Banjo

Hi Jasper, Seems like You have checked everything, but the car being quite olde, there are a number of issues. If You go look on the internet, it does turn up a few possibilities, for your Daihatsu. ____________________________________________________________________________________________________________________________________________________________ If a 1982 Daihatsu Charmant LE fails to idle properly after all standard manual tests have been performed, the issue typically stems from specific aging components or internal carburetor failures that are not easily detectable through standard diagnostic routines. Hidden Carburetor and Fuel Issues Because the 1982 Charmant uses a carbureted system, standard tests may miss internal mechanical wear: Clogged Internal Idle Circuits: The fuel path for the idle circuit is highly convoluted, often traveling above the fuel level in the float bowl before descending. If the car has sat with old fuel, these internal passages can clog in ways that surface cleaning or "blowing through" cannot fix. Worn Needle Valve Tip: Microscopic wear on the needle valve tip can allow fuel to leak past into the engine, causing a rich mixture that cannot be corrected by the idle mixture screw. Malfunctioning Economizer Valve: If the idle is too rich and cannot be leaned out via adjustment, the economizer valve may be feeding fuel into the engine incorrectly. Sticking Automatic Choke: On older carbureted engines, a choke that is even slightly sticking or misadjusted will disrupt the air-fuel ratio enough to prevent a steady idle. Vacuum and Air Leaks Standard manual checks often focus on visible hoses, but hidden leaks can be more subtle: Carburetor Base Leak: Vacuum may be leaking around the base of the carburetor or through worn-out gaskets rather than the hoses themselves. Intake Manifold Gasket: Worn or damaged gaskets can cause a vacuum leak directly at the engine block, which is difficult to see without specialized testing like a smoke test. Clogged EGR Valve: Carbon buildup inside the EGR valve or its passages can cause stalling at idle. Mechanical and Ignition Timing Ignition Timing "Out of Whack": Even if set to manual specs, wear in the distributor or on-board computer components can cause timing to drift, resulting in a poor idle. Worn Valve Seals or Piston Rings: In older Daihatsu models, worn internal engine components can lead to reduced performance and erratic idling that external adjustments cannot fix. ____________________________________________________________________________________________________________________________________________________________ Hope somewhere in amongst that, You spot something, You've not checked as yet. Having had a couple of 4K engines, the Intake manifold gasket leaking, is quite a common issue. I'd be taking off both inlet & exhaust manifold together, & then run a straight edge across them. As it's been happening for some time; You may even see evidence of the leak, with discoloration of the metal surfaces, each side of the gasket. Let us know, if any of these suggestions, bring your idling issues to a sucessfull resolution. Here is a post on this forum, written 15 years ago, where the issue, I'm suggesting was discussed. https://www.rollaclub.com/board/topic/38875-4k-intakeexhaust-gasket/ Cheers Banjo

Hi Gino, Welcome aboard ! Many members on RollaClub run aftermarket ECUs, of various types; but You are the only person, who has advised they are running a Speeduino; & a V0.4 at that. I haven't actually got any road miles under my belt as yet, with my Speeduino. All my experimenting, with various trigger systems, has been on the bench. My Speeduino, will be running my 5K initially, which is a running working test engine, on a stand in my garage. That will be to prove the synchronizing of overlapping CAS signal, with the crankshaft single pulse per revolution. I've had it running on the 5K a while back, & then it was off to bench, for lots of triggering experimentation. Then it will transferred to my KE-30 2 door sedan. As I won't then have a test Speeduino, on the bench, to further experiment with; I've recently finished a Speeduino V0.4.3d. It is a nice ECU, as it's board is 30% smaller than my V0.3.7 Speeduino. The V0.4.3d has Bluetooth & idle control, so looking forward to that. This is what it looks like now, as this week I hooked up a 40 screw down block, so that all 40 pins on the Speeduino V0.4.3d, can easily be hooked up. Interested to know what type of triggering system you used, & how You implemented it. Post some pics of your setup in the your car, with some details of the setup, including ignition coils or COPs etc. I'm presuming that it is currently controlling "ignition only" ? I've actually got a second test engine in my garage, which is a 4K with a 7KE fuel injection setup. That will be the final step, so a lot more fun, & learning, before I get to that point. Cheers Banjo

Further to my post yesterday, I should warn, that what I'm doing, may not work for everyone, as trigger signals to various industry COPs, can vary in voltage & polarity. The COPs I am using are Denso 4 wire COPs, widely used in current & recent Toyota model Corollas, Yaris & Echo vehicles. The fourth wire I am not utilising; as it is a feedback signal to the Toyota factory ECU, to indicate whether a COP has fired correctly or not. Most COPs, are triggered by a +5v pulse from the ECU. However, there are COPs that apparently use a 0V pulse. There are apparently; also COPs, in the world that use a +12V signal. I discovered this by doing a bit of research last night, on Google AI. Here is what I asked, & how Google A.I. replied. Do all COP (Coil Over Plug) automotive ignition coils trigger off a +5V DC signal being applied to them, by the ECU; or do some brands use a negative going pulse, to trigger them ? No, not all Coil-On-Plug (COP) ignition coils trigger off a positive 5V DC signal . While a 5V positive pulse is common, other triggering mechanisms exist across different vehicle makes and models. The specific triggering signal depends on the coil's design and how the ignition amplifier (ignitor) is integrated into the system: Integrated Ignitor (3 or 4-wire coils): The majority of modern 3-wire and 4-wire COP systems have the ignition amplifier built directly into the coil assembly. These typically receive a low-voltage 5V trigger pulse (usually positive-going) from the Engine Control Unit (ECU). The coil fires when internal electronics are signaled by this voltage pulse. External Ignitor (2-wire coils): In older or different systems using 2-wire coils, the switching (triggering) is done by an external ignition amplifier module or directly by the ECU's internal driver. In these cases, the signal wire from the ECU/module essentially provides a negative-going pulse by grounding the primary coil circuit to build a magnetic field, then opening the circuit to induce high voltage (the primary voltage briefly drops close to ground potential before skyrocketing when the circuit is opened). The voltage spikes on this primary negative side can reach several hundred volts. Other Voltages: A few vehicles might even use full charging voltage (approx. 12V) as the trigger signal amplitude, though this is less common than 5V. Ultimately, the required signal type and voltage level are specific to the vehicle manufacturer and the particular ignition system design. -------------------------------------------------------------------------------------- There is another easily fixed issue with the polarity of the narrow pulse emanating from my listed circuit. I added four (4) opto couplers to the circuit in yesterday's post. The MOC5007 opto coupler has a open collector output, which means it pulls the output low, (to ground/chassis potential). I advised that the Denso COPs accept a +5V pulse, to start the COP charging the primary coil inside. This is easily fixed. The little MOSFET ignitor module I depicted in the previous post, have a positive & a negative terminal & an opto coupled input, to the MOSFET circuit. If the trigger signal, to charge the COP coil is +ve; then connect it to the +ve input terminal on the MOSFET module & tie the -ve terminal to ground. If the trigger signal, to charge the COP coil is -ve going; then tie the +ve terminal to +5V, on the opto coupled module, & the -ve terminal to the output of your circuit. P.S. There is one glaring flaw with my suggestion above. By simply reversing the polarity, of the opto-coupler connections, the width of the resulting pulse will be much greater, & does not allow intelligent COPs, to cease charging the coil, when saturation is reached. In the case of the Denso COPs, I am utilising; this betwen 3-4 m/sec. Cheers Banjo

