Thermostat white paper

E-Type
41

You’re way ahead of my writeup, but I’ll give you some previews. It’s not that I’m saying you’re wrong, but you’ll have to defend.

Around 1962 or 1963, Jaguar finally abandoned the bellows thermostat and went with a reverse poppet wax thermostat from Western Thompson, C20766 (74C). This thermostat had a stepped cap that allowed the sleeve to tuck into the thermostat cover. The sleeve telescoped into the cap, so that it had plenty of clearance for the slot, even given that it had a “long” sleeve. You’ve probably never seen one of these, so here it what they looked like, closed and open:


This thermostat should substitute for any sleeved bellows thermostat. Since it’s a reverse poppet wax thermostat, it’s possible to use better water pumps and high pressure caps. The only problem is that Western seems to have dropped production fifty years ago. You’ll never get closer to one than these photos, unless someone comes along and starts making them again ;-).

C3731/a as currently offered by several vendors is an attempt at replicating the functionality of this thermostat. It’s actually cut down from a Land Rover series IIa thermostat, which for some reason is abundantly available. The LR thermostat is very similar to C20766, except that the cap isn’t “necked” so it won’t fit into a Jag housing without modification. C3731/1 cuts 3/8" out of both cap and sleeve. Although the remaining sleeve is narrow, it should still be wide enough to block any slotted bypass. Again, it doesn’t matter that it moves backwards. If you have actual measurements that indicate that it’s mismatched to the port on pre-150 cars, I’d like to see your work, and I’ll include that in my writeup. Here are photos of C3731/1 open and closed, as well as the donor landy stat:



It’s not a perfect solution, but it’s not awful. There ARE two issues with C3731:

A) The landrover has a narrower flow path than C20766.

B) Since these skirts are made by a drawn cup method, they taper slightly. The part that’s cut off is 48.5mm, the part that remains is 47mm. So it’s far from a perfect seal, but the design is weak in theory, and can never really seal anyway.

You’ll not that these photos look nothing like the photo of 3731/1 on Bob’s page. This part had been undergoing development for many years, and there are a few different versions. The Robertshaw adaptation he speaks of hasn’t been available for many years, the current version is based on the Land Rover stat. The version sold today is probably the zenith of this concept, any improvement will require a completely new design.

For early cars, the Moss Motors bellows thermostat is probably a better bet, and certainly less controversial.

Here’s another tidbit: it’s really not all that important whether or not the bypass seals. I’ve spent six months visiting shops and measuring any manifolds I could find. Not one of the manifolds I’ve examined had a sleeve thermostat. Not a bellows thermostat, not C20766, not C3731. They’ve all been using single poppet thermostats, probably for decades. This works just like any other open bypass system: Newton rules, and the coolant flows straight through the thermostat rather than making a hard right turn into the bypass. The 4.2 manifold is especially well suited to single poppet thermostats, due to it’s very straight flow paths. Earlier manifolds placed the thermostat at right angles of flow to reduce lifting.

Finally, there’s no “minimal flow thru the closed thermostat to the radiator flow path to aid in a more rapid opening”. If the thermostat is in good shape, it seals tightly when closed. Bypass flow bootstraps opening.

1 Like
42

Very impressive, science-based work, Mike… keep’em coming!

43

to Michael…thank you for your work…Exactly the continued information to help clear up the “Thermostat Enigma” , which is for most of us, simply comes to…“what thermostat do I put in there ?”. .Would it be correct to say, that in an otherwise efficient engine/cooling system the operation temp of the engine coolant should be the spec temp of the thermostat plus or minus a few degrees as it cycles, noting the coolant will be cooler as it leaves the radiator and warmer as it leaves the engine…thus the cycling of the thermostat as it battles its enemy from both sides… coolant too cool…coolant too hot. Th XK120 had hot day and traffic cooling issues from the start and this is well noted in the articles and road test reports of the day. Why Jaguar did not make any changes until the XK140 is a mystery. With this issue, with the exception of quite cool days, it is likely the XK120 thermostats opened…and stayed open, with bypass closed, the radiator being unable to cool to a temp lower than the thermostat specs,

44

Just took a fresh look at the thermostat issue by measuring/observing the parts I have: 140 manifold, 150/E-type/MK2 manifold, C3731/1 as supplied as a replacement for original. On the 140, the bypass slot lies only .225" below the thermostat mounting face and is about .3" wide.

