Stewart Warner "Wings" water temp inaccurate

I've been working with Brad on his problem with his water temp gauge. FIRST be sure you have a gauge/sender/voltage problem before trying to correct it. I replaced a lot of stuff before I checked the voltage at the gauge with a good (Fluke) DVM and it was slightly over 15v. As somebody else correctly pointed out the temp gauge is nothing more than a volt meter reading voltage thru a circuit varied by a resistor (sender) to ground. As the water temp goes up/down the resistance changes noted by the pre-calibrated meter movement.
If you change the circuit, such as wrong sender, or wrong input voltage your reading will be off.
I went so far as to buy another SW 100-240 gauge / sender and install it in the truck. They both read very close to each other, both high. I had this extra gauge cobbled up hanging under the dash which is why I checked the input voltage and noted it was over 15v. Checked my wiz bang internal reg one wire alt and yup, over 15v out. Note when first fired up at idle it was 14v but once you bring up the rpm the charging circuit kicks in and the voltage went up to 15. Yes, I could go buy another alt, this came from Napa, and assembled off shore probably, but it isn't in warranty. So I decided to spend about $5.00 to fix the problem.

BTW -to answer another question you cannot solder the copper conductor in the analog gauge, it's actually a capillary tube sending a varied pressure back to the gauge similar to a analog oil pressure gauge. The sender (bulb) cannot be separated from the gauge.

 

Mechanical temperature gauges work off a capillary tube that is connected to a bulb filled with Ether. The ether boils and build up pressure in a curved cylinder inside the gauge, this bends and causes the needle to move. It can be adjusted if you know what you are doing. I repair SW mech temp gauges starting at $100+ depending on how much work they need. Its tedious and time consuming work, you are working with a tube that has the opeing the size of a needle and you can't get anything in there to block it. One small hole and you lose all the ether.

 

Buy a mechanical gauge. I never checked one against any higher tech, but the ones I have used have all indicated normal engine running near thermostat temp and have only shown big numbers when things like steam coming out of the hood verified them...they always squared with what should have been going on and with my perceptions. And you can check them in the kitchen before you install them or even unroll the catheter to be sure.
If you add in some complexities to adjust on a gauge with what seems like thoroughly non-linerar response, i think you will just move the non-linear response around and correct at a single chosen temp...
Temp gauge is one of the few things really necessary on the dash...get one that works

 

Lots of educated advice has been provided for my question, and I appreciate the response. At this point I'm still going with BobF's repair/fix, even though I can see the gauge indication is "non-linear" so would probably only be correct at one specified temp, similar to a watch that has stopped. Its 100 % correct once a day....Maybe I am chasing rabbits but the idea of adding a diode and a resistor intriques me and I prefer to give that a shot. I can purchase a mech gauge, although I don't know if SW makes a "Wings" mech gauge (havn't checked yet). I too feel it is important that the temp gauge be fairly accurate. I know I'm not getting hot, but I also don;t like driving around with a 230-240 indication, even when I know its not a correct reading, so I'm going to try the bridge diode/resister theory, just for the hell of it. BobF said it worked for him, and I can tell from his post that his experience/knowledge with electronics is much greater than mine. Went by Radio Shack this morning and they weren't open so I'm heading back over now. I'll post any results, whatever they may be.
--------------------------

BOutlaw

 

As mentioned before, you get what you pay for. Look at certified aircraft parts.

My dad is an EE and I'm an aspiring ME, and its been covered pretty well here by a few people. If you want to try to fix the problem you can probly add a resistor since its reading high.

Spent all the time thinking about that and forgot to ask you one question...Is it a chip or is it an analog circuit? If you cannot get to the reference resistor in the "bridge", you may not be able to do anything.

Just to clarify, these senders are not good for measuring wide ranges of temperature and are designed to "work properly" at a given range say like 180-230 or so. They are also not linear so think of it as a power band... It is expensive to make an engine with lots power everywhere, and the same goes for these things.... They are made to give you a reading just like the one you are getting...

Honestly your time is worth more than trying to fix it in my opinion. The capillary tube ones are more fragile but the cost is worth piece of mind. Not trying to sound like a smart ass, I just want to help the tinkerers of the world..

 

Looks like others beat me too it! Im a "hunt and peck" typist.

As mentioned the adjustable resistor is a great idea and probly the most fun to play with. Also try using a regular acohol thermostat and foil/heat tape it to your outlet water hose.

 

This may have been said already.I ordered one the other day and just rechecked what I bought.Range 100 240 and the sender needed from Summit was same 100 to 240. Boiling water should be about in center of gauge 180 degs.Will Test when I get it.

 

Some thought's on gauges and infrared digital thermometers: We don't trust the $30 gauge, but we trust a $30 (most likely) Chinese built infrared thermometer which only reads accurately if the emissivity value of the surface being measured is the same as what the thermometer is calibrated for.