The dual MOSFET ignition output drivers, for powering COPs & external coil ignitors, (model TC-4424), are mounted on the Speeduino PCB. I therefore cannot use these for triggering the 4 off COPs, so the Speeduino PCB, remains "unmolested". The outputs of the four (4) off 4075 ICs, on my switching module, are highly unlikely to have enough current capacity, to trigger the COPs. I therefore used external MOSFET opto isolated modules, to switch the four (4) off COPs. I've hooked up the outputs of my Sequential to Wasted spark switching module, to four (4) off these small MOSFET modules. That's the MOSFET on the LHS, & the little IC on the RHS; is the opto coupler. These are available on ebay, & are very economical. I paid $ 1.62 each, off memory. You can even purchase a dual version of these, with two (2) MOSFETs mounted on the one PCB. For about $ 8.00, You can have four (4) MOSFETs, to switch your four (4) COPs, or ignitors. However, the dual ones, don't appear to have opto-couplers, for the input signal. I might just add four (4) opto-couplers to my switch module output, so that it doesn't matter if the MOSFET module used; does not have an opto-coupler built-in. The Opto-Coupler I generally use, is the MOC5007. This Opto has "hysteresis" built in, which makes the system, I've created; very "noise tolerant". The beauty of the both MOSFET modules, I depicted above; that they both have a little LED on the board; so You can see that a signal is being received from Sequential to Waste Spark switching module. I've only been able to find one aftermarket ECU that incorporates Waste Spark functioning, during cranking. That's the high end MaxxECU race ECU. Apparently, LINK ECU mentioned way back in 2021, that they were considering including it; in a future upgraded specs; but I don't believe that has been implemented as yet. Cheers Banjo

Good news on the "start engine" in Wasted Spark mode; & drive in sequential, there after. It appears that a Speeduino ECU, can only be changed from Wasted Spark to Sequential; or visa versa; by changing the mode of operation in the project file setup screens, within TunerStudio. There appears to be no way to change it, "on the run/fly". Well I'm pleased to advise, I have my Speeduino, running on the bench, in sequential mode; but can switch it in & out of "wasted spark" mode, at will, on the run. This was done, without making any modifications to the Speeduino, whatsoever, or changing it's setup tables. The olde dissy cap, with four High Tension leads & plugs, never really had an issue, with timing or starting, as the rotor button in the distributor, will always fire the next spark plug & cylinder, in the sequence, irrespective of where it is, in the two revolution cycle; or where in the cycle, the engine finally came to a halt. With COPs, & no rotor button, & four (4) COPs under direct control of the ECU, there is a better way of possibly getting the engine to fire earlier; (in the first rotation of the crankshaft), while cranking. Start the engine in Wasted Spark ! The solution was quite simple. Take the four (4) ignition outputs from the ECU, (IGN1, IGN2, IGN3, IGN4) & run them through a set of logic gates, which pass them through as "Sequential", or convert them automatically to "Wasted Spark", as a result of an external signal, supplied by a switch on the dash, possibly by the extra wire on some TPS sensors, or instigated by the start position in the ignition barrel. When You hold the ignition key in the spring loaded position, whist cranking the engine; this signal can apply the extra signal to the logic gates to create two outputs for each sequential input. ie: When the ECU supplies a signal to fire cylinder no: 1, then the logic gates are set, so that COPs 1 & 4 fire together. When the ECU supplies a signal to fire cylinder no: 3, then the logic gates are set, so that COPs 3 & 2 fire together. When the ECU supplies a signal to fire cylinder no: 4, then the logic gates are set, so that COPs 4 & 1 fire together. When the ECU supplies a signal to fire cylinder no: 2, then the logic gates are set, so that COPs 2 & 3 fire together. The Speeduino ECU has four little LEDs, indicting sequential or wasted spark mode. In Sequential mode, the ECU simply fires the four (4) Ignitions outs one after another. IGN1; IGN2; IGN3; IGN4. It's up You, or the installer; to have to wire the four outputs to the appropriate COP, or external Coil Ignitor. I've hooked up four (4) LEDs, on the output of the small logic gate PCB, & I'm sitting here watching the LEDs on the Speeduino ECU board, in individual sequence, & that replicated on the logic board LEDs. However, as soon as I provide the additional logic signal to the small logic gate PCB, the four (4) LEDs on small logic board,; instantly change to 1 & 4 on together, & 2 & 3 on together. There is no delay; it happens instantly. Here is the logic circuit, if anyone wants to replicate this. No micro-processor, or programming required. Speed or frequency is not an issue. There are just 4 off "integrated circuits" (ICs) involved, that can be purchased at any electronics store, like Jaycar; for about $5-$6 total. The whole thing is powered by 5V DC, which will come from the ECU. A small 12V relay, powered by the start position on the ignition key switch, should be used to switch the +5V signal, to the logic board. I might add an opto-coupler there, with hysteresis built in; so there is no chance of "chattering" contacts, from a relay. Just have to wait now, until this is all in the KE-30, & I can try it out, on the road. Cheers Banjo

I did a little exercise late yesterday, & asked Google AI, to find me pictures or photos of Toyota engines, in which the distributor was depicted. AI duly supplied many ! I had a look through them, & came to the conclusion, that Toyota would have been happy if the engine did not have a distributor at all; as the dizzy location always seemed to be an "after-thought". Granted these variety of engines for RWD vehicles were used in later model vehicles, where they were adapted to FWD cars, but the "dizzy position", did not appear to gain a lot of attention. The K series engine had perfect positioning. Halfway along the block, on the sloping back side of the engine. If, for instance; the K series were ever to be modified to fit in an engine bay, for a FWD car; the dizzy would be facing the radiator, with good access & cooling. Toyota seemed to often have lots of self-inflicted issues; when designing position of dizzies; always fitting dizzies, close to very hot exhaust headers & extractors, & the like. Big burnt holes in the dizzy cap, were quite common ! Cheers Banjo

