A/C bc27-28 readings

[89RedDarkGrey]

This is why I asked why can't our systems just used a clipped on thermocouple to the evaporator like many others use. Easy and cheap, easy access to evaporator, and the original plug comes from the BCM any way, right? Would that work?

[2seater]

45 minutes ago, 89RedDarkGrey said:

This is why I asked why can't our systems just used a clipped on thermocouple to the evaporator like many others use. Easy and cheap, easy access to evaporator, and the original plug comes from the BCM any way, right? Would that work?

I'm guessing the calibration might need to be changed a bit and I would guess the response time wouldn't be as fast not being in contact with the actual refrigerant, but may have possibilities as a fall back idea. There are two ideas, including the one above, that sound like a good garage project. As the Brits say, "get thee to the shed".

[2seater]

1 hour ago, 89RedDarkGrey said:

 

If 2seater can make a Flow Bench, TEVES accumulator tester- why not a "Thermistor tester"? Tubing, fittings, an evaporator, etc.- maybe a gutted junkyard minifridge? A NEW one is $60 at Wal*Mart, used even cheaper. All the needed components- just smaller.

 

It'd be cool to tinker with

I wondered about checking the response of the sensor also but the impression I got from drtidmore was they are pretty dependable and the connection may be more problematic. It looks like we have reasonable confidence we can predict and test the sensors at normal ambient temperatures and apply an external correction of needed. A hi and low temperature chamber could be pretty small, and may have other uses, but I don't think there is much low hanging fruit regarding the sensor we are concerned with at present. JMHO

[89RedDarkGrey]

Did you check this place?

http://www.durexindustries.com

 

They advertise "made to order" and "automotive".

[89RedDarkGrey]

(more precisely)

 

http://durexindustries.com/immersion-resistance-temperature-detectors

[89RedDarkGrey]

I just filled out their contact form with info. If I hear anything- I'll post it ASAP.

[drtidmore]

11 hours ago, 89RedDarkGrey said:

I just filled out their contact form with info. If I hear anything- I'll post it ASAP.

Bob,

Trying to get around the low temp sensor on the Reatta's is about like saying you want to get around the fuel injection.  Sure, it is possible, but the system is what it is.  The thermistor itself is NOT a big issue, even thought its impact can certainly be considered an issue!  GM stopped selling replacements and the 3rd party market has yet to fill the vacuum.  Thermistors are extremely reliable, especially in lower temp environments like the A/C system (both low & high).  GM used a disc thermistor with two small solder joints and as I have stated, solder becomes brittle after countless thermal cycles and as it becomes brittle, it becomes resistant to the flow of current until it essentially becomes an open.  The problem with an evaporator clip-on thermistor is one of hysteresis (where the state of the system is dependent on its history).  The thermal mass of the evaporator is such that it severely lags in temperature response compared to the actual refrigerant.  Yes, GM could have engineered a system based on a something as low tech as the old capillary thermostat that was used in aftermarket installed A/C systems years ago but GM was working to maintain as close to suction throttle behavior as they could as that was the reference standard for GM and had been for literally 3 decades AND the system on the Reatta WAS the one designed by and for Cadillac. 

 

What 2Seater and I have done is establish the curve for the thermistor by exploiting the fact that the software in the BCM was written around the ideal curve for the part GM incorporated, so it is a fixed reference.  Now that we know the curve, we can PREDICT, with ACCURACY, what the thermistor should read at ANY temperature.  That allows us to either use the other temp sensors in the car as a group normal to compare against on a car that has sat overnight, OR we can measure the temp of the suction line adjacent to the thermistor with an infrared thermometer to determine the ACTUAL sensor temp.  Then we can measure the resistance of the thermistor and see how much it has drifted which allows us to correct it with a parallel resistor, assuming that the drift IS in the thermistor NOT the solder joints.  Since we have limited experience with the longer term success of trimming to correct drift, trimming is likely to turn out to be a short term fix at best.  However, trimming COULD be useful to correct a fully functional thermistor that is just a tad off optimal, but NOT drifting per say.  I am not advocating that this is as desirable as being able to run to your local FLAPS or even a GM dealership and buying a new low temp sensor, but it is certainly NOT difficult. 

