Banjo

Hi Graeme, The easiest check on this fault is to cut it in half, when it happens next time. To me the half way point is the starter motor relay, which I think in your girl's KE55, is down behind the driver's kick panel. If you remove the kick panel, you can locate it by sound or feel, as it a bigger than normal 12 volt relay, so has a good "clunk". If you turn the key to the start position, & the relay does (audibly or "feel wise") not operate, then the problem will be the relay itself, the ignition barrel, or an intermittent 12V supply to ignition, or the wiring between ignition barrel & starter relay. If the relay does operate, then it is more than likely that the fault lies beyond the relay, in wiring to the Starter motor solenoid, or possibly in the starter relay contacts themselves. If however, you can hear the starter solenoid come in, but no starter rotation, then the issue will be either in the heavy duty cabling or connections between battery & starter, or the brushes in the starter itself. A quick check you can do, is on a cold start, measure the 12 volts at the starter motor itself, to the engine or starter motor frame. Couple of long leads with a couple of little alligator clips will do. This may show up any potential issue with the battery to starter motor cabling or terminals, if the reading is very low. P.S. Prevention is always better than, cure on a cold winters might on a back road in Buccan ! Cheers Banjo

Hi Graeme, Intermittent electrical problems are always the worst ones to uncover. Open circuits, or short circuits are easy. I agree with what Jeremy has said about battery terminals, as starting the car from a cold start, is the highest load you can ever put on the battery & associated wiring. Unfortunately, it will happen again, as it unfortunately is not a "one off event". I would be inspecting all the heavy duty terminals between battery post & the starter motor. Don't discount the starter motor itself, either. The brushes on them do wear, & they can get to the point where they just make contact, with little pressure on the armature. There are two other items, that could cause the problem in a 40 year old KE55. The starter relay, which is hidden down behind the drivers kick panel, off memory. Not unheard of, for them to fail. usually contacts, because they handle a fair current to drive the starter motor solenoid. Never forget the ignition barrel switch. They do get tied after many years, & I have seen a couple in my time, where they worked sometimes, & not others, just as you describe. Probably a couple of hours work there, but you will solve the riddle eventually, which may be better than the fury of a teenage daughter, as I have one just like yours. Cheers Banjo

Hi Graeme, Yeh, hope it is something simple, & not the Accuspark unit. Was there a ballast resistor on the coil before you added the Accuspark ? If not, what was the primary impedance of the ignition coil you have fitted, as Accuspark unit can only handle coils with primary resistance of 3.0 ohms, without ballast resistor. If a ballast resistor is used, the total series resistance of coil & ballast should be no more than about 3.0 ohms. Low resistance primary coils, should not be less than 1.5 ohms. Cheers Banjo .

Hi Graeme, Good to hear you got it all up & going OK. Initially, you probably won't see any more improvement, than if you had just replaced a new set of points. However, unlike points, which start deteriorating from the moment they are fitted, the Accuspark unit, will not experience any drop off in performance, over time, EVER. However, the major way of improving the ignition performance of your girl's KE55, is with the advance curve. The Accuspark unit does not affect that. It is simply an electronic points replacement/substitue. Advance is still provided by the springs & bob weights below the dizzy mounting plate. These should always be kept clean & oiled lightly. I've been running my Accuspark for a couple of years now, without issues at all. I have seen a great improvement in performance, but I locked the dizzy advance mechanism up, & provided a far more aggressive advance curve, with a Jaycar programmable ignition controller. The Accuspark unit is purely used as a trigger for the Jaycar unit. If you ever want to go down that path, give me a yell, & I'll come across to Buccan, & show you what is involved. Cheers Banjo