Don't want to "butt in" on this wonderful conversation, but I could not help myself, when I did a "double take", when I saw this picture, of a distributor on a twin overhead cam Toyota engine. Was this someone's attempt to remove the dizzy altogether, from it's factory position; too close the exhaust manifold; where the "bakelite" dizzy top, got too hot, to the point of breaking up ? Cheers Banjo

Hi David, Is your dash instrument cluster, similar or like this one I found on ebay, that has a number of photos, including the rear ? https://www.ebay.com.au/itm/326234860459 Cheers Banjo

Yes, The Toyota TPS sensor, which apparently was utilised on Camry & Corolla vehicles, definitely has 4 wires. I know I played with it a bit, when I received it, after buying it on line; but couldn't quite work it out. That did not concern me, as I wasn't going to use it, anyway. However, I did wire it up, with 4 wire cable, so I have the idle wire, that I could play with, & add a piggy back board, as it is already on the 4K-U engine. Not sure that will do what I want; as if I use that to "start the engine", it will drop back into "waste spark mode", every time, You pull up at a set of traffic lights, & take your foot off the accelerator. I really need it only during the engine "starting process". If You research it, a lot of countries will not allow vehicles with their engines running in waste spark format. Apparently, this is because of emission regulations, as waste spark, produces slightly higher emissions than sequential operation. Apparently, there were other factors that came into it; especially for the Americans, with a preponderance of V8 & 6 cylinders engines. The dual HT output coils used on GMH vehicles here in Australia, were widely used, & very popular. Simon / TAZ used them on his K Series engine. No rotor button, or dizzy cap with individual spark plug leads. I've always believed in sequential control, as You can tune each cylinder slightly differently, & pick up & control "knock" in an individual cylinder. Apparently, cost comes into it also. No CAS signal required from the camshaft. You don't even need a separate crank angle sensor. Just use a 24, 36, 60, or 72 toothed crank wheel, & remove one or two teeth, to become your crank angle sensor. Super simple; super cheap ! Initially, I will be using my ECU, as ignition only. My technique, whilst starting the car, has always been the same. Never have to use the choke, here in Qld. Before turning the ignition key, two quick pushes of the accelerator pedal, to activate the pump squirt function in the carby, to enrich the first suction cycle of the cylinders. Then I press the accelerator slightly (to open the throttle butterfly; & turn the ignition key to the start position. Perfect starts, every time ! I'll have a play with the TPS, later today, but I don't think it would work for me, in my situation. With COPs, it should be easy, as I'd only be redirecting the Ignition trigger outputs 1 or 4, to fire COPs & 1 & 4 together. Signals on ignition trigger outputs 2 & 3, would switch fire COPs 2 & 3 together. Remember, with COPs, I'm only talking about redirecting 5V DC trigger signals, as each COP, has it's own ignitor within/onboard. I'll try it here on the bench, with the Speeduino ECU & four spark plugs. I'll let You know how I fare. If You haven't worked me out yet; I just "love experimenting", & seeing whether ideas are "goers"; or "dead ducks". You'll never know, unless You try it out. Here is what AI dug up, about the pros & cons of Waste Spark, & Sequential ECU control: Here is what AI produced regarding emission standards requirements. Cheers Banjo

My KE30 2 door Corolla, has always been a very fast starter, since I installed my very first electronic basic ECU; which was a Jaycar project, from back, nearly a decade ago. Many people who have seen my car, have remarked how quickly it starts. It still has that Jaycar ignition system installed, (but not for much longer). The triggering is an olde K Series Denso dizzy with the springs & bobs weighs all removed, & the shaft "locked up". The points were removed, & in their place, I fitted one of those Hall Effect sensors, with an ignitor built in. I described it back in this thread, that Graeme started, way back in 2017, when His daughters Rolla was going through points regularly. https://www.rollaclub.com/board/topic/73743-electronic-distributor/#comments My setup was possible, because our ignition switch, has a start & a run position. In the start position, the starter relay to drive the starter motor solenoid, is activated, & the "ballast resistor" in series with the coil is effectively shorted out. The rotor button directs the HT from the coil to the very next cylinder's spark plug, in the "timing format". The Jaycar module was not in circuit, during the starting process, & the timing was the static timing, of 10-12 degrees BTDC. I would like to do a similar thing with my Speeduino ECU. However, not running a rotor button & distributor cap & leads, the best You can achieve is start the engine in "wasted spark" mode. I believe some modern day cars use this technique. However, I'm not sure whether the Speeduino ECU has the capability of carrying out this action. I'm currently awaiting an answer on that question. However, it may well be possible, using the start position on the ignition key switch, to close a relay, that mates up the paired cylinders in "Waste Spark" mode. (1 & 4) & (3 & 2) I'll try a couple of experiments on the bench here, & see if it is possible. P.S. When an engines stops, it usually comes to a spot in the 720 degree rotation, that is in one of two spots. I asked Google AI about this, & this is what it came up with . . . . Cheers Banjo

Hi David, It's a sad story with fuel level sensors in olde Corollas. They are not particularly accurate or linear, in their display. I've tried all sorts of things over the years. Best bet is to find something similar, that will fit in your tank, with it's float; & then massage the resistive signal at the dash end until it reads close to what it should. The single biggest problem I've come across with these fuel gauges, is the DC power supply regulator, fitted to the back of the dash unit. They regularly burn out. a regulator is required, as the 12 volt supply can go up & down, & the gauge would like increase of decrease. The regulator, is usually a 6, 7, or 8 volt DC one. I'd pull out the dash, & with the fuel sendor unit disconnected; measure the DC supply, with a multimeter, at the dash. I think I covered this in more detail once, somewhere on here. I'll see if I can't find it, & add a link in my reply. https://www.rollaclub.com/board/topic/76685-cleaning-inside-of-fuel-tank/page/2/#comment-723267 Cheers Banjo