 

With Ronnie discovering that his failed low temp sensor was actually no longer mechanically secure between the support leads, that was the smoking gun for me.  I had a hunch regarding the solder joints, but no evidence outside of years of dealing with damned solder joints in electronics.  Ronnie has mailed me his failed low temp sensor and we shall see what I find.  

 

If my hunch is correct and I can reflow the solder successfully, then THE solution to this issue is to fix the sensor, NOT try and go around the way the system was designed.   

 

ANY external thermistor that you might find will NOT work correctly in the Reatta system as again, the system was designed to directly monitor the refrigerant temp immediately after the orifice with ZERO hysteresis.  The first problem you are going to find is thermistor with a value of 2600 ohms @ 25C simply does not seem to be available (I have done a pretty extensive search and found nothing).  The second problem is that mounting it externally is going to swing the evaporator between freezing up and too warm due to hysteresis of the evaporator compared to the refrigerant itself.  There is NO WAY to simply change the way the BCM code is written. The code in the BCM is ALL binary, the source code and the development environment is LONG gone and attempting to find memory locations, let alone interpret what is at those locations while not totally outside of technical feasibility is a matter that is obtuse and VERY prone to error.  After 30 years, we still don't have a complete understanding of the even the ECM and nothing on the BCM.  But remember, we are NOT talking about a failure of the BCM, but of a simple sensor. 

Edited June 22, 2017 by drtidmore (see edit history)

[drtidmore]

12 hours ago, 89RedDarkGrey said:

 

If 2seater can make a Flow Bench, TEVES accumulator tester- why not a "Thermistor tester"? Tubing, fittings, an evaporator, etc.- maybe a gutted junkyard minifridge? A NEW one is $60 at Wal*Mart, used even cheaper. All the needed components- just smaller.

 

It'd be cool to tinker with

We have such built into the Reatta.  Just let the car sit overnight, then go into the ECM and BCM and get the temp readings from all the various temp sensor and compare them.  There is NO value in testing the thermistors in our car over a range of temps as they either work or they don't.  It is their resistance vs temp that is of value and again, comparing the sensors after sitting overnight gives us that information as we only need ONE measurement point as long as we have other sensors of the same type exposed to similar temps that are reporting.  

Edited June 22, 2017 by drtidmore (see edit history)

[drtidmore]

I have thinking about HOW, if possible, to bypass the OEM design regarding the compressor clutch cycling.  

 

I pulled out the FSM and studied the schematics closely.  First a few things to note.  The ECM, NOT the BCM, is actually the one that ultimately controls the compressor clutch relay. The BCM computes the desired state of the compressor clutch and sends a signal via the data network, based on the low pressure and low/high temp sensors, that the ECM reads and then using logic in the ECM controls the compressor clutch relay state (i.e. engaged or disengaged) . The power steering pressure switch, the throttle position, and the engine coolant temp also feed into the ECM as part of the compressor clutch control logic so even if the BCM sends the engage signal along the data network, the ECM MAY decide to NOT engage the compressor clutch based on its other inputs (i.e. wide open throttle, engine too hot, power steering load).   The BCM also feeds the A/C programmer with a PWM signal (direct wire) that the programmer in turns converts into a signal to the blower control module (blower speed).  Also, the BCM computes the refrigerant high pressure value (there is NO high pressure switch in the system) using the low and high temp sensor inputs and uses that computed value as a safety to kill the compressor if the computed head pressure exceeds safe limits.  Of course the compressor itself has a mechanical high pressure relief valve as the avenue of last resort against exploding something in the A/C system.   Additionally, the ECM, which is responsible for the fuel injection and IAC on the throttle body, uses the compressor clutch state in its computations as it controls the fuel and idle air controls. 

