Bad sender or bad temp gauge?

[SoCalCoupe]

Nearly stock 41 flathead V8 and 6v electrical. Temperature gauge reads 3/4 hot immediately after a cold start. After a long drive a high quality contact temperature sensor shows only 140 deg when the temp gauge is slammed all the way to Hot. No steam, no percolation noise, no overflow. This is enough information for me to conclude I don't have an overheating problem but rather an indicator problem.

Could buy both a new transducer and a new gauge for about $40 each but I'd rather troubleshoot and replace only the bad part. The 32-41 shop manual says test the temp gauge with a fuel level float. Don't have a spare float.



What's another way to test the temp sender and temp gauge?

[fortyonerag]

There are two sender units on the 41, one in each head. They are linked together.

One provides the over temp indication - full scale deflection on engine overheat. It has two terminals and is on the drivers side (I think).

The one on the passenger side (I think) provides the actual temp reading.

Disconnect the line from the gauge to the drivers side sender unit and earth it. You should get a deflection on the gauge. This will indicate whether the gauge is working.

I'm not immune to memory fade, so if someone says I'm in error, fair enough.

Being a braces and belt type of guy, I installed a pair of SW mechanical gauges. Apart from more accurate temp reading, I have much better oil pressure than I thought!

[Frank Miller]

Here is how I believe it works. Open to corrections.
when the key is turned on the temp gauge goes from H to C. The two terminal sensor is a switch that opens at 212 or so. The one terminal is the analog sensor that varies resistance with temperature. I believe you need close to 0 ohms at rest. If your ohm meter suddenly rises with initial startup then your sensor is at fault. If not then it is the gauge. I'm leaning towards the sensor as the gauges are usually pretty solid.

[SoCalCoupe]

Originally Posted by fortyonerag

There are two sender units on the 41, one in each head. They are linked together.

Hmmm. I've heard that before. It appears something's not stock. I have a sensor on only one head. There's no port at all on the passenger side head.

[Ggmac]

Can you post or email a picture. Does the sender have one post or 2 ?

[SoCalCoupe]

Sponsored Links (Register now to hide all advertisements)

Originally Posted by Ggmac

Can you post or email a picture. Does the sender have one post or 2 ?

It's a one-post. It may actually be stock 1940. This car is a Ford Special, a real odd duck that Ford sold stripped down to cut costs. I've heard they used leftover 1940 parts where they could get away with it. It could be a 1940 engine.

[SoCalCoupe]

Originally Posted by fortyonerag

Being a braces and belt type of guy, I installed a pair of SW mechanical gauges. Apart from more accurate temp reading, I have much better oil pressure than I thought!

I'd like to have both stock gauges and the mechanical gauges under the dash. Is that what you're doing, both? How are you doing two temp sensors? Two oil senders?

[Floyd]

the 40 and 41/42 Fords only had one temp sensor. The port on the other head was plugged or used for the accessory hot water heater. The 59AB heads had two ports per head so the second sensor could be used.
The one post sensor works just fine without the second 2 post sensor switch. Why 2? The engine has two separate cooling circuits ,each with their own pump. The only element common to the two systems is the radiator. Therefore for the engine temp monitoring, we would like to know what is going on for both the left and right bank of the engine. Instead of having two temp measurements and thus two gauges, the system uses an "off-on" switch for one bank and a actual analog temp sensor for the other bank.
To explain this, one has to understand the King-Sealy method of instruments and that means that they do NOT use resistance or variable resistance as the measurement means. Put away your VOM as it does not apply to this system. Do not even think about resistance or Ohms. This is current driven with a heated bimetal breaker point system. (same for the gas gauge and oil pressure)
For a fail safe mode, the dash gauge reads full hot when off or no current runs through the circuit. That is why the gauge reads "hot" until you turn on the iginition and apply power to the circuit and the current flows "pulls" the gauge back to cold area.
In operation, the variable temp indicator sensor is a one terminal device like yours. It connects to ground through the brass body when operating and provides the variable sensing signal that the gauge decodes. this works perfectly with just the one sensor, however we are only measuring the temp of one bank of the engine, in your case the driver side. However we do not know the temp on the passenger side. Since the radiator is common to both, the inlet water temp is the same providing the water pump is actually pumping. In order to provide info about the passenger side, all we need to know is if the temp is too high for operation (as if the water pump on that side has stopped working (rare)). For this information, all we need is a switch that opens at approx206 degrees F to tell us it too hot, otherwise the switch will remain closed and when connected to the one terminal sensor in series, The system will display variable temp data from the other bank which is indicitve of the radiator cooling water temp. Should something happen in the other bank that drives the temp up to 206+ F, then the switch opens and therefore the gauge immediately goes to full hot on the display. It is an early version of an "idiot" light combined with actual temp measurements.
However the system works perfectly well without the the two terminal sensor.
I have only the one terminal sensor in my 40 pickup as that is the way all 81A engines were fitted
The proper part number for the single terminal sensor is 99A-10884 Which is made by King Seely is 6 Volt and used for all flatheads. The aftermarket sensors usually work OK but are not and will never be accurate. The dash gauge is very reliable and unless some turkey has burned it up with 12 Volts, they last forever. I think you have a sender problem. Find a real one and you will be happy. The repro's are not accurate.
By the way, 140 degrees is way too low for a proper operating temperature. That will kill an engine with water vapor from combustion gas washing the lubricant off the cylinder walls