Hi Keith, Well my 16 x 1.5M hose adaptor arrived today, so I can now install a radiator bypass hose, & see what effect that has on the various logged temperatures in the coolant system. I like your idea of using an ABS sensor to provide a signal that the car has stopped. It is not a very important sensor, for our application; unlike the real ABS application, where you need a toothed wheel, so you can measure rate of change of speed, as well as actual speed. I've been thinking about how I could mount a sensor, without too many issues. ABS sensors are pretty cheap on ebay. It's probably a bit difficult to modify a front KE hub, & fit a sensor. I was thinking of maybe attaching the sensor to the diff housing, and have the sensor measure the tail-shaft to diff flange going round. Probably no need to fit a toothed wheel. Maybe just sense the heads of the four (4) flange bolts whirring around, with a Hall Effect Magnetic sensor. After all, it's just a Go/No Go condition we are looking for. However, we could actually bring on the thermofan, above stopped, if you are crawling along in traffic, on a hot day, by switching the fan on when the speed is between 0 - 5 kph. Actually, whilst writing this I just remembered I've got a little "potted" reed relay, that is used for sensing a tiny magnet clipped to the spoke of a bicycle for driving the bike speed module. Could hook that up, real easy me thinks. So when the ignition is on, and there are no pulses, that would indicate the engine is running, but the car is stopped, & the thermofan could be switched on. You don't want the fan coming on, whilst you are doing a cold start first thing in the morning, so could add another condition, that the temperature has to be above some reasonably high limit. That would limit the fan to automatically come on, only when it is really needed. Lots of possiblities there. I'll let you know how I go. Cheers Banjo

Hi Jasper, It could well be ignition or carburation, but it is a bit hard without hearing the motor under load. 1. Does the flat spot feel like it is load related ?. If so, it could be running on 3 cylinders, or a spark plug or lead breaking down under load. 2. Does this shudder feel like the ignition cutting out altogether, intermittently ? 3. If it is fuel related, it will feel more like a "starving" condition, without jerkiness (dying sound). Have a look at the engine HV electics in the complete dark,, & look for any break down around or inside the leads. Have a good look for leaks in the inlet manifold, where it may be leaking more air in. Cheers Banjo

Hi Keith, I can assure you the rear head & lower radiator temp sensors are hooked up to the correct data logging inputs. It is however, a bit difficult to get your head around the rear of the head being cooler than the coolant at the bottom of the radiator. If it stayed that way, then there would certainly be something wrong. However, it doesn't, as & there are long periods, where the rear of the head coolant temperature is greater than coolant exiting the bottom of the radiator. This changeover, only occurs when the car stops, & the air is no longer passing through the radiator, & the top radiator hose coolant temperature is not yet high enough, to switch the thermofan on. You can clearly see this in graphs a few posts back on this page 10. The problem is, that as soon as the car stops, the coolant at the bottom of the radiator rises in temperature very, very quickly. The rear of the head has to play catchup. This sort of things always occurs in "systems" that have "transport" lags. First I am going to see what difference the radiator bypass creates. You will notice, that this occurrence, only happens in stop start traffic, but not on a long straight runs. I was thinking, I could flatten the response right out, by fitting a N.O. relay contact across the thermoswitch, with the relay automatically coming on, whilst ever the ignition was on, (engine running), & the car was stopped. That would result in air passing through the radiator at all times, and should reduce, & hopefully almost illiminate the sudden rises in lower radiator coolant temperature. I can test this out, without fitting the relay, as I have this switch on the dash, that can bring the fan on, irrespective of the conditions. I'll log a run in traffic, but every time I pull up, I'll switch the fan on. Likewise as soon as I move off again, I'll turn the fan off. The resulting graph, should clearly show the difference in coolant system performance. All fun, but learning lots at the same time. Cheers Banjo

