Car ignition system, Hall Sensor failure in 10 min Why?

[GoceKU]

Today my little niva die just as i pulled out of my street. I lost spark at the spark plugs, It has an electronic ignition, that i installed about a year ago, so everything is OEM parts and almost brand new. After troubleshooting i wasn't getting signal from the hall effect sensor in the distributor. I had a new in the box sensor so i replaced it, they are identical, car run but still issues running like randomly dies, and after 10 min it wouldn't run for more them 5-10 seconds, after that i replaced everything else, known good ignitor, new ignition coil, another good igniter, another coil, disconnected the tachometer, check the power everything is fine. One thing i noticed the tachometer was bouncing when i was cracking the engine. So its the Hall Sensor failure again but there must be something burning them out, what to check. The Hall Sensor is a standard type like in the picture below. Any experience or advice is much appreciated.

 

[markba633csi]

Two things: Automobiles are a hostile environment for electronics. The power supply is usually quite noisy from the alternator and ignition system.
You might need some additional filter caps near the ignitor and perhaps even at the sensor. Check for poor grounds

Second thing could be arcing in the distributor from the rotor to sensor.

 

[Deleted member 69715]

First, I am not familiar with your car or that particular ignition. When you originally installed that ignition did the instructions mention a specific air gap to clearance the pick-up at?

 

[great white]

Confirming the grounds are electrically sound (ie: corrosion, loose, etc) is always a good first step when chasing electrical “gremlins”…

 

[Bi11Hudson]

Up front, I am not familiar with your car or that particular ignition.(quoted) However, a "Hall Effect" sensor simply replaces the "points", contacts, on the line side of the coil. The "flyback" from a 12 volt ignition coil will smoke a transistor, often within seconds. You might look to see what sort of "condenser" (capacitor) is used and if external that it is properly grounded. It also would be possible to use a condenser from a points fired automotive ignition system to connect the "Square Wave" output of the Hall Effect sensor to ground.

There are many types of "snubbers" used with electrical equipment. They are simply capacitive and resistive devices rated for a higher voltage but otherwise essentially the same as automotive condensers. Electro-Magnetism is a theory that gets very deep very quickly. In short, magnetism is energy in a stand-by mode. When a switch is closed, initially current flows but then becomes just exitation current. When the switch is opened, and the magnetic field collapses, the magnetic field "tries" to maintain its' strength. Since the switch is open, there is no current flow. In its' attempt to get a current flow, the magnetic field will create a very high voltage. I have seen Current Transformers in industry that have literally exploded recorders and buried parts in a concrete wall. We estimated the voltage peaked around 150 KV based on the damage. A magnet on an overhead crane has a similar effect. The "discharge controller" of even a small (36") magnet has several grid resistors and a goodly sized capacitor. When power fails on the hot rails, the discharge controller won't work. Usually, the controller guts end up on the catwalks and the magnet doesn't work any more. An automotive coil is but a smaller magnet.

Semi-conductors (transistors) can handle high current for a short time. But when the rated voltage is exceeded they will let the magic smoke out immediately. Like 3 minutes ago fast. . . A "condenser", a "snubber capacitor", allows the current to flow as the magnetic field collapses. That keeps the voltage to a reasonable level. This is not intended to be a lecture on electricity. It is simply an overview aimed at a novice. We are talking many years experience here. Schools and books only cover the very basics of the theory, This discourse could not take place with even a graduate electrical engineer new to the field. He would need to have suffered a CT blowing up or a magnet discharge controller failing in his face to fully grasp the theory. A good example of a CT is the donut shaped sensors on the power lines to a large commercial customer. It is the non-electronic equivilent of a hall effect sensor.

You have spoken at length about the small car. I have followed some of the rebuild but didn't comment. As a Russian built component, I cannot comment on the sensor. But under today's circumstances, quality control is often offset by price or time constraints. It is quite possible that a faulty batch of ignition modules managed to get to market. If there is an external condenser somewhere in your system, it has quite likely failed or is very near failure. Using a condenser very near the coil could well alleviate your problem.

EDIT: Added some text and an afterthought, if I may. A modern tachometer, unlike the older mechanical tach on my diesel tractor, is merely a volt meter that reads the delay time (or lack of) of a constant pulse train from the points in the distributor. Hence the need for an aftermarket tach to need a ranging resistor for 2, 4, 6, or 8 cylinders. The speed of the pulse train, not the duty cycle, determines how fast the engine is running. As in a Volvo(?) 3 cylinder 2 stroke engine from the '60s would need to be connected as a 6 cylinder 4 stroke engine to a tachometer. 3 engine pulses per revolution. . .