[fortyonerag]

Originally Posted by SoCalCoupe

I'd like to have both stock gauges and the mechanical gauges under the dash. Is that what you're doing, both? How are you doing two temp sensors? Two oil senders?

On my 41, there is a sender port on each head at the base of the water neck (are you sure you don't have one on each side?). I connected the SW mechanical unit to one head and used the original sender on the other. So, stock on pass side, SW gauge on drivers side. The stock gauge shows just right of centre when the SW gauge registers 160C.

For oil I decided to replace the stock sender unit. The stock gauge was registering around 5 PSI, whereas the new SW unit shows some 40PSI. I don't like looking at a dead gauge, but I didn't have a lot of faith in the stock sender technology. Having said that, there is an oil boss in the block just to the side of the stock oil sender that I think could be used so both gauges can be run.

[SoCalCoupe]

Originally Posted by fortyonerag

On my 41, there is a sender port on each head at the base of the water neck (are you sure you don't have one on each side?). I connected the SW mechanical unit to one head and used the original sender on the other. So, stock on pass side, SW gauge on drivers side. The stock gauge shows just right of centre when the SW gauge registers 160C.

For oil I decided to replace the stock sender unit. The stock gauge was registering around 5 PSI, whereas the new SW unit shows some 40PSI. I don't like looking at a dead gauge, but I didn't have a lot of faith in the stock sender technology. Having said that, there is an oil boss in the block just to the side of the stock oil sender that I think could be used so both gauges can be run.

Thanks for the reply! That's what I thought. I'm 100% certain that I don't any ports on my passenger side head. If I get really curious, I'll try to identify what year it was made. Believe I've seen water temp senders that tap into radiator hoses; might think about that.



Pretty sure that oil boss is intended for the optional oil filter. I have that one too. I may add an oil filter but suppose there's a tee fitting available enabling both the filter and an extra sender.


I'm with you; don't like looking at a dead gauge and have questionable faith in the stock sender. Recall sitting a stop light and watching my oil pressure go to -0-. That was a real pucker moment. I revved the engine a couple of times and drove away from the stop light and the indication returned to normal. Never found anything wrong and it never happened again.

[SoCalCoupe]

Originally Posted by Floyd

the 40 and 41/42 Fords only had one temp sensor. The port on the other head was plugged or used for the accessory hot water heater. The 59AB heads had two ports per head so the second sensor could be used.
The one post sensor works just fine without the second 2 post sensor switch. Why 2? The engine has two separate cooling circuits ,each with their own pump. The only element common to the two systems is the radiator. Therefore for the engine temp monitoring, we would like to know what is going on for both the left and right bank of the engine. Instead of having two temp measurements and thus two gauges, the system uses an "off-on" switch for one bank and a actual analog temp sensor for the other bank.
To explain this, one has to understand the King-Sealy method of instruments and that means that they do NOT use resistance or variable resistance as the measurement means. Put away your VOM as it does not apply to this system. Do not even think about resistance or Ohms. This is current driven with a heated bimetal breaker point system. (same for the gas gauge and oil pressure)
For a fail safe mode, the dash gauge reads full hot when off or no current runs through the circuit. That is why the gauge reads "hot" until you turn on the iginition and apply power to the circuit and the current flows "pulls" the gauge back to cold area.
In operation, the variable temp indicator sensor is a one terminal device like yours. It connects to ground through the brass body when operating and provides the variable sensing signal that the gauge decodes. this works perfectly with just the one sensor, however we are only measuring the temp of one bank of the engine, in your case the driver side. However we do not know the temp on the passenger side. Since the radiator is common to both, the inlet water temp is the same providing the water pump is actually pumping. In order to provide info about the passenger side, all we need to know is if the temp is too high for operation (as if the water pump on that side has stopped working (rare)). For this information, all we need is a switch that opens at approx206 degrees F to tell us it too hot, otherwise the switch will remain closed and when connected to the one terminal sensor in series, The system will display variable temp data from the other bank which is indicitve of the radiator cooling water temp. Should something happen in the other bank that drives the temp up to 206+ F, then the switch opens and therefore the gauge immediately goes to full hot on the display. It is an early version of an "idiot" light combined with actual temp measurements.
However the system works perfectly well without the the two terminal sensor.
I have only the one terminal sensor in my 40 pickup as that is the way all 81A engines were fitted
The proper part number for the single terminal sensor is 99A-10884 Which is made by King Seely is 6 Volt and used for all flatheads. The aftermarket sensors usually work OK but are not and will never be accurate. The dash gauge is very reliable and unless some turkey has burned it up with 12 Volts, they last forever. I think you have a sender problem. Find a real one and you will be happy. The repro's are not accurate.
By the way, 140 degrees is way too low for a proper operating temperature. That will kill an engine with water vapor from combustion gas washing the lubricant off the cylinder walls

