Welcome to our resource for MG Car Information.



MG parts spares and accessories are available for MG T Series (TA, MG TB, MG TC, MG TD, MG TF), Magnette, MGA, Twin cam, MGB, MGBGT, MGC, MGC GT, MG Midget, Sprite and other MG models from British car spares company LBCarCo.

MG MGB Technical - Cooling Question

Car is a 72 BGT, basically stock as far as I know. It is a recent purchase, and is now my daily driver.

Coming home from work this afternoon turned out to be a bit of an adventure. There is a heat wave here in Southern California and also a major accident on I-8, with the result of stop and go traffic for what could have been an hour or more in temperatures in the high 90s. After about 20 minutes stuck in this I noticed the temp gauge was above what I consider "normal". I have a 195 thermostat in the car and the temperature gauge is about 1 needle width below the H block when the thermostat opens. Under most conditions it doesn't rise more that 1/2 a needle width above this position. It did today. Anyway, I managed to get off onto I-15 and bypass most of the congestion.

Now to the question part. When I got home, and before shutting the engine down, I took my IR thermometer and checked the radiator temperatures. Top tank consistent at 209, left side of radiator over 200, middle of the radiator (horizontally) in the 170s and the left side was down in the 150s. Anyone want to guess what is happening? Do I have some blocked tubes? Or is this normal?

And yes, recalibrating the temperature gauge system is on my agenda. I have done a resistance vs temperature test on the temperature sensor (on two) but haven't done any measurements on the gauge yet.


1972 BGT
Larry Hallanger

Larry - The teperatures you measured on your radiator sound like they are a normal distribution of temperatures, until you get to the "second" left side. The center of the radiator is going to read lower because that is where the air is being drawn through by the fan. The 150° reading would indicate a possible blockage on that side (now water circulating through, no temperature rise like the rest of the radiator. Cheers - Dave
David DuBois

1 needles width below the *H* block is normal? That sounds very high to me, or did you mean the 'N'?

195F is a stat for cold climates, in fact even higher than the highest specified for the MGB, is there any reason why you have this stat? A high temperature stat will make your car run hot, cars in hot climates normally use a cooler stat of 165F.

If anywhere in the core of the radiator is cooler than the outlet, then that does indicate blockages. The action of the fan can distort the readings, easy to disconnect an electric fan, with a mechanical you would probably need to put some card one side of the rad or the other and test the other side.
Paul Hunt 2


I have a temperature gauge "problem" in that at about 195 deg F the needle is at the "one width from the H block" position. I am still trying to determine what the real problem is. I have 2 new senders and have calibrated both with nearly identical temperature-resistance curves. I plan to test the gauge using a variable resistor as the "sensor" element and see what results I get. Probable solution is to add a fixed resistance between the gauge and the sensor. Will "publish" the results when I gat all of the data.

I run the hotter thermostat on personal preference. Higher temperatures give higher efficiencies, and better heater performance in the winter (the few times I will need it).

Larry Hallanger


Just a couple ideas re: your high reading temp gauge.

a) Do you also have a "high reading" fuel gauge? (If so, change out your voltage stabilizer.)

b) Is there any chance the temp gauge, though it would look similar, is out of a '75 or '76 B? If so, the "early" sending unit (GTR104) sending to this gauge will read 'high'. (Likewise, a later unit (GTR101-- don't ask me why it's an "earlier" number) mated to an early gauge reads low.) From my own experience (see other thread!) 209F is about 2/3rds between the N and the H (toward the H). I think that would be around 3 needle widths from the H. :-)

For my own $.02, I'd take that winter thermostat out of there and go with either the 165F (what we run here in North Carolina) or the standard 180. It's for instances like what you just experienced-- creeping up due to traffic. With the 180, you get a "buffer" of 15 degrees.

Or you could add some kind of overflow tank to your system (the later '77 style can be added rather easily). This saves the progressive coolant losses due to "spiking" when you park the car, and also helps the car run 5 to 10 deg cooler due to larger volume/less air pockets/chance for expansion and contraction.

Best of luck,

Max Fulton

If the coolant really is at 195F then that should be lower on the gauge than 1 needles width below H, the older numerically calibrated gauges went up to about 230 in what became the H position, 195 is only about 1/3rd the way between N and H. However 209F would be higher than that, but still less than half-way between N and H.

