3rd Generation Specific (1993-2002) 1993-2002 Discussion including performance modifications and Technical Support Sections.
Sponsored by:

testing the OEM knock sensor, shop manual info incomplete... help please.

Thread Tools
 
Search this Thread
 
Old 09-11-06, 02:36 PM
  #1  
Racing Rotary Since 1983

Thread Starter
iTrader: (6)
 
Howard Coleman's Avatar
 
Join Date: Oct 2001
Location: Hiawassee, Georgia
Posts: 6,097
Received 519 Likes on 289 Posts
testing the OEM knock sensor, shop manual info incomplete... help please.

i am really scratching my head reading and re-reading the shop manual on what should be a very simple test proceedure:

here's the manual:

Knock Sensor inspection (on vehicle)

1. connect voltmeter(-) terminal to the MEN terminal of the data link connector.
2. connect the data link connector terminals TEN and GND by using a jumper wire.
3. turn ign switch on.
4. lightly tap on the engine hanger with a hammer.
5. verify voltmeter indicator moves.
6. turn ign switch to off.

i get nothing on the voltmeter.... but doesn't the voltmeter positive lead need to be connected somewhere????

i have gone thru two knock sensors in a short period of time. any ideas?

thanks for any help.

howard coleman
Old 09-11-06, 02:54 PM
  #2  
Junior Member

 
foamfan's Avatar
 
Join Date: Aug 2006
Location: Rochester, MN.
Posts: 46
Likes: 0
Received 0 Likes on 0 Posts
IDEA: Knock sensors are usually Piezo-electric devices that produce voltage on a certain frequency range. It is unlikely that you can go through two of them over a short period of time. It is more likely that your voltmeter is not sensitive enough or you need to amplify the voltage produced by the sensor.

And Yes the Positive terminal of your voltmenter has to be connected.

Last edited by foamfan; 09-11-06 at 03:00 PM.
Old 09-11-06, 03:52 PM
  #3  
Racing Rotary Since 1983

Thread Starter
iTrader: (6)
 
Howard Coleman's Avatar
 
Join Date: Oct 2001
Location: Hiawassee, Georgia
Posts: 6,097
Received 519 Likes on 289 Posts
thanks for the help.

i run a power fc and a datalogit and do log knock routinely. in the last few weeks my knock readings have dropped to the default (1) reading which is the same as if the sensor was disconnected. there is a shielded wire around the sensor which communicates to ground thru the sensor body. a one prong connector situates between the sensor and the ecu. the shielded wire on both sides (not connected thru the plug) communicates w ground.

i did splice (soldered) part of the wire between the ecu and the connector and used aluminum foil to shield the spliced in wire. it shows continuity to ground. the first sensor ran fine for 500 miles and the second (i replaced only the sensor and it immediately worked) lasted around 150 miles.

i still need help as to where to place the positive lead from the voltmeter.

howard coleman
Old 09-11-06, 03:56 PM
  #4  
Rotary Enthusiast

 
kuroi FD's Avatar
 
Join Date: Jan 2005
Location: orlando/st. petersburg
Posts: 769
Likes: 0
Received 0 Likes on 0 Posts
those suckers arent cheap either, are they oem? I would call the person you bought them from and see what they can do for you
Old 09-11-06, 04:08 PM
  #5  
Lives on the Forum

 
DamonB's Avatar
 
Join Date: Feb 2001
Location: Dallas
Posts: 9,617
Likes: 0
Received 8 Likes on 7 Posts
Originally Posted by howard coleman

i still need help as to where to place the positive lead from the voltmeter.

howard coleman
It's got to be to the positive post of the battery.

Since the instructions say to read the negative side from the MEN terminal that must be the reference output from the knock sensor. That output must be with respect to ground since the instructions say to place the negative lead of the voltmeter there. Given that the positive lead must go to the positive terminal of the battery in order to create a potential in the circuit.

I think foamfan is right in that the knock sensor is going to produce some tiny reading so you'll most likely need a digital voltmeter to even notice it.
Old 09-11-06, 04:25 PM
  #6  
Tequila? ..it's like beer

iTrader: (1)
 
NewbernD's Avatar
 
Join Date: Mar 2001
Location: Woodbine, MD
Posts: 1,124
Likes: 0
Received 0 Likes on 0 Posts
Originally Posted by DamonB
so you'll most likely need a digital voltmeter to even notice it.
I would have guessed the opposite, that an analog meter would be better for identifying tiny pulses (unless you have an o-scope.) Most digital meters react too slowly to pick up momentary readings like these, but with a sweeping needle you can see it bump.
Old 09-11-06, 04:43 PM
  #7  
Rotary Freak

 
mad_7tist's Avatar
 
Join Date: Jan 2004
Location: tampa
Posts: 1,899
Likes: 0
Received 0 Likes on 0 Posts
i would agree that you need a Good meter fluke etc something with some resolution/speed or a good old school analog you would use to set dwell etc.

i have never done this but i assumed that the knock sensor would output a + voltage as it is a piezoelectric item. so i would say the neg lead to the men terminal and the pos lead to the neg battery terminal

i would agree with damon about the pos terminal but the sensor is not a variable resistace. i thought all pizo devices output voltage when compressed or changed shape when voltge is applied...

Last edited by mad_7tist; 09-11-06 at 04:45 PM.
Old 09-11-06, 06:53 PM
  #8  
Racing Rotary Since 1983

Thread Starter
iTrader: (6)
 
Howard Coleman's Avatar
 
Join Date: Oct 2001
Location: Hiawassee, Georgia
Posts: 6,097
Received 519 Likes on 289 Posts
THANK YOU ALL for the guidence. you sort of confirm my suspicions that the other lead would be to the battery. it would seem the sensor should output positive voltage (the sensor body is grounded) so maybe to the neg side. i do have a fairly good digital multimeter so i should pick up a signal. i will post tomorrow as to my results.

thanks again,

howard coleman
Old 03-18-10, 09:14 PM
  #9  
Senior Member

iTrader: (20)
 
EvilAreXeven's Avatar
 
Join Date: Mar 2008
Location: LosT Angeles
Posts: 416
Likes: 0
Received 0 Likes on 0 Posts
Howard, I'm sorry to bring this back from the dead, but this thread is the closest I've gotten in search.

I'm currently trying to test my knock sensor, and the shop manual leaves a lot of gray area.

In which way did you test the sensor:

1) Positive probe to positive battery terminal, negative probe to MEN connector, jump 10 + GRND, hammer engine hanger.

or

2) Positive probe to MEN connector, negative probe to good ground, jump 10 + GRND, hammer engine hanger.

If you remember, could you please let me know which way you tested the knock sensor, and what types of values you saw?

Although the manual makes vague reference to option 1), I used the second option because connecting the negative probe to the MEN output didn't make much sense...however, I'm no expert, which is why I'm asking you.
Old 03-18-10, 09:38 PM
  #10  
RX-7 Bad Ass

iTrader: (55)
 
DaleClark's Avatar
 
Join Date: Jan 2002
Location: Pensacola, FL
Posts: 15,399
Received 2,438 Likes on 1,509 Posts
First off, the shop manual's test procedure will only work with a stock ECU I do believe. The PowerFC doesn't have a lot of the diag stuff the stock ECU does - for example, you can't pull engine codes and the like.

If you intend on rewiring the knock sensor, you need a shielded wire. Radio Shack has a shielded wire, it has a foil braid around it to reduce interference. The knock sensor puts out such a weak signal that any interference will show up as knock. Stock, the shield is grounded - that's part of how the shield works.

Dale
Old 03-18-10, 09:42 PM
  #11  
Senior Member

iTrader: (20)
 
EvilAreXeven's Avatar
 
Join Date: Mar 2008
Location: LosT Angeles
Posts: 416
Likes: 0
Received 0 Likes on 0 Posts
Thank you for that information, but more importantly, I'm looking for information on the testing procedure and values as the shop manual is very ambiguous in its description.
Old 03-18-10, 11:12 PM
  #12  
Form follows function

iTrader: (8)
 
Speed of light's Avatar
 
Join Date: Oct 2004
Location: Now in Arizona
Posts: 1,203
Received 33 Likes on 21 Posts
When you check a pizo sensor, you need to use an AC voltmeter, on a low millivolt scale to see the sensor output. These sensors produce their highest output in their bandpass region, which is approx. 3-->4kHz, IIRC. It does not output a DC voltage, per se. That being said, the signal is usually superimposed on a DC bias--but you still need to use an AC scale to properly see the sensor output. Your best bet is to reference your DVM or VTVM to ground.
Old 03-19-10, 01:18 AM
  #13  
rotorhead

iTrader: (3)
 
arghx's Avatar
 
Join Date: Aug 2004
Location: cold
Posts: 16,182
Received 429 Likes on 263 Posts


that's from the 91 Technical data sheet for series 5 FC turbo cars. The FD technical data does not list a frequency range. The knock sensors are interchangeable between FC and FD.
Attached Thumbnails testing the OEM knock sensor, shop manual info incomplete... help please.-knock_frequency.png  
Old 03-19-10, 11:28 AM
  #14  
Senior Member

iTrader: (1)
 
ttmott's Avatar
 
Join Date: Aug 2006
Location: Space Coast Florida
Posts: 398
Likes: 0
Received 8 Likes on 4 Posts
A VOM is the minimum test to see if the sensor is even producing a signal. To understand what the sensor is really doing is to put an O- scope on it. With the scope you can see the duration and amplitude of the signal. The scope will also determine if the shield is doing it's job.
Related Topics
Thread
Thread Starter
Forum
Replies
Last Post
trickster
2nd Generation Specific (1986-1992)
25
07-01-23 04:40 PM
efini7
Test Area 51
1
09-04-15 02:59 AM
zuesskroph
2nd Generation Specific (1986-1992)
0
08-13-15 06:05 PM



Quick Reply: testing the OEM knock sensor, shop manual info incomplete... help please.



All times are GMT -5. The time now is 09:59 AM.