Floyd, thanks for all the great information! Know where I could get one of these NORS senders?


As to 140 deg, thanks for the tip. My temp sensor isn't an infrared unit, it's a contact sensor. On the surface after cooling down for a couple of minutes, it's 140. May be hotter inside but that's cool enough to let me know it likely wasn't close to 212 deg. I'll investigate this some more. That's why I need a good temp gauge.

[tubman]

The reproduction senders are usually junk and go bad quickly. If you can't find a NOS unit, I'd be on the lookout for a good used original sender. I have no experience with NORS units. I might have a good used one; I'll look when I get back to my shop.

[Graeme / New Zealand]

Here's the bumph on how to test senders. Pages out of order sorry

GB

[Floyd]

For measuring liquid temps, a good kitchen quick reading thermistor type works well but you need to have the radiator cap off. The good kitchen units are about a hundred dollars. This is OK if you have the stock unpressurized cooling system. An Infared gun , even the cheap Harbor Freight one like we all have works very good on solid objects, but can be suspect on clear liquids. A really good way is to use the good old fashion thermcouple which generates a tiny voltage and is measured by a millivolt scale on a VOM. Many VOM's now have a scale for temp and a thermocouple to use with it. For not much money ($25) you can get a dedicated temp reading instrument with thermocouple. You can drop the thermocouple into the radiator on slightly close the cap on the wire. If we are interested and possibly concerned about cooling or even oil temps, we need to measure accurately and properly what we have or just quit and go home.
As I stated above, you have one port on the left side head facing aft for the temp sensor and a port on the right side facing forward which is for the valve for the hot water heater or if not fitted, will be plugged. Ford did not use two sensors on the 81A engines. You do not need two sensors as the chance of a problem on the right bank will show up somewhat on total water temp and be picked up by the left side. Millions of these cars worked perfectly from the factory with only the one sensor. Anyone so paranoid about this kind of stuff should not be out driving around.
Also, the original Ford gauges and senders , when operating and wired properly are far superior and more reliable than any aftermarket stuff or junk that has been cobbled up by home mechanics.
First: measure what your coolant temp actually is; 2) check the coolant sensor that you have with a pan of boiling water and a thermometer/thermocouple and plot the results based on the gauge reading. Do not try to measure anything electrically because it won't help.
If you find the sender (most problems are actually the wiring and connections) is bad, then a NOS is worth finding. Worse case is a new reproduction one ,knowing it will not be accurate.
If looking at used senders, ( most of them are still good) the ONLY way to test them is with a Ford dash gauge, 6 Volts and a pan of water and a thermometer. Many good used temp and gas senders were thrown away because people tried to test them with an Ohm meter and could not get any results.
Report back

[Floyd]

A little added info and a clarification. Above I noted that using an Ohm meter is not the way to test these sensors or gauges. That is true. However for those that know concept of how they actually work, an Ohm meter CAN be used to see if one is dead but will not tell if it actually works. The resistance or continuity between the terminal and the brass body will read 13 Ohms +/- .3 ohms, if the heater wire circuit and the contact points are closed. This means it is not dead (dead = no continuity). However we don't know if the points are stuck, the heater circuit and bimetal works under load etc.
So, an Ohm meter can verify a full dead one but nothing else.
While on the subject, I have attached a rough sketch of the temp gauge reading for 1940 using a NOS sender, original Ford dash gauge, 6.33 V battery an electric kettle for heating the water and a $100 insta- read thermistor, elevation -sea level.
You will note that the range between the end marks starts at 125 F and goes to 212 F this is to the edge of the marker line. When the gauge is "off" the needle goes past the end marker. When below 125 F, the needle is below the "C" marker line. You will note that "N" is 175 F. The short marker lines are centered at 150 F and 200 F. This would relate to the acceptable operating range of actual coolant temp as the Ford engineers have noted. 150 F would be considered warming up but not yet fully warm. the 200 F would be acceptable for hot day, traffic ,climbing a hill ,etc. At the 212 F line, we have a boiling situation (unpressurized system) and as the gauge notes we are "HOT".
See the rough(very rough) sketch of real numbers. I will run this sometime with an aftermarket sender , we know it will be different.

