we dont have out here even monitors individual cylinder egt's to ensure even better economy and lower emissions. By triming individual timing and injector pulse widths they get even better results.
Cant wait for that stuff out here!!!
|
|
|
Registered:
March 2005
|
|
Re: knock sensor location ??
|
Sat, 19 March 2005 01:06
|
|
|
does anyone have a diagram that shows the locations for both knock sensors on a 7mgte?
|
|
|
I supported Toymods
Location:
Australia
Registered:
November 2003
|
|
|