 

I think you should now be getting a better idea of how the sensors, modules and logic are all intertwined and co-dependent on each other.  This is also why the BCM will signal for the compressor to be disengaged if the low or high temp sensor go way out of value as they are central to the control logic. 

 

As the data network is how the ECM and BCM communicate, it is NOT possible to spoof the ECM so this is where unintended consequences begin to unravel things and things get worse if we try to bypass the ECM as well. 

 

The ECM is computing the correct fuel injection signals based on the engine LOAD which is of course significantly impacted by the compressor, so if we bypass the ECM control of the compressor clutch, we will run into idle issues as the ECM currently adjusts the fuel injection and IAC to compensate appropriately.  In the old days, there were kick up pots controlled either electrically or with vacuum that mechanically forced the throttle open slightly more when the compressor was active.  In our Reattas such is done with the IAC and fuel injection which are under the ECM control and the ECM needs to KNOW about external loads. 

 

The more I investigate this, the more obvious it becomes that even attempting to bypass the OEM control of the compressor is a bad idea!  Even IF we did bypass the ECM/BCM control of the compressor clutch, the problem would remain that the ECM would be getting data from the BCM indicating desired compressor state control and the ECM would in turn adjust the fuel injection and IAC to compensate EVEN THOUGH, the actual compressor state would likely NOT be in agreement.  So the idle would be all mess up and might even stall the engine under certain conditions as the compressor clutch could be engaged when the ECM thought otherwise and vise versa, not to mention engaging when other conditions would contraindicate. 

 

So I really don't see a viable way to control the compressor outside of the OEM design which brings us back to the living with the OEM design.  We WILL solve the low temp sensor issue.  I have already discussed why an external clip-on evaporator thermistor is a bad idea (throws off all the compressor control logic in the BCM due to hysteresis).  I am not trying to shut down discussion either as such is invaluable, but hard facts are hard facts and we need to understand HOW the GM design functions before we can begin to discuss work arounds.  I hope that this helps explain in a bit more detail HOW the A/C system works.

Edited June 22, 2017 by drtidmore (see edit history)

[89RedDarkGrey]

4 hours ago, drtidmore said:

Trying to get around the low temp sensor on the Reatta's

 

I'm not trying to "get around" a component.

 

4 hours ago, drtidmore said:

GM stopped selling replacements and the 3rd party market has yet to fill the vacuum.

 

I'm attempting to see if parts can be made, that's all.

4 hours ago, drtidmore said:

ANY external thermistor

 

I simply asked if and why it didn't have a thermocouple- like I've seen in other systems. I'm not trying to "work around" or reengineer anything. That- would be way beyond my pay grade.

 

Calm down.

[DAVES89]

89 rdg you are trying too hard. Go easy kemosabee...

[Ronnie]

David,

Good explanation of how the A/C system on the Reatta functions. When searching for the reason my A/C was performing poorly (as a result of the bad low temp sensor) a few years ago it would have been nice to have your explanation in front of me instead of having to figure it on my own by searching the FSM. The A/C, like a lot of the other systems in our Reattas, is complicated. It's always better to have a good overview of how a system works instead of having to start from scratch in the FSM.

[drtidmore]

20 minutes ago, 89RedDarkGrey said:

I simply asked if and why it didn't have a thermocouple- like I've seen in other systems.

 

FYI, what you thought were thermocouples are typically thermistors.  Thermocouples are used in HIGH heat (thing furnaces) as they generate a small electrical voltage proportional to the temperature. That is how the old household heating thermostats used to work as the thermocouple held open a safety valve that kept the pilot lit AND provided the power to open the main valve when the wall thermostat closed the circuit.   A thermocouple is not a good fit for vehicle A/C controls.  What used to be common in car A/C was a capillary tube thermostat where the expansion/contraction of a working fluid in the bulb at the end of the capillary tube created pressure to move a mechanical piston in the thermostat to make or break a set of contacts. These are still used in some some simple applications but even those applications are moving to thermistors and circuitry as the cost of the capillary tube thermostat has become the more expensive option. 