However, the later cars (HD carb set-up) had a bypass slot which began about .525" below the mounting surface and is only about .125" wide.

20190929_085913
20190929_085913.jpg4032×3024 2.78 MB

The C3731/1 moves about .48" to full open, so clearly it allows the bypass slot to be open on the 140 manifold, yet closed on the later manifold. The diameter of the replacement 'stat is also an issue as it is fully 3mm smaller than the opening, 47 vs 50mm. I wish I had a 120 set-up to evaluate but maybe some other lister could take the time to make notes on that set-up with respect to the suitability of the C3731 or C3731/1 on the 120.
20190929_085415
20190929_085415.jpg1512×2016 885 KB
[20190929_085716|375x500](uplo
DSCF0159
DSCF0159.JPG1920×1080 513 KB
DSCF0163
DSCF0163.JPG1920×1080 525 KB
C37311%20open
C37311 open.jpg1512×2016 806 KB

45

Ok I’m a little slow. You’re saying that the thermostat will work for the later manifolds because it does shut off the bypass when warm, but won’t on the 140 because the bypass will be open as the sleeve has moved all the way past it?

46

Bingo!, although it is very close on the later manifold since the sleeve is just about flush with the slot when fully open. I do believe the biggest problem is the rather small diameter of the sleeve vis a vis the thermostat cavity…3mm of difference there.

47

I just wonder about the use of the Jaguar part number…C3731 to descibe any particular thermostat…as in the day…that meant a Smiths bellows/sleeve with one of the applicalble Smiths part numbers…most common maybe X85025/ xx. But now…when one uses that part no…what is the actual thermostat…as said prior…any vendor can call it a C3731…but what is it really? Is it this Smiths? None are…except old stock real Smiths. So they are something else… Snut=: smiths%20thermostat

48

That’s good info. I’ve measured a lot of manifolds over the last few months. And until your 140 measurement, I would have said that all OE manufacturers cut the slot between 3/8" and 3/4" depth. So a 3/8" wide sleeve should be pretty much a universal fit. I’ve found some aftermarket manifolds set up like your 140, but haven’t considered that this was anything other than a dumb error. And in fact, after some consideration, I think this is also and error, made by Jaguar. Here’s why:

Each of the bellows thermostats in my collection has a hard stop which limits the travel of the sleeve. This makes sense, as you wouldn’t want the sleeve to travel past the “sweet spot”. It’s hard to be absolutely sure from just cold measurements (and I’m not quick enough to measure them hot), but I’m pretty sure that each of them limits the sleeve at about 3/8" from the flange. Jaguar’s “universal” C20766 is a reverse sleeve that begins at a depth of .382", ready to move into the sweet spot. So that wouldn’t work either.

I’m wondering if one of the thermostats in Bob’s collection has a different stop height. It’s pretty strange if you think about it. There were maybe half a dozen standard thermostats in the Smith’s line at the time. To place the slot higher, Jaguar would have had to commission a special thermostat just for the 140. For what purpose? It would be the same diameter and depth, close at the same temperature, and would fit in the same place in all other respects. My feeling is that whether this was a one-off machining error or an error in specification, it was still an unintentional blunder. Either way, there’s no thermostat in my collection that would work, and I have a whole lot of thermostats here. What might work for you is to use a Moss and file a little off the end of each stop. It would take some trial and error to get it right.