Hi Keith, Although initially it sounds strange, that top & bottom of the radiator could be at the same temperature. When you think about what is happening, it is not so strange at all. Here is a graph of that small period of the previous graph. The Rolla had only been stopped at the Dentist for about 30 minutes, after the trip there, so the coolant was still at an elevated temperature above ambient, at around 48 deg C, when the car is started. The fan is not running at all, but the water pump is. However, by the time the temperature reaches say 87 deg C, the thermostat is fully open, & as water pump is running, so coolant flow through the radiator is constant, & therefore top & bottom radiator coolant temps are at the same temperature. However, as soon as the temperature of coolant at the thermostat area reaches 92+ deg C, the electric thermofan switch automatically kicks in. The radiator then does it's thing, & the bottom hose temp instantly drops rapidly. That cooled water then feeds through the engine block/head, and within 30 secs. the top hose temp starts to drop off also, to then be controlled by the thermostat, & held at between 80 & 90 deg C. All good, as should be. Note the fan ran for only 40 seconds, before it dropped the top hose temp; to the point where the fan was not required any longer. Your earlier point about the fan hardly ever required, is proving very, very true. What a waste of power driving that old plastic fan all the time. Ideally, every component in the system should be matched in terms of performance. However, in this case in mild ambient conditions, the aluminium radiator appears to be more efficient than needed at that time, & therefore the system cycles a bit, at around 4-5 minutes as you describe. As the ambient temperature rises, to say a very hot 40 deg day, I would expect the 4-5 minute cycling to decrease, & the top hose temperature to be fairly flat. The only way to prevent this "cycling", would be to fit thermostatically controlled louvers in front of the radiator, as are common in big trucks, to control the amount of air through the radiator; or fit an electric water pump, as are common in some racing cars, so they can control engine temperature to a very tight controlled point. It will be interesting to see what reinstating the radiator bypass hose does in terms of control, & "the cycling" you have identified. Just waiting on a 16 x 1.5M barbed 14mm hose adaptor to arrive, so I can reinstate the bypass function. Actually, I have an old TPS sitting around here. It wouldn't be too hard to hook it up & log how hard I push the accelerator pedal. P.S. As the bypass hose is very close to the point where I am measuring the "top hose" temperature, (under the thermostat), I thought I might move the measurement point to the bypass hose, which would allow it to react quicker, like the rear head sensor, which has a hose each side of it. Cheers Banjo

OK, so went to the Dentist & back, & recorded the trip at 1 second intervals. Here is the complete trip. This following graph is the trip home. The differential between the front & rear of the head is still around 9 deg C. However, despite that, the thermostat is working perfectly, maintaining the coolant temperature between 80 & 90 deg C. The graph below shows that clearly, and the thermofan kicking in, whenever the car physically stops, & the temperature shoots up over 90 deg C. Next step will be to reinstate the radiator 15mm bypass hose, & see what transpires, after a run with that. P.S. Still can't believe the lower radiator hose temperature drops so quickly to around 50 deg C, after stopping the car. Cheers Banjo

Hi Scott, The fuel guage not working makes sense, & is what you would expect with a polarity reversal. The DC supply feed to the fuel guage & temperature guage is usually a regulated voltage of around 7 - 10Vdc, depending on the model. These regulators in the Starlet, are most probably solid state, & it has been made U/S from the reversal. My guess is the ECU / ECM is U/S also. Off memory, the Starlet EFI only uses just a single injector. You could take the injector out, & take it to a injector cleaning service, which should have the equipment to test whether it is working. Cheers Banjo

Hi Keith, Good morning ! One thing I have not fitted on my coolant system, during these tests, is the thermostat bypass hose. My theory has always been, that here in the northern half of Australia, where it rarely gets cold, that the bypass hose wouldn't do anything much that was useful. To me, the bypass flow, would only lengthen the period of initial engine coolant warm up, which you want to be as quick as possible. Currently, without a bypass, my graphs indicate that period is about 8-12 minutes, from a cold start. Once the thermostat cracks open, the bypass hose function is basically redundant, to my mind. However, in reading up on it overnight, there appears to be several reasons for it's inclusion. One is to stop steam pockets forming in the head, during the warm up period, and the other is to prevent the water pump impellor from cavitating, which could happen during warmup, whilst the thermostat is completely closed. So this morning I thought I might remove the "block off caps" I currently have on there, and insert a piece of hose. It was then that I realised I have a mismatch between lower thermostat house bypass outlet point, & water pump bypass inlet point. I have a lower thermostat housing with a bypass 8mm outlet & a water pump with bypass 15mm inlet, on the LHS of the engine. It appears that somewhere during the K series development, Toyota in their wisdom changed the size of the heater bypass hose from 8mm to 15mm ID. I noticed there were a couple of pics in a thread you contributed to back in early 2015. This one clearly depicts the 8mm bypass hose on the LHS of the water pump, viewed from the front of the engine. However, on later K series engines, the bypass was increased to 15mm ID hose, & fitted to the RHS of the water pump, as depicted in this pic. I have the option of fitting either size to my existing setup, but I am wondering what was the reason Toyota increased the bypass hose size, & what existing issue with the small bypass hose size prompted this. My guess is that, the smaller 8mm hose just simply blocked up, as I have witnessed over the years, but I was wondering whether it was to increase flow, & why. Can you throw any light on this, from your experience ? I'm just a bit concerned, that if I fit the 15mm bypass hose, whilst running the high flow thermostat, the thermostatic control, might not be as good. I suppose, the proof will be in the eating. P.S. Going for a run shortly for about an hours trip, with the data logger set at 1 second logging rate, & see how that looks. Cheers Banjo

