Incorrect Temperature Guage: a fair assumption?

[Guest]

Hello Folks,

One more question. I had to have some head work done on my car last week, as I had overheated it and fried two valve guides. (I was on the way to the shop to have my cooling system modified.) It is fine now, and the temperature is staying down with my upgrades (bypass filled and drilled to 1/4", upgraded water pump, and 160º thermostat).

I have a Celsius temp guage. Like all good Americans, I HATE Celsius. As you know, the lines on the Farenheit guage at 185 and 230 (line in between is 210) equate to the 85 and 110 on the celsius guage (line in between is 98). Since Ed did head work, he used a probe to see what the temperature ACTUALLY was, compared to the guage. The guage read 85, which is 185 Farenheit, and the probe showed 150º!!!! So, I was actually running 35 DEGREES COOLER than I thought I was.

Here is my question to you. Can I assume that if the guage is 35 degrees too hot at 85 degrees that it will be 35 degrees too hot the entire range of the guaage? If this is a valid assumption, I can put some label tape on my dashboard that says "150 - 175 - 195." What do you think?

Muchas gracias.

 

[TR3driver]

1) I don't think you should make that assumption. Depends entirely on why the gauge is reading high.

1a) Was the probe in the thermostat housing (where the gauge probe is), or in the radiator? The radiator should be much cooler ...

2) Is your gauge mechanical or electrical?

 

[tdskip]

I'd be careful about making that assumption. You could check various points of the engine/cooling system with an infrared thermometer and see if the gauge scales linearly, or just get a new known to be good gauge.

 

[angelfj1]

Hello Folks,

One more question. I had to have some head work done on my car last week, as I had overheated it and fried two valve guides. (I was on the way to the shop to have my cooling system modified.) It is fine now, and the temperature is staying down with my upgrades (bypass filled and drilled to 1/4", upgraded water pump, and 160º thermostat).

I have a Celsius temp guage. Like all good Americans, I HATE Celsius. As you know, the lines on the Farenheit guage at 185 and 230 (line in between is 210) equate to the 85 and 110 on the celsius guage (line in between is 98). Since Ed did head work, he used a probe to see what the temperature ACTUALLY was, compared to the guage. The guage read 85, which is 185 Farenheit, and the probe showed 150º!!!! So, I was actually running 35 DEGREES COOLER than I thought I was.

Here is my question to you. Can I assume that if the guage is 35 degrees too hot at 85 degrees that it will be 35 degrees too hot the entire range of the guaage? If this is a valid assumption, I can put some label tape on my dashboard that says "150 - 175 - 195." What do you think?

Muchas gracias.

De nada!

Jorge: I may be wrong, but didn't TR3B's have electrical temp. senders and gauges as opposed to the capillary tube type of earlier cars? The electrical temp gauges require a regulated source of dc voltage and Triumph used a device known as a <span style="font-style: italic"><span style="font-weight: bold">voltage stabilizer</span></span>. This was a crude type of voltage regulator which maintained approx. 10 volts for use by temp and fuel gauges.

Further details here: Smiths Voltage Stabilisers

If you have this type of system and the stabilizer is missing or not functional, your temp reading will vary with the voltage available at the temp gauge, e.g. for 12 volts, the temp would read 20 percent high. Check this out. It may be the answer.

 

[tomgt6]

If your voltage regulator is working then I wouldn't assume the 35 degrees. I find most of the time is works as a % off. So at 130 it is 10% off (13 degrees off) and at 180 it is still 10% off (18 degrees off) but the number is bigger so you have to be careful because this is how it is most likely working.

 

[TR3driver]

Just for clarity, the original "voltage stabilizers" do not supply anything that even faintly resembles 10 volts. What they do supply is either 0 volts, or full battery voltage; and they switch between those two extremes to make the gauge read as though the supply was a regulated 10v. If you hook a test light or voltmeter into the circuit (eg across the sender terminal and ground), you can easily see it flash on and off (if the VS is working).

However, my limited research indicates that TR3Bs were not fitted with electrical gauges or voltage stabilizers by the factory; even though the TR4s built before the TR3B did have. If George does have an electrical gauge, IMO it is because someone replaced it with the 'wrong' gauge; making it even more likely that the VS is not present.

Also worth noting that the VS must be grounded through it's mounting point for it to work properly. The gauge itself doesn't need a ground, but the VS does.

 

[TR3driver]

Oh, one other point : If the problem is a defective, ungrounded or missing VS, how high the gauge reads will depend on the battery's state of charge. You'll get things like the gauge reading going down while idling at a stop light, even though the temperature is actually going up.

 

[6TTR3A]

Here's an easy way to tell whether it's a capillary type or an electrical one.