Many of You; like myself, spent yesterday afternoon, watching the final Supercar race, for the season; in Adelaide. There was a lot riding on the result, & the big question was whether Broc Feeney could pull it off. It was not to be, & He was devastated, during his interview after the race. His car was not performing well, & He'd had a similar issue during the race the previous day (Saturday). https://www.supercars.com/news/supercars-news-2025-broc-feeney-technical-issue-adelaide-grand-final-engine-sensor-triple-eight https://www.supercars.com/news/supercars-news-2025-triple-eight-suspects-repeat-broc-feeney-saturday-misfire-engine-issue-explained-grand-final It was suspected that the issue was a misfiring engine, due to the VR or Hall Effect crank angle sensor, which is fitted at the rear of the engine, & apparently picks up it's signals from a toothed ring, mounted in front & adjacent to the flywheel starter teeth around the edge. I've been playing with Hall & VR sensors for some time now, & have never had a failure. I did a bit of searching on the nett last night, & there was some interesting, if not confusing data. There were statements that the Supercars used VR sensors; while other articles stated Hall Effect. Maybe, there is a mixture of sensors utilized ? You would assume; that as Broc's team had experienced the fault during Saturday's race, that they would have changed it out automatically, overnight. Maybe regulations made that messy or not practical ? My reading this morning, indicates that some Hall Effect sensors are "temperature sensitive". This guy in the UK, found that that is; & can be an issue. https://www.youtube.com/watch?v=kAOlPEnuTMw I've always fitted mine down on the crankshaft pulley at front of the engine. Altezzaclub, has advised me that rally cars usually don't like fitting them on the crankshaft, as if the engine is mounted north/south, then the crankshaft pulley, is in the danger zone; as sticks/stones/dust & debris, often finish up is area. However the Supercar engines all being V8s, there would be a lot of heat, in the area the sensor is fitted, from the exhaust headers & pipes; which no doubt glow red during the race. Could not find a photo or picture of it on the nett, but presume, is is probably similar to this picture. I was interested to find in my reading, that SuperCars all run in "waste' spark mode, rather than "sequential". This surprised me, as I would expect you would need sequential operation, if you were tuning each cylinder to the "n"th degree, to get the most out of it. It also stated that they are not allowed, uder the regulations, to fit a CAS sensor or signal. They still need a signal, to determine the position of the crankshaft, so the teeth or openings on the ring adjacent to the flywheel ring gear, must have one or two holes or tabs missing. No doubt there will be more discussion on this particular event yesterday, in the days to come, & some of answers to the questions I've posed & pondered, will be answered. Please add to this topic, if You are aware of an answer to some of my unknowns. Most VR or Hall Effect sensor pulse outputs in ECU controlled engines, use "pattern recognition" expectation, & can signal or substitute, if a pulse does not appear. I'm interested, as I'm not sure, at this point in time; whether the Speeduino ECU I'm using is sophisticated enough, to have those timing pulse fault detection features. I will pursue. I'm also guessing, if Supercar race engines only run i waste spark mode; that the flywheel mounted punched hole or tab gear is in a 60-1, 60-2, or 72-2 format. I'd like to be able to start my K series engine in waste spark mode; & then switch to sequential once it is running. This makes starting quicker, as it only needs one revolution of the crankshaft, for the ECU to know it's position, whereas waste spark, can require up to 2 revolutions, to gain the same positional information. Please add, answers You have to questions I've raised, above. A very interesting subject. Cheers Banjo

Hi Graeme, Thank You ! Well it took me a while to find this MCM video You mentioned. The video, is less than a week olde ! https://www.youtube.com/watch?v=Z34a0estBNI I believe this is it. The bit about modifying the dissy, is right at the very end of the video. It appears Marty got the bits He needs from a "Timmy" in Japan, from I6 Industries. I stopped the video when He got to the point of holding up the toothed wheel. It appears to maybe have 36 teeth, with one missing (36-1). The toothed wheel looks like it is aluminium; because of it's colour, but is probably iron. That means the sensor could be a VR sensor, or maybe a Hall sensor, with a magnet built in behind it. Have to await the next episode to find out. I gather it will be hooked up to, a Haltech ECU ? I have taken this a step further, because I what to synchronise the CAS signal, produced by the camshaft, with a crankshaft pulse, which will produce a pulse, that appears, as if it was produced by the crankshaft, that is "rock solid", timing wise. Cheers Banjo

Hi Graeme, No, I haven't ! I used to follow them religiously, in the early days, but haven't watched them, for a while now. I'll certainly go take a look in the next 24 hours. My exercise with the 24-1 punched hole S.S. disk & the Mitsubishi module; is to be able to compare the performance of both, & see if there is any perceptible bettering of performance with the high resolution 360 slot disk, with 2 degrees resolution over 720 degrees of rotation of a full engine cycle, as opposed to the 24 slot disk, with just 30 degree resolution over 720 degrees of rotation. I'm always trying to find the limits of "things electronic", in things automotive applications, as breaking down "crossing the "Hay Plains", or similar long drive, could result in a very expensive tilt truck ride. I've never had to be towed in my entire life;; so not planning on starting now. Strangely, I came across an issue this morning with my setup, that has me baffled. I had the Speeduino running on the bench with my motorized camshaft simulator. I then pulled the CAS plug input, & was surprised that the Speeduino ECU continued to run perfectly. The readings on all the guages on the Tuner Studio dashboard, still read correctly. However, when I shut down the ECU, & restarted it, the ECU would not run, until I plugged the CAS sensor back in. There is no indication in Tuner Studio, that I can see; that indicates, that there is no CAS signal being received. So in real life, I'm driving across the "Hay Plains" & unbeknown to me the CAS signal Hall Sensor falls off, or becomes defective. I have no idea this has occurred. The engine continues to run perfectly. The half way across the plains, I stop for a break & a drink, & let the engine cool a little (it's the middle of Summer !) I hop back in to continue my trip. The engine will not start ! It was running perfectly when I stopped. This seems unbelievable; as up until I stopped for the break, there was no indication that anything was amiss. I need to add a bit of circuitry, that monitors the CAS pulse presence, whilst the engine is running, & alert the driver, if the CAS signal drops out. So still not believing that this is possible, I ask Google AI the following question. Google AI did it's homework, for me, & Here is the answer it came up with. A Speeduino ECU requires a Cam Angle Sensor (CAS) or Crank Position Sensor (CPS) signal to establish its initial synchronization and properly fire an engine. However, it can continue to run after the signal is lost because it uses historical data and prediction to maintain engine timing. This is common behavior for many ECUs, not just Speeduino. Why the engine needs the CAS signal to start Establish a baseline: Before the engine can start, the ECU must determine the exact position of the crankshaft and pistons. It uses the CAS signal to find a known reference point, typically a "missing tooth" on a trigger wheel, that corresponds to the engine's top dead center (TDC). Calculate ignition and injection timing: Once the ECU knows where the engine is in its rotation, it can accurately calculate the correct moment to fire the ignition coils and fuel injectors. Without this initial reference signal, the ECU cannot establish the engine's position and will not send any ignition or fuel commands. Predictive timing: While the engine is running, the ECU receives a continuous series of CAS pulses. If this signal is suddenly lost, the ECU doesn't immediately shut down. Instead, it predicts the upcoming pulses based on the engine's most recent speed and acceleration data. "Keep it alive" mode: This predictive, or "limp-mode" function, is a built-in safety feature in many ECUs. It allows the engine to continue running for a period, though often in a degraded state, which can prevent dangerous situations like sudden stalling in traffic. Sequential vs. wasted spark timing: The type of ignition timing can also affect how the engine reacts. With a sequential setup, which requires a CAS signal for cam position, a lost CAS signal would cause a major disruption. With a wasted spark setup, which only needs crank position, the ECU has an easier time predicting timing based solely on the last known crank speed. Degraded performance: The predictive timing is not as accurate as live sensor data. While the engine may keep running, the ignition and fuel timing will likely be suboptimal, which can lead to reduced performance, rough idle, and misfires. What happens next Once the engine with the disconnected CAS sensor is shut off, it will behave exactly as it would if the sensor were disconnected from the start: it will refuse to fire because the ECU has lost its synchronized state and cannot establish a new baseline for timing. _____________________________________________________________________ Google AI, advised it visited 10 off sites to detect the answer it provided. I'll go & have a look at these sites, & see what they say. Actually the Tuner Studio website was one website I saw listed. As an aside, I asked Google AI a couple of weeks ago, a question related to what I'm working on. I assumed Google AI would not source it's info from forums, but only recognised websites. Imagine my surprise, when one of the sites it visited was RollaClub. Even more surprising, was that the piece of info it had sourced; came from a post I placed on this subject, back in 2022 ! Cheers Banjo

