1937 Coolant Temp Issues - Again

[John_Mc]

For a very long time, I've had erratic or no temp gauge at all on my '37 Zephyr.  I was hoping with the new engine, the issue would have been fixed.

I have an in-head sending unit from a 1948 engine I believe and as i recall it's nearly new.  First, I grounded the sending unit and with the ignition on, the gauge needle does not move, it is pegged at cold no matter what I do.  I do have continuity from the sender unit to the gauge.  Must be the gauge, correct???  Any test I can perform before buying an new dash gauge?  Thanks. 

 

 

Edited April 23, 2018 by John_Mc (see edit history)

[hchris]

5 hours ago, John_Mc said:

First, I grounded the sending unit and with the ignition on, the gauge needle does not move, it is pegged at cold no matter what I do.  I do have continuity from the sender unit to the gauge.  

Not sure if I'm reading you correctly, but what you should be doing is, with ignition on, remove the wire from the sender and ground it. Your guage should read max temp, if not then the guage is faulty.

 

To check the sender, connect it with an ohm meter and note its cold resistance value, then run the engine and look for a change in the resistance as it warms up. The specific resistance values applicable to temp changes vary from one make to another and you would have to search for your values, nevertheless if the resistance changes with heat then the sender is working,  how accurate it is may be questionable. 

[38ShortopConv.]

John
If you can get access to the dash temperature gauge terminals you should have 6 volts at the gauge terminal with  the ignition switch on. [The dash gauges should be fed directly from the ignition  switch, not the  ignition resistor bank ] If you  ground the other terminal at the gauge  the dash gauge needle should rise to the top. Don't leave it on more than a few seconds. Do all this with the sender wire disconnected at the sender end, isolated from ground
. If the dash gauge is working after these tests ground the sender wire at the sender end, if the gauge still works then the sender is not compatable with the gauge. If you have the ignition switch on for any length of time while testing its best to disconnect the two ignition coil  wires at the coil to avoid the coils overheating and draining the battery especially if the radio is on.  Could overload ignition switch as well in this situation and cause early ignition. switch failure.
Cheers Roy  

[John_Mc]

Excellent info, thanks guys.  For some reason I am not show 6 volts at the sender with the ignition on so that might be my issue.  The gauges are wired right off the ignition switch.  Thanks again, I will post my results.

[John_Mc]

Update:. I have roughly 5.5 volts at the gauge and when I momentarily ground the hot lead, the needle still has no movement.  Time for a new gauge, correct?

[Ray500]

John...if you can pull the gauges out of the dash or get behind the dash to look at the terminals on the temperature gauge, you will find the 'hot' terminal that goes to the ignition switch, and measure the voltage at that point.  It should be in the range of 6 volts negative since the positive terminal of the battery is grounded to the chassis/body.   The other terminal of the temperature gauge is a wire that goes out to the temperature sensor on the engine.  You can ground that terminal directly to the dash metal or some other good ground in the vehicle and view the gauge to see if it goes to the other side of it's reading usually the HOT reading. Be sure and don't ground the hot terminal from the ignition!  If it doesn't move the needle then there's a problem with the gauge.  You can remove the temperature gauge from the dash cluster and either find a replacement for it or there are places that do repair of them.  They advertise in Hemmings books and on line, just have to search.  Some of the  suppliers of parts on LZOC might have a gauge  to replace it.  Good Luck!

[Colin Spong]

John... The sender unit for a 1936-38 LZ is mounted in the top tank of the radiator and looks like a small cigar.

As a point of interest it is the same as the one used on the 1937/38 Willys. The temperature gauges used

on 1936/38 LZ's "rest" on COLD but the 1939 and possibly later "rest" on HOT. Therefore I seem to remember that

the cylinder head sender was incompatible with the 1937 dash unit.  Of course you may well have a defective

dash unit but this could be part of your problem.  Colin

[38ShortopConv.]

John, correct!
looks like a new gauge then may  be a compatible sender. Good point from Colin.
Roy 

Edited April 27, 2018 by 38ShortopConv.
spelling plus further coment (see edit history)

[John_Mc]

7 hours ago, Colin Spong said:

John... The sender unit for a 1936-38 LZ is mounted in the top tank of the radiator and looks like a small cigar.

As a point of interest it is the same as the one used on the 1937/38 Willys. The temperature gauges used

on 1936/38 LZ's "rest" on COLD but the 1939 and possibly later "rest" on HOT. Therefore I seem to remember that

the cylinder head sender was incompatible with the 1937 dash unit.  Of course you may well have a defective

dash unit but this could be part of your problem.  Colin

Thanks Colin, yes my block is from 1941 and I'm not using the radiator based sender.  That might be my problem.  I'm sure the original type senders are available and maybe I'll go that way before replacing the gauge.  My radiator man was fearful of removing the remains of the original thermostat for fear of damaging the thing, but now I kind of wish he would have tried to remove it.

Edited April 27, 2018 by John_Mc (see edit history)

[John_Mc]

11 hours ago, Ray500 said:

John...if you can pull the gauges out of the dash or get behind the dash to look at the terminals on the temperature gauge, you will find the 'hot' terminal that goes to the ignition switch, and measure the voltage at that point.  It should be in the range of 6 volts negative since the positive terminal of the battery is grounded to the chassis/body.   The other terminal of the temperature gauge is a wire that goes out to the temperature sensor on the engine.  You can ground that terminal directly to the dash metal or some other good ground in the vehicle and view the gauge to see if it goes to the other side of it's reading usually the HOT reading. Be sure and don't ground the hot terminal from the ignition!  If it doesn't move the needle then there's a problem with the gauge.  You can remove the temperature gauge from the dash cluster and either find a replacement for it or there are places that do repair of them.  They advertise in Hemmings books and on line, just have to search.  Some of the  suppliers of parts on LZOC might have a gauge  to replace it.  Good Luck!

Ray, I grounded the lead going out to the sending unit and.................the temp needle jumped right up to HOT as I and hoped!  So, it would appear that the issue is NOT the gauge, but is either a bad sending unit or an incompatible sending unit for the gauge.  Also, I decided to run a parallel wire from the output of the dashboard gauge post out to the sender and fired up the engine.  This time the needle did move very slowly up to hot and stayed there on a barely warm engine.  It took a full minute or so for this to happen.  When grounding last evening, I grounded the hot from the ignition switch, only for a second so hopefully no damage has occurred.

Edited April 27, 2018 by John_Mc (see edit history)

[Ray500]

12 hours ago, John_Mc said:

Ray, I grounded the lead going out to the sending unit and.................the temp needle jumped right up to HOT as I and hoped!  So, it would appear that the issue is NOT the gauge, but is either a bad sending unit or an incompatible sending unit for the gauge.  Also, I decided to run a parallel wire from the output of the dashboard gauge post out to the sender and fired up the engine.  This time the needle did move very slowly up to hot and stayed there on a barely warm engine.  It took a full minute or so for this to happen.  When grounding last evening, I grounded the hot from the ignition switch, only for a second so hopefully no damage has occurred.

Yeah John, I think it's an issue with the temperature sensor, being usually in the radiator on those vehicles, but you said you have a 41 engine, but still the sensor might not be compatible with the gauge.  Finding the correct sensor and seeing if you can mount it on the engine or just go back to the radiator position if that makes it work properly.  You do need to know the temperature of your engine, as some V12s tend to run hot.  I have baffles behind my water pumps to help push the coolant toward the rear of the engine to get better cooling.  I do use 100% ethylene- glycol not reduced with water to get better cooling, especially since I live in the tropics where it's mostly hot!  

[hchris]

20 hours ago, John_Mc said:

but is either a bad sending unit or an incompatible sending unit for the gauge.  

 

Ok good news with grounding the sensor wire now you know the guage works. For sure the sender unit needs to be compatible with the guage, so if you are not using original components you're not on a level playing field. 

 

The sender unit is at the receiving end of the 6v line passing through the guage to ground, it is a large resistor which varies it's electrical values when exposed to heat, as such the changes in resistance affects how the guage reads, if the sender resistance isn't matched to the guage then the guage won't work or will be inaccurate. 

 

To check the sender, disconnect the sender/guage wire and connect a multimeter between the sender and earth whilst cold, noting the resistance value.  Keep the multimeter connected between sender and earth and run the engine, as the coolant temp rises the resistance value should change, this at least confirms that the sender is working. 

 

Assuming that this works you now have to determine if you have a matching sender/guage resistance, so a bit of research will be required or fitment of a known value sender.

Edited April 28, 2018 by hchris
Words rearranged (see edit history)

[19tom40]

John, before you try to remove the sending unit from the radiator, you should see if you can get another. You don't want a hole that you can't fill.

 

The test that hcris gave you will work if the sending unit is not the King-Seely type used in the Fords. If it is the Ford type, you will read a continuous short if the unit is good, no matter the coolant temperature. The Ford type uses a bi-metal strip with contacts for current flow. As the coolant gets warmer the contacts are open longer and less current flows in the dash unit - sending unit circuit and the bi-metal strip in the dash unit moves the needle towards hot. The Zephyr unit may be the same type, but uses more current to move the needle towards hot.

 

Put some penetrating oil on the sending unit in the radiator every day for a couple of weeks and then try to remove it. If it doesn't move, take the radiator out and take it to Lake Zurich Radiator and have Jim hot tank it and remove the sending unit. The unit is probably brass and the radiator tank is also brass, so it should come out without damaging the tank.

 

I don't know the size or how your gauge mounts, but I have a good 1940 gauge with a bad face and maybe you could exchange the face from your gauge and use the sending unit in the head. If you want to consider this option, send me photos of the back of the gauge and the face. Also send me some measurement of the mounting stud separation and the diameter of the body. The measurements just need to be approximate to see if this is possible.

[hchris]

11 hours ago, 19tom40 said:

. The Ford type uses a bi-metal strip with contacts

 

Well done, that's where product familiarity is great.

 

Given the change in current flow, couldn't you still use a multimeter to ascertain if the sender is working ? accuracy may be another matter, but at least you can prove/disprove its function.

[19tom40]

Good question, I have never tried it, I have an extra gauge that I use for testing. I will try using a voltmeter to see if the average voltage changes when the engine temperature changes and let you know.

[John_Mc]

Tom, Hchris and Ray, in the 1932-1948 green service manual it says that all Mercurys and all Lincoln Zephyrs use the same sender that will vary resistance based upon engine coolant. temperature. It also states that with the current turned off the gauge should be in the HOT position.  Only when current is supplied will the needle move down to a normal position.  So am I right in assuming that as the engine gains heat, the resistance increases, so in essence, the lower the current, the higher the gauge reading?  My gauge without power shows cold, so I don't know what to think at this point.  I will check today the resistance at the sender, with the power wire disconnected, to see if the value changes from cold to warm.

[hchris]

So let me "fess up here" - I'm not a Lincoln person, nevertheless temp senders generally fall into two categories, resistance based or voltage based.

 

For the purpose of your exercise you only need to establish if the sender value changes with heat, so lets first assume it works on resistance, do the previously mentioned resistance measuring with the multimeter as the engine warms up.

 

If you get no response with this then reconnect the wire to the sender unit, set your multimeter to volts and take a voltage reading between the guage and sender unit as the engine warms up, looking for a voltage change consistent with temp change (it's worth first measuring the cold voltage).

 

If you get a response in either of these exercises you have at least established if the sender is working, or not. If you get lucky enough to find the sender is working then your next step would be to see if it's the correct one for your guage.

 

Having so far established that the guage is working, it's logical to see if the sender is working before pulling anything apart.

[Bloo]

 If this is a King Seely style gauge.....

 

The gauge itself is thermal. A heater in the gauge heats a bimetallic strip to move the needle. Meanwhile, the voltage for this gauge is supplied by a device that is in essence a thermal flasher. The ground is supplied through the sending unit. The flasher sits there flashing all day, hitting the heater in the gauge with pulse after pulse of current. How much current is determined by the resistance of the sending unit. With more current in each pulse, the heater (and bimetal) in the gauge get hotter, changing the needle position.

 

You wont get anywhere at all trying to test this type of gauge live with a multimeter, you will just see it turning on-off-on-off all day long.

 

Checking the sending unit by itself for resistance should work ok.

[hchris]

2 hours ago, Bloo said:

Checking the sending unit by itself for resistance should work ok.

 

OK so there we have it, out with the multimeter and check resistance  whilst warming up the engine

[John_Mc]

Will do, you guys are just fantastic, thanks you for all the replies.

[19tom40]

Using 1 multi-meter on voltage, the voltage at the sending unit varied as the internal contacts open and closed. You could not see any difference in the hot or cold. My true RMS read a steady voltage of about 3.2V hot or cold. The difference in time that the contacts are closed does not seem to vary much from cold to hot.

 

As Blu said, you won't get anywhere testing the sending unit with a multi meter.

 

The pre-1939 cars had a different sending unit and dash unit as Colin said. Your gauge should rest on cold with the ignition off. The sending units differ internally with the earlier one supplying more current to the dash unit as the engine warms up and the later one supplies less current to the dash unit as the engine warms up. The dash unit has the spring wound differently so that the pre-1939 spring tension pulls the needle to the cold position.

 

It is possible that your 37 sending unit is OK and the problem is inadequate grounding of the radiator unit. To check this, wire the radiator sending unit up normally, turn on the ignition switch and check for voltage at the sending unit wire with your red meter lead on a good ground (engine stud, engine ground strap or firewall bolt). Then compare the reading with one taken between the sending unit wire and the sending unit case. They should be the same. You can do another check on a warm engine by connecting the clamp on your red jumper lead to the POS terminal of the battery and the clamp on the other end to the case of the sending unit. Turn on the ignition switch and check the reading of the gauge.

