4K EFI Test Rig - IAC Valve

[ToyoCoro80]

Banjo, it's surprising that the Speeduino keeps running even after you disconnect the CAS pulse. Hopefully switching to rusEFI gives you the rock-solid continuous syncing

[Banjo]

Hi Allan,

              You'd be surprised how many modern ECUs operate that way.  There is a setting in TunerStudio, which allows You to set it to Resynch, but it doesn't tell You, that sometimes the firmware code has a mind of it's own.  There is another setting that allows you to set the resynch as "Weak", "Medium" or "Aggressive", but that has no effect at all.  There is a Synch lamp on the TunerStudio dashboard, that is green if it is synch'd & red if it is not; but I've never seen mine ever go red.  It appears that is because as far as the firmware is concerned; it is synch'd via it's prediction guessing code.  Why when you've got a rock solid CAS signal once per 720 degree cycle, would you not use it at all; other than for starting the engine. Even if you resynch'd once every few minutes; using the CAS pulse, it would be better, than driving for hours & hours without using it at all.  When I asked A.I. to investigate this matter, it advised that Speeduino believe their method is more accurate, as it can correct timing inbetween CAS pulses more accurately. I can understand that requirement is motor sport applications, where the acceleration can change significantly, in 2 revolutions of the crank.  To ignore a guaranteed C.A.S. pulse, is not what I want or need for my 5K engine, thank you.  Apparently even some of Haltech's latest model ECUs, still use the C.A.S. for resynching.   In my case, I understand that the CAS pulse can jitter & become "time unstable", due to camshaft sprocket & chain, & belt drives. However, I've logically "AND'd" the camshaft one with a single CPK pulse per revolution, from the crank pulley, & the resulting pulse is absolutely "rock solid".  I've not been able to find any settings in Speeduino trigger setups, that allows me to use the C.A.S. as a regular synching input. Beats me !  I'll keep on researching. At least I've got an alarm now, to alert me to the C.A.S signal no longer being there. However, that is acedemic; as the C.A.S. pulse is not being used at all, whilst the engine is running. However, it does alert me that it is no longer present; so I won't be stopping for a rest or cuppa, or photo shoot, whilst I'm crossing the Hay Plains; as I know that once the engine stops, it won't restart, without a C.A.S. pulse.  I rest my case.   In this case, I'd love to be "proven wrong", & that I've missed something.  So if anyone has additional info or experience, regarding this matter, pleased to hear your thoughts or experience.

https://speeduino.com/forum/viewtopic.php?p=74149#p74149

Cheers Banjo    

Edited March 22 by Banjo

[Banjo]

Recently discovered that there is a module produced by rusEFI in the USA, which includes the high speed SMT32 processor, which is simply a plug & play replacement, for the Arduino Mega2650 microprocessor board. It is called the MEGA144, & runs at speeds many, many times faster than the Arduino.

There is a couple of matters that need attention, as it runs on 3.3V where the Arduino Mega2650 runs on 5.0V, so I may need to upgrade, my N type Mosfets, so they can be switched by the SMT32 micro's 3.3V outputs.   It's almost unbelievable that it sells for just USD49.00. As The AUD vs USD exchange rate is pretty good atm, I ordered one from the rusEFI store yesterday.   That will overcome the issue I'm having with the Speeduino not having enough grunt, to continually using the CAS pulse to synch the ECU, after the engine has started.  The rusEFI has firmware that You download from their website; which does use the CAS pulse to resynch continuously.  If that works well for me, I might invest in rusEFI's model UAEFI Model standalone ECU. The specs of the ECU is amazing, for USD 175.00 ea.  This is not a kit. It is assembled ECU, ready to go.  It even has a wide band controller onboard for wide band O2 sensor. There are metal & plastic cases for it, & it has provision for Molex quality connectors, for loom connections in your vehicle.

I'll keep You posted, once I receive the MEGA144, & get it up & running.

https://rusefi.com/

Cheers Banjo 

[Banjo]

Update on my issue with Speeduino ECU's non use of the CAS pulse to resynch, after the engine starts.  The rusEFI Mega144 high speed microprocessor board, with a SMT32 microprocessor on board, should be here shortly, via UPS from the USA.  In the meantime, I addressed over the weekend; a consideration, that is "the bain" of many ECUs;  Electrical NOISE !