49

Just to make sure there is no confusion here, the thermostat I have (bought from one of the usuals) is supposed to be a correct replacement for a factory unit and is sold as part # C3731/1…it is NOT the bellows type AND the sleeve does not move toward the direction of water flow. I can see the stop(s) you are talking about as preventing the sleeve from getting within .225" of the stat’s mounting surface…rendering it ineffective for the 140.
What I also find curious is the tech article mentions a non-heater thermo as different from a heater-equipped car. Was the owner expected to change thermostats with the season? After all, there is a provision for turning off the flow of hot water to the heater core, thereby rendering the car a non-heater version. What could possibly be different between the two types based on “heater”, or non heater"? Did the “heater” version thermostat have a restriction built-in (lower lift) to ensure water was forced through the heater core, reducing flow that would otherwise go to the radiator?

50

Good! Because I had already ordered one from SNG this morning having used a standard thermostat all these years.

51

Moss Europe has what they call a bellows/sleeve thermostat…but they do not list it as for early Jaguar…or any Jaguar… I do not know dimensions. XKs lists the same part # Superstat they have listed for years…090160…with photo on their website.but now also with part C3731/1…well…when Jaguar used that part number it was a Smiths bellows sleeve type, and now from any of the parts vendors it is most certainly not a Smiths type, tho they do have a sleeve, and the sleeve does move to close the bypass.090160%20therm%20superstat The thermostat for the xk120 does not install in the intake manifold…it is in the inlet of the radiator.
Nick

52

That article by Bobine shows a bellows-type right next to a 140 manifold…most of the way toward the bottom of the page. The article also mentions the 150 and up part as a completely different Jaguar part # from earlier models…this makes perfect sense since the intake system was wholely new after the 140. Looks like the C3731/1 as is easily available is just fine for the 150 cars and up, and MAYBE for the 120(need verification), but NOT for the 140. And the opening temp appears to be the only difference between the C3731 and C3731/1…non-heater, heater.

53

Beyond that, since there is a hard stop on those bellows-type, what happens as the temp rises above whatever temp it took to fully open the stat? Does the increasing internal pressure cause the bellows to bulge, or break?

54

On the XK120 thermostat housing, the bypass port also lies about .225" below the mounting face. However, the bypass is circular measuring about 1/2 inch by 1/2 inch.

T-Stat%20(1)
T-Stat%20(1)1230×923 418 KB

The first photo is the XK120 thermostat housing and a Smiths X85025/80 bellows thermostat with bypass sleeve.

T-Stat%20(2)
T-Stat%20(2)1447×1446 504 KB

The second photo is looking into the bypass port with the thermostat in place. At the bottom of the opening you can just see the top 1/16 inch of the bypass sleeve.

T-Stat%20(3)
T-Stat%20(3)612×816 173 KB

The third photo shows the hard stop of the thermostat at the top of the bypass port. Because the bypass port is 1/2 inch tall and the sleeve moves only 3/8 inch, the bypass port does not fully close when the Smiths bellows thermostat fully opens.

55

Good info…of course, the round opening means the total area available for water flow is relatively small even with the resulting 1/8" gap. What is the diameter of the sleeve vs the housing opening?

56

I think it not productive to use the Jaguar part number…the C3731…althouh the vendors have assined it to the later type thermostats they now have, as it does not really provide info as to what thermostat is in hand. If it is a Smiths…cite #, same for AC or QH or other. the stat would be old, old stock. None of those are currently made. It remains to be seen if the Moss Europe bellows sleeve for the Triumph…has measurements that will fit the XK Jaguars. So the vendors use the C # but provide whatever thermostat. Bob K states: for the Smiths bellows/sleeve appropriate for the XK: the sleeve lifts 10mm, (Jaguar said 9.5mm) it is OD48 and 15mm in height (I measured 48.3) . ,…the thermostat is ID 50mm and OD at top is 54mm. The thermostat can be damaged by pressures much more than 4lb…All the XKs had/should have a 4lb radiator cap. That pressure in a sound system may never be reached…the expansion of water…if the level is at normal op level…and not fill to the neck will not create the pressure…but it does if overfilled…there is a SB about that.,. In a sound system, if you fill to the neck, then take a run, whatever water is expelled via the overflow…which comes into play when the 4lb cap is overcome, is normal, so you then have the normal operation coolant level. (There is a TSB on this). With a 4lb cap there is virtually no provision for coolant retention over the un-pressurized boil point. It is not lineal…but you can calculate a 3 Degree C lower boil temp per 1600 ft of elevation, or for each psi of pressure. .50 50 antifreeze has a boil point of 223F. A 4 PSI cap will provide approx 226F boil point of 50 50 antifreeze…215F for water, at sea level. As you know a boil makes steam and steam will find a way to expand, the 4lb cap is the safe to engine way out… Nick