Bob

 

The cheapest and easiest fix is to just do what I did. Put one of the Mr Gasket radiator caps on because they ARE accurate (according to my heat gun) and then just know your dash gauge is off by so many degrees, and not worry about it. When I am going down the road and my dash gauge reads 210 I know my engine is really at 180, so I just use that 210 reading as a high water mark. If it were to go to 230, then I would start looking at the Mr Gasket one to verify and worry about it then.

Same with my gas gauge in my 27, it hasn't been right in 23 years, but I just make mental adjustments and when it gets to empty I know I still have 3 gallons in the tank.

These things are never going to be like our brand new high dollar daily drivers, where the dash readout tells you exactly how much fuel you have, what the exact temperature is, how many mpg you are getting, etc. Old hot rods are supposed to have some idiosyncrasies, part of their charm.

Don

 

It IS part of their charm, and I think its worth it. I KNOW my engine is not hot, and could certainly mentally explain that to myself as I'm driving down the road looking at 230-240 degrees and know its not really that hot. I also realize that an EE or ME has forgotten more about this stuff than I will ever know and I do appreciate your knowledge and advice. In spite of all that I still plan to install the diode/resister, just because I want to try it, and I've already picked up the components. Working on that right now.
Regarding Chinese infared thermometers, I have a Fluke, which I have always considered to be a fairly reliable well made device, and I have a HF Chinese brand infrared heat gun. They read within about 10 degrees of each other. I was surprised. I also realize that a mech water temp is more accurate but I just hate the damn capillary tube that goes with it, with respect to routing etc. But, if the little experiment with the Zener diode and 20 ohm resister does not work, thats what I'll be going to I suppose.
Thanks again guys...

BOutlaw

 

So can anybody recommend an electric temp gauge that does have good accuracy?

Like boutlaw I want to get away from the capillary tube.

 

Well, early on in this thread, I asked if the gauge was an old or new one. I'm having the same problems with my old one. I got it in a trade with another HAMBer..he basically gave it to me, so I sent it off to that gauge guy in Ohio (John Wolf & Co.) for cleaning, a (supposedly) compatible sender & God knows what other sort of arcane shit he did with it. In all fairness, it is a 6V gauge, that is now on a 12V car. He didn't elaborate, but told me that I should not be using a 12V to 6V reducer. I had already bought one, so I experimented with both. With the 12V, the gauge pegs, but with the 6V, it only goes to about 135. Being an old stock 283 sbc, it should be at 180. My direct screw in the manifold Moon gauge shows the proper temp of 175-185, depending on the seasons. As a sidebar, I was running a Chinese Sun Pro, which was accurate until it (ironically) crapped out shortly before I installed the refurbished S/W Wings gauge. I contacted the gauge guy, but he didn't have any answers. I let him off the hook by admitting that I could live with the 135 if I knew it was a constant.

I've read the explanations here, but I will admit that this is definitely not my strong suit. I think I understand the basic theory, but that's as far as it goes. An idea of an inline rheostat appeals to my simple thinking, because the answer is probably somewhere in the middle. I guess Walmart doesn't carry one.

 

Do a search on here, Someone gave a good way to fix a mechanical gauge with dry ice and a capillary tube off cheap gauge.



Ago

 

Fleetside, I apologize for not answering your question regarding old or new. They are new, I bought the gauges while building the 31 coupe over the last 2 years, as I liked the "Wings" design. After the purchase I began reading about the problems with the temp gauges. Since my coupe is a rear radiator car, I expected problems with cooling from the onset and initially assumed it was running hot, but have since discovered it is not hot at all. I just don;t like the capillary supplied gauges, acurrate or not. Its just hard to hide the damn tube and then roll up 3 ft of it under the dash.
Regarding repair of the mech gauge, just like Ago said, I saw a post on repair of the tube, something I would never have expected, but I learn stuff on the HAMB all the time.
I got the 2 10 ohm 1 watt resisters and the Zenor diode wired together and into the car this afternoon, but the weather moved in and I decided to close shop for the day. Tomorrow I'll test it to see what happens. About $4.00 worth of parts and about an hour of time. Thanks BobF for the details and schematic. If it works, and I expect it to, I will post some pics.
BOutlaw

 

Just in case somebody has found an electronic gauge they like, I'll ask again:
Can anybody recommend an electric temp gauge that does have good accuracy?

 

SUCCESS......at least so far.
I thought I would post a couple of pics of the diode/resistor lead string I fabbed up per BobF's schematic. The first pic is of the resistor pack and diode pack I picked up from Radio Shack. The second pic is of the diode/resister "string" I soldered up. I removed the + terminal from the instrument and left the terminal intact so if this little mod does not work I can reattach to instrument. I added a terminal to the resister end of my "string" and attached the two terminals together, covered with a terminal protective sheathing. The other end of the string, from the diode out goes to ground. The wire lead connection between the resistors and the diode is now attached to the + terminal of the water temp indicator.

 

I started the engine and let it warm up about 10 minutes. Normally the gauge would be reading 160 degrees at this time, rather than the 140 or so indicated by the infared heat gun. First pic is of the gauge reading and the second is of the IF gun reading taken immediately after the gauge pic.

 

Let the engine run about 15 more minutes and the indication is just below 150, and the IF gun reads 146.

