Road Speed Sensor problem - genius required!
Discussion
Happy New Year everyone, in spite of what’s happening in other parts of the globe!
My speedo’s been failing intermittently (when the engine’s hot) and soon after this happens the engine doesn’t run properly - I suspect the Road Speed Sensor (in the SD1 gearbox) as this is used to drive the speedo and it also feeds a signal to the ECU which adjusts fuel ratio etc.
I checked the voltage at the sensor connector and found the feed to be 12 volts, and the signal to be switching between 0.7 volts and 6.9 volts when the back wheel was turned. Aha, says I, that’ll be the problem because it should switch between 0 and 12 v approximately (according to the bible) (or 0 and 8.5-10 v according to a Haynes “Fuel Injection and Engine Management” book . There are only two wires going to the sensor and the unit seems to be self contained. It has a plastic helical gear on a freely rotating shaft with a sealed aluminium housing (about 30mm diameter). When I shove my finger in the hole from whence it came in the gearbox, all I can feel is a steel helical gear at the bottom which drives the plastic one. This steel gear is about 15mm away from the units housing, so I think it’s too far away to be sensed with an inductive sensor that may be inside the housing. There are no other metal rotating parts in the vicinity. I concluded that the unit must be self contained and the switching mechanism must be inside the aluminium housing.
Now here’s the tricky bit. I thought the problem might be a high resistance through the earth return through the body so I did a bench test and put 12v onto the feed, 0v onto the body, and measured the signal voltage. I got a constant 12v with no switching when the gear was rotated, even when a lump of steel was brought close to the sensor. I re-fitted the unit into the gearbox and re-measured the 0.7-6.9v.
Can anyone explain why the thing doesn’t behave the same way on the bench as in the gearbox?
Steve.
My speedo’s been failing intermittently (when the engine’s hot) and soon after this happens the engine doesn’t run properly - I suspect the Road Speed Sensor (in the SD1 gearbox) as this is used to drive the speedo and it also feeds a signal to the ECU which adjusts fuel ratio etc.
I checked the voltage at the sensor connector and found the feed to be 12 volts, and the signal to be switching between 0.7 volts and 6.9 volts when the back wheel was turned. Aha, says I, that’ll be the problem because it should switch between 0 and 12 v approximately (according to the bible) (or 0 and 8.5-10 v according to a Haynes “Fuel Injection and Engine Management” book . There are only two wires going to the sensor and the unit seems to be self contained. It has a plastic helical gear on a freely rotating shaft with a sealed aluminium housing (about 30mm diameter). When I shove my finger in the hole from whence it came in the gearbox, all I can feel is a steel helical gear at the bottom which drives the plastic one. This steel gear is about 15mm away from the units housing, so I think it’s too far away to be sensed with an inductive sensor that may be inside the housing. There are no other metal rotating parts in the vicinity. I concluded that the unit must be self contained and the switching mechanism must be inside the aluminium housing.
Now here’s the tricky bit. I thought the problem might be a high resistance through the earth return through the body so I did a bench test and put 12v onto the feed, 0v onto the body, and measured the signal voltage. I got a constant 12v with no switching when the gear was rotated, even when a lump of steel was brought close to the sensor. I re-fitted the unit into the gearbox and re-measured the 0.7-6.9v.
Can anyone explain why the thing doesn’t behave the same way on the bench as in the gearbox?
Steve.
Hi Rick.e.
when I tested the sensor in the gearbox, it was connected up to the ECU and speedo etc. When I tested it on the bench, it was not connected to anything (apart from my DVM).
Would a pull up (or down) resistor explain why the thing switches when in the car, but not on the bench, or does it just limit the current to the ECU input? I'm afraid my electronics knowledge is shaky, much to my frustration.
Steve
when I tested the sensor in the gearbox, it was connected up to the ECU and speedo etc. When I tested it on the bench, it was not connected to anything (apart from my DVM).
Would a pull up (or down) resistor explain why the thing switches when in the car, but not on the bench, or does it just limit the current to the ECU input? I'm afraid my electronics knowledge is shaky, much to my frustration.
Steve
Them speed sensors are funny things and are often driven by constant current sources or use changes in current to indicate the signals which requires a special drive circuit or at the very list pull up /down resistors. This means that it can only be really tested in situ. Some may require a magnet to be present to trip and not simply a bit of metal.
I don't know what you tested the system with but as stated in the bible, many multimeters average the signal out and so the low readings may not be accurate and simply due to the way the voltage is measured. A scope is the only real foolproof method.