I've built up another interface board, so I can still investigate this optical dizzy system on the bench, with the new Speeduino & a second optical dizzy I've built, with a 24-1 punched stainless steel disk. I ran into a few issues, as the pulses coming out of the Mitsubishi optical module, I fitted inside a K Series dizzy body, had me scratching my head. Mitsubishi Optical Distributor Module: There are a number of issues that I have come across, with the above modules. This is largely due to the fact; that little specific & detailed information is available; as these modules are sold as replacement parts, & expected to be simply replaced, & the engine wiring harness, just plugged back in. For those of us, that are experimenting with them; or using them in other ways, to process their trigger & CAS output signals, it is not entirely clear, even with A.I. querying. 1. Are the modules powered by a +12 volt or + 5 volt DC supply ? 2. What is the arrangement of the four (4) connections to the module ? 3. Are output pulses, in a switched +12 volt or +5 volt format; producing positive going pulses ? 4. Are the output pulses simply a “open collector” output where the output goes to ground, for the duration of the negative pulse, & requires external fitting of a suitable load resistor, to either, a +12V or a, +5V DC supply ? 5. Is the pulse produced, when I.R. (Infra Red) light passes through the punched disk, or when it is blocked ? 6. Is there any internal logic, that prevents the leading or trailing edges of the CAS & trigger pulses, to coincide ? ECU microprocessors that are fed these pulses; usually are only interested in the leading & trailing edge of these pulses. This is because they are usually connected to “interrupter inputs”, on the microprocessor. This technique is designed to remove the reading of these pulses from the constant running code. The code to read the interrupter input, is not in the program, running continually. When an input changes state (either hi to lo, or lo to high), on one of these “interrupter inputs”; then the running code stops & jumps to the programming code that handles this input. The program then resumes to the constantly running cyclic code, after it has finished reading & acknowledging the pulse has arrived. 1. My research has indicated that most, if not all optical modules; are +12 volt DC powered. 2. The second consideration is to find out what the four (4) connections are, as to get it wrong, when wiring it up; could possibly result in the module becoming inoperative, (defective, U/S). I still do not know, if all optical distributor module manufacturers, all use the same arrangement ? However, the one I am using, (model 93740928) suitable for a Mitsubishi engine is connected as follows. Holding the module, with the IR reader head facing vertically, & facing the 4 pin connection moulded plug ; from left to right the four (4) pins are as follows . . . . . Pin 1: +12 volts DC. Pin 2: 0 volts/Ground Pin 3: Trigger pulse output. Pin 4: CAS pulse output. 3. It appears, that few if any modules actually produce a positive (+) going output pulse, for either the trigger or CAS pulse. 4. This is easily checked, by either connecting a multimeter, or a LED test probe, to the output, & turning the dizzy slowly by hand. If there is no positive voltage output, it is more than likely that the output is a “open collector” format, where the output pin is tied to ground, for the duration of the pulse. If this is the case, the other end of an external “load” resistor can be connected to either +12v or +5V, from each dizzy optical output. 5. The next question was . . . . Is the pulse produced, when I.R. light passes through the punched disk, or when it is blocked ? One would expect that when the I.R. passes through the disk’s punched hole, that a pulse would be produced. However, this is apparently not the case. It is impossible to look into the narrow slot on the I.R. header reader, & see whether the light beam is blocked or open. The light beam is infra red, so cannot be seen visually. The only way, is to remove the punched disk completely, then slide a piece of thin “opaque material” into the slot, to block the light, whilst monitoring the output. The answer; is in the common name for these devices, where the word; “interrupter” is used. ie: when the light beam is “interrupted”, the pulse is created / provided. This is the opposite to what maybe expected; that a pulse is provided, when the I.R. light passes through slot in the thin stainless steel rotating disk. Not the case ! One could assume, that the single CAS pulse & triggers pulses, are both produced exactly the same way. If You don’t want to remove disk, then monitor the CAS pulse output & rotate the dizzy slowly, until the output pulse appears. As there is only one CAS pulse, per 360 degrees of dizzy rotation, it is easy to see the result. You can pretty safely assume that both CAS & trigger pulses are both “light interrupted” outputs. Not the case ! As there being only one CAS pulse per revolution, of the dizzy; that would result be a very long pulse. The CAS pulse output, therefore has to be inverted, in post processing, to produce the C.A.S. & trigger pulses, in the exact same format (negative going). As far as I know, all commercial ECU, use & expect negative going trigger & C.A.S. pulses. 6. The stainless steel punched disk I am using with this optical reader, (24-1) is purposely punched; such that the leading & lagging edges of the slot for a single C.A.S. pulse, & the trigger pulses, do not coincide. (see photo of punched disk below) This is to prevent two interrupt commands being generated to the micro-processor, at the very same time. Micro processors, are very fast devices; but they are basically confined to carrying out one instruction/action at a time. This is much harder to achieve, when there are 360 punched slots around the outer edge of the punched disk; as on my Nissan optical dizzy module. The interface circuit I've built between the dizzy optical module & the Trigger & CAS board, also inputs a third pulse generated by a Hall Effect sensor, mounted adjacent to the crank pulley. This interface board creates a single C.A.S pulse by logically "Anding" the CAS pulse generated by the disk above ; & the single crank pulse per crankshaft revolution. Here are a few of the oscilloscope screen grabs of the pulses generated by the interface board, as fed into the Speeduino RPM2 & RPM1 inputs. The vertical divisions on the oscilloscope screen are 2 volts; so the negative pulses depicted are 5 volts in amplitude. The top trace is the single CAS pulse generated by the dizzy, & synchronised with the stable, wider pulse, derived by the crankshaft (not depicted). The bottom trace; is of the negative going pulses, generated by the 24 slots around the disk. Note that the "transition" of these two streams of pulses; do not occur at the same time. This can be clearly seen in the following screen grab. Here is the little interface circuit, I've built to process the Mitsubishi optical module outputs. Working with CAS & trigger pulses, can always be fraught with noise issues, in the automotive environment under the bonnet of a car. Careful use of shielded cables to transmit CAS & trigger pulses between the dizzy & interface board; & interface board & ECU, should always be installed; grounded at one end only. In addition, I have incorporated in this interface board; optical IR couplers, at both input & output of this circuit. In addition the switching functions in both the 4093 CMOS quad Nand gate, & the MOC5007 opto-coupler, both incorporate hysteresis switching, which is a great way of reducing noise in the system. Hysteresis, simply is two levels of switching, with a dead band in between the two switching points. In other words, a 0-5V dc signal switching from zero to 5V, will not switch, until the voltage reads say 4.0 volts. Conversely, a 0-5V dc signal switching /transitioning from 5V to zero; will not switch, until the voltage decreases to say 1.0 volt. For input voltages in between 1 & 4 volts, nothing switches. This is the hysteresis band, which is very useful in reducing noise in inherently noisy control systems. Cheers Banjo