 

[drtidmore]

Well, I thought I would drop an update on this matter.  Ronnie shipped me his failed low temp sensor and I was able to determine that the failure mode WAS solder related, but unfortunately the stress from the solder failure actually pulled the soldering surface on the disc thermistor away and therefore the thermistor was destroyed.  Bummer!  FYI, by setting my soldering pencil at the minimum temp that would melt tin/lead but not silver solder I determine that good old fashioned, low melting point,  tin/lead solder was used to attach the OEM thermistor between the sensor leads, so using such in a repair will not degrade the sensor longevity as we know that they lasted well over 25 years as designed and built.   Silver solder has a much higher melting point and there is really no need for such plus subjecting thermistors to heat beyond their operating range does present the risk of a shift in their value. 

 

With the info that 2Seater gleaned from his trim pot vs BCM reported values, we knew the curve of the OEM thermistor.  The bad news is that I have been TOTALLY unable to find an exact replacement.  I then backed off a bit and thought about what we needed and decided to ditch fitting the curve and focus on the value at 0C which is where we WANT the BCM to hold the evaporator.  I then went on a quest for any and all thermistors that would be close, regardless of packaging.  Thanks to Mouser.com, I was able to narrow the search considerably.   I finally found a SMD (surface mount device), NTC thermistor that came within 0.5C of the mark (ie at 0C actual, the BCM will report between 0C and 1C).  The curve on this thermistor is a bit different so at 25C ambient, the BCM will report closer to 30C but that is immaterial as again, we are really only interested in a value at 0C.  While this thermistor was originally designed to set on a PC board with flow soldered connections at each end, mounting it between the low temp sensor leads will be straightforward as its length is ideal to fit between those leads.  Fortunately this was a part where units of 1 could be ordered (I found a closer fit, but the minimum order was 15,000!).  I have ordered 10 of these little gems to test and I will report back once I have something to report. 

[89RedDarkGrey]

This site looks promising- and one of their Stockers is that Mouser company. Does our OEM appear to be a "glass encapsulated" type? You could easily use tweezer-type heat sink when soldering. Are they bidirectional or polarized?

 

http://www.ussensor.com/products/ntc-thermistors

 

And maybe fill out the techno details on this, and see what they reply? I'd do it- but I lack the info. If you could read through it, PM me the info- I'll submit it, just to see.

 

http://www.ussensor.com/contact-us/sensor-application-form

Edited July 4, 2017 by 89RedDarkGrey (see edit history)

[drtidmore]

Bob,

The OEM sensor is a bare thermistor disc soldered to two substantial leads sticking down into the refrigerant stream (the leads are substantial to provide mechanical robustness against the raging stream of refrigerant.  Glass encapsulated thermistors would NOT be my first choice for a number of reason but mainly due to the lack of a good way to mechanically secure the thermistor (ie the thin wires going into the glass enclosure would be the weak link).  There are other reasons why but that one alone makes it a non starter from my perspective.  FYI, Mouser is a wholesale distributor to the electronics industry and as such represents a massive range of vendors.  The vendor you located does have an SMD thermistor that would work as well, so I will keep them on file should I need another source.   FYI, I really don't need to submit anything to them as their data sheets gives me everything I need to know.

Edited July 4, 2017 by drtidmore (see edit history)

[89RedDarkGrey]

1 hour ago, drtidmore said:

The OEM sensor is a bare thermistor disc soldered to two substantial leads sticking down into the refrigerant stream (the leads are substantial to provide mechanical robustness against the raging stream of refrigerant.  Glass encapsulated thermistors would NOT be my first choice for a number of reason but mainly due to the lack of a good way to mechanically secure the thermistor (ie the thin wires going into the glass enclosure would be the weak link).  

 

Oh? I was going from foggy memory of an OEM sensor Ronnie posted, and I thought it looked the same.

 

 

1 hour ago, drtidmore said:

The vendor you located does have an SMD thermistor that would work as well, so I will keep them on file should I need another source.