Steve
www.tvrbooks.co.uk
>> Edited by shpub on Sunday 2nd January 23:15
I don't know what you tested the system with but as stated in the bible, many multimeters average the signal out and so the low readings may not be accurate and simply due to the way the voltage is measured. A scope is the only real foolproof method.
Steve
www.tvrbooks.co.uk
>> Edited by shpub on Sunday 2nd January 23:15
I think you guys are onto something here! I re-fitted the sensor into the box and connected the 12v supply, but left the signal wire unconnected. I then measured the signal voltage which was a constant 12v, even when the wheel was turned - just like I measured on the bench. I then reconnected the signal wire and once again measured the 0.7-6.9v when the wheel was turned. By the way Steve, I get a constant 0.7v when the wheel is stationary, and then I turn the wheel until the voltage switches to 6.9v and remains constant until the wheel is turned further, so I assume that there is no averaging going on.
So it seems that the damn thing only switches (or more accurately, you can only measure a switching voltage) when it's connected to the ECU.
So my dilemma now is - is there a problem with the speed sensor, or is there a problem with the ECU circuit driving the sensor? How do I find this out?
Steve
>> Edited by red griff 500 on Monday 3rd January 21:07
So it seems that the damn thing only switches (or more accurately, you can only measure a switching voltage) when it's connected to the ECU.
So my dilemma now is - is there a problem with the speed sensor, or is there a problem with the ECU circuit driving the sensor? How do I find this out?
Steve
>> Edited by red griff 500 on Monday 3rd January 21:07
Unfortunately I can't take the car anywhere as it's propped up in my garage with the exhaust off etc (need to do this to get the sensor out).
Having found that you cannot test the sensor with a voltmeter when it's disconnected from the ECU, I found that you can test it with the "diode" setting on your DVM. I suppose this isn't surprising when you realise that the thing is not just a basic sensor but has some kind of transistor driven output. Anyway, I heated the sensor in the oven and re-tested the output at elevated temp and found that it changed by a reasonably high amount and so I've ordered a new sensor (4 pulse per rev) from David Geralds. Hope to get that fitted before the weekend, and engine, gearbox and diff oil changed, fire her up and burble around a bit. Can't wait.
Incidentally, I checked the transmission bolts and found two out of four bolts on the propshaft flange (gearbox end) were loose enough to be rotated by hand.
Imagine the carnage if that came flying off at speed!!! Make sure you or your service guy checks these regularly!
Steve
Having found that you cannot test the sensor with a voltmeter when it's disconnected from the ECU, I found that you can test it with the "diode" setting on your DVM. I suppose this isn't surprising when you realise that the thing is not just a basic sensor but has some kind of transistor driven output. Anyway, I heated the sensor in the oven and re-tested the output at elevated temp and found that it changed by a reasonably high amount and so I've ordered a new sensor (4 pulse per rev) from David Geralds. Hope to get that fitted before the weekend, and engine, gearbox and diff oil changed, fire her up and burble around a bit. Can't wait.
Incidentally, I checked the transmission bolts and found two out of four bolts on the propshaft flange (gearbox end) were loose enough to be rotated by hand.
Imagine the carnage if that came flying off at speed!!! Make sure you or your service guy checks these regularly!
Steve
That does make it difficult....
Not surprised that the diode value changed when heated though. Tis the laws of physics operating as the diode junction voltage is proportional to the diode's temperature. Used as a way of measuring chip temperature. Yes I know Intel think they invented it but Mullard were doing it in the 1970s! One of my first jobs was to temp characterise silicon by measuring the diode junction.
Not surprised that the diode value changed when heated though. Tis the laws of physics operating as the diode junction voltage is proportional to the diode's temperature. Used as a way of measuring chip temperature. Yes I know Intel think they invented it but Mullard were doing it in the 1970s! One of my first jobs was to temp characterise silicon by measuring the diode junction.
Oh bugger! Steve, are you trying to tell me that I've wasted me hard earned cash on the new sensor. Ah well, it's fitted now, and the voltage is virtually identical to that which I measured with the old one i.e. 0.7v low and 7.0v high (when connected to the ECU). I guess that's the correct reading then. Haven't taken the car for a run yet, so I don't know that it's fixed the problem which occurs when the engine's hot.
Got me fingers crossed though.
Steve
Got me fingers crossed though.
Steve
Gassing Station | Griffith | Top of Page | What's New | My Stuff