The Temperature Gauge - Problem Solving

The temperature gauge set-up in the Z Magnette is unusual and for that reason often causes owners a disproportionate amount of trouble. It also means that if a unit fails, it is almost impossible to find second-hand replacements, so owners often have to resort to alternative set-ups using parts that are easier to obtain. Two previous contributors to the Register’s publication have researched this problem, the late Roy Smalldridge and David Dubois. Their articles appear below, grouped together into this single location for ease of reference.

First, David’s explanation of the technical background:

“Looking through the bulletin board messages, I see a number of questions about the operation of the Magnette temperature gauge. I would like to take this opportunity to clarify some points about its operation. 

The temperature-sending unit in the late MGBs and all modern cars is a negative temperature coefficient device, i.e. as the temperature goes up, the resistance decreases allowing a greater current flow. This means that as the temperature of the engine increases, current flow through the gauge increases and the needle deflects toward H. 

Temp_SenderThe temperature-sending unit and gauge in the Magnettes do not operate like this. First of all, notice that when the ignition of the Magnette is turned off, the temperature gauge reads maximum H. When the ignition is turned on, the gauge swings down to C and then works its way toward H as the engine warms up. To operate this way, the sending unit consists of a bi-metallic switch, with the switchblade wrapped with resistance wire. The switch contacts are closed when everything is cold. When the ignition is first turned on, current flows through the resistance wire, the switch contacts and the temperature gauge, causing the gauge to move to the C position. The current through the resistance wire in the sending unit heats the bi-metallic switchblade and causing it to bend away from the stationary contact and opens the switch. When the switch is open the bi-metallic blade cools and straightens, closing the switch contacts. This cycle repeats as long as the ignition switch is on. This is exactly the way a flasher unit in turn signals works. Everything is factory calibrated--that is what the adjusting screw on the switch is for--so that the switch opens and closes at the required rate for the gauge to indicate C. 

This bimetallic switch action also acts like a voltage stabilizer so changes in battery voltage will not cause the gauge to fluctuate. When the engine heats up, the heat is transferred to the bi-metallic blade in the sensor unit causing it to stay open for a longer period of time, interrupting the current flow to the gauge for longer periods of time. Since the gauge also consists of a bi-metallic blade that deflects with current generated heat, moving the needle to C, the longer periods of no current flow allows the needle to move back toward H. In effect, they have made the temperature sensor act as a positive temperature coefficient unit (the higher the temperature, the higher the effective resistance causing less current flow). This is all a rather complicated method to accomplish a relative simple procedure and one that has a lot of areas for failure. 

I suspect that those of you who are seeing the gauge work backward have a car in which the sensor failed and someone has replaced it with a sensor from a late MGB (rather than pay the astronomical price for a Magnette sensor). The sensor in my Magnette failed shortly after I bought it (it failed on the freeway between Portland and Seattle, scaring me out of my wits when I saw the temperature gauge climb up to and beyond the H). After checking the prices and finding out how the system worked, I purchased a late MGB sensor and built an add-on circuit to reverse its operation. After calibrating the sensor-converter-gauge combination, the temperature gauge in my Magnette now works the way it was intended to and if the sensor goes bad, it is relatively inexpensive to replace. 

If anyone is interested in doing the same thing on their Magnette, see below the diagram and instructions on how to build the circuit and calibrate it with the temperature gauge. This is a fairly involved project and requires a good working knowledge of electronics and construction, soldering and calibration practices. I would not recommend anyone without this skill and knowledge undertaking this project.” 

David DuBois,USA

Secondly, David’s instructions on how to build an alternative circuit:


Dl        9.1 volt zener diode 1N5239 or equivalent

R1       12 ohm 2 watt resistor

R2       1000 ohm 1 watt trim pot

R3       100 ohm 1 watt trim pot

Q1       NPN 500 ma transistor 2N4400 or equivalent (2N4402 for +ve ground cars)

Temperature sensor late model MGB sensor Moss p/n 760-180

                                     Cold resistance = 900 ohms (approx)

                                     Hot (212°) resistance = 47 ohms (approx)

                                     Resistance at 170° - 180° = 100 ohms (approx) Temperature gauge

Magnette gauge

                                     Resistance = 21 ohms (approx)

                                     Current for cold reading = 195 ma (approx)

                                     Current for hot reading = 43 ma (approx)

 >> = External connections to temperature sensor converter box

Diagram is drawn for negative ground system. If using on a positive ground system, reverse the direction of Dl and use PNP transistor (2N4402 or equivalent).




















To calibrate:

 1. Turn ignition on

2. Adjust R3 for reading of C on gauge.

3. Place sensor in boiling water and allow its temperature to stabilize.

4. Adjust R2 for reading just below H.

5. Remove sensor from boiling water and for reading of C.

6. Repeat the above steps as many times to read C when the sensor is at ambient temperature and just below H when the sensor is in boiling water.