When the meter is "jumping around" it is indicative of spurious voltages coming from somewhere. Again, if what I suspect is correct, the snubber for the coil is occasionally missing a pulse, allowing the flyback from the coil into the system. This would likely be a random, short term, voltage spike that would be essentially dampened by the battery. But for a short duration, on a local system, could develop spikes of well over the probably 36 volt tolerance of the hall sensor. As a Hall Effect sensor is a semiconductor, one can easily see where it would be disrupted by failure of the snubber.

.

 

[Jake M]

Today my little niva die just as i pulled out of my street. I lost spark at the spark plugs, It has an electronic ignition, that i installed about a year ago, so everything is OEM parts and almost brand new. After troubleshooting i wasn't getting signal from the hall effect sensor in the distributor. I had a new in the box sensor so i replaced it, they are identical, car run but still issues running like randomly dies, and after 10 min it wouldn't run for more them 5-10 seconds, after that i replaced everything else, known good ignitor, new ignition coil, another good igniter, another coil, disconnected the tachometer, check the power everything is fine. One thing i noticed the tachometer was bouncing when i was cracking the engine. So its the Hall Sensor failure again but there must be something burning them out, what to check. The Hall Sensor is a standard type like in the picture below. Any experience or advice is much appreciated.

How long do you have to leave a spare tire in place for it to imprint it's lettering like that?


It's hard to tell you exactly what to check for, as your vehicle doesn't exist in my world, so I have no idea what makes it tick.... To get started, you want (and it sounds like you have?) a very good overview of the ignition control system layout.

The first thing I'd want to do is check the power and the ground, under load. That is, engine running and (hopefully) acting up. Make sure they're solid and steady. Best referenced directly to battery negative, and not other places that "should" be the same potential.

The second thing I'd want to do might not be so accessible. I'd want to see the return signal. (The square wave). If you can get your hands on an oscilloscope, DSO, graphing meter, or anything that could present that square wave to you, you could probably make a LOT of deductions from that information.
Short of that, a digital (not analog) volt meter, on the AC scale should show a steady AC voltage for the return signal. The actual voltage number will be meaningless, but the question would be, is that voltage constant at a steady RPM? Catching the car when it's acting up would alter that voltage reading if the signal were dropping out, and if the signal were not dropping out, you could not guarantee, but be very suspicious that the signal was not dropping out.

If you can correlate the tachometer fluctuations to an actual dropout in the signal, you've guaranteed that the loss of the distributor pickup signal is the actual cause. If you can correlate fluctuation in the power/ground circuits (While they're under a load) to the tachometer fluctuations... On the other hand, if you find a steady return signal from the hall sensor, and the tachometer still fluctuates, you can move on to the power stage/amplifier/igniter module. Or the Engine controller? I'm sure if you have one of those it isn't doing much, but again, the system layout becomes important.

You're in a spot that makes the general public hate repair professionals... A repeated failure (presumably at this point), where you are no longer trying to "prove that a part is bad". You're trying to prove that literally, everything else is good. That's a hard and frustrating thing to do. (And of course, expensive if you're paying retail....).

Before tossing any more parts, do validate the power and ground, and at least the steady AC voltage of the return signal WHILE THE CAR IS RUNNING, and hopefully get it to act up with a meter attached. That is the best direction.

Short of that, if you're gonna toss parts into the diagnostic shotgun (Hey, we've all been there...), validate that the air gaps to the sensor and the wheel are correct. Make sure there's no cracks in the wheel. Look carefully at the distrubutor cap, rotor, the carbon button, the plug wires, and the plugs themselves. Make sure all of that is in excellent condition, and there's no unnecessary restrictions to the secondary ignition that might make it want to seek it's ground via an alternate route (perhaps through the distributor housing).

 

[Lo-Fi]

They're generally pretty robust... Though often fed 5V via a regulator somewhere. Check it's being fed the correct voltage as that might kill it. Tacho bouncing implies some kind of weird ground issue though, so maybe a return path isn't quite as solid as it should be? Alternator not doing something funny and spiking up to silly voltage because the reg is poorly?

 

[homebrewed]

With the sensor starting out working OK and then failing in a few minutes suggests voltage spikes being induced onto the power, ground or signal lines. It's possible that something else failed, like a transorb or voltage snubber, bypass capacitor or a shielding problem, possibly large voltage transients introduced by nearby circuits that are the result of a totally unrelated failure. Have any ignition wires been re-routed that might come close to the sensor? Some work that might have caused a floating ground (an unconnected wire)?

I think it's likely you can replace that sensor lots of times and they all will fail until you get to the actual root cause of failure. It may be that the original sensor failed due to to the same problem.

A hack that might at least allow you to have a usable car would be to place ~15V zeners right at the sensor power/ground terminals, and on the output/ground terminals. They would clamp any voltage transients to a survivable level -- maybe -- depending on the magnitude of the transients. If extreme, they might smoke the zeners, too. A big capacitor between the sensor's Vcc and Ground terminals might help too. Don't put one between the senor's output and ground, that will mess up the signal needed to fire the ignition system. The only thing you can do for that pin is to clamp the voltage (as previously mentioned) or insert a series resistor to limit the current flow in that line. All these suggest some electrical surgery so they would be in the "if everything else fails" category.