Noise can affect any electronic equipment, via two basic avenues. One is through the atmosphere, which most people describe as "interference".  It's why a radio, or TV may crackle, whilst a storm is in the vicinity.  This type of noise; can, quite often; be protected, by fitting the ECU, inside a metal case, which is grounded to the chassis of the car.  Although the chassis ground is not a "true" earth or ground, as the vehicle has rubber tyres, that insulate the vehicle's chassis from the ground/earth. You will be aware, that many cars used to have a rubber strap, (with wires therein) that hung below the car, & was attached to the chassis/body of the car.  It dragged on the roadway, beneath the vehicle. This was used to reduce charge built up in the body of the vehicle; that was believed to cause "car sickness". The strap "shunted" the charge to ground effectively; although quite often only when the car stopped.

 The problem with electronics in the car, is that most circuits use the vehicle's chassis as the return path for most circuits. Example:  All lights only require just  one "switched" wire to turn the light on.  The return path to the -VE terminal of the battery, is the chassis of the vehicle. That is why there is simply one big thick cable, connecting the battery -VE terminal to the chassis, & why it is so important. So the result of this fact; is that an ECU, which runs on 3.3V or 5V, shares the same return to the battery -VE terminal, as does the starter motor, which pulls sometimes hundreds of amps, & generates a lot of noise, due to it's carbon brushes & commutator. I hear You say;  Hold on; the ECU runs on 3.3V or 5V.  Yes they do, but the the 3.3V or 5.0V supplies, are derived by a regulator that "steps down", the 12V power, & therefore the 3.3V or 5.0V supplies; "share" the same ground or return path for the 12V battery.  My thought was; if I could find away, to isolate the 3.3V or 5.0V power supply from chassis ground completely, then electrical noise in the system, could be reduced, or nearly eliminated entirely.  I wondered how many other ECUs, on the market, used a technique like this.  I thought it might be common, in top end ECUs; especially, those used in motor sport applications.

My research indicated there were two manufacturers of commercial ECUs, that address this specifically.  One was Haltech, (although only on the R3, R5, S2 & S3 series) The other was the Motec M1 Series, although apparently not as well isolated as the Haltech series above..

So I set about building an electronic power supply, which has a 12V input, & a 5V output, but absolutely no connection between them at all.  All possible using a tiny little "Meanwell" isolator module with a  12V input; & in my case a 9V output. I then used the 9V output, to feed the input of a +5V or +3.3V regulator (LM7805) precision regulator.  The 9V input is exposed to drop off, as the 12V battery voltage, can easily drop to 10V or lower; during cranking, on a Winters morning.  However, the Meanwell isolator at that level, still produced enough output; for the +5V regulator to still provide a consistent & stable +5V supply. 

Exactly the same technique could be used with a 3.3V regulator.   If I eventually change over completely, to the rusEFI model uaEFI ECU, which happens to run on 3.3V DC. 

Because my system is "compartmentised" into three (3) blocks of electronics, I've actually built three of these isolated +5v power supplies, so the load on each one, is quite conservative. There are other ways, I'm hoping to reduce noise in the system.  One is to use screened wiring, with the shield, tied to "chassis ground", at one end of the cable, only, on all cables out in the engine bay, feeding signals to COPs or injectors.

The other technique is to use "opto couplers" throughout the circuit, to feed the very narrow ignition signals to their respective points on the engine. The "opto coupler"; is again a device where a light beam inside the opto-coupler feeds the signal from input to output; without any hard wire connection in-between input & output terminals.  Bear in mind, that engine bay in any vehicle is a "hostile environment", at the best of times. Heat, vibration, wind & water are always possibilities. In addition, there are parts of the engine bay may have high levels of radiated noise, from spark plugs arcs/injectors; & starter motor commutator & brushes  etc.  

One potential issue I found that upset my whole concept of isolating the 5V ground circuit from the 12V ground, was the C.O.P.s.  The C.O.P.s have two return paths requirements. One for the return path, for the +12V coil charging circuit, which can quite commonly carry 8-10 amperes. The other is the return path for the +5V trigger signals to the C.O.P.s.  The real issue here is, that most C.O.P.s have a single "shared" return wire, which services both the 12V charging of the C.O.P. & the +5V trigger pulse. This is how the DENSO C.O.P.s I've been using work. I chose DENSO originally, as my late model Toyota Corolla uses them.  However, a bit of research uncovered an AC Delco C.O.P. which does have a separate return wire for the +12V coil charging, & a separate wire for the +5V trigger signal.  I also found a compatible C.O.P. to the AC Delco C.O.P.s made by ICON, & available here in Australia on line, at a better price.  My tests to date with them have proven very good.  Just have to be careful, that You don't provide them too much dwell, as they internally monitor the charging current, & when the coil reaches saturation, they automatically fire. I've set my dwell range to 2.0 to 3.0 m/sec.  Last thing You need or want; is your COPs running "hot", & either dying, or suffering a shorter life.