 

Thanks- I'm glad my ramblings are sometimes useful

Edited July 4, 2017 by 89RedDarkGrey (see edit history)

[drtidmore]

An update.  Received the prospective thermistor and it soldered in nicely.  It is MUCH smaller than the OEM disk thermistor and actually just marginally thinner than the support leads. Its width is virtually identical to the OEM disk so again a really nice fit.  I determined earlier that the OEM thermistor was soldered in using good old fashioned tin/lead electronics quality solder (melting point between 230-250C).  The thermistor under test was specifically designed to accept the same exact solder.  There is a significant amount of solder connection on each end compared to the overall area of the thermistor, so it is mechanically secure between the leads.  As the orientation of the thermistor inside the sensor can be set independent of the compression nut, I feel that it would be good to orient the thermistor in line vs across the flow of refrigerant to lessen the strain on the thermistor chip itself (ie the high speed flow of refrigerant in the line across the thermistor area...remember that the thermistor sits IN the stream of refrigerant)

 

 

As I had posted earlier this thermistor is NOT an exact temp vs resistance fit to the OEM thermistor curve encoded in the BCM (thanks again to 2Seater for collecting the data) but what IS important is that it is VERY close at 0C which is the where we want the evaporator to run and therefore important that the BCM have a fairly accurate reading at that temp.

 

Resistance of the thermistor installed in the sensor housing and immersed in a fairly large, stable ice water bath (distilled water and ice mixed) comes up with a resistance that the BCM would register as 1C.  That is in our favor as that means that the BCM would allow the evaporator to pull a tad lower than -2 before cycling the compressor.  At 17C (indoor ambient when I tested) the thermistor value would correspond to about 19-20C so again in line with what I saw at 0C.

 

I also ran the numbers on the power dissipation when installed in the BCM circuit and we are looking at about 3 orders of magnitude between actual and max so self heating will NOT be an issue. 

 

Then I plugged it into the BCM harness and with my calibrated K-type thermocouple showing an ambient temp of 23.5C the BCM BD028 registered 27C which again is due to the curve of this thermistor not being an exact fit to the OEM curve in the BCM.  Then, once again, I used an ice water bath and the BCM registered between 2-3 degrees.  The fact that it did not indicate 1C as expected is likely due to my having to use small container of the ice water mixture due to space constraints and it likely was not stabilized as well as when I did the same test indoors.

 

This rudimentary testing in the BCM circuit at least confirms that the thermistor is running in the same ballpark as expected.

 

The next test will be to actually install and run it.  My A/C system has been running fine for almost 4 years on R-134a with no leaks and at optimal charge as set by the BCM method, but I am considering purging the R134a and switching to R-152a as its heat properties are much close to R-12.  I am currently involved in my transmission project, but I may make the switch to R-152a once that is complete and install the restored sensor for further testing. 

 

In the meantime, if anyone runs across either low or high A/C temp sensors on Riv/Reatta/Toro/Eldo/Caddy-in-general of our vintage, grab them regardless of condition as it is the housing that is going to be the commodity that can't be replaced as the salvage supply dwindles (mainly do to being crushed along with the cars).  If pulling from salvage I recommend that you use a tubing cutter and remove a short piece of the refrigerant line with the sensor remaining installed in the cut away section so as to protect the internal sensor leads (got that idea from Ronnie)

 

I am now confident that I have a workable restoration of the low temp sensor.  FYI, I have 9 more of the thermistors on hand (more can be ordered) so I can do more if anyone is interested.  Remember that this sensor stops our A/C unit dead in its tracks as it fails and the workaround for a failing sensor is at best temporary and there is NO workaround for a failed sensor!

 

 

[Ronnie]

I think you have found a good repair for the temp sensors. Thanks for your hard work on it. I would be happy to try that sensor if I was having problems. I see no reason why it shouldn't work. If the reading in isn't spot on when the sensor is installed it could be easily tweaked with a resistor if needed to get optimal cooling. Perhaps just rotating it would make a difference in making it read just right?

