Generic Oilfield Engine

[cfellows]

This is turning into a real head-scratcher.  Terry Mayhugh came over a couple of nights ago and we spent 3 hours trying different things.  We tried relocating the ground wire and making sure all the other leads were well separated.  No luck, we zapped another hall sensor.  We verified the continuity of the ground wire between the engine frame and spark coil ground.  We also hooked a ground wire from the hall sensor ground to the engine frame, soldered in a new hall sensor and tried it again.  No joy, the sensor stopped working almost immediately.

I then brought out two of my other engines, all wired the same, and, using the same ignition module and battery, tried each of them several times.  Both started and ran with no problem.  I switched them off and restarted them each repeatedly with no failure. 

We then went back to the first engine, soldered in a new hall sensor, and switched spark plugs with one of the other engines we had just tested.  Before plugging in the ignition module, I plugged in my test circuit with an LED and spun the engine over to make sure the sensor was working.  The LED flashed when the magnet rolled past the sensor as it should.  So we disconnected the test circuit and reattached the ignition module, turned it on and flipped the engine over.  This time the engine never even started.  After flipping it over several times, we disconnected the ignition module and reattached the test circuit.  The sensor was dead. 

So, now I've gone through 5 sensors trying numerous things to no avail.  The same ignition module runs my other engines and their hall sensors are none the worse for wear.  I've ordered 10 more sensors from Newark.  I'm going to keep trying different things.  I really want to get to the bottom of this.

Chuck

[smfr]

My (very uneducated) guess would be be on some high-frequency current induced in the sensor circuit by the spark plug. If you have an oscilloscope I suppose you could look at the voltage across the Hall sensor. Perhaps a small capacitor fitted somewhere would quench the HF?

Simon

[cfellows]

My (very uneducated) guess would be be on some high-frequency current induced in the sensor circuit by the spark plug. If you have an oscilloscope I suppose you could look at the voltage across the Hall sensor. Perhaps a small capacitor fitted somewhere would quench the HF?

Simon

Thanks, Simon.  Terry had also mentioned that possibility.  I'm going to be working with Roy Scholl at S/S Engineering who makes the ignition module to see if we can wrestle this thing to the ground.

I guess my main question is, with everything external to the engine being the same, why would one engine fry hall sensors and the others don't.  I just don't see how the ignition module could be at fault.  The only other culprit I can think of is the wiring harness from the hall sensor to the connector on the ignition module.

[Roger B]

As another currently baffled engine builder can I offer a few off the wall thoughts:

Is the hall sensor on your other engines mounted on an earthed (grounded) metal bracket or a plastic one?

Are you using an insulating or metallic head gasket (assuming the plug is in the head)?

Do your other engines run faster than this one? If the current through the sensor is too high it may not have time to 'cook' it on a high speed engine.

[fumopuc]

Hi Chuck, until now you have not mentioned if the magnet positioning/orientation  north and south to the hall sensor was checked.
May be this is the only remaining difference between two of your engines ?

[Don1966]

Chuck when you say you used the same ingition and Hall sensor and battery. Do you mean the same kind or the same ignition module and Hall device? If not the same ignition module try switching it. You may need to check the current by placing the magnet if front of the sensor and leave it there till you read the current. Is the Hall sensor you ordered the same number as the ones in your other engines? Is the Hall ground wire seperate from the ignition wire? How much distance is the HV lead from the other wires? Does the HV stay on when you keep the magnet in front of the Hall sensor? This last question is to see what kind of circuit you have. Some electronic ignitions have and internal oscillator and the Hall Device turns it on. In which case could cause HF noise into the battery leads and could burn out your sensor. LIKE Simon said a capacitor could filter this out. It would have to be put as close to the ignition circuit as possible. Not at the end of the leads. Something from .01 to  .047 UFD ceramic with a high voltage rating.

Don

[Mayhugh1]

Chuck,
         Another thought I just had was maybe the high voltage isn't the thing killing the sensor; and, in fact, maybe the sensor is really not being killed. Maybe the sensor is oscillating. I believe the sensor has a Bipolar output stage and sometimes the complex wiring impedance (stray capacitance and inductance) can be reflected over to the input of the driver as a negative resistance causing the driver to oscillate. If it is oscillating, then the CDI module would not get a chance to charge its capacitor, and the result would be no high voltage and an assumption that the sensor was dead. I can't remember if we tested the sensor every time with the low voltage test circuit that we thought we had blown it. Even if we had, it is possible that the sensor continued to oscillate during that test since the portion of the sensor harness remained connected to it. Remember in our testing we swapped everything between the two engines except for the wiring harness. If the driver was oscillating there are several possibilities: 1) nothing was actually killed and the frequency of oscillation was so high there was no high voltage being created in the CDI nor led indication being seen in the test circuit, 2) the sensor was killed by a power dissipation issue in the chip itself, or 3) the sensor was being killed by the CDI module through some circuit interaction we don't understand due to a lack of circuit information about the CDI. In any event, a possible solution is to place a 100 ohm resistor in series with the white lead of the sensor and to physically place this resistor close to the sensor itself. This resistor will overwhelm the complex impedance due to the stray wiring impedance and maybe kill the oscillation. -Terry