That indicates an over-reading gauge. This can be caused by a faulty voltage stabiliser - if you have the original type on the light-green/green at the gauges, or at the wire at the sender (disconnected from the sender) you should see 12v switching on and off about once per second. If you have an after-market stabiliser this may be electronic in which case you would see about 10v steady. There are also different senders, and used with the wrong gauge can give incorrect readings, as well as the sender itself being faulty.

What 'winter' temperatures do you get? If the heater system is up to snuff (and there are at least ten aspects of this that can contribute to poor performance) then I would have thought it would be adequate (albeit not as good as a modern car) at least down to freezing. I know mine are, delivering 140F at an ambient of 50F.
Paul Hunt 2


I will have to check the gauge. Are there any numbers on the gauge that would indicate which one I have? The car is a relatively recent purchase, and the "restoration" done on it before I got it could very well have used the wrong gauge.

Any data on the resistance vs temperature characteristics of the GTR101 and GTR104?

Fuel gauge reads correctly.

On the 195 thermostat, temperature generally holds within a half needle width once the thermostat opens. Only time it has gone higher was on the way home last week (see previous post). Just a personal preference. Will see how things go long term. I do have an overflow tank added on. I also plan to have the radiator checked and rodded out if necessary as soon as I get the A back on the road.


Larry Hallanger

The 195 stat was specified in California in later years, as running the engine a little warmer reduces emissions and improves fuel efficiency. (a car engine is a heat engine, after all.) If your cooling system is working properly (i.e. capable of holding the temperature steady) the 18V engines are proven to run happily at this temperature, in California, in the summer. I'd call this thermostat a good choice.

If the cooling system can't hold your temperature steady, that's another story. You have the IR gun, so scan your radiator for hot and cool spots. It should be a nice fairly even cooling as it goes from top to bottom. If not, make it so.
Sam Good

I forget if you are using the stock engine driven fan or not, but I found that temp control much better and more consistent after installing an electric fan to replace it. I have included a photo of a double 10" fan set up (not my own) simply to show that it can be done. You may not want to go this route, but even my single 10" fan's ability to pump air thru the rad while stopped or in traffic as you describe made a noticeable difference the system's ability to hold a lower temp.

Bob Muenchausen


I see we think along the same lines. Cooling system holds temperature very well normally.


I am using the stock engine driven fan. Will consider going the electric fan route but first I think I need to sort out the radiator issue, if it actually exists. Next step is to get out the IR thermometer again and rerun my "scans" of the various surfaces. Thanks for the picture.


I do need to put a meter on the supply to the gauge. Fuel gauge appears to read correctly but I have not actually measured the voltage.

Larry Hallanger

Before you try the aforementioned tricks, I'd suggest that you try something much more simple and basic. Take the car to a competent radiator shop and have the components of the entire system, including the engine, radiator, and heater core, flushed and descaled to remove the 35+ years’ accumulation of muck, rust, and mineral deposits that act as insulators that keep heat from being dispelled by the cooling system. You will be surprised at how much cooler the engine will run in the summer and how much warmer the heater is in the winter.
Steve S.

Just re-checked the radiator readings when I got home this afternoon. Car was idling with the hood open and had idled for about 5 minutes before the test while I got the IR thermometer. Took 4 sets of measurements at the right, center, and left (car right to left, my let to right from in front). The header tank read 201, 199 and 184. Fins 1/4 if the way below the header tank read 164, 158, and 157. Next pass, half way between header and bottom tank, read 155, 149, and 150. Last pass was 168, 157, and 150. Given the variability of the way the thermometer reads the fin temperature (angle, dirt, etc. etc.) I figure that this is relatively uniform but with the highest temp at the upper hose and lowest at the bottom hose. It does appear that my earlier reading, that started this thread, was an anomaly.

I also put a meter on the sensor wire and got a fluctuating voltage a bit below 10 VDC. Fluctuations were probably 4-6 per second. Voltage below 10 makes sense with a 10 VDC supply and the resistance of the gauge in series.

Engine was rebuilt before I bought the car. Not many miles after the rebuild but it didn't get driven much so time is a bit longer than would be expected for the miles. I am getting a rusty "deposit" in the overflow tank and plan to flush the system myself in the next couple of weeks. Next step will be to pull and flush the radiator. But will have to wait until I get the A back on the road since the B is my daily driver.