57

Here are the measurements I took of the manifold and three thermostats I have.
In the right diagram it shows the slot in the manifold measured from the thermostat seat.

image
image.jpg603×610 89.9 KB

The left diagram shows the thermostat in the closed position, with sleeve measurements from the mounting flange.

Below that are the findings of each of the three units when fully open.

Note the position of measurement B in relationship to the right edge of the slot.

Unit one is right there and the other two have gone past by .04 and .09

Given that the edge of the sleeve is tapered/rounded, I think there is a lot of room for hot coolant to continue to flow.

image
image.jpg579×592 103 KB

image

58

Whew. That’s a lot of questions. So here are my answers in no particular order:

  1. The Moss pn 434-156 in my illustration is listed for MG, but should work on any British manifold that’s set up for a bellows sleeved thermostat. It has a 2 1/8" flange, the sleeve is 48.5mm. It WILL NOT work in a 4.2, because the 4.2 manifolds were deliberately machined not to accept bellows stats of any kind. It won’t work perfectly on your 140, but I’m not sure that there are any thermostats that would. And it looks like 120’s had the same problem

  2. I’ve always been under the impression that C3171/1 was a non-Jag part number made up by SNG,or maybe long ago by British Auto USA. The old version was built up from a Robertshaw, by adding an outer ring. The new version is cut down from a Land Rover Series IIa thermostat, as noted. The surprise for me is that there’s a bellows thermostat with the same part number. The SNG part should work on any sleeve bypass application, including 4.2’s, but will be a loose fit due to the limitations of the approach, as noted above. Your early XK’s seem to be an exception. Again, I don’t know of any thermostat that will work correctly for these applications, new or old.

2a) This is what the older version of C3171/1 looked like. It was based on a Robertshaw thermostat. the sleeve was soldered to the sides of the standard thermostat’s shutter:

This version was a little better in that the sleeve was a full 48.5mm, but the flow characteristics of the native thermostat were compromised by the mod. I’ll be doing a long writeup on Robershaw thermostats as part of this series, just for completeness and because it’s a really interesting design. They really have no Jaguar application.

  1. The bellows hits the hard stop near the end of its range. If you recall what I said in the very first article, almost all the expansion occurs as the working fluid transitions from liquid to gas. After transition, there will be relatively little expansion, but at the stop, internal pressure rises in the bellows. The bellows itself is made of bronze, and has a spring temper which allows it to absorb some “discomfort”. I suppose if it became superheated, it would split open. The good news is that the bellows is a normally open spring: the only reason it ever closes is because a vacuum forms as the alcohol fill cools. So if it splits, it will fail open. The thermostats I’m showing you aren’t unusual in having the stop, this is the way all bellows thermostats are made.

  2. Some early cars had inline heaters circuits, where the heater was controlled by a manual valve on a hose mounted housing. I’m not aware of any Jaguar that had this arrangement, but I’m interested if you turn anything up.

  3. Winter/summer. Hotter thermostats were used in winter because return water temperature depends on ambient air temperature. When the weather is very cold, the thermal shock of the return water can affect performance and driveability. In the old days, there was a LOT of maintenance involved with changing coolants and thermostats with the seasons. Some folks would use a winterfront or simply put a piece of cardboard in front of the radiator rather than go through the trouble of semi-annual drains and refills. After the industry moved to wax thermostats, this changed because higher temperatures/pressures were possible. When I publish the chapter on bypass systems, you’ll see some ways this is addressed in modern applications.