[Ian S C]

Chuck, the only thing I can see that might be of trouble, the LT wires and the HT wire coming out the same hole in the base.  There may not be a transfer of current via leakage through the insulation, but the capacitive effect of the wires lying along side each other.  Keep the HT as far away from the LT as you can.       Ian S C

[cfellows]

Still waiting on some parts.  Roy Scholl is sending me an interface circuit to plug in between the CDI module and the hall sensor.  He's also including a couple of new hall sensors which will give me a chance to test a different wiring harness.

Roger, the hall sensors are mounted in plastic on all 3 of the engines tested.  There are no head gaskets on any of the engines (good thought, though) and I tested the continuity between the base of the spark plug and the end of the ground wire that plugs into the CDI module.  All engines run at about the same speed although one of the engines that doesn't fail is a hit n miss which fires the ignition every crankshaft revolution.

Achim, the magnetic orientation is good.  The hall sensors work for a bit, then fail.

Don, I am using the exact same CDI module and battery for all the engines.  I simply unplug the wires from one engine and plug them into another engine.  We've tested the faulty engine with all the wires widely separated and it didn't make any difference.  One of the other engines which doesn't cause failure has the hall sensor and high tension ground wires in close proximity same as the problem engine.  I believe the operation of the CDI module is as follows.  It has an oscillator which turns on and begins charging a capacitor when activated by the hall sensor.  The oscillator turns off and discharges the capacitor through the coil when the hall sensor turns off by moving the magnet out of range.  Terry had mentioned the possibility of the oscillations inducing current into the battery leads since the CDI module and the battery case are in close proximity.  However, I don't see how this would cause a problem on one engine and not the others.  Finally, I've tried hall sensors from 3 different sources.   One of the engines that doesn't cause failure is using the exact same sensor (Honeywell S440R) as the problem engine.  We also burned out 2 hall sensors that Terry had brought over, but I don't know what make or model they were.

Terry, here is the content of an email I received from Roy Scholl yesterday...

Here's what I'd like to do. I'm going to make up an interface circuit board that will plug into the hall sensor connector. I am going to put the 200 ohm resistor in the feed line, a 5.1 volt Zenor diode in both the red and white leads. If this does not cure the problem then I'm in very deep do-do. I'm 99% sure it will. I can send a couple spare sensors also because if it does cure the problem I would like you and or Terry to remove (one at a time and test run) each of the Zenor diodes. I figure we have a 50/50 chance of finding the culprit lead on the first test. What I would expect is that the red wire is seeing a large voltage spike for some reason. When you remove that diode I would expect to see it fail a sensor. Of course there is also the Murphy factor so it could only be the white wire. If you remove both diodes and it does not fail then I would suspect it is a current issue and the 200 ohm resistor takes care of that.

Chuck

[cfellows]

Got all my new parts, including the hall sensor / CDI interface today.  Tried a bunch of different configurations.  Now I can't get the Hall sensor to fail, even if I go back to the original configuration.  Don't really have anything else to try except running it until something happens...

Chuck

[Dave Otto]

Hi Chuck

Maybe you just got a bad batch of crappy Chinese sensors.

Good to hear it's working better now.

Dave

[Don1966]

Chuck I a firm believer of flipping all the rocks when it comes to trouble shooting electronics and electrical and I have to say something is missing and getting a bad batch could be one of them. I am glad to see it's now doing fine. You showed just how much you have that ability my friend. Nice work all around Chuck.

Don

[b.lindsey]

Chuck, did the failed hall sensors come from Roy also? I guess he could get a bad batch of them too but I suspect he buys what he things are good quality ones. Seems like you have checked most everything else, so maybe it was a simple as a bad sensor. At any rate, glad you have it working now.

Bill

[cfellows]

I burned out three sensors that I had bought from Newark electronics.  Then I burned out 2 or 3 more that terry had brought over.  Chances are pretty good that his were from a different source and we're probably a different part number.  The new ones I just bought, that haven't failed, are also from Newark electronics and are the same part  number as the previous ones that failed.  I've never had any of Roy's sensors fail.  I now have one of his sensors on my "sensor killer" engine and it's still running strong.

Chucks