[SoCalCoupe]

Status Report:


- With the ignition off the gauge reads full hot as it should.
- With the engine cold, turning the ignition on, the needle moves towards cold but barely detectable, maybe 1/32 or 1/64 of a full sweep colder
- removed the lead and wire-brushed the screw, washer, lead and sender. Reassembled and no change.

- measured the resistance between the lead and the screw: 0.1 ohm.
- The washer is just a split lock washer rather than a star washer but I couldn't find a star washer of the right size; another trip to Home Depot.
- The resistance between the brass body and a known ground is 0.2 ohm.
- The resistance between the brass body and the screw is 11.5 ohm.


This all seems pretty inconclusive and doesn't prove or disprove a bad sender.



- The contact sensor I've been using is a 'K' type thermocouple.
- Going for a long drive today and will take the thermocouple with me.
- Ordered a budget priced infrared temperature gun that will arrive tomorrow.
- Found a NOS 51-52? Lincoln temp sender FAA-10884 that I'll probably take a chance on as compatible


Looks like replacing the sender will be a little messy. I suppose it's in direct contact with coolant and that coolant will flow out of the head when I remove the sender if I haven't already lowered the coolant level through the radiator drain?

[Floyd]

Well, we are getting somewhere. The gauge going full tilt when OFF is good , but this is just a no current / open circiut condition. With the key (switch) ON and the gauge moves only slightly back to about the HOT line as shown by my children's sketch above of the gauge face means that the current is not sufficient in the circuit to energize the heater in the dash gauge to pull it back to "Cold". You have done the under hood checking very well and the connections there are OK. Now the next adventure is a trip under the dash to check the connections on the back of the instrument panel. Loosen, wiggle, remove the wire, clean the post and terminal, etc and reconnect. You don,t need star washers, good bronze split washers with tiny brass flat washers makes a perfect connection. Check the voltage at the power terminal on the gauge with the switch on and verify you have 6.3 or more Volts. Crawl out from under the dash , take a deep breath and recheck to see if the gauge moves back to "Cold". If there is no joy, then we can test the dash gauge itself to see if it is the culprit.
Assuming your original gas gauge works, ( they never fail unless they were murdered), we can use the dash gauge for this step. Back under the dash, remove the lead from the temp gauge and the gas gauge and connect the temp lead to the gas gauge. Switch on the key and see what happens. The gas gauge, like the oil pressure displays opposite direction from the temp, but will display movement. If you get a reading and the needle moves more than before, it could be the dash gauge. Give this a try. Dash gauge failures are very rare but can happen. NOS ones are cheap and plentiful because they work forever. The temp circuit problem is 70% wiring connections ,25% sender unit and 4.7% dash gauge and .3% bad ground circuit between engine and chassis..
Easiest way to remove the sender is to drain down the radiator to a level that is well below the cylinder head. Save the coolant if new and good or just go ahead and change it out if more than two years old.
Your experiment with the thermocouple will tell us what the actual temps are and when we get the gauge situation sorted out ,the numbers will be exactly the same.
Report back!

[drolston]

Gauge test:
1) disconnect wire from sender on the engine

2) Ignition on; gauge does not move, indicates full hot.
3) Ground the wire that goes to the sender; that sends full current through the gauge and it should head quickly to someplace near Cold. If it does, your gauge is good. Do not leave the wire grounded, once you establish that the gauge is good. If it does not get close to Cold, the gauge is bad and should be replaced. They are cheap, plentiful, and really not that hard to replace, if you can bend yourself in a way to see the back of the instrument panel.

[SoCalCoupe]

Originally Posted by drolston

Gauge test:
1) disconnect wire from sender on the engine

2) Ignition on; gauge does not move, indicates full hot.
3) Ground the wire that goes to the sender; that sends full current through the gauge and it should head quickly to someplace near Cold. If it does, your gauge is good. Do not leave the wire grounded, once you establish that the gauge is good. If it does not get close to Cold, the gauge is bad and should be replaced. They are cheap, plentiful, and really not that hard to replace, if you can bend yourself in a way to see the back of the instrument panel.

STATUS UPDATE:

That's pretty slick! Still on HOT with ignition off. Disconnected the lead from the sender and jumped it to ground. Turned the ignition on and within a couple of seconds the reading went to COLD.

Now I know everything works correctly and is well connected from the gauge all the way to the sender. Must be a bad sender. My NOS Lincoln sender should arrive by the end of the week; pretty sure it will work.

[SoCalCoupe]

Oh by the way...


Both heads are 81A 6050 heads. One source says they were made from 1938 - 1942 so I'd guess they're both original to the car.


Since it's a Ford Special (stripped down) with only one tail light, one horn, one windshield wiper and one sun visor, it doesn't surprise me that it has only one temperature sensor.