 

All I can say is that the gauge is reading way more accurate than prior to adding the diode/resistor. The chance of rain is pretty high here right now so I don't want to drive the car just yet. It probably won;t be until tomorrow before it clears up enough to test it at normal driving temps but I am very satisfied with the results so far.
Below is a pic of the diode/resistor string attached to the water temp gauge before being secured. Kinda difficult to tell the routing from the pic. BobF's schematic to follow.



4Woody....to answer your question, I don;t know of any accurate electrical water temp gauges but this little mod sure seems to make the SW "Wings" gauge read pretty darn accurate, thanks to BobF.

 

Wow, this sounds promising. I'll be tuning in to see how you fair out with your extended road test. I'm still can't quite understand the wiring hook up, despite your fantastic pics. I'm such a freakin' dim bulb when it comes to this stuff. I definitely want to try it...it can't make things much worse. When the time comes, would you mind if I harassed you for details via a P.M.? ...Greg

 

Could you please post BobF's diagram and any "For Dummies" instructions.

 

Here's BobF's diagram. My Wings gauge has 3 terminals, (I) which is the + 12V source, (S) Sender, and (GND) which is Ground. BobF's diagram just shows a wire in and a wire out, but his gauge case is the ground. The diode/resistor string I soldered together is shown in one of the preceeding pics. Power comes into one end of the resister/diode "string", passes through the 2 - 10 ohm resistors (in series) and continues straight through to the Zener diode to GROUND. Be advised the Zenor diode, Radio Shack p/n 1N4742A, has a stripe on one end. That stripe goes towards the resistors in the string. Between the resistors and the diode, solder in another wire. This wire goes to the (I) or + terminal on the Wings gauge. The (I) terminal is the center terminal on the SW Wings gauge. The remaining wire goes FROM the diode (opposite the striped end) to chassis ground. So, BobF's diagram is just a little different when used with a 3 terminal gauge, but is still basically the same. Hopefully this explanation is clear enough to understand, if not, pls feel free to PM me if you have additional questions. Also, BobF said it was very important to check the alternator voltage prior to insertion of the resistors. My Voltmeter read 14 volts, but the alternator was actually putting out 14.6 volts. He also sent me a web page where the actual voltage you are getting from your alternator can be inserted into a provided formula along with a milliamp (mA) draw taken from the gauge/sender circuit. I didn't do that as my overvoltage was very similar to what he had on his gauge. I feel sure he will chime in on this thread as I may not have this interprtation exactly correct. Anyway, I am very sarisfied with the indication accuracy so far. I will drive the car tomorrow to check the reading at normal operation. Thanks BobF AND everyone else who helped me with this issue. I'll post again tomorrow after I get another reading.

 

Oh, by the way.... Fleetside, you got a very cool dash going on there, and feel free to contact me if I can provide any additional info, but remember, I'm a dim bulb myself.

 

Hope you didnt use teflon tape on sender. May affect the reading.

 

I got my gauges today didn,t check with 12 volt to gauge but tested sender.Room temp about 400 ohms then warmer 250,then in boiling water on stove 100 ohms,so hotter under pressure when it go close to 20 ohms it should go the full scale 240 degs.
So adding resisters lowers full scale.
I don,t think that gauge is for testing water temp maybe for testing Pizza oven.

 

Update, drove the car today and got it to normal operating temp. I went to my buddys body shop. Usually, when I arrive the temp indication is 230-240. Today, the water temp showed about 191 and the IF gun reading was 182. Still a discrepancy between the two, but no where near 30-40 degrees, so I am very happy with the results of the diode/resistor modification. Thanks BobF and all others that commented. I will continue to check it as the weather gets hotter to see if there are significant changes.
Adios...
BOutlaw

 

I'm curious about the Zener diode. I understand it's function from working on British motorcycles for decades, but why can't the reading be lowered simply by adding more resistance between the gauge and sender? Why is it necessary to strive for exactly 12 volts? Seems simpler just to correct with resistance.

Bob

 

If you increase the voltage you will increase the current, for a given resistance. Depending on whether or not the gauge uses a bridge circuit and how it is wired, your needle may bounce around as the voltage changes. It is just an easy way to reduce the chances for error. It may or may not help depending on the design of the circuit.

Basically a cheap voltage regulator.

 

Makes sense. Thanks.

Bob

 

Thanks to everyone that took the time & did the work to educate the rest of us. Like anything else, the more you study it & do it, the clearer it becomes, even if it's not one's strong suit. As soon as my strained back gets a little better, I'm going to get on this.

By the way, on the subject of teflon tape, I was warned that using it could possibly affect things. In my case, it didn't seem to change anything, one way or the other. I ran a ground directly from the sender body to a good ground source on the car & it didn't make any difference. Also, my old SunPro gauge (that I knew was accurate) did have teflon on the threads of the sensor (which I put on). I've always noticed that when you remove the sensor, the ridges of the threads are always bare, meaning that it is getting some ground contact. Still, the whole teflon tape thing should still be a concern, since there are different degrees of ground.