Temp sender not working?

[Righty]

The electrical conductivity of the coolant is not involved with the type of sending unit with which you are dealing. The thermal conductivity of the coolant, and its surface tension, can affect response time of the sending unit to a negligible degree. The coolant itself is not part of the circuit path, which flows as shown in the following:

Battery Positive (+) --> Voltage Limiter --> Gauge --> Sending Unit --> Engine Manifold/block --> Battery Negative (-)

The temperature sending unit in question is a negative temperature coefficient resistor. The impedance of the unit is essentially and wholly a direct current function of resistance only.

Replacement components can be bad, and modern units have a much worse record than older aftermarket and original units. For testing the sending unit in operation, the following cardinal points and resistance values are of interest:

• Ambient temperature at about 25 degrees Celsius, before running engine, gauge needle at lowest point, sending unit resistance = 360 Ω
• Engine at ~49 degrees Celsius - Needle moves to Cold mark on the temperature gauge, sending unit resistance = 72 Ω
• Engine at ~88 degrees Celsius - Needle moves to middle on the temperature gauge, sending unit resistance = 22 Ω (approximate range of ~18 to ~26 Ω)
• Engine at ~110 degrees Celsius - Needle moves to Hot mark on the temperature gauge, sending unit resistance = 9 Ω

Note that some gauges have intermediate marks above Cold and before Hot and others have actual numeric markings from 120 to 170 to 230 to 250 degrees Fahrenheit.

Also note that Chrysler revised the actual numeric equivalence to the hot and cold marks a couple of times, but the resistance values of the sending units and their correspondence to the positions on the gauges themselves are essentially the same with a slight variance for peak temperature. This correspondence is true for the temperature sending units as well as those for fuel level and oil pressure which operate with most of the King-Seeley Constant Voltage type gauges Chrysler used from about 1960 through 1989 on domestic rear wheel drive vehicles.

When checking the resistance of the temperature sending unit, probe the unit terminal post and body to check the unit itself. Probe the terminal post to manifold to check body to ground, and further check terminal post to the battery negative (-) terminal for a more end to end post to ground test. These tests should show the same effective resistance. Direct continuity between the body and the points can also be tested directly by probing between the unit body and the ground points. The resistance should be effectively zero.

You tested the wiring to gauge set up by shorting the terminal lead to ground and watching the gauge response. This confirms the circuit is operational, and although not a fully thorough check, usually sufficient. Other functional checks can be made by using a voltmeter between the sending unit post (connected) and ground or a test light between the sending unit post and wiring terminal. With the key on, watch for cycling voltage on the voltmeter (an analog meter is better in this case), or for the flashing of the test light as the voltage limiter cycles. As mentioned, a quick check of the sending unit as it warms can be performed with the test light leads between the unit post and +12 volts at the battery or other point and watch for the intensity to increase as the resistance decreases as the engine warms.

A quick test lamp can be made from an 1156 bulb, a couple of wires, and a couple of clips. For testing between the sending unit and sender wire, a 6 volt bulb, such as 1129, better matches for the voltage limiter output. Test lights, are quick and simple tools, but are more qualitative than quantitative. You have a meter which can do the quantitative checks as well.

Although it sounds like your gauge and associated wiring is okay, if desired to check cardinal points of the gauges, use the following resistance values:

Chrysler Specification:

• Cold = 72 Ω
• Middle = 22 Ω
• Hot = 9 Ω

Gauge position tolerance at these points is listed in the service manual.

Miller Special Tool Specification (used by Chrysler for testing):

• Cold = 75 Ω +/-5% -> 71.2 - 75.8 Ω
• Middle = 22 Ω +/-5% -> 20.9 - 23.1 Ω
• Hot = 10 Ω +/-5% -> 9.5 - 10.5 Ω

The Miller C-3826 tester can be used to test these points, or the point resistances can be made using fixed resistors or a combination thereof, variable resistors, decade resistance boxes.