OK, so yesterday, I finished replumbing the return coolant line, from the back of the head, to join the exist coolant from the front of the head, mixing directly under the thermostat, in the lower thermostat housing. You can see the mods in the following pics. So prior to even fitting the Echo aluminium radiator & integral electric fan, I had consistently obtained differential coolant temperatures between the front & rear heat exit water points of 6.2 - 6.5 deg C. I was hopeful that replumbing the rear coolant flow to under the thermostat, would get these two temperatures, even closer together. Well, to my surprise, the complete reverse resulted, with a differential of almost double that at 10.7 deg C. Hmmm ! So back to the drawing board. Would love to be able, to have a look inside the block & head water jackets, & see where the water flows actually travel. As I can't do that, then the only thing is to change things, & see how it affects the graphed results. I might "throttle" the flow in the rear head return line, & see what affect that has. Maybe it requires a separate thermostat for each head exit coolant point. However, that getting away from the KISS principle. Cheers Banjo

Hi Scott, Welcome aboard ! it sounds like the wire you replaced from the battery, was probably the "fusible link". If it was thicker, you would probably have described it as a cable. Not sure whether a 34 year old Starlet had any form of reverse polarity protection, built in, on things like the EFI controls. Generally, electronics does not take kindly to have reverse polarity applied. You need to go through everything electrical in the car, with a check list, of what still works, & what does not. That may provide some additional clues. Horn, lights, gauges, fans, etc. My guess is that it will electronics failure, most likely. As the main electronic device is the ECU, & it is the engine that will not run, then that is probably the culprit. Remember, the EFI Starlets had conventional distributor ignition, so that will be why you obtain a spark. The ECU, to my knowledge, has no control at all over the ignition system. The only quick & simple way to know whether the ECU, is your issue, is to substitute with a known working one. P.S. If wires going to the injectors are hot, then either the injector coil has failed or the driver that fires the injectors have failed & are permanently on. Cheers Banjo

Agreed, the more you find out, the more questions it raises. With a couple of heads working on this, I'm sure we will be able the nut it out. A thermostat certainly is a "damped device", in terms of it's opening & closing reaction time, to changes in coolant temperature. Like you surmise . . . if you drop a fully closed thermostat at room temperature into a pot of boiling water, it does take in excess of 10 seconds to fully open. However, The whole coolant system has a response time also, which is controlled by the thermal mass of the block & head, & the flow rates provided by the water pump, along with the efficiency of the radiator itself, in terms of removing heat from the coolant. I suggest, the coolant response time is much greater than 10 seconds, as the block in particular, acts like a great big heat sink. I suspect, I will see some changes, once I have rerouted the rear head return line. At present, I am running the data logger at 1 reading per 8 seconds. I will run it in a "fast read mode", at 25 readings per second (40msec), & we should see a picture of the upper & lower hose temperatures, with a greater resolution. If the ripple effect is still there, I do have the opportunity, of lowering the Echo radiator efficiency slightly, by blocking air movement throughput, with a 100 mm strip of cadrboard, across the front lower section of the radiator core. Let you know what this turns up. Please keep your ideas coming. P.S. Just did a test on the kitchen stove with a new thermostat, suspended in a pot of boiling water. Soaked it in boiling water, until the valve was fully opened. Took it out, and watched it close. As soon as it was fully closed, I dropped it back into still "boiling water, (which was probably about 95+ deg C(, and started stop watch. It was fully open in 21 sec. Reset stop watch, & I then removed it, to room temperature (30 deg C), and started stop watch again. It was fully closed after 30 seconds. Therefore, the differential temperature that was applied to the thermostat, on "dunking" it in "boiling water", was about 15 deg C. (97 - 82 deg C) The differential temperature that was applied to the thermostat, after taking it out of the "boiling water", was about 65 deg C. (95 - 30 deg C) In real life where the differential temperatures would be around 10 deg C (90 - 80 deg C) the response time would be much slower than those measured above. Cheers Banjo

Hi Keith, Yes, there are still a couple of unanswered questions, which I still need to look into. That's why I have done the mod. in stages, to see exactly what effect, each change has made. At the moment, the return line from the rear of the head, uses the heater return pipe along the side of the head, behind the dizzy. This line does not return coolant from the rear of the head to the top of the radiator, but directly to the inlet of the water pump. This makes sense for a heater return line, but not for the rear head coolant return, which should be returned below the thermostat. I may even get around to doing this mod today, as I have another "lower" thermostat housing with a large spout on the LHS, that can be easily plumbed with the rear head coolant return line. I'll then have to run a separate, heater return line to the water pump inlet. I'll then see what difference that makes. The sudden fall in lower radiator hose temperature after the engine is switched off, is interesting. It actually tumbles, very quickly, once the engine stops. It is like once flow stops, the hot & cooler water separate quickly. Hot water moving up, & cooler water moving down. I think it is called segregation. As to why the variation in lower radiator hose temperature is so erratic, by comparision to the top hose temperature, is still a bit of a mystery. I have a theory, which I will look into. Although I have labelled the temp sensor as "top hose", it is actually not in the top hose, but fitted to the lower thermostat housing, in a existing spare 16 x 1.5M threaded hole. The sensor being mounted in the metal housing, dampens the response of the sensor, as it effectively has more mass. However, the lower radiator sensor, is mounted directly into the lower hose, using one of those "Red Devil" type adaptors, you referred to. The adaptor is made of aluminium, and has little mass. It therefore reacts more quickly to changes in temperature, & therefore has a greater "swing range". These adaptors were, I believe, designed for racing applications, so that faster response times of water temperature could be achieved. The graph below is a small portion of the graph above, between 9:40pm & 9:50pm on the 28th March. The graph just displays only the top/rear head & bottom radiator hose temperatures. You can see that head front (top hose) & rear head temperatures mimic each other, with a differential of about 4-5 deg C. The "top hose" temperature is flatter than the rear head coolant temperature. This is because the rear head coolant temperature sensor is not in the head itself, but mounted in a "Red Devil" type hose mounting in the return coolant rubber hose, directly behind the head. Again, like the bottom radiator hose, the response time for temperature changes is quicker, & this is reflected in the graph above. So where the front head & rear head coolant temperatures mimic each other, the lower radiator hose coolant temperature is a complete reversal. ie: when the top hose temperature goes up, the lower hose temperature goes down. My theory is; that the ripple effect in the graph of the top hose temperature is actually the thermostat action, constantly opening & closing off slowly, to maintain a consistent temperature. If that is the case, then the lower hose temperature reverse swing makes sense. Top hose temp increases; thermostat opens slightly more; increased water flow; radiator has greater cooling effect on more water, & return coolant temperature to the inlet of the water pump drops. The frequency of the "ripple" effect observed in the top & rear head coolant temperatures, appears to be fairly constant. My guess is that this is a result of slight mismatch between the dampened response time of the thermostat, & the transport lag of water movement around the engine. It takes a finite time, for the a specific sample of coolant to physically move from the outlet at the bottom of the radiator, until it exists the head into the lower thermostat housing, where the thermostat, can measure & react accordingly. So until, some other piece of data, jumps out at me, & dispels my current theory, that's what I am currently thinking. More to come, no doubt ! Cheers Banjo