I have one other area, I'll look at in the next day or so, which is sensors, that feed information to the ECU.  Example:   The coolant temp sensor, is commonly screwed in the engine  somewhere; where it's return path is the block or head of the engine, which is actually at the 12V ground potential. This would effectively connect the +12V ground & +5V grounds circuits together, outside the EC; after I've gone to so much trouble to separate them.  A simple way, I'll try in the coming days, is to turn the coolant sensor into a 2 wire device. I'll gut it mechanically, & then glue a two wire thermistor in there, with two wires coming out. This will result in a simpler solution, than building another electronic isolator for any external sensors, that use the chassis as a return path. To some extent, the same situation, may be an issue with the Intake Air Temperature (I.A.T.), but that is much easier to isolate from chassis ground.  Talking about chassis ground; the termination of the strap/cable between the 12V battery negative terminal, & the chassis is most important.  It should be clean on all mating surfaces, & be a very firm tight joint.  When starting the engine on a cold morning, the cranking of the starter motor can draw substantial current from the battery, via that connection joint. I have measured voltage drops of 1-2 volts, across that joint on starting. 

Cheers Banjo

Edited April 23 by Banjo

[Banjo]

Most modern vehicles apparently now use 2 wire temp sensors for Air Intake Temp, & Coolant Temp; so no need for me to gut my olde sensor that screwed into the thermostat housing.  I managed to order them both, in the last day or so, from the same supplier, for a good price. Bargain; as they come with the 2 wire plug & tails also.  

 

 

Cheers Banjo

Edited March 31 by Banjo

[Banjo]

There is one other input into an ECU, that is important; but posed a problem, as it also "shares" the +12V chassis ground, with the +5V ground return path.  That is the battery voltage signal.  This important in an ECU as lots of things have to change there settings, based on what the battery voltage is at the time.  ECU setting like dwell time for ignition, & injector timing, are critical, & are usually adjusted by map settings in the ECU.  However, the ECU needs to know whether the +12V battery voltage is 14.4V whilst charging; or +7.6 volts whilst cranking on a cold Winters morning. Normally You would simply scale down the battery voltage with two resistors in series, such that the lower resistor had exactly +5V across it, when the battery voltage was at it's highest (eg: +14.4V)  The only problem with this is, that this action would then tie the  cars +12V chassis ground, to the +5V isolated ground, I had so carefully devised; so it sort of defeats the purpose.  A bit of research indicates there are precision linear opto-couplers that you can feed on the input side with a +12V signal & obtain an isolated +5V signal on the isolated side, where the +5V signal represents the actual battery voltage. So todays exercise is to experiment with this & get this working, & then move on.   My rusEFI  processor MEGA144 board should be delivered today or tomorrow, from the USA; so I'll be excited to get that up & running with my Speeduino, & put my lack of CAS signal usage issues with Speeduino code behind me.

Cheers Banjo   

[Banjo]

So a little bit more to do as yet, but very excited that my rusEFI Mega32 processor has arrived, & will soon be integrated into the Speeduino V4.3d PCB, in coming days.

 

 

Most modern vehicles apparently now use 2 wire temp sensors for Air Intake Temp, & Coolant Temp; so no need for me to "gut" my olde sensorhousing that screwed into the thermostat housing.  I managed to order them both; on ebay, from the same supplier, for a good price. Bargain; as they come with the 2 wire plug & tails also.  They've arrived & I've checked them on Speeduino board, & they are reading spot on, at room temperatures. If I want them to be more accurate at their operating temperutues; it is simply a matter of holding them at that temp, & measuring their resistance; and then changing the voltage divider resistor on the PCB to match, so that at operating temperature the reading is approximately half the supply voltage           (5V /2 = 2.5V)  

 

 

 

Cheers Banjo

 

 

Edited April 23 by Banjo