Thank you very much further to some of the info before that is really comprehensive. When I "tested" the wire from the gauge I used a wire to the engine block, maybe I should have checked it to the manifold to see if it triggered the gauge and if it did was it as strong a response, something that is very clear on FABO discussions about any electrics is how important good continuity to ground is.

 

[Righty]

Just making sure I understand

You shorted the sender lead to the block and the temp gauge moved to HOT fast. You removed the grounded sender lead from ground and the gauge slowly went back to cold. Is that correct.

If you kept the sender lead grounded for any amount of time you MAY have burnt out your gauge.

What is the sender part number you bought?

If you ground the sender lead again does the gauge move like before?

Yes quick swing up the gauge and then slowly dropped back down to rest, as described by my beautiful assistant (the wife).
The contact was very brief.
Standard Motor Parts TS17 which is as far as I can see the only sender that is claimed to be compatible on various parts checkers.
I did a second test and yes it moved as before, so hopefully a sign I didn't burn out the gauge.

 

[Righty]

I might have missed it, but did you chk connection at bulkhead?

No I didn't I will do, RHD car with a fair bit of user wiring so I'll have to fumble my way along the wire to find the connection.

 

[Righty]

I thought I would check the continuity to ground after reading Syleng's post. Manifold has ground strap to firewall. Manifold to firewall = 0.8Ω, manifold to cylinder head bolts = 0.8Ω, manifold to radiator support = 0.8Ω. (Multimeter has settings for up to 200Ω / 2kΩ/20kΩ etc so I'm using 200Ω setting.)
So a quick test before work with a cold engine.

Hex of sender body to manifold = 0.8Ω
Hex of sender body to firewall ground = 0.8Ω
So I'm guessing continuity to ground is good, the 0.8Ω seems a constant so that's probably my multimeter or probes.

Stud of sender to manifold lead disconnected = 17Ω
Stud of sender to manifold lead connected = 17Ω ignition off
Stud of sender to manifold lead connected = 37Ω ignition on
Stud of sender to manifold with engine running just after the thermostat opened (upper hose warm) I just couldn't get a steady reading jumping around from 30Ω to 140Ω to no reading.
Would the stud readings indicate a bad sender, either off the shelf or damaged when I installed it?
I suppose all i can do is buy another sender and try that and be as careful as possible installing it.

 

[RedFish]

BINGO and you didn't need to pull the inst' panel. Good job

 

[Vaanth]

I thought I would check the continuity to ground after reading Syleng's post. Manifold has ground strap to firewall. Manifold to firewall = 0.8Ω, manifold to cylinder head bolts = 0.8Ω, manifold to radiator support = 0.8Ω. (Multimeter has settings for up to 200Ω / 2kΩ/20kΩ etc so I'm using 200Ω setting.)
So a quick test before work with a cold engine.

Hex of sender body to manifold = 0.8Ω
Hex of sender body to firewall ground = 0.8Ω
So I'm guessing continuity to ground is good, the 0.8Ω seems a constant so that's probably my multimeter or probes.

Stud of sender to manifold lead disconnected = 17Ω
Stud of sender to manifold lead connected = 17Ω ignition off
Stud of sender to manifold lead connected = 37Ω ignition on
Stud of sender to manifold with engine running just after the thermostat opened (upper hose warm) I just couldn't get a steady reading jumping around from 30Ω to 140Ω to no reading.
Would the stud readings indicate a bad sender, either off the shelf or damaged when I installed it?
I suppose all i can do is buy another sender and try that and be as careful as possible installing it.

It sounds like the sending unit is indeed bad, although at 17Ω cold, the gauge should read about 3/4 scale, unless it's a bogus reading due to the unit. But with it varying, and all else okay, it is bad.

The "key on" reading of 37Ω is not valid if the battery is connected and the circuit powered. DC resistance measurements cannot be made with the circuit powered. The additional current flow along with the meter's two-wire constant current source for resistance measurement causes an erroneous voltage drop and subsequent wrong resistance reading.

The 0.8Ω resistance reading is good for a continuity check, and part of the reading is your test lead resistance. The test lead resistance can come from the lead wires themselves or the end connects at the probe or meter end, or a dirty probe tip. You can account for this lead resistance.