Method 1 = Compare it to the oil pressure gauge. (Assuming the oil is a Smiths
Jaeger): The needles and the hole through which the needle protrudes should be EXACTLY the same width, length and size respectively. If the needle is wider on the temp gauge it is electrical. Also, the hole is much larger on the electrical gauge.
Method 2. = Take a close look at the spot on the thermostat housing where the
temperature sending unit enters the housing. On the electrical unit there is a male spade connector on the sending unit, sometimes covered with a rubber boot to
hide the fact that it is not a capillary tube.
If yours proves to be a capillary type, they can be calibrated but it requires
removing the gauge from the vehicle and sending it off to a professional.
Hope this helps,
FRank

 

[Guest]

Thanks, guys. I just talked to a engineer and he confirmed what you said, "Never trust ANY guage toward the ENDS of the dial. The middle is where it is most accurate.

So, I will belabor along with the 185 mark actually being 150, but from there, I will wing it. O lord...

 

[angelfj1]

Just for clarity, the original "voltage stabilizers" do not supply anything that even faintly resembles 10 volts. What they do supply is either 0 volts, or full battery voltage; and they switch between those two extremes to make the gauge read as though the supply was a regulated 10v. If you hook a test light or voltmeter into the circuit (eg across the sender terminal and ground), you can easily see it flash on and off (if the VS is working

Not really. It's just a matter of semantics. In signal theory we often refer to the average value and dc value of any complex waveshape, the assumption here that this is a waveform whose characteristics are periodic or repeating. So average value and dc are the same thing. So the guys at Smiths should be lauded for producing a very simple and cheap way of getting a regulated 10 volts from more or less 12 volts. Of course thay had to do this without the benefit of zener diodes or regulator IC's.

I recall seeing an oscillograph of the input and output of a voltage stabiliser. Unfortunately I cant find a copy so I have drawn the following from memory.

A voltage stabiliser with these characteristics is what I recall from the magazine article. With a 12 volt input, it produces 10 volts.

Assuming the input is approximately 12 volts from the battery, a voltage stabilizer would produce an average 10 volts if it turned on and stayed on for 1 sec and turned off for 0.2 secs. Very simple and very easy and cheap.

I intend to try to duplicate this in my lab/shop.

 

[TR3driver]

Not really. It's just a matter of semantics. In signal theory we often refer to the average value and dc value of any complex waveshape, the assumption here that this is a waveform whose characteristics are periodic or repeating. So average value and dc are the same thing.

Yeah, signal processing folks are kind of weird ... dunno how they manage to talk without knowing the Fourier transform of the English language.

But most folks would not look at a test lamp flashing every second and say it was "half lit" even if the duty cycle was 50%.

<div class="ubbcode-block"><div class="ubbcode-header">Quote:]I intend to try to duplicate this in my lab/shop. [/QUOTE]Why?

 

[angelfj1]

<div class="ubbcode-block"><div class="ubbcode-header">Quote:]I intend to try to duplicate this in my lab/shop. [/QUOTE]Why? [/quote]

<span style="font-style: italic"><span style="font-weight: bold">Just because it's fun and being retired I have the time.
</span></span>
Incidentally, I agree with your comment concerning signal processing types. Not my niche at all. I have spent most of my career in the industrial and commercial power systems arena, you know IEEE IAS and PES. My forte is power systems design for co-generation. Have been doing lots of work with CAT-Solar, down in San Diego. However this little exercise was so basic I had to explain that the average voltage produced by that 50 cent stabiliser is based on very sound theory. Although something tells me that your not convinced.

 

[angelfj1]

really, so easy a caveman can do it!

 

[TR3driver]

However this little exercise was so basic I had to explain that the average voltage produced by that 50 cent stabiliser is based on very sound theory. Although something tells me that your not convinced.

Oh, no, it's not that. I understand very well that the average voltage is 10 volts, and even tried (in my own way) to explain that above. It's just that IMO, most people are not accustomed to talking about a long term average when dealing with a "voltage". If you say "it's 10 volts", they expect to be able to connect a meter to it and read 10 volts. After all, what they really want to know is "Does this part work?", preferably without tensor calculus

I do have a 'scope in the garage, so if I really wanted to, I could just hook it up to the Stag that is also in the garage. Turn the key on, set the sweep to maybe .5 seconds/div, take a photo. But it doesn't seem worth the effort to me; so I'll be happy to let you do it.

 

[Deleted member 8987]

Most IVR/CVR or "stabilizers" are nothing much different than a turn signal flasher.

Fords, Dumb Cooks (Stupid Bakers), Willys, all sorts of earl cars, used "King-Seeley" type gauges, slow-acting.
If you watch closely, you could see the needles waver ever so slightly.
Yeah, the average was 6-10 volts (depending on what vintage).
General Misunderstandings, not so.
FAST acting, needed gauges match to vehicle voltage or you got more smoke than a Lucas system.

If you have electric fuel, temp and oil, and the others are known to be close, probably not the stabilizer.