two(or more) ecu's for higher smaple rates?

[barana]

Gday, I have a question that has been bubbling to the surface in my mind these past months.....

If I would like to have a greater sample rate for an engine managed subsystem that the delco is able to give,( and I read how you can increase the sample rate for a sensor by 'switching off' some sensors , thereby freeing up clock cycles to allow Higher sample rates for the remaining sensors) Could I , or would it be possible to hook up a second delco to just manage my subsystem - with a very high sample rate - and then hook it up to the main delco ecu and enable the main ecu to accept (2nd delco's) input towards the main 'program'?

Ill give an example - the subsystem is a vvt camset. one or two to 4 variable valve timing cams.hydraulically actuated cam phasors ... or even solenoid stepped cam like porkers(porsche's).

[j_ds_au]

From what I've read, you have Delco ECU's with 68HC11 MCU's with two different clock frequencies and 68336 based ECU's. That's quite a performance spread. Why not start with a 68336 based ECU, surely that has all the "grunt" you need.

Joe.

[barana]

Dont get the details of what ur saying. Whats this m68k ecu ur talking of? Is it oem from something?

[Jayme]

you are barking up the wrong tree I think... the speed you are talking about is the speed of logging sensors on your laptop / scan tool. the limiter there is data bus bandwidth, not CPU speed. the less sensors you log the quicker you can log them up the data bus. the main program running on the PCM polls the sensors much much much faster than the slow old data bus... there is no gains to be had there. no wasted clock cycles to free up, just a slow data bus that cant keep up.

[barana]

Ahhhhhhhhhh! Thx m8! So what speed do the slowest 68hc holden ecus poll sensors @? Also what speed do they actuate solenoids etc @? Or where can I dind out these things myself?

[VL400]

The main code loop runs at 160Hz, so this is the maximum sample rate of some sensor reeadings. Other sensors get sampled on every reference pulse from the crank/dizzy trigger - so sample rate is RPM dependant.

But your question is difficult to give a definitive answer for all sensors/actuators though. Sample rate and actuation differs for each the type, no point reading a coolant sensor every 6.25ms. Some sensors are sampled and then filtered too, so the response is slower but it filters out the noise.

I dont really understand what you are trying to do though, logging or controlling a VVT system?

But as to where to find these things out, have a look at various ECU code hacks.

Bit of google work nets .. http://www.diy-efi.org/gmecm/papers/prog101b.txt
The aussie ECUs compute at these rates too, although MAP is also sampled on a reference pulse.

Code: Select all

-<><><><><><><><><><><><><><><><><><><><><><><><><><><><><><>-
$From: "Bruce Plecan" <[email protected]>
$Subject: Misc 101 Information
$Date: Mon, 4 May 1998 12:01:33 -0400
In case you were ever curious, how often the ecm "looks" at
things.
The ECM never stops polling the sensors.  Under certain conditions, it might
ignore a measured value though.  Here's how often the XYZ
ecm samples the inputs:
MAP is sampled 160 times a second.
O2 is sampled 160 times a second.
VSS (MPH) is sampled every VSS pulse (up to 80 times a second.)
RPM is computed 80 times a second (in the spark code.)
TPS is sampled 80 times a second.
Battery voltage is sampled 10 times a second.
CTS is sampled 10 times a second.
ALDL is sampled 10 times a second.
MAT is sampled 10 times a second.
MPH is computed 10 times a second.
Fuel and spark are computed 80 times a second - on alternate 160 Hz beats
Most of these sampled items are filtered, so the value used for
further computations won't change as fast as the sample rate
Cheers
Bruce
-<><><><><><><><><><><><><><><><><><><><><><><><><><><><><><>-