S2 XJ6 engine, gauge shows overheat

The S2 XJ6 motor in my 420G, coolant temp gauge showed “max” after a 10min drive to get petrol

It did not show any signs of coolant boiling

An IR gauge showed RH side of engine and rad, hose, thermo housing, manifold at ~70C

LH rad lower at ~45C lower, upper at 70, BUT LH cam cover showing higher at up to 100C

I decided not to let the motor run and see if the temp of the coolant or RH cam side rises,
rather that I have one of 2 problems;

  1. A failed sender, it measured ~40 Ohms hot…anyone know the specified “range” ?

  2. thermostat failed, not opening

I am thinking of firing the parts canon and replacing both these items, but the sender isnt immediately available, although I get an old one out of a S1 manifold…thread isnt all that good though, took 3/4" drive gear to get it out, still screws tightly back into the hole I took it out of

I will probably put it in anyway tommorrow

Thinking as I write, should just bite the bullet, pull and test the thermostat first ? (I can cut new gaskets), I had one do this once before and the symptoms were similar, except I think the coolant did boil

all opinions welcome

I am not saying this is the problem in your case but I once redid my whole S2’s cooling system trying to get the temperature gauge to read in the normal range. The problem on mine was not the gauge or anything in the cooling system. My problem was a failing voltage regulator behind the dash that modulates the power getting to the temp gauge. It had failed and so the gauge was reading higher than it should.

David

1 Like

Try here for test results on the temp sender.

https://forums.jag-lovers.com/t/xj-temperature-measurements/195957

David

I would think that the cam cover would show a little lower than oil temperature as the cams will splash them with hot oil. If the water rail is at 70° (did you measure at a non-reflective area?) and the radiator gets hot as well, look at two things: coolant level and the sender.
With low coolant I noticed that the gauge will suddenly shoot up in temperature when driving downwards, and then suddenly cool again. Maybe there’s some sort of bubble forming at the sender.

You probably have a spare sender or gauge somewhere?

thats handy…at 40 Ohms measured for my sender in manifold, with coolant full and not boiling, that would indicate a bad sender (from your figures)

I might try swapping that out with the S1 item I have, which measured 165 Ohms at air temp, a little easier than pulling the thermostat

Tony,

with water temp needle at max and manifold at 70°C something’s wrong with the indicator, not with the coolant system. Even 100°C at the LH cam cover don’t seem exaggerated to me. Do you use the SII standard heat shield over the exhaust manifold?

I would leave the t-stat alone for the moment. If your rad heats up to 70°C upper and 45°C lower the t-stat obviously opens the full radiator water circuit.

David’s idea about the voltage stabilizer is good, but I’d guess that your fuel guage is powered through the same voltage stabilizer, so your fuel guage should rise in line with the temp guage.

The sender is the cheapest and most probable solution - I’d start there. It is a standard part - probably the same as in a Spitfire or MG - and will run around 10 EUR.

Good luck

Jochen

75 XJ6L 4.2 auto (UK spec)

Hopefully able to report a happy ending.

I screwed in the S1 XJ sender I had, after Ohming them out at air temp ~ 28C one was half the value.

I gave it a longer road test than the day before, and it behaved as I expect, rising up to “M” in NORMAL, and stabilising there, not rising while idle

The spare sender did not have very nice looking threads, I used nickel anti-sieze, and screwed in in not too tight. No risk of it coming out though. Bi-metallic corrosion is the issue

Getting the Thermostat out looks like quite a nasty job, so I will be glad if I dont have to do that

I repeated the IR tests and it was much the same story, albeit a bit hotter, as I ran it twice as long. The LH cam cover appears to be hotter due to exhaust manifold radiant heat

I am having the same issue with my Series II. It all started when I flushed the coolant system and replaced the thermostat because the engine never got up to temp. which is a real PITA in the winter months. So I snapped off some bolts, practice my skills at using easy outs. Replaced the thermostat, coolant reservoir cap& and the coolant filler cap. Fired it up and was tickled pink I finally got some heat!
then come summer time, the temp guage registered just past “Normal” which isn’t normal. busted out the thermal temp gage and aeverything is in spec with the execption of the temp guage. I figured it was an issue with the after market cheap Uro-parts temp guage. I ordered one from SNG and got the same results. It didn’t really bother me since I knew the engine temp is as it should be. I have bigger issues to take care of.