Thanks to all for the inputs. They are truly helpful! I will continue the temperature gauge issue in the temperature gauge question thread.

72 BGT
58 A
Larry Hallanger


You've got a 35 year old British car, a southern California heatwave, a one-hour traffic jam - and you didn't overheat?



As far as I could help, you can easily fix a blocked rad. core. Give the car a trip to warm it up enough. Stop it and immediatly check by hand the rad front area. If ever you get a cold area on it, you can be sure that it's blocked, in normal condition it should be hot everywhere; I once experinced this to sort out that kind of problem.

I would have expected a lot lower than 10v with the gauge near its maximum deflection, that is where the sender is at its lowest resistance and the greatest voltage dropped across the gauge, hence the least across the sender, also much slower fluctuations than that.

I've checked my fuel gauge and on 3/4 tank it shows just over 3v wrt ground, and pulses on briefly about once per second. That is with the engine off. With the engine running the system voltage will be higher, which will mean the 'on' periods are shorter, but the off periods should be about the same. Whilst this will mean it will pulse faster with the engine running, it will only be about 20% faster and not 4-6 times faster. I know this isn't a temp gauge, but it is about the same era as my 75, and AFAIK the characteristics of fuel and temp systems are about the same. It would be interesting to swap over the temp and fuel senders and see just what happens.

Also there isn't a 10vDC supply for the gauges. The factory stabiliser outputs system voltage which can vary between 12v and 15v, however it switches it on and off in a varying duty cycle so that it *averages* about 10v whilst the system voltage varies as above. If you had an aftermarket stabiliser which regulates to 10v, then it would be a *constant* 10v, not fluctuating. With the stabiliser switching on more frequently the gauge will read higher. In theory this should be compensated for by switching off more frequently as well, but as your stabiliser seems to depart so much from what I have observed with both mine, i.e. it is probably operating outside its design spec, then there is every reason to expect that the duty-cycle is out as well and it is averaging higher than 10v.
Paul Hunt 2

Since the OE stabilizer is electro-mechanical and some newer designs are solid state, am I correct in assuming that you are saying that a steady 10V out is with a solid state stabilizer? If so, that would appear to be of some advantage, perhaps generating more accurate readings??
Bob Muenchausen

Bob - yes, a solid-state stabiliser should give a constant output voltage, usually given as 10v. However I think they will result in *less consistent* readings than the factory stabiliser. The gauges have a heating wire wrapped round a bi-metallic strip and the amount the pointer moves depends on the heat applied to the strip. But the heat comes from two sources - the heating wire and ambient temperature, which can vary from 0F to 140F and probably more. Solid-state stabiliers are usually based on voltage regulators, which are designed to give a constant voltage output under a wide range of temperature conditions, which means the gauge will read higher in high ambients and lower in low ambients, for the same volume of fuel. However the factory stabiliser also consists of a heating wire wrapped round a bi-metallic strip which is subject to the same ambient variations as the gauge. In high ambients that will make the gauge read higher, it will also make the stabiliser 'switch off' sooner, which will tend to make the gauge read lower, so compensating for variations in ambient temperature. For whatever reason, my Celica with solid-state instrument regulator would regularly indicate one fuel level when I arrived at a destination on a warm afternoon, then next morning when cooler would indicate a gallon or more less!
Paul Hunt 2

I have the feeling, especially with Paul's observation, that I really need to check out the Voltage Stabilizer. Now, where is it physically located?


Larry Hallanger

Larry, It's on the firewall just above the pedals.

Clifton Gordon


the voltage stabilizer is as Clifton says, on the firewall just above the pedals. But INSIDE THE INTERIOR, UNDER THE DASH. It is somewhat difficult to reach. Being a man of size I like to pull the driver's seat out before I lay on my back to retrieve it!

Dave Braun

Thanks Clifton and Dave

Will go "floor diving" this weekend and try to measure the output voltage.

I have plotted the temperature-resistance data for the sensor from 3 sources and will post on the "Temperature Gauge / Sensor Problem" thread.


Larry Hallanger

This thread was discussed between 30/08/2007 and 12/09/2007

MG MGB Technical index

This thread is from the archive. The Live MG MGB Technical BBS is active now.