Regarding test lead resistance when measuring low resistance values:
Minor test lead resistance is not much of a concern at higher resistances, but when measuring low resistances, such as the ground to ground continuity, account for the resistance of the test leads connected to the measuring device. This can be done by shorting the leads together, noting the resistance, and subtracting it from the measurement readings, or use the meter's relative setting (if equipped) similarly with the lead shorting.

Some meters have a function to measure conductance and convert accordingly for low resistance values. Others, typically lab or some bench type meters can use a four wire ohms measurement to overcome any test lead effects.

---

Although largely beyond the realm of basic automotive shop work, except for the meter zeroing usefulness, some extra details provided here as FYI with some test equipment eye candy...

The following images are some examples of the aforementioned meter techniques where I show a relative zero of a meter and I am measuring a 1Ω decade box setting and 10Ω lab resistor (at ambient conditions) with a couple of four wire measuring meters. additionally showing a direct difference of two wire versus four wire readings and an example of a meter that measures conductance.

Checking test lead resistance. This meter has good leads and connections. The 0.1Ω should be deducted from any readings.

Or, for the same meter and lead, with the leads shorted, use the Relative (REL), or zero, button to deduct the lead resistance from subsequent readings in the meter itself:

Measuring 10Ω on the lower meter, and 1Ω (scaled in kilohms) on the upper meter with four wire leads:

Measuring 1Ω on the lower meter with four wire leads:

The decade box and lab resistor used for the foregoing with the leads used. The test leads have about 0.1Ω and 0.5Ω lead resistance, which is essentially removed from the reading by the four wire measurement.

The next two images are showing the difference in a two wire versus four wire measuring at very low resistance, using a shorting bar, which demonstrates how the resistance of the bar itself is effectively removed from the reading of a direct short.

Two-wire:

Four wire:

A meter which used conductance measurement to measure low resistance:

 

[Righty]

It sounds like the sending unit is indeed bad, although at 17Ω cold, the gauge should read about 3/4 scale, unless it's a bogus reading due to the unit. But with it varying, and all else okay, it is bad.

The "key on" reading of 37Ω is not valid if the battery is connected and the circuit powered. DC resistance measurements cannot be made with the circuit powered. The additional current flow along with the meter's two-wire constant current source for resistance measurement causes an erroneous voltage drop and subsequent wrong resistance reading.

The 0.8Ω resistance reading is good for a continuity check, and part of the reading is your test lead resistance. The test lead resistance can come from the lead wires themselves or the end connects at the probe or meter end, or a dirty probe tip. You can account for this lead resistance.

Regarding test lead resistance when measuring low resistance values:
Minor test lead resistance is not much of a concern at higher resistances, but when measuring low resistances, such as the ground to ground continuity, account for the resistance of the test leads connected to the measuring device. This can be done by shorting the leads together, noting the resistance, and subtracting it from the measurement readings, or use the meter's relative setting (if equipped) similarly with the lead shorting.

Some meters have a function to measure conductance and convert accordingly for low resistance values. Others, typically lab or some bench type meters can use a four wire ohms measurement to overcome any test lead effects.

---

Although largely beyond the realm of basic automotive shop work, some extra details provided here as FYI with some test equipment eye candy...

The following images are some examples of the aforementioned meter techniques where I show a relative zero of a meter and I am measuring a 1Ω decade box setting and 10Ω lab resistor (at ambient conditions) with a couple of four wire measuring meters. additionally showing a direct difference of two wire versus four wire readings and an example of a meter that measures conductance.