[2seater]

I think the solution found is awesome!

 

I pulled the two sensors from my somewhat derelict '89, primarily to check their response to find a good one, but that is likely moot now. The one thing I noticed is the hex size is 27mm, or about 1 1/16", from the '89 which look like the one in these posts. As noted before, the electrical connector on my '90 is completely different (aside from the small difference in location), and the hex is also smaller, 24mm or 15/16", so I am not certain they are interchangeable? Maybe it is like the three different types of crank bolts that all fit the same space?

 

I do have a small question. I know the response is relatively close but for a non-optimized system, that swings between -2*C and 10*C, is there an estimate on how far off the mark they will be?

[drtidmore]

3 minutes ago, 2seater said:

I do have a small question. I know the response is relatively close but for a non-optimized system, that swings between -2*C and 10*C, is there an estimate on how far off the mark they will be?

Based on what I know at this point, this thermistor will result in the evap temp actually swinging for more like -3 to around 8 which would be slightly colder on the average.  If you use the BD028 method of optimizing the refrigerant charge you could either shoot for 2C (probably closer to .5-1C) or just a tad lower.  The good news is that the BCM will report a temp about 1C higher than the actual temp in at the sensor using this thermistor, SO we potentially can gain an extra 1-2C lower evap temp before the BCM cuts out the compressor.  

[Ronnie]

7 hours ago, drtidmore said:

(got that idea from Ronnie)

Have to give Dave the credit for that idea. That sensor was shipped to you the same way Dave shipped a good replacement sensor to me.

[drtidmore]

3 hours ago, Ronnie said:

I think you have found a good repair for the temp sensors. Thanks for your hard work on it. I would be happy to try that sensor if I was having problems. I see no reason why it shouldn't work. If the reading in isn't spot on when the sensor is installed it could be easily tweaked with a resistor if needed to get optimal cooling. Perhaps just rotating it would make a difference in making it read just right?

Rotating the sensor in the refrigerant stream will not make any real difference in the registered temp but WILL reduce the mechanical load that the refrigerant stream places on the thermistor element itself.  The small amount of difference that this thermistor will make could almost considered immaterial especially since it gives us an extra 1-2 degrees of cooling before the BCM will cycle the compressor off.  I see no reason to trim the value as we are talking such a small offset and again it is in our favor when trying to optimize the cooling out of the evaporator

 

[89RedDarkGrey]

I wonder if some type of refrigerant-tolerant screen could be made to house the probe, just to protect it from the current pressure of the flow? It just looks so flimsy. I wonder what the actual current flow on it is?

[DAVES89]

I could go to Gibson's anytime and pick up at least 10 each of them. As it stands there are 3 Reattas, 7-8 Rivs, 2-3 Troferos and that doesn't count the Cadillacs, Olds, and Pontiacs of that vintage. Then there is that yard near West Bend that had the 1990 Red/Tan that has 3 Reattas, 4 Rivs, and numerous Cadillacs, Olds, and Pontiacs.

 Let me know how many you want and I will grab some. Remember I don't get them for free, but I would forgo any profit in selling them to you David as you have the repair parts and would be the one to solder them in and could sell them to others. I would expect a repaired one gratis for my efforts. Let me know what you want to do.

[2seater]

Dave, if you pull a bunch of them, it will be of interest to me to see if you find two different hex sizes and if they will interchange.

 

As I mentioned before, I pulled the sensors from the '89, and testing them indicated they are pretty close to the curve, and each other, but I cannot plug them into the car to see what the BCM thinks :((  Just for reference, the high side sensor is 2.891k @ 24.3*C and the low temp is 2.93k @ 24.2*C.  Temp. gun laser on the sensor disc from 6" away but I don't know how tightly aimed the actual sensing beam is?

[drtidmore]

8 hours ago, 89RedDarkGrey said:

I wonder if some type of refrigerant-tolerant screen could be made to house the probe, just to protect it from the current pressure of the flow? It just looks so flimsy. I wonder what the actual current flow on it is?

