How to test water temp gauge

[VetteLT193]

I have a brand new Gaffrig water temp gauge and I can't get it to work.

All the gauges move just a bit when power is turned on except for this and the oil temp gauge. The oil temp gauge isn't hooked up so I don't expect it to work (no sender yet). But, this one is hooked up and I get nothing. I haven't run the boat for too long either, just 2 minutes max, so that could be why it hasn't moved.

I'd like to test it... can I test using regular electrical tester, or some other quickie hack to see if the needle moves? This is also the only gauge that has water in it so I'm thinking it could be a brand new dud.

[techspecial]

I put my sender in a pan of boiling water. This is your 212 degree standard (at sea level). The water will maintain the 212 for quite a while until you get to the sender.Check wiring...Senders are cheap but rarely fail in my experience. I'm puzzled by your statement that your gauge has water in it though...Good Luck!!

[VetteLT193]

I put my sender in a pan of boiling water. This is your 212 degree standard (at sea level). The water will maintain the 212 for quite a while until you get to the sender.Check wiring...Senders are cheap but rarely fail in my experience. I'm puzzled by your statement that your gauge has water in it though...Good Luck!!

what I mean by water is the gauge fogged and now has water drops inside of the glass. I've experienced this on plenty of boats but it has always been in multiples. this time this is the only gauge that has the fogging/water problem.

I'll try the sender thing if no one else chimes in with an electrical hack. I figured there would be an easy way to spoof the gauge into just pegging in order to see if the needle moves.

[RickR]

Momentairly touch the sender wire to ground. Gauge should peg.

[Conquistador_del_mar]

Momentairly touch the sender wire to ground. Gauge should peg.

:yes:

[VetteLT193]

Momentairly touch the sender wire to ground. Gauge should peg.

worked perfect... THANKS!!!

After the old feel test and a few minutes of running, looking, and feeling I think it might actually be running too cold. There is condensation on the crossover and the inside of the carb. none of the hoses are even warm except for the ones that connect the manifold to the tails.

[Ghost]

I don't know what I'm talking about.

But having a vague notion of how a gauge works when I saw this, I started trying to dig a little deeper. What I found makes sense to me, and seems consistent with everything above, and seems like it explains the why/how. Figured I'd share it in case it is any use to anyone (I find it easier to remember stuff I understand, rather than memorizing stuff).

Most people here probably know all about this stuff, and if anything here is wrong, please point it out. And if it's hopelessly off, I'll delete the post.

I think a water temperature gauge uses a sender with a thermistor, a device that varies in electrical resistance with temperature. I suspect the gauge itself registers the temp by measuring either the current that 12V puts through the sender (essentially the gauge is an ammeter) or by measuring the resistance of the sender (essentially the gauge is an ohmmeter).

I found two things (attached) that struck me as useful in thinking about it, but they are in no way specific to Vette's scenario. Just a random thermistor spec that looks like it's for water temp measurement (32-212 degrees) and a temperature gauge wiring diagram from an ultralight plane of all things. (Edit: and I don't think it's that great of a diagram, but the basic notion is a gauge has three contacts: 12V hot, ground, and a wire to the sender.)

Looks to me like the advice above in the thread above simply corresponds to the short condition in the thermistor spec, that would impersonate max temperature and peg the gauge.

Anyhow, I don't know if any of this is actually accurate, much less useful to anyone. But it seemed like it made sense. If anyone spots anything incorrect, please chime in. I don't want to spread bad info.

Also, if anyone knows more about the gauge basically being an ammeter or an ohmmeter (or if both of those basically work the same way, and there are just different mechanisms for different types), I'm curious about that. A pure ohmmeter like most of us are used to of course needs to be able to supply some power in some cases, where an ammeter can use the power making the current flow already, but I don't honestly know if they are otherwise largely doing the same things behind the scenes, and just reading out differently. But something makes me think that technically there might be some subtle differences, and that the gauge would be classified as one or the other. Just a hunch.

FWIW.

[MOP]

Did you install the matching sender? If so did you use Teflon tape when you screwed in the sender? That can ruin readings, always use liquid Teflon so you get metal to metal contact.

[Conquistador_del_mar]

I don't know what I'm talking about.

But having a vague notion of how a gauge works when I saw this, I started trying to dig a little deeper. What I found makes sense to me, and seems consistent with everything above, and seems like it explains the why/how. Figured I'd share it in case it is any use to anyone (I find it easier to remember stuff I understand, rather than memorizing stuff).

Most people here probably know all about this stuff, and if anything here is wrong, please point it out. And if it's hopelessly off, I'll delete the post.

I think a water temperature gauge uses a sender with a thermistor, a device that varies in electrical resistance with temperature. I suspect the gauge itself registers the temp by measuring either the current that 12V puts through the sender (essentially the gauge is an ammeter) or by measuring the resistance of the sender (essentially the gauge is an ohmmeter).

I found two things (attached) that struck me as useful in thinking about it, but they are in no way specific to Vette's scenario. Just a random thermistor spec that looks like it's for water temp measurement (32-212 degrees) and a temperature gauge wiring diagram from an ultralight plane of all things. (Edit: and I don't think it's that great of a diagram, but the basic notion is a gauge has three contacts: 12V hot, ground, and a wire to the sender.)