My last post above, was the theory of how You can stabilise the CAS pulse, that is produced by the camshaft driven dizzy, but now it was time to turn the theory into practice. I built an interface "circuit board", that slotted between the dizzy Nissan optical distributor module & the Speeduino ECU. The interface board accepted 360 pulses per dizzy rotation, as well as the single CAS pulse from the dizzy. In addition, there was a single "crankshaft generated" pulse to line up with the CAS pulse from the dizzy, every second revolution of the crankshaft. There are twelve teeth on the cog on the bottom of K Series distributor; so you choose one, where the two pulses basically will align; & the dizzy is located, so the 4 way electrical plug & socket faces away from the engine, so it is accessible. I positioned the crankshaft at an angle where it's pulse occurs, at a point before top dead centre, greater than the "very maximum advance", You are ever going to use. Anywhere between 60-110 deg BTDC No: 1 will suffice. N.B. A lot of ECUs actually note this point as being ATDC. So 60-110 deg BTDC becomes 250-300 deg ATDC. Then all you have to do is rotate the dizzy by hand, until both pulses align. I fitted a couple of LEDs on the interface board I built, so this was easily accomplished. As per the diagram below, in my previous post, it is important, that the pulse created by the CAS in the dizzy, is wider than the pulse generated by the crankshaft. The crankshaft pulse cannot move around. The dizzy CAS pulse can move around, dependent on wear & slop in the camshaft sprocket & chain; as well as the helical gear between the bottom of the dizzy & camshaft. This exercise was to measure the angles of overlap, to ensure the CAS pulse was extremely stable. This needs to be checked with a degree wheel on the crankshaft pulley; or the flywheel, (which is exposed on my engine test bed arrangement). I was a bit concerned as the 54mm dia. slotted stainless steel disk in the Nissan module is extremely thin (0.5mm) & I did not think I would be capable of filing out a tiny hole to make it wider, (if it was necessary) without destroying a disk; whose replacement cost can be around $ 100 ! As it turned out, that was not necessary. The single pulse width from the crankshaft was 4-5 deg of crankshaft duration. The single pulse from the dizzy CAS pulse was 15-16 degrees of crankshaft duration. I then carefully rotated the dizzy, until the crankshaft 5 deg pulse occurred right in the middle of the 15-16 degree duration of the of the dizzy CAS pulse. That means that the CAS pulse generated by the dizzy, can move around + & - about 5 degrees, without having any control over the stability of the resultant pulse, when these two pulses are "logically", AND'd together. So next step will be to fire the engine up, & put the dual channel oscilloscope on both pulses & the resultant "AND'd" pulse, & see whether the theory all plays out in practice. Each of the 360 & single CAS pulse from the dizzy, are clean & square, as demonstrated in the picture above, on the oscilloscope. Those vertical axis squares on the screen are 2 volts each; so the pulses are exactly 5V in amplitude. The development of something like this, is not something that can easily be accomplished on a running engine. I built a little test stand to mount the K Series Bosch dizzy, with the Nissan 360-1 trigger S.S. disk, so I could work on it's design, in the comfort of my office/workshop. A little 3500 RPM electric motor drives the dizzy, as depicted below. The electric motor is fully controlled by a knob from a "dual gang" potentiometer. One gang, controls the electric motor speed, & the other sends a TPS (Throttle Position Sensor) signal to the Speeduino ECU. As the electric motor will run right up to 3500 rpm, then I can effectively test the ECU at engine speeds up to 7K rpm, which is a lot more than my little 5K engine will ever be asked to do ! The Speeduino ECU is connected to TunerStudio software; which has a dashboard with all the guages You need for fully testing the ECU. TunerStudio, also incorporates a trigger pulse data logger, where you can view the stability & duration of all of the trigger pulses. Both diagnostic functions are really appreciated, when developing something dynamic like this exercise. P.S. When I get this running in my KE30, I will still need a Speeduino ECU for further bench testing. I don't need the hassle of pulling my ECU in & out of the KE30 constantly; so I managed to purchase another Speeduino PCB very cheap ($10). I've just about finished populating it with components, so I will be soon have two (2) to work with. The later version (V0.4.3d) printed circuit board, is about 2/3rds the size of my other board, (V0.3.7) & has a single 40 way plug in terminal block, so removing it & refitting it, to the vehicle wiring loom is much easier. P.S. When You look at the pricing of the big brand commercial ECUs; the cost really prevents the likes myself & others, from "dabbling" in this way. The Speeduino ECU, is the most low cost effective ECU system, I have come across, & is "open sourced", so it is always getting better & better, & has a very active forum, where You can always get assistance & information. Cheers Banjo

I'm am amateur welder; but am "in awe" of people that can produce precision cutting & welding, like you depict there. However, it was something I spotted in the top right corner of one of your great set of photos, that brought a smile to my face. I gather you sit on "the Esky", & put your feet on the two little "steel sticks" out the side. I can just see a handle bar & steering mechanisms on the front wheel. However, it's the yellow & black "motor at the rear that caught my eye. It looks to me, that it might be a chainsaw motor, with blade & cutting chain removed ? ? ? Cheers Banjo