In my experience connection problems are the most-likely source of problems like this. Not that it can be easy to diagnose -- due to vibration or constant flexing wires can break inside their insulated jackets, making it hard to track down. My VW Golf regularly starts acting strange because the wires going into the door break due to the flexing due to opening/closing the door; and they most always break _inside_ the jacket. What fun to diagnose, huh. A major clue is to bend individual wires and see if they curve uniformly or have an abrupt bend. The latter is a smoking gun.

 

[whitmore]

I'd check to see that all the metal around the sensor was adequately ground-bonded; auto
electrics do sometimes depend on chassis metal for power connection, and that
metal has some... interesting history on this vehicle. Second thing to check, is
the alternator wiring.

 

[GoceKU]

Up front, I am not familiar with your car or that particular ignition.(quoted) However, a "Hall Effect" sensor simply replaces the "points", contacts, on the line side of the coil. The "flyback" from a 12 volt ignition coil will smoke a transistor, often within seconds. You might look to see what sort of "condenser" (capacitor) is used and if external that it is properly grounded. It also would be possible to use a condenser from a points fired automotive ignition system to connect the "Square Wave" output of the Hall Effect sensor to ground.

There are many types of "snubbers" used with electrical equipment. They are simply capacitive and resistive devices rated for a higher voltage but otherwise essentially the same as automotive condensers. Electro-Magnetism is a theory that gets very deep very quickly. In short, magnetism is energy in a stand-by mode. When a switch is closed, initially current flows but then becomes just exitation current. When the switch is opened, and the magnetic field collapses, the magnetic field "tries" to maintain its' strength. Since the switch is open, there is no current flow. In its' attempt to get a current flow, the magnetic field will create a very high voltage. I have seen Current Transformers in industry that have literally exploded recorders and buried parts in a concrete wall. We estimated the voltage peaked around 150 KV based on the damage. A magnet on an overhead crane has a similar effect. The "discharge controller" of even a small (36") magnet has several grid resistors and a goodly sized capacitor. When power fails on the hot rails, the discharge controller won't work. Usually, the controller guts end up on the catwalks and the magnet doesn't work any more. An automotive coil is but a smaller magnet.

Semi-conductors (transistors) can handle high current for a short time. But when the rated voltage is exceeded they will let the magic smoke out immediately. Like 3 minutes ago fast. . . A "condenser", a "snubber capacitor", allows the current to flow as the magnetic field collapses. That keeps the voltage to a reasonable level. This is not intended to be a lecture on electricity. It is simply an overview aimed at a novice. We are talking many years experience here. Schools and books only cover the very basics of the theory, This discourse could not take place with even a graduate electrical engineer new to the field. He would need to have suffered a CT blowing up or a magnet discharge controller failing in his face to fully grasp the theory. A good example of a CT is the donut shaped sensors on the power lines to a large commercial customer. It is the non-electronic equivilent of a hall effect sensor.

You have spoken at length about the small car. I have followed some of the rebuild but didn't comment. As a Russian built component, I cannot comment on the sensor. But under today's circumstances, quality control is often offset by price or time constraints. It is quite possible that a faulty batch of ignition modules managed to get to market. If there is an external condenser somewhere in your system, it has quite likely failed or is very near failure. Using a condenser very near the coil could well alleviate your problem.

EDIT: Added some text and an afterthought, if I may. A modern tachometer, unlike the older mechanical tach on my diesel tractor, is merely a volt meter that reads the delay time (or lack of) of a constant pulse train from the points in the distributor. Hence the need for an aftermarket tach to need a ranging resistor for 2, 4, 6, or 8 cylinders. The speed of the pulse train, not the duty cycle, determines how fast the engine is running. As in a Volvo(?) 3 cylinder 2 stroke engine from the '60s would need to be connected as a 6 cylinder 4 stroke engine to a tachometer. 3 engine pulses per revolution. . .

When the meter is "jumping around" it is indicative of spurious voltages coming from somewhere. Again, if what I suspect is correct, the snubber for the coil is occasionally missing a pulse, allowing the flyback from the coil into the system. This would likely be a random, short term, voltage spike that would be essentially dampened by the battery. But for a short duration, on a local system, could develop spikes of well over the probably 36 volt tolerance of the hall sensor. As a Hall Effect sensor is a semiconductor, one can easily see where it would be disrupted by failure of the snubber.

.

Couple of things you've mentioned seem to happen at once, why the first sensor failed is still a mystery, the second one seems to be defective from the factory, i've checked the wiring no defects. Installed an used sensor it works, also i've run out of LPG and on the way to the fuel pump the alternator belt shredded, so it may had voltage spikes. I've driven 15km today it seams to be fine.