Checking test lead resistance. This meter has good leads and connections. The 0.1Ω should be deducted from any readings.
View attachment 1716444745

Or, for the same meter and lead, with the leads shorted, use the Relative (REL), or zero, button to deduct the lead resistance from subsequent readings in the meter itself:
View attachment 1716444747

Measuring 10Ω on the lower meter, and 1Ω (scaled in kilohms) on the upper meter with four wire leads:
View attachment 1716444746

Measuring 1Ω on the lower meter with four wire leads:
View attachment 1716444749

The decade box and lab resistor used for the foregoing with the leads used. The test leads have about 0.1Ω and 0.5Ω lead resistance, which is essentially removed from the reading by the four wire measurement.
View attachment 1716444744

The next two images are showing the difference in a two wire versus four wire measuring at very low resistance, using a shorting bar, which demonstrates how the resistance of the bar itself is effectively removed from the reading of a direct short.

Two-wire:
View attachment 1716444755

Four wire:
View attachment 1716444754

A meter which used conductance measurement to measure low resistance:
View attachment 1716444748

I forgot about using rel to account for the lead resistance, they are cheap leads, tbh its a cheap unit.
Fascinating stuff in your post, I'm afraid my brain operates on the 'creative' side of the hemispheres, I can draw you a pretty picture, but I find myself needing a sit down in a dark room after trying to grasp this type of thing.

 

[George Jets]

^ So damn funny...



The comments on the YouTube video are just as funny...


☆☆☆☆☆

 

[67Dart273]

I thought I would check the continuity to ground after reading Syleng's post. Manifold has ground strap to firewall. Manifold to firewall = 0.8Ω, manifold to cylinder head bolts = 0.8Ω, manifold to radiator support = 0.8Ω. (Multimeter has settings for up to 200Ω / 2kΩ/20kΩ etc so I'm using 200Ω setting.)
So a quick test before work with a cold engine.

Hex of sender body to manifold = 0.8Ω
Hex of sender body to firewall ground = 0.8Ω
So I'm guessing continuity to ground is good, the 0.8Ω seems a constant so that's probably my multimeter or probes.

Stud of sender to manifold lead disconnected = 17Ω
Stud of sender to manifold lead connected = 17Ω ignition off
Stud of sender to manifold lead connected = 37Ω ignition on
Stud of sender to manifold with engine running just after the thermostat opened (upper hose warm) I just couldn't get a steady reading jumping around from 30Ω to 140Ω to no reading.
Would the stud readings indicate a bad sender, either off the shelf or damaged when I installed it?
I suppose all i can do is buy another sender and try that and be as careful as possible installing it.

Simple. You can confirm the .8 by FIRMLY shorting the probes while on the same resistance scale. They will likely read .8. YOU CAN NOT check components while "in circuit" that is with the sender lead connected. I have never made that reading with the sender lead connected, so I don't know whether your reading would be reasonable or not, but it is not meaningful.

If you could find the curve for the thermister in the sender, you could actually build a resistance vs temp curve for it and in a round about fashion, use your multimeter for a thermometer. (Many multimeters HAVE a temp scale, but I think most of them are for thermocouples and not thermisters)

 

[67Dart273]

I am disappointed that a couple of you keep on this nonsense with test lights. Test lights vary as to lamp type, or "lamp at all" (LED.) What this means is, that test lamps draw DIFFERING AMOUNTS OF CURRENT and cannot be generally depended on for test results in something like a sender. So when "Jim" claims that under one condition the test light should be dim, and under some other condition should be "bright," and then "John" comes along with a DIFFERENT test lamp, drawing way less current, it may be "quite bright", so is that "dim?" Or what?

This is especially disturbing to me, because I remember when multimeters WERE NOT sold on every corner, and good ones were very expensive, and high impedance solid state meters DID NOT EXIST

Nowadays, ANY HD or Lowes, Horrid Fright, or our local Ace hardware, and in some cases the hardware isle of large supermarkets have some form of meter for sale.

 

[Righty]

Simple. You can confirm the .8 by FIRMLY shorting the probes while on the same resistance scale. They will likely read .8. YOU CAN NOT check components while "in circuit" that is with the sender lead connected. I have never made that reading with the sender lead connected, so I don't know whether your reading would be reasonable or not, but it is not meaningful.