A number of people have sent me PMs, asking about comments I have made in the past, about errors & stability of pulses generated in my experiments, from various trigger wheels & CAS arrangements. These comments are only related to those, who want to control their ignition and/or injection, using a "sequential system". (injecting & firing one cylinder at a time; (1-3-4-2 ) This results from the crankshaft having to rotate through 720 degrees, (2 revolutions); to fire all cylinders in sequence. In "wasted spark" mode, two cylinders are fired together every revolution. (1 & 4, on the first crankshaft revolution, & 2 & 3, on the second crankshaft revolution, although there were some olde British designed & manufactured 4 cylinder engines which had a firing order of (1-2-4-3). https://www.youtube.com/watch?v=ZCUftAm3Qag As any engines get older & older, they inevitably wear; & timing chain/sprockets & timing belts stretch & loose. This results in timing errors, in anything that is "camshaft driven; such as valves, & distributors, which require precise timing. It is impossible to correct valve timing, except by replacing camshaft sprockets, chains, & tensioners. Other items that are camshaft driven; items like the fuel pump, & oil pressure pump, don't need precise timing considerations at all. Injection & ignition timing is however critical, to get the best performance from your engine. There are a couple of places where this slop, wear, & inaccuracy of timing can occur. One is obviously the sprockets on the crank & camshaft; & the chain that joins them. The distributor, is driven at right angles to the camshaft from a set of helical gears & a dog joint to the oil pump, in our K Series engines. The dog joint & slot is a wear point. Next time you have you distributor out, take a look at it, & you'll see the tell wear mark, on the bottom of the dizzy shaft. However, this dog wear does not effect the timing; as the dizzy is driven by the meshed gears on the bottom of the dizzy shaft, & the camshaft. I recently did a test on my 5K engine, which had the sprockets & chain replaced about 4-5 years ago. I jammed the "locked" distributor internally, so it could not turn at all. I moved the crank in a clockwise direction & took a reading, from the large degree wheel attached to the flywheel; & noted the reading. I then turned the crank in the opposite direction; & again took a note of the degree wheel reading. I then subtracted one from the other, & the result was nearly 4 degrees. Bear in mind, that this was a static test, in that the engine was not running, & there was no oil pressure to drive the timing chain tensioner, to take up the "slop" in the timing chain. Dynamically, I'd expect the error would be about say 2 degrees. It could be much greater, in a tired olde K Series engine. A better way to see this "slop" in the camshaft chain & sprocket setup, is to run the engine at idle, & watch the timing light, illuminate the TDC mark on the camshaft sprocket cover. If it is rock solid, then your engine is in a good state; (timing wise) If it is wandering about, back & forth; then there is wear. I have come up, with a system, that overcomes that timing error, in a "sequentially fired & injected engine". It does this by synchronizing the varying C.A.S. (camshaft angle sensor/pulse) which is moving around; with a crankshaft generated pulse, which cannot move around, as it is rock solid, because it comes from either the crankshaft pulley, or flywheel; both of which, are one with the crankshaft. The following rough sketch, describes this graphically. The top line (A) of the C.A.S (Crank Angle Sensor) pulses are created by, & generated in the distributor. Timing wise they will move around relative to the crankshaft, as depicted on line "A". The camshaft angle sensor (CAS), will produce this once every two revolutions of the crankshaft. The second line "B" is the crank angle sensor, which is rock solid, as it cannot move around horizontally, but does turn up every revolution. However, if we logically "AND", or "NAND" these two pulses together in a logic IC (integrated circuit), we produce an extremely stable single pulse, every two (2) revolutions of the crankshaft, which becomes our new stable C.A.S signal, generated by the crankshaft, in co-ordination with the camshaft. The only proviso of this arrangement; is that the distributor produced pulse must be wider than the crank pulse, so as it moves around laterally/timewise; that the two pulses will always overlap, & allow the logically generated C.A.S. pulse on line "C" above; to be produced. So where do we need to place this crank pulse, operationally, described above as line "B" ? Most ECU manufacturers suggest somewhere well before the most advance one could expect to require. Very few engines, or ECUs, would require an advance, under any condition; greater than say 40 deg BTDC No: 1 cylinder. I've always set mine somewhere between 90 - 60 degrees BTDC No: 1 cylinder. Be warned that many ECU manufacturers use & stipulate a number that is after TDC; so the setting in the ECU, may actually be 270 - 300 deg ATDC (which is 360 - 90 or 360 - 60). Unless you are using a distributor that distributes the spark via a rotor button & individual spark plug HT wiring, where the positioning of the distributor, is critical, the setup is pretty easy. Set the crankshaft in the desired position, where the pulse for line "B" above is produced. Then simply insert the distributor, & rotate until, the CAS pulse (line "A") above is also present, at the same point in time. Ensure duration of "A pulse" overlaps pulse "B", & pulse "C" will be produced cleanly & reliably. Hope that is all clear. Cheers Banjo

Well sounded like a really good idea, at the time; but the realisation of it, was like going down a "rabbit burrow"; with so many twists & turns. The Nissan vehicles were one of the first to introduce the 360 slot disc, although it had to be 54mm in dia.; whereas all the others; including my 24-1 S.S. disc, were 50mm in dia. The optical reader module was also slightly larger, to accommodate the 54mm disc. This posed a problem; as the module for the 50mm disc, just fitted inside a "gutted" the larger K Series 'Bosch dizzy housing, so how was I going to squeeze in the 4mm larger Nissan module, therein. However, I was buoyed & encoraged by Taz-RX, who had previously done this. https://www.rollaclub.com/board/topic/10940-sis-5kte-ke55/page/67/ It required a lot of cutting & filing; drilling & tapping; to generally get the module, concentric around the center shaft. It also required a completely new base plate to be fashioned, to attach the optical module to. There are two critical measurements, when attempting this project. One is that the center shaft of the dizzy, is completely concentric with the big hole it passes through, in the module. This results in all the slots in the punched S.S. disc, passing directly under the infra-red source & reader head, on the top right corner of the above pic. The second critical dimension is that the disc pass clearly through the middle of the slot in the reader head. The gap in the reader head, is only 1.5 mm wide. The disc is 0.5mm thick. That only leaves a gap on either side of the S.S. disc of 0.5mm. So just when I got it all right, & was fitting the dizzy cap I had made; I found the reader head on the module , fouled with the inner edge of the dizzy cap; requiring some delicate work with a rotary cutter & drill. Anyway; looked good when finished & now I had to make it work. I then made up a new stand for the dizzy, so I could test it, on the bench, as with 360 slots around the edge of the S.S. disk, You don't need any vibration to upset the "cleanness" of the delivered pulses to the ECU. I was dissappointed to find that the Speeduino firmware does not cater for the Nissan 360 slot disc I had, which had 4 off CAS slots per revolution. One slot which is the true CAS pulse; is wider that the remaining three (3); which are all the same width. Looked like I might have to revert to using a /2 frequency divider, to reduce the effective trigger pulses to 180 per camshaft revolution; & also block off the three (3) CAS slots I did not require. (tiny little slivers of black insulation tape worked well) I then came up with an idea of cheating, by setting the firmware trigger settings, to advise the ECU that this dizzy's outputs where generated by the crankshaft; & not the camshaft. That also allowed the teeth per revolution to be set to 180, as Speeduino firmware will not accept a trigger number greater than 255. Well it worked, & I now have some serious testing to do on the bench, to ensure that it works under all conditions, & scenarios. P.S. Not having to divide the trigger output of the module by two (2); resulted in a trigger resolution of 2 degrees, which is excellent. Cheers Banjo

I was in Canberra at the weekend, to see the Floriade Flower Festival, & view for the first time, a house my son & His partner have recently purchased, On Sunday, the 5th, My son took us to the lovely little country town, of Boorowa, for the Irish Woolfest. https://irishwoolfest.com.au/ Was very glad to see a KE Corolla, in the main street parade ! There was also a lovely Hilman Imp ! And a Bellett A couple of very olde classics . . . . . . And of cause; some tractors ! And a couple of exotics . . . Lovely town, with a lot of history; & a lovely day out ! Cheers Banjo

Altezzaclub put me onto a Camry TPS that He has used; & that are reasonably priced on ebay. The only disadvantage is, that they are not rotated from the center. Center driven ones are available. This Honeywell TPS I found online, has a 6mm dia. shaft in the middle of it; however, at $ 220.00 ea., its a bit rich. Link to Above Honeywell specs. Big issue mounting a TPS on a carby K series engine; as it has to sit out there in the middle of nowhere. I made a mounting bracket out of 3mm thick aluminium, as below. It is important, that the center of that big round hole, where the Camry TPS is mounted, is perfectly in line with the carby butterfly shaft center. The two top holes arrowed, attached to the top of the carby. The bottom single mounting point, was the extension on the inlet manifold, which normally, will support the standard air filter housing leg. Add an EFI Solutions 4W plug & harness wire back to the ECU under the dash; & it looks like this. Works really well ! Only three of the four wires are used in a standard TPS, ECU installation. Twas time to eventually fit my COPs mounting frame, to ensure I can fit the Bosch dizzy in there, underneath it, as I'm now not using the trigger disk, on the crankshaft pulley. So all good, but a couple of return coolant lines to reroute first. Getting close to the first fire up, as I now have an acquired a Nissan CAS dissy, to fit into a Bosch K Series dizzy housing, which Altezzaclub sourced at His local wreckers, ThankYou ! This particular one has a 54mm dia. S.S. disc, with 360 slots around the edge. That will provide excellent resolution,; for the ECU, for the position of the crankshaft, at any time. Cheers Banjo

Good to hear You are "back on the road" ! Chers Banjo

Hi Pete, So sorry to hear You have "a health issue". Join the club ! Getting back behind the steering wheel, will certainly help. I did a recent trip to Sydney & back, from Brisbane, & it was really good for the moral. I find, when You are on the open road, travelling close to the speed limit, You are concentrating so much on the job at hand; that all other issues seem to be forgotten about, for the time being. I've tried a few coolant water additives, over the years, but not Thermocure Evaporust. Had a quick look on Google last night, & found a utube guy with a BMW, who did a clean, & the results were really good. https://www.youtube.com/watch?v=aq3brzIA3e0 What I like about this product, is that it keeps giving for quite a while. In another video, a guy didn't throw away, the waste from the coolant system. He left it in a plastic tub, & just threw in any metal tools that had a bit of rust on them. After about a week, He took them out, & rinsed them, & they looked like new. Really a problem in Brisbane & Qld. in general, where our humidity levels are higher. P.S. What is it about BMW water pumps ? I had to change mine, in my BMW; that had a failed bearing like this guys. The seal breaks down, & once there is coolant in there; it is "bye bye bearing". Cheers Banjo

Water flow issues at the back of the head, have always been a problem. The original moulds for manufacturing K Series heads, were made in a couple of styles. Whether it is a block or a head; there is a need to remove the sand inside the finished product, before the bare casting starts to get machined. It's those holes that get filled with "Welsh Plugs", in the block. The head on the other hand, already had a large hole right in the front of the head, where the thermostat housing attaches, & the coolant exits the engine. It needed a hole in the casting, to remove the sand at the rear of the head. There were some K series aluminium heads that had the equivalent of a large Welsh plug. The best head was one that had a bolt on plate, with four (4) off 10mm bolts. This is a reasonably large hole, which you can release a lot of crud through. It is also an ideal spot to fit an additional outlet to; to feed coolant back to the thermostat housing, from of the rear of the head & engine. There were some early model Corollas, that took the cabin hot water heater feed from this point. I also have a 4K-U engine, with a factory fitted coolant feed point from the rear of the head, back to thermostat housing at the front; running through a factory fitted pipe running down the side of the engine, just below the spark plugs. I've also modified my 5K engine, to have a return pipe, running just behind the inlet manifold. Tests I've carried out before & after fitment, indicated the real reduction in temperature at the back of the head. I think I documented it on here somewhere; but off memory, it was about a 15 deg C reduction. Cheers Banjo

After a lot of time spent, investigating different triggering systems for my 5K engine, I've come up with a simple arrangement, by fitting an optical CAS distributor module, inside a K Series Bosch dissy, which is a lot wider than the more common Denso model dissy. I'm really happy with the Speeduino's performance on the bench, so it is now time to fit it to my 4K-U in my KE30, 2 door sedan, daily drive. I've just been checking all the external temp sensors on the car today, to hook into the ECU. However, I've just realised, my KE-30, doesn't have a TPS on it. It will need one, for the Speeduino to function properly. I have a TPS of "unknown origin" hooked up on my 5K engine on a test bed in the garage, but don't want to pilfer that. It's a bit crude, but it has worked well on the 5K on the test stand, for a few years now. This olde TPS rotates from the center of the carby butterfly shaft. Just looked on ebay & there are plenty of TPS units there for many makes of engine. They don't however indicate which direction they rotate in. Currently I just want to test this ECU, using ignition only. Once that's sorted, I'll add the 7K EFI inlet system I have, complete with injectors. Has anyone made up a TPS arrangement for a K Series carby engine, that works OK, & could pop a photo up, of how they have mounted it, & what TPS unit they used. The stainless steel timing disc I currently have, has 24 camshaft trigger pulses per revolution, & one CAS pulse. I would like a S.S. disk with 36, 60 or 72 holes around the disk. Chased everyone here in Australia, & nothing available; so might have to bring one in from the USA. There is one, that was used on Nissan vehicles which has 360 slots punched around the perimeter. Heaven knows how they do it; as the disks are only 50mm in diameter. It anyone has one of these disks they want to part with, for $$$, then give me a yell. The 360 slot disc, on the camshaft, would provide a crankshaft RPM resolution of 2 degrees, whereas my current 24 tooth disk, provides, just 30 degrees. If I do get hold of a 360 hole disk, I'll have to divide down the frequency a bit, as Speeduino ECU setup, will not accept a number greater than 255, for the number of trigger teeth. Cheers Banjo