The sensor leads are actually quite robust.   I don't see ANY need for a screen and besides there simply is NOT sufficient room for such without impeding the flow of the expanding refrigerant.  As I mention, I am going to recommend that the sensor be oriented such that the sensor itself is inline rather than broadside to the refrigerant flow (the sensor alignment runs parallel with the external connector so it is easy to orient).  This way the sensor leads will take the brunt of the refrigerant stream.  I really don't think there would be any problems even oriented broadside to the stream but it is so easy to orient it to be inline that there is no reason to not do so. 

[drtidmore]

12 hours ago, Ronnie said:

Have to give Dave the credit for that idea. That sensor was shipped to you the same way Dave shipped a good replacement sensor to me.

This is what Ronnie and I have been discussing as to ship the sensor without damage.  Great idea Dave!

 

 

Also, here is a couple of shots showing the orientation that I have been describing

 

[Ronnie]

Good photos. Anyone looking at the 3rd photo should take note that the hex nut that tightens the sensor to the fitting can be turned without moving the sensor body making alignment of the sensor easy. You already stated this but the photo makes what you were saying easier to understand for someone that hasn't had one of the sensors in hand.

[89RedDarkGrey]

1 hour ago, drtidmore said:

 As I mention, I am going to recommend that the sensor be oriented such that the sensor itself is inline rather than broadside to the refrigerant flow (the sensor alignment runs parallel with the external connector so it is easy to orient).  This way the sensor leads will take the brunt of the refrigerant stream.

 

Whose to say that when installed on xyz car- the threads will hopefully allow for alignment? The flats where disc is parallel should be clearly marked (yellow paint) and instructions for proper installation orientation.

 

 

 

[Ronnie]

12 minutes ago, 89RedDarkGrey said:

Whose to say that when installed on xyz car- the threads will hopefully allow for alignment? The flats where disc is parallel should be clearly marked (yellow paint) and instructions for proper installation orientation.

 

 

That's a real nice drawing Bob. I think you might could have saved the time you spent on it if you had read my previous post closely. OR, maybe my post didn't make sense?

 

 

 

Edited July 10, 2017 by Ronnie (see edit history)

[89RedDarkGrey]

Ronnie- for some reason- your and drtidmore with photos and your post (#108, #109) didn't show up in my stream? Sorry about that

 

3 hours ago, Ronnie said:

Anyone looking at the 3rd photo should take note that the hex nut that tightens the sensor to the fitting can be turned without moving the sensor body making alignment of the sensor easy

 

Oh? I wasn't aware the hex part was a collar nut, and the main body of the sensor was independent. It will have to be held in position-alignment while tightening. Another good reason to have all O-rings and threads smeared with silicone paste when assembling AC components.

 

drtidmore- could you setup an "exchange program" like Jim Finn has for headlight switches- or is direct repair/return better? Have you figured a fee yet?

[89RedDarkGrey]

[DAVES89]

On 7/10/2017 at 6:35 AM, 2seater said:

Dave, if you pull a bunch of them, it will be of interest to me to see if you find two different hex sizes and if they will interchange.

 

As I mentioned before, I pulled the sensors from the '89, and testing them indicated they are pretty close to the curve, and each other, but I cannot plug them into the car to see what the BCM thinks :((  Just for reference, the high side sensor is 2.891k @ 24.3*C and the low temp is 2.93k @ 24.2*C.  Temp. gun laser on the sensor disc from 6" away but I don't know how tightly aimed the actual sensing beam is?

 They are two diferent sizes. I pulled 8 each from 1988/89 and 2 each from a 1990. I will be boxing them up and sending them off to David to test/fix/swap out as he finds neccesary.

[Ronnie]

46 minutes ago, DAVES89 said:

They are two diferent sizes. I pulled 8 each from 1988/89

Are there two different sizes used in the 1988/89 models or does the size change occur only in later years?

[DAVES89]

I did not look at a 1991 Riv [there are no 1991 Reattas in the yard] The 1990 size is smaller.