 

Lots of people have used Teflon tape to seal the threads on sending units and Teflon tape is a pretty good insulator. The sending units have pipe threads and do not need any sealant. If you can remove the sending unit from the radiator, clean the threads on the sending unit and the radiator to ensure a good connection. Check that the radiator has a good connection to the chassis or body.

 

You coolant should always be in contact with the sending unit. Make sure that you have enough coolant to ensure this.

[John_Mc]

Tom, thanks and i will go thru the steps you've outlined above as soon as I am able.  In the meantime here i'm not sure how to interpret the results but here are the numbersfrom my tests today.  The continuity tests were performed with the hot lead disconnected.

 

Ohms  17.7 when cold

Ohms 20.07 when hot

 

Volts 3.24 when cold

Volts 4.58 when hot

 

Also, I discovered a broken wire from the output of the gauge to the sending unit; no wonder I was having no needle movement!!! .  Once repaired I turned on the ignition switch and the needle did move slowly up towards HOT, in fact it pegged and HOT and beyond and fearful that I would burn out the gauge, disconnected the wire.  It was similar to needle movement if I just grounded the hot lead.  So now at least I have a gauge that apparently works and my sending unit must be the issue.  Of course originally in 1937 the sending unit was in the radiator as opposed to the one I have now screwed into the head.  I do not have teflon on the threads.  I keep wondering if my gauge is not calibrated to work with a hood based sending unit.

[38ShortopConv.]

John, now that you have repaired your sender wire  and the dash gauge moves to the top when you ground  it, it sounds like you just need a sender that works and  is  compatable to  your  gauge.
You can  test sender units in a jug of boiling water under the hood suspended from the sender wire. Of cause you need to remove them from radiator or head first  and ground sender body. Don't put sender wire connection into water. Be aware of escaping radiator water when removing senders. Start with warm water coming to a boil and watch gauge rise slowly, if the sender is compatable and working.
  Beats running motor all the time,  try not to overheat the ignition coils if both the ignition points are closed at once which they sometimes are when motor is not running with ignition switch on, could disconnect  ignition  wires at coils.
Frequent short start ups on new LZ  motors can glaze bore , then they smoke, best to drive   briskly non stop for  100 miles to bed rings in, don't know how many miles yours has done

.Cheers Roy  

[19tom40]

John, do not use the sending unit in the heads with your dash gauge. You will damage the gauge. Do ALL of your testing with the radiator sending unit. Replace the head unit with a  1/2" pipe plug.

 

Your test results using the head sending unit are what I would expect. The head unit allows more current to flow in the  circuit when the engine is cold and that would peg your gauge.

[John_Mc]

17 hours ago, 38ShortopConv. said:

John, now that you have repaired your sender wire  and the dash gauge moves to the top when you ground  it, it sounds like you just need a sender that works and  is  compatable to  your  gauge.
You can  test sender units in a jug of boiling water under the hood suspended from the sender wire. Of cause you need to remove them from radiator or head first  and ground sender body. Don't put sender wire connection into water. Be aware of escaping radiator water when removing senders. Start with warm water coming to a boil and watch gauge rise slowly, if the sender is compatable and working.
  Beats running motor all the time,  try not to overheat the ignition coils if both the ignition points are closed at once which they sometimes are when motor is not running with ignition switch on, could disconnect  ignition  wires at coils.
Frequent short start ups on new LZ  motors can glaze bore , then they smoke, best to drive   briskly non stop for  100 miles to bed rings in, don't know how many miles yours has done

.Cheers Roy  

Thanks Roy, I will do as you have outlined, but as Tom has described, I need to focus on a sending unit based on the radiator installation.  I’m glad I disconnected the wire before I destroyed the gauge or had a fire.  I’m going to plug the head port and drive this car.  I have had a little smoking issue with this newly rebuilt motor and after speaking with the rebuilder he also suggested to get it out and drive under load and at a high rpm.

I’m working on finding a radiator sender right now.

[John_Mc]

On 4/27/2018 at 5:38 AM, Colin Spong said:

John... The sender unit for a 1936-38 LZ is mounted in the top tank of the radiator and looks like a small cigar.

As a point of interest it is the same as the one used on the 1937/38 Willys. The temperature gauges used

on 1936/38 LZ's "rest" on COLD but the 1939 and possibly later "rest" on HOT. Therefore I seem to remember that

the cylinder head sender was incompatible with the 1937 dash unit.  Of course you may well have a defective

dash unit but this could be part of your problem.  Colin

Colin, apparently the usual suspects with Lincoln Parts is having a very tough time finding a radiator based sending unit.  Can you recommend a Willys dealer over hear that might have this sender???  Thanks in advance.