I did talk with a Jag specialist and he said it has to do with the new temp senders don’t have enough resistance to register correctly. He suggested to install a resister between the sending unit and the gauge in order to get the gauge to read correctly. I’m not real hip on this idea. I haven’t tackled this one yet. I have other projects.

Keep us posted on your findings.

Cheers!
Mark

**
The IR gauge is correct, Tony; 70C at the water rail means your coolant temp is indeed 70C and the engine is not overheating. As the gauge reading is taken at the sender; the temps should be read at the base of the sender - for comparisons with the gauge. As David says; a failed voltage regulator may be the cause of the gauge misreading - but the temps sender may be faulty, or the wrong type; unsuited to the gauge.

Someone may know the relevant resistance for the gauge type. With a variable resistor you may check out what resistance gives a gauge reading relevant to IR temp measured?

Arguably, 70C is on the low side, and may imply a faulty thermostat - a separate problem from the gauge reading. Nominally a failed thermostat may cause overheating - but will certainly show up as extended time for the engine to warm up; 5 minutes of driving should give thermostat opening temp…

Higher temps on metal parts not cooled by coolant is perfectly normal…

Frank
xj6 895 Sov Europe (UK/NZ)
**

1 Like

After a 5-10min drive, not even fully warmed up, more like 80C today.

The thing that pointed to not the thermostat was that the rail, thermo housing, coolant casting, rad hose and upper RH rad were all at approx same temp, so couldnt logically be the thermo stuck shut or you would expect a differential in temp.

Btw, this sensor is Whitworth size…(same as 18mm)

I know Ray was working on a set of very excellent gauges

I would make mention digital WP gauge setup is available, not expensive., am very satisfied with my digital Evo Oil pressure gauge

**
Indeed, Tony - a stuck thermostat would typically cause temperatures to vary with load…

However, the thermostat only controls the temperature within the block/head/water rail - it exerts no direct control of temps in the hoses and radiator. With the thermostat closed; coolant is recirculated through the block to rapidly increase engine temps to the thermostat spec. Typically in some 5 minutes of driving - though slightly more with low initial block temps…

The radiator and hoses remains at ambient temps until the thermostat opens and the radiator/hose temps starts to rise. It takes some time for all the coolant to heat up, but with the thermostat fully open, engine warmed up, one would expect the upper radiator/hose to show rail temps.

The difference in temp between the top and bottom hose reflects the cooling capacity of the radiator - which naturally will vary with ambient air temp. However, a properly working thermostat will maintain (approximately) thermostat spec temps throughout. (Except; in very cold weather, the heater may cause enough cooling to prevent the engine from reaching operating temps).

A failed thermostat, usually ‘half’ open, may cause overheating in ‘hot’ weather due to restricted circulation,. And running cold in cold weather, excessive circulation for the load - with delayed warm-up in either case. The temp differences in the various parts of the system will be, sort of, constant, but the average temps will be higher or lower mainly dependent on ambient air temps and thermostat’s stuck position…

Evaluation the cooling system with an IR thermometer is a good start - but interpretations is not necessarily straight forward…:slight_smile:

Frank
xj6 85 Sov Europe (UK/NZ)
**

Jaguar itself did exactly that on at least some Series III XJ6s. The resistor is very clearly shown in the parts catalogs. Discussions 15-20 years ago failed to reached any conclusions as to which cars got the resistor, bringing me to believe it was done only on an as-needed basis.

Questionable sending units are not exclusive to old Jags. I work on lots of old Corvettes and muscle cars and often come across incorrect temp readings on the gauge. The consensus is that modern day replacement sending units are not manufactured to the correct spec and/or are manufactured to be somewhat correct for a too-broad range of applications.

To correct the too-high gauge readings I’ve used a few of these…

…which are a nifty and easy (and over-priced!) solution. The twist-on wire nuts are a bodge, of course, and are tossed in the trash bin. Ugh.

Cheers
DD

1 Like

I used a similar wire nut to loosen the breakaway headed bolt securing my ignition switch.

Might have another use. making electrical connections, not one of them. although a larger version is commonly used in house wires…

Perhaps on the bench for test purposed temporary circuits??

Carl

Great for temporary use, yes. And great in house wiring…which isn’t subject to the constant vibration of automotive wiring

Cheers
DD

Tony,

that does sound like a happy end - of this episode that is;-)

Indeed, with the faulty sender identified you can start getting the
sender issues right - with all the hassle you describe regarding your SI
sender, you might wish to order that correct SII sender and replace it,
just for peace of mind.

Then, you’d probably have to get back to the t-stat issue. With a known
good water temp guage you can do serious test runs with +20mins of
decent driving at 3000 RPM and find out whether water temp changes. On
one hand the needle should stop somewhere at a quarter to twelve, on the
other water temp measured at the rail should be around 80°C. Do you run
an 82 or 88°C t-stat?

The operation of the t-stat is - as Frank has written - confirmed by a
period of approx 5 mins it takes for the water temp needle of the
driving car to reach its “usual” quarter to twelve position.

If the t-stat proves to be shot or wrong, because the engine keeps
running too cool, the swap is not that difficult if the engine has been
run with antifreeze. Otherwise at least one of the bolts that runs
through the water channel tends to corrode and snap upon removal. I used
some rust penetrant, let it work for an hour, and had no problems.

Beware of replacement t-stats: only the second new one I received really
operated at the desired - and imprinted - temperature.

Good luck

Jochen

75 XJ6L 4.2 auto (UK spec)

At least put a FAIL-SAFE 190F thermostat since your working on it.
This way if it over heats its not the thermostat.

I had one stay closed on me once, but it happened in the garage.
Walter

I dont know what thermostat is in it, as its a transplant engine, previously rebuilt by a pro shop.
In Australia, most people reccommend a lower temp Thermostat

Sounds like a good idea, but I’m not sure that it is specified when I look at what is for sale ?

I’m not a big thermostat changer, leaving them alone if they work, unless I am doing a major cooling system rebuild, on an older vehicle, dont always trust a new one to work

I will post a pic up later to ascertain what Parts need to be removed to get it out, seems to be a hidden bolt. This S2 engine was transplanted by me into a 420G and there is a mixture of parts

So to make a long and interesting thread summary: the high gauge reading was due to a faulty sender!?!
My 4,2 Mk10 has the same problem. Having tested the engine, rad and hoses with a digital IR device, I just ignore it, but will try another sender, as it is nice to have the gauge read correctly. (I also like the clocks to work, and thanks to Mike Eck, most of them do)

**
If the temp stays at some 70C, Tony - a 72C (190F) is likely fitted…

Your numbers otherwise indicates a cooling system in fine fettle - there is no need to avoid a 82/88C thermostat. With sufficient cooling; a 72C thermostat will run the engine at 72C - an 82C at 82C, and an 88C at 88C…

Basically; overheating is caused by insufficient cooling - and the thermostat cannot do anything about that. The given reason for using a 190F thermostat is that it temporarily slows the overheating - giving sort of a cool buffer for a brief time when engine is producing more heat.

The 190F is not a ‘failsafe’ - it cannot counter overheating. Past an engine temp of 190F this thermostat exerts no control of engine temps - temperature will vary according to load. As it will with any thermostat when cooling capacity is insufficient…

A characteristic of engine cooling is that the cooling effect increases with engine temps. And indeed, up to a point, an engine is more efficient and economic at higher temps; the cooler the engine the fatter the mixture is required…

If there is enough cooling at 82C, or for that matter 88C; such thermostats will be the right thing to use…

However; even with the general imprecision of the dash gauge; it presents a valid indication of changes(!) in engine temps. Which is a very important warning actor - if coolant is lost, rising temps is the first, vital, and possibly only, sign before the engine fails…:slight_smile:

For that reason; fix the gauge, one way or another, to give a meaningful reading. It’s more important than musing on your cooling system - which seems to be working well…

Frank
xj6 85 Sov Europe (UK/NZ)
**

I am going to try and do some work on the car, but the cooling system seems ok with the used sender.

The only thing worse than a failed sender is if the fault is real, only had a stuck on thermo once, and it would get very hot on the gauge and boil steam in almost no time.

dont intend to pull the thermo, lots of other jobs to do.

I made a new DS doorcard…took me hours…but it doesnt fit right, and there is no alternative but to start from scratch. Been putting off that job :frowning:

will give it a local test drive…when I did this yesterday, someone spoke to me, the car gets a lot of attention, however, this guy seemed like he might be abusing me…odd times