How to test a Knock Sensor (P0328) - Subaru

Video Statistics and Information

Video
Captions Word Cloud
Reddit Comments
Captions
okay 2000 Subaru Legacy with a 2.5 we got a Pio 328 knock sensor code in memory we have no drivability problems just a code and a tech engine light and what I want to do is I want to walk you guys through how I handle these knock sensor codes every car is a little bit different and we're going to go to the sensor next and do a direct measurement real quick here's the wiring diagram for it this is a mental wiring diagram a couple things we want to look at pay attention to one is the knock sensor is grounded on the block itself so it's only got one wire the other thing to point out in this diagram is this little dotted line around here that is an indication of a shield this is a shielded circuit and the reason they will shield circuits like this is to prevent outside radiofrequency or magnetic interference so when we go to the car you'll probably notice that this wire is going to be a thick single wire component just something to point out follow the wire it goes over to the engine computer up here so it's just a single wire knock sensor again a couple different ways we can do this we can adopt in back probe the connector use a t pin use our scope and we can look at the signal while we're tapping on the block with a blunt object pry bar hammer next to the knock sensor and will actually make the knock sensor produce a sine wave zoomed out it would it would it would be something like this so as you've not as you hit on the block you would see an AC voltage on here we're going to do something a little bit different with this first and this system I believe uses a seven volt bias circuit on the knock sensor and we're going to do some text at the sensor itself see what we have and then I'll plug the theory in behind it okay we're down at the knock sensor give you a little perspective of where we are this is driver's side there's a throttle plate right there throttle cables right below that there's a bolt holds a knock sensor in place there are two connectors for this knock sensor one is a gray connector it's plugged right into the sensor and the other one's the white connector that we're back probing that is kind of the sub harness they call it that actually comes with the knock sensor so we're on the sub-continent back probing the knock sensor we're on a graphing multimeter right now and we're reading about two and a half volts on this knock sensor and what that line is it is a that is a bias line that this sensor uses and I misspoke in the last segment it's actually a 5 volt bias on this not a 7 so the computer sends 5 volts to the sensor the resistance of the sensor actually pulls it to ground a certain amount two and a half volts is pretty good that's what we want to see so a couple different texts we can do take a pry bar and I'm going to tap on the block next to the knock sensor we want to see activity on the screen when we do that you see that AC voltage is actually riding over top of that DC if I want to see that more detail I'm going to get out of the graphing meter I'm gonna go to my lab scope so I see a little bit better time base and I'm actually going to take my trigger and move my trigger up on a level that got to freeze this picture when it hits that level hopefully with my trigger there's your a CSUN wave right there riding overtop of a DC that's what you're looking at that's how the knock sensor works so this is actually a functional knock sensor right now is what that test tells you is this some intermittent fault the light comes and goes it does come and go so it's not constant no so a couple different problems we could have computer wiring sensor that's where we're at obviously the sensor is functional right now we're going to focus back on this bias voltage now you can see what the signal looks like when it's producing a or when the engine would be knocking the computer would see something like that and by the way that's not what sets a trouble code on these you get a knock sensor trouble code it's not because the sensor is picking up an engine knock you get a trouble code on this sensor from an open shorted circuit or sensor that knock sensor would be doing its job if it the computer saw something like that that would not set a code so engine knocking is not an issue here we're worried about the wiring or integrity of the knock sensor so one of the things that you can use would be this bias line if you understand how it functions and we actually have this if you guys remember signal circuit integrity at the end of my signal circuit integrity chapter in my book I gave you an example of a knock sensor where uses a 5 volt bias and how the computer uses it and how we can quickly identify circuit integrity so we're going to use that page in my book and plug it in right here and see what we got so I'm going to unfreeze this picture and take my time base we could probably go back to the graphing meter it doesn't matter I'm going to go to five second screen if we want to look at a longer time period now and what we want to do is we want to focus on the wiring of this system one of the things you can do is to unplug the sensor and this is what I'm showing in my book you unplug the sensor what you're going to read with the sensor unplugged is the computers bias line and we just buried our scale let me readjust this will go to 10 volts with an unplugged I'm reading four point nine seven volts basically five volts with the sensor unplugged so what that tells you is the wiring from the computer to the sensor is good and I'm not concerned at least initially of that wiring in between now we could wiggle the harness and see if we have any drops or anything like that that wouldn't be a bad test to do if that signal changed when you were wiggling the harness that would indicate a wiring problem between the computer and sensor this code is generated from a resistance change in this circuit we plug the sensor back in I'll show you this sensor is going to have a specific resistance that's going to pull this bias voltage down and in this case it's around two and a half volts so whatever the resistance is of this sensor that's what's pulling it down this is very similar to a coolant sensor circuit and how it functions as far as the bias line internal resistor external resistor voltage between two resistors is dropping the circuit down now if I take this I'm going to wiggle the connector on not the sub not the connector we're back rubbing but the sub harness connector at the knock sensor it's gonna wiggle this I want you to see we had had some decent changes here you see that all I'm doing is taking this pry bar and I'm wiggling very gently on that knock sensor connector at the sensor itself that is a sensor problem right there you're going down the road and at times that resistance is changing Computers going to see that bias line change and that's what's set in that trouble cut that test right there it tells you need an offset now you might be thinking well maybe it's the sub harness connector I'm actually stressing the pins on that and what I've seen on these when you're moving that connector you're actually moving the male pins on the sensor and that's what's dropping that out certainly wouldn't be a bad practice to try to tighten the pins up but in this case you're getting a sensor with a new connector you try one more time see if we can make it worse actually pretty good now that I've labeled it certainly don't want to break it I can't recreate that again but what you can do is this kind of testing if you had a hard fold on a knock sensor go down to the sensor measure your bias voltage see what it's reading it's reading five wiring is good all the way down your sensors bad there's no reason to even do a resistance check on the sensor manufacture is going to give you resistance specs for this if you go down to the sensor and you're fixed at zero volts on this design you unplug the sensor and it jumps to five again your harness is good you need a sensor so a little bit of a lesson on a bias voltage and we could definitely see resistance changes wiggling that connector that is where our problem is on this car I'm going to feel comfortable putting a knock sensor in this not worried about the harness not worried about the computer you
Info
Channel: ScannerDanner
Views: 34,458
Rating: 4.9463806 out of 5
Keywords: DIY auto repair, auto technician training, P0328, how-to, bias voltage, scannerdanner, engine performance diagnostics, knock sensor test, auto computer systems, auto repair, Subaru knock sensor, scope test, Subaru Legacy (Automobile Model), ASE L1, technician training, car repair, knock sensot tap test, 2000 Subaru Legacy Knock Sensor Test (P0328), how to test a knock sensor
Id: 2-JBTwOvD9w
Channel Id: undefined
Length: 11min 5sec (665 seconds)
Published: Tue Apr 15 2014
Related Videos
Note
Please note that this website is currently a work in progress! Lots of interesting data and statistics to come.