If you could find the curve for the thermister in the sender, you could actually build a resistance vs temp curve for it and in a round about fashion, use your multimeter for a thermometer. (Many multimeters HAVE a temp scale, but I think most of them are for thermocouples and not thermisters)

Ok that’s interesting and I was going to ask Vaanth about that because he said the same thing effectively in his post.
So my measurements that are relevant would be my general continuity checks, the sender body continuity check and my one test of the stud with the lead disconnected and ignition off 17ohms. The other 3 tests were with the lead connected and the last 2 with ignition on. I will run the engine up tomorrow and repeat the cold, warm, hot tests with the lead disconnected.
Obviously it won’t change the fact that the gauge doesn’t move, with either the new or old sender but the readings done properly might reveal something.
This might be a trick of the eye but I was watching the gauge earlier and I almost felt there was a barely perceptible back a and forth of the dial by a couple of hairs breadth, but that could be in my head! Could the voltage regulator be failing so the zero resistance of grounding the lead to the block triggers a response, but with the higher resistance of the sender in the path the voltage is too low to swing the needle?

 

[RedFish]

I wish someone would stick a meter to a good sender in hand or where ever and take picture(s) what a good sender looks like.

 

[George Jets]

T-Series
#TS17T

Went out to the shed this afternoon and did 2 tests on this new aftermarket temp sender.

Working

Both showing a dim bulb. 12 volt power +, temp sending body grounded - to negative battery post, test lights ground wire > to post on the sending unit.

#194 bulb in side marker, both test lights showing same amount of dimness at 90° outside air temp.

As temp rises they will get brighter.


☆☆☆☆☆

 

[RedFish]

View attachment 1716444911

T-Series
#TS17T

Went out to the shed this afternoon and did 2 tests on this new aftermarket temp sender.

Working

Both showing a dim bulb. 12 volt power +, temp sending body grounded - to negative battery post, test lights ground wire > to post on the sending unit.

#194 bulb in side marker, both test lights showing same amount of dimness at 90° outside air temp.

As temp rises they will get brighter.


☆☆☆☆☆

Needed to report steady reading with or without actual ohms but we are getting there. Now Righty, try Rockauto again or take a ohms meter to the local parts store? There are some things better sourced at local NAPA. I guess light bulb fragile could apply to temp senders. Don't know.

 

[Dana67Dart]

Stud of sender to manifold lead disconnected = 17Ω
Stud of sender to manifold lead connected = 17Ω ignition off
Stud of sender to manifold lead connected = 37Ω ignition on
Stud of sender to manifold with engine running just after the thermostat opened (upper hose warm) I just couldn't get a steady reading jumping around from 30Ω to 140Ω to no reading.
Would the stud readings indicate a bad sender, either off the shelf or damaged when I installed it?

you can not test Ohms with voltage on the wire ( and you could burn out your multimeter)

How an Ohm meter works (very basics explanation)

the Ohm meter passes a small voltage from its electronics down one lead to the sender then up the other lead to the electronics. in the electronics there are devices that compare the sent voltage to the received voltage and converts to display ohms.

 

[Dana67Dart]

I wish someone would stick a meter to a good sender in hand or where ever and take picture(s) what a good sender looks like.

you mean like this...

 

[Dana67Dart]

Cold sender should be in the 300-350 ohm range with all leads disconnected.

If you were measuring with the sender wire connected you were measuring the resistance of the Gauge.

 

[67Dart273]

View attachment 1716444911

T-Series
#TS17T

Went out to the shed this afternoon and did 2 tests on this new aftermarket temp sender.

Working

Both showing a dim bulb. 12 volt power +, temp sending body grounded - to negative battery post, test lights ground wire > to post on the sending unit.

#194 bulb in side marker, both test lights showing same amount of dimness at 90° outside air temp.

As temp rises they will get brighter.


☆☆☆☆☆

WHY do you have SO MUCH HATRED for doing things IN AN ACCEPTED WAY and simply USE A MULTIMETER which are now so available. GEE ZUZ. When I was a young man YOU COULD NOT GET multimeters in every dime store, and then they were no where near accurate at lower resistance readings. Please just go bang your head against a wall.