  4. On a related topic, early ethanol antifreeze solutions required lower thermostat temperatures, to prevent the alcohol from boiling off. So the winter temps were actually lower.

  5. As Lee’s photo shows, bypass control probably wasn’t very accurate. One mitigating factor for the 120 setup is that the top of the sleeve of an 85025 is slightly dished, so a bit more of the bypass may be covered, but it still won’t quite close. The diameter of the sleeves on these old bellows stats is 48-48.5mm. The diameter of the main passage is 50mm. So there’s always a bit of room for leakage. The c3731/1’s sleeve is about 47.5mm, since it’s cut back from a tapered cup. So a little more room for leakage. Still, I wouldn’t hesitate to use it unless you have a better alternative.

  6. I’d appreciate being able to use the photos of the 140 and 120 rails that you’ve posted here. I won’t do it without your permission.

  7. A bit more on open bypass. Contrary to a lot of the useless, intuitively pleasing crap you read on the web, an automotive water pump doesn’t create pressure, or at least, not in the way you may think it does. What the pump does is transmit kinetic energy from the belt system to the water jacket. And this creates flow. Flow is a vector, it’s described by magnitude and direction. Pressure in relation to pump output is what you would feel if you try to stop this flow. So if you hold your hand in front of the discharge port, you would sense the force.You know what? You’d likely feel about the same force if you tried to hold the pump pulley with your bare hands (don’t try this at home). Let’s call this dynamic pressure. But this is distinct from static pressure, which is an omnidirectional force associated with gravity and fluid expansion. Let’s say that pump flow is directed through a straight, frictionless passage and we insert two different pressure probes. One probe is simply a tube place at right angles to flow. The second tube has a bend so that it points directly into the flow path. This is what you would be able to sense in terms of pressure:

At port A, you can only sense static pressure. This is because the port is orthogonal to the flow vector. At port B, you sense the total of static and dynamic pressure. Pd is the difference between static and total pressure, and is the dynamic force associated with flow. Dynamic pressure is what pushes the coolant through the open thermostat and into the radiator.

Now here’s the fizzicks part: static pressure is very nearly uniform throughout the cooling system. So if the bypass hose is at right angles to both the thermostat housing and the pump inlet, it’s exposed to the same static pressure at both ends. If the flow path is unobstructed, open thermostat, there’s no pressure differential pushing coolant through the bypass. Supposing we place a closed thermostat at the end of the flow path. Energy is conserved. So the kinetic flow energy becomes potential energy, which is to say, static pressure increases. But it finds relief through the bypass port, and the redirected flow vector carries coolant back to the pump.

If you use a single poppet thermostat in a Jag manifold, there will only be flow through the bypass to the extent that the thermostat and radiator create a restriction in the flow path. So the only flow that goes around the block is flow that would be wasted anyway. What happens if you have a proper sleeved thermostat? Static pressure would rise locally because even a high flow thermostat is restrictive. This would raise head pressure, which does help control boiling in the head, which is good. But some of that excess energy forces coolant to leak through the imperfect bypass seal. And some of that energy goes to raising engine temps. And of course, that dang bellows is pressure sensitive, so if local pressure ever gets too high, it may just close up on you. So between a single poppet thermostat and a proper sleeved thermostat, it’s hard to say which is better. And that’s why so many Jag engines are puttering around with Chevy thermostats and uncontrolled bypasses.

Not as easy as you thought? Now you know why it’s an e-book.

59

Michael, feel free to use my photos of the XK120 thermostat housing. Let me know if you need additional photos. (I also have an XK140 intake manifold with the built in thermostat housing.)

I agree with your measurements under # 7. I measure the diameter of the sleeve on my Smiths X85025/80 thermostat at about 48.3 mm. The diameter of the main passage of the XK120 thermostat housing is about 50.3 mm. Thus, there is about 1 mm clearance between the sleeve and the wall of the thermostat housing.

60

I never knew I knew so damn little about thermostats!

:+1:t2: