SVX Network Forums Live Chat! SVX or Subaru Links Old Lockers Photo Post How-To Documents Message Archive SVX Shop Search |
IRC users: |
#1
|
|||
|
|||
Internal ATF Temp sensor & then some
I've been toying with the idea of adding a Tranny temp gauge. Searched a few posts to decide where to put the sending unit (seems the fluid out to filter/cooler line is most likely), then started thinking about the internal sensor.....
Alldata reports the sensor (nominally) sends to the TCU the following: at 20C (68F) 3.0 to 3.5V or 2.3K to 2.7K ohms at 80C (175F) 1.0 to 1.3V or 280-360 ohms. OK, my questions: Does anybody have any other data points so I could determine the curve a bit better, or at least what voltage/resistance lights up the dash indicator)? My intention is to use some circuit (TBD) and an analog voltmeter to show at least some general indication of ATF temp (realize this is probably not linear). With switching could double as a voltmeter - wow, what will I think of next - to keep tabs on the charging system. Anyway, searched archives for this data and didn't find any. So, anybody know anything more about the sensor in the valvebody of the tranny? |
#2
|
|||
|
|||
Im really interested in thsi subject. Someone please figure this out, it'd be *****in.
- Ca$h |
#3
|
|||
|
|||
Quote:
-**** |
#4
|
||||
|
||||
I agree with what you said Mike!!!!
__________________
. Earl .... ... .... ><SVX(*> Subaru Ambassador [COLOR=”silver”]1992 Tri Color L[/COLOR] ~45K (06/91) #2430 1992 Dark Teal LS-L ~184K (05/91) #0739 1992 Claret LS-L ~196K (05/91) #0831 1992 Pearl LS-L ~103K (06/91) #1680 1992 Pearl LS-L ~151K (06/91) #2229 1992 Dark Teal LS ~150K (07/91) #3098 (parts car) 1992 White LS-L ~139K (08/92) #6913 1993 25th AE ~98K (02/93) #164 1993 25th AE ~58K (02/93) #176 1993 25th AE ~107K (02/93) #215 1993 25th AE ~162K (02/93) #223 1994 Laguna Blue Pearl LSi ~124K (1/94) #2408 1994 Laguna Blue Pearl LSi ~144K (10/93) #1484 1994 Laguna Blue Pearl LSi ~68K (10/93) #1525 1994 Barcelona Red LSi ~46K (02/94) #2624 1994 Pearl LSi ~41K (12/93) #1961 1995 Bordeaux Pearl LSi ~70K (02/95) #855 1996 Polo Green LSi ~95K (03/96) #872 1997 Bordeaux Pearl LSi ~55K (08/96) #097 2003 Brilliant Red LS1 Convertible ~29K (04/03) #8951 1999 Magnetic Red LS1 Coupe ~33K (04/99) #6420 My Email | Old Locker | New Locker | Picture of 15 of the 19 |
#5
|
||||
|
||||
Lee, good and worthwhile idea.
Measuring the actual internal temp. of the transmission is the way to go. It would appear that the transducer comprises a thermistor. As there appears to be no specs. available the way to go would be to get hold of a transducer and do some tests and measurements. Establishing a calibration curve would take some work but the project would be interesting. Any one got one on hand or in a trans. not in use ?
__________________
Trevor, New Zealand. As a child, on cold mornings I gladly stood in cowpats to warm my bare feet, but I detest bull$hit! |
#6
|
|||
|
|||
Quote:
22630AA071 Temperature Sensor $21.07 + $8.00 S+H USD |
#7
|
|||
|
|||
getting longer and lost"er"
I've been cruising around the net trying to find some info. No solid leads, but here's the closest I've found.
Reviewing our thermistor, Alldata says: 68F = 2.3 to 2.7K ohms 175F = 280-360 ohms Sample from a web site for an ACI/2252 NTC thermistor 68F=2.8K ohms 175F=283 ohms 200F=177 ohms 230F=115 ohms 260F=77 ohms 300F=42 ohms So I think this is a close enough curve given the general non-linear nature of thermistors. On to my next step. Several sites suggest a whetstone bridge can be modified to employ a fixed resistor where the normally variable unit is used. That way when the temperature rises to some point defined by the fixed resistors, the bridge will become sufficiently unbalanced to allow enough current to flow to trigger a relay. This site specifically discusses using the modified bridge as a temp alarm http://www.atpsensor.com/ntc/ntc_app...ions_main.html I'm assuming at lower temps the resistence is high enough to limit current flow, also inhibiting the relay although the bridge is still unbalanced. This is what I now think the TCU does. I'm not an engineer, anybody out there care to comment on if this makes sense so far? I was originally thinking of simply putting a voltmeter across the lines into the TCU and sensing temp that way even though I knew it would be inverse and not linear. Now I wonder if I would have to buffer somehow to avoid screwing up the bridge. If I'm right (??), a more adventurous soul than me could open the TCU and change the bridge resistor values to get the light to come on at a lower temp (something I've seen asked for in earlier threads) without screwing up any other parameters of system operations. I'd still like to employ the built-in sensor for some unfathomable reason vice just buying the kit thing from Summit or ???? so now soliciting info AND circuit ideas. |
#8
|
||||
|
||||
It is very unlikely that a thermistor would be used in a bridge configeration. It is more likely that the variable voltage from the thermistor is arranged to operate one or more schmit triggers at set temp. points. ( a schmit trigger comprises solid state devices and is available complete in many forms as an I.C. ) However there are other possible voltage sensitive arrangements.
I think you are on the right track with a volt meter across the existing circuit. Provided that the meter has a high internal resistance there should be no practical effect on the existing set up. A before and a after voltage check could be made using a sensitive digital multimeter to prove this. A 0-50 micro amp meter ( typical internal resistance 3500 ohms ) could be used with a suitable multiplier resistor in series giving full scale at the desired voltage. Panel meters of this type are available with a blank scale which can be scaled using stick on lettering. (Letraset ?) But will the voltage curve provide a suitable scale even if non linear ? THat is the question and a spare transducer would prove this but you are on the right track studying typical thermistor specs. provided one is sure that a thermistor is used. Personally I am confident this is the case. The problem is that one can hardly build a fire under the car, pop a thermometer down the tranny filler, connect a VM and observe ! Don't give up as there is an answer and your thinking is on target.
__________________
Trevor, New Zealand. As a child, on cold mornings I gladly stood in cowpats to warm my bare feet, but I detest bull$hit! Last edited by Trevor; 08-28-2002 at 02:49 AM. |
#9
|
||||
|
||||
This is a subject of interest to me also. If you have a trans schmatic check out the flow form the trans. It is after the TC. The line out of the trans has all the heat from the converter. The sump is a better place to measure.
Lately I have been thinking about adding a thermocouple to the pan. Just put a hole in it and add a thermocouple with a seal in the hole. Actually a multi channel digital readout with afew more temperatures elsewhere. |
Thread Tools | |
Display Modes | |
|
|