I'll give you a little clue on something else, in this import/ T.I Whan/ Chineseoationized world we live in, and that is that if you buy a bulb that has 194 emblazoned on it, there is some chance that it JUST MIGHT NOT MEED the specifications that a no. 194 bulb was supposed to meet back in say, 1960 or 1970

 

[Vinny]

The temperature gauge in my 69 Barracuda doesn’t work, as the Queen once said recollections vary about whether it ever did. (I’ve had the car about 7 months).
I ordered a replacement from Rockauto, I popped out the old one and grounded the lead to the block to check that the gauge was responding which it did - needle swung up pretty fast and slowly returned to rest. So I went ahead and fitted the new sender, the coolant was up to the top of the hole so I’m sure it’s sat in the coolant, but still no joy at the gauge.
So not sure what’s happening here, the lead to the gauge and the gauge and voltage regulator would appear good based on testing it to ground? I guess I could just be unlucky and have a bad sender out of the box, but the one that came out looked pretty new too so I’m wondering if there is another issue?

If the gauge moves with the wire to ground, the problem is the sending unit. Short and sweet! Don't go down the rabbit hole of ripping your car apart and bench testing the dashboard gauge cluster, It's not necessary. Try another sender. Why don't you just go to local auto store for a sending unit, so you can return it easily if the new one is defective. I would not use Teflon tape either on the threads of the sender. You may be interfering with the proper grounding of the housing.

 

[Righty]

If the gauge moves with the wire to ground, the problem is the sending unit. Short and sweet! Don't go down the rabbit hole of ripping your car apart and bench testing the dashboard gauge cluster, It's not necessary. Try another sender. Why don't you just go to local auto store for a sending unit, so you can return it easily if the new one is defective. I would not use Teflon tape either on the threads of the sender. You may be interfering with the proper grounding of the housing.

Yeah sadly here in the UK decent parts stores are thin on the ground, especially carrying parts like that. I’ll probably have to order another from Rockauto, even with shipping and taxes it will be cheaper than from a supplier here and will get to me quicker.

 

[Woody500]

Yes, I would check all grounds (engine to frame, frame to body and body back to engine) and make sure the sender has good contact on the threads. Sounds silly but the dang liquid teflon they throw on the threads or teflon tape can keep a good circuit from working.

I’ve had this happen to me years ago and it stumped me. “Why was the rest of the engine stuff working?”The previous owner had plastic washers on the intake bolts. I was moving to fast to see the electric choke was inop also.

They don't coat all the threads, on purpose, so a ground can be had. But, if you don't chase dirty threads or don't tighten sender enough, you may never gain that connection.
Or if threads totally coated by mfg(production error or stupidity) of course it won't ground.
Same as Teflon Tape or Permatex tube of thread sealant, or Aviation, or ANY sealant, never cover all the threads.

 

[Righty]

They don't coat all the threads, on purpose, so a ground can be had. But, if you don't chase dirty threads or don't tighten sender enough, you may never gain that connection.
Or if threads totally coated by mfg(production error or stupidity) of course it won't ground.
Same as Teflon Tape or Permatex tube of thread sealant, or Aviation, or ANY sealant, never cover all the threads.

Well there would be no harm me making doubly sure the threads are clean, I removed all the tape off the new sender to eliminate that as an issue, but the thread in the manifold could be contaminated.

 

[Dana67Dart]

Well there would be no harm me making doubly sure the threads are clean, I removed all the tape off the new sender to eliminate that as an issue, but the thread in the manifold could be contaminated

with tape or sealer, as long as you have continuity from the sender body to the battery negative you're fine.

 

[Righty]

If you have continuity from the sender body to the battery negative yoyrvfine

Yes, agreed, it's got to be a bad sender.

 

[Vinny]

Yeah sadly here in the UK decent parts stores are thin on the ground, especially carrying parts like that. I’ll probably have to order another from Rockauto, even with shipping and taxes it will be cheaper than from a supplier here and will get to me quicker.

I Apologize. I didn't notice the UK location. That is definitely an issue. I have a good friend in NZ with the same issue as you do.