Note 1: There are some adjustments on the temperature gauge that affect the range from C to H, but they have to be made very judiciously.

Note 2: I set the reading of the temperature gauge to just below H when the sensor is immersed in boiling water, as it is not uncommon for engine temperature to go up to 230°and not be considered overheating. You may consider a lower temperature to be overheating for your car, so set the gauge to read H at whatever temperature you feel comfortable with. Just keep in mind that boiling water will always be 212° at sea level (if you are doing this in Denver, you will have to compensate).

Note 3: It is an extremely involved procedure to build this converter and calibrate the system. I will not accept any responsibility if the conversion does not work in your car or causes any damage from being done incorrectly.


Lastly, here is Roy Smalldridge’s piece, written after dismantling a sender and experimenting with it:

 "We all like the fact that our Magnettes have full instrumentation, i.e. fuel, oil pressure, ammeter and temperature gauges. However, the Z Magnette is unique in that the temperature measuring system is like none other I have come across. So we do not want to get rid of them when they go wrong.It is almost impossible to buy a new temperature transmitter and this, together with the mention of these items in last year‘s Magnettics, prompted me to look more deeply into the situation. My first job was to borrow a stock of old transmitters and gauges from Warren Marsh.

The Problem

The engine, as we know, works better when it is hot but not when it is boiling, which is often a condition indicated on the temperature gauge, caused by a defective transmitter. If a correct radiator cap is fitted and no water or steam is coming from the overflow then the temperature gauge is misleading us and the water is clearly not boiling.

Gauge and Thermal Transmitter - How They Work

The gauge and transmitter are of the hot wire type. They both have a coil of fine cotton covered wire wound around a bimetal strip. As current flows through the wire it heats up and causes the strip to bend. This happens in both the gauge and transmitter at the same time. Everyone will have observed that the gauge on the Zs reads hot with the ignition off and engine cold, i.e. in open circuit. This is by design: it could easily have been the other way round. My tests on many transmitters and gauges show they all have a resistance of approximately 25ohms. When the ignition is switched on, the gauge needle goes to the cold position which is determined by the resistance of the gauge and transmitter. Both are adjustable. The gauge is electrically connected at all times, i.e. it is a continuous wire from one connection to the other. This means it seldom goes wrong.





It is a different story for the transmitter; the drawing shows what is inside and how much can go wrong. The brass sleeve is in two parts, the centre flange used to hold it in the engine by a threaded sleeve nut. The brass sleeve can be split by heating the solder joint at the flange, with a suitable iron. This reveals three very vulnerable points:

(a) the wire connected to the outside screw relies on a scraping spring to make a connection.

(b) a brass spring is used as another scraping connection to make the circuit to earth through the engine.

(c) a pair of contacts which open when boiling point is reached, making the gauge read hot, (same condition as ignition off).

The transmitters were obviously of older design than our Zs and manufactured by Smiths Instruments before the first Z left Abingdon factory. Air was not evacuated from them, so over many years a film builds up on the inner surface of the sleeve causing resistance at the scrapers to rise. Thus the gauge reads incorrectly. Also a resistance builds up between the points so that they can show open circuit, when the gauge will read hot all the time.

What Can Be Done About It?

I calibrated the gauge using a travel jug, industrial thermometer, two multi -meters and a 12-volt battery (See sketch). The ammeter shows the current flowing, and the resistance meter (on ohms) is only connected when the battery is disconnected.


Heat water to normal engine temperature  – I used 80 to 85 degrees.  On the rear of gauge find and remove two blanking plugs. Now the needle can be set to normal (this is done when manufactured). Only a slight turn with a tiny screwdriver will move the needle. However, this operation is unlikely to be necessary as the fault is almost certainly to be found in the transmitter.So, to the transmitter. Be brave, take courage in both hands and split the sleeve into two parts by using a large soldering iron. Remove the inner part. Clean both scraping connections, including inside the sleeve and cheek resistance. If the latter is still high, clean the points, but not with an abrasive as they do not burn or pit. Card should do the trick.

The resistance of the transmitter rises, as it gets hotter. When it reaches 100°C the points open and the gauge needle goes fully to hot, (our open circuit situation). Adjustment is by means of a small screw. Turning it in (clockwise) does not make the gauge read colder. It will stay the same because the reading depends on the resistance of the wire, due to the temperature of the water. Turning it anticlockwise will mean the points will open more easily, hence an earlier warning of over heating. There is an anti vibration lock to stop any variation after setting up. By the way, cleaning usually does the trick, so don’t get carried away with the adjusting. Finally the transmitter must be assembled and the sleeve being re-soldered. It goes without saying, before you take any of the actions described, check that it is really the instrumentation at fault and that the engine is not actually overheating, but I could go on forever about that and I don‘t want to tell you what you already know.

Maybe someone out there has a drawer full of brand new transmitters marked “Smiths Motor Accessories Ltd, England, part no. TT1200/01-12v“ and save us all the trouble.”

You have no rights to post comments