Looks to me like the advice above in the thread above simply corresponds to the short condition in the thermistor spec, that would impersonate max temperature and peg the gauge.

Anyhow, I don't know if any of this is actually accurate, much less useful to anyone. But it seemed like it made sense. If anyone spots anything incorrect, please chime in. I don't want to spread bad info.

Also, if anyone knows more about the gauge basically being an ammeter or an ohmmeter (or if both of those basically work the same way, and there are just different mechanisms for different types), I'm curious about that. A pure ohmmeter like most of us are used to of course needs to be able to supply some power in some cases, where an ammeter can use the power making the current flow already, but I don't honestly know if they are otherwise largely doing the same things behind the scenes, and just reading out differently. But something makes me think that technically there might be some subtle differences, and that the gauge would be classified as one or the other. Just a hunch.

FWIW.

Instead of the temp gauge being an ohmmeter, you can think of it as a voltage gauge with temperatures as the measurements. The sender acts like a simple potentiometer with varying resistance according to temperature. Bill

[Ghost]

Instead of the temp gauge being an ohmmeter, you can think of it as a voltage gauge with temperatures as the measurements. The sender acts like a simple potentiometer with varying resistance according to temperature. Bill

Bill, thanks for taking the time on this. No need, but if you don't mind, can you help me along a little more here?



If the temp gauge is a voltmeter, between what two electrical nodes is it measuring the voltage drop? I would *think* based on what the wires and components in the system actually are, that the gauge has terminals that connect to:

the 12+V node of the battery/alternator
system ground
the lead to the non-ground side of the sender
IF this is the case, and the the gauge is actually a potentiometer, I'm picturing it works fundamentally like my crudely attached circuit diagram. The two resistors, Rg in the gauge, and Rs in the sender, are in series. Then, knowing the current (I = v/R) in the two is the same, because they are in series:

v (gauge resistor) / R (gauge resistor) = v (sender) / R (sender)

And with:

the gauge measuing the voltage across the gauge resistor, and
the resistance of the gauge resistor known, and
the voltage drop across the sender's resistor being the voltage of the source (known by measurement) minus the voltage measured by the gauge (known by measurement)
It would seem possible to calculate the resistance of the sender's resistor, and thus know the temp. Calibrate the gauge properly for its internal resistor and the sender's resistor, and you have a potentiometer functioning as a temp gauge.

The key being that the two main resistances in the system are IN SERIES, and the gauge is at its core a potentiometer. It has at its direct disposal the overall voltage between the system hot and ground, as well as the ability to measure the voltage drop across its internal resistor. (R sub g in my crude diagram.)

Anyhow, if the gauge is basically a potentiometer, as you say, and I don't doubt at all, this is how I imagine it would work. (Further, for all I know, most ammeters and ohmmeters work in similar ways, and all the rest is calibration, scales, and supplying power for resistance measurements on unpowered items, like a foot of wire. Been a LONG time since I thought about such things in detail.)

I always had heard that most gauges were essentially voltmeters, and it was only when I started thinking about the stuff above in the thread that I considered that they might be measuring current or resistance directly, rather than voltage. Your saying that it really is voltage forced me to think about how it would really work if that's true, and this is what I've come up with. In short that there are two main resistors IN SERIES, one in the gauge, and one in the sender. And as before, tapping the sending lead to ground would just convince the system that the sender resistance was near zero, and the temp would peg at maximum.

Anyhow, I assume like all EE work, it's not QUITE this simple, but at it's core, this might describe the basic approach. Any confirmation, denial, or refinement is much appreciated.

Thanks,

Mike

[txtaz]

between what two electrical nodes is it measuring the voltage drop?
Mike

First off, this should be a separate thread in the tech section. Of Fuzzy one, can we have a move here?

Mike, in my EE classes back in the day, it was referred to as differential potential since ground is not absolute, but voltage drop is the same thing.

In your scenario, the resistance can be series or parallel. The difference would be the resistance curve vs temp. Series would have a steeper slope (rise over run) and a longer curve.

Now I have long forgotten that exact details of the internals but the meter measures the differential potential between sensor and ground. The +12v would be used to stabilize and condition input to the meter movements stator windings for a constant fulx field. The variable field is the armature (the part that moves or pointer) is powered by the drop over the temp sensor.

Make sense?

Da Taz

[Ghost]

Wes,

Thanks, I think I get everything but the series or parallel part. I was thinking it had to be series and not parallel. Reason being, if there is no effective gauge resistor in series with the variable sending unit resistor, the voltage drop measured by the gauge (measured from the the sending wire to ground) will pretty much always be the full differential between the system's voltage source and ground. What will fluctuate as the sender changes resistance will be the current passing through it, and of course the resistance itself, but not the voltage drop. Hence, as temperature changed, a "potentiometer" for a gauge would never see the voltage drop across the sender depart from the voltage drop from the system power source to ground. (They're parallel, so the voltage drop across them is the same.) And thus a potentiometer wouldn't serve as an analog to the temperature.

Whereas if the gauge has a resistor in series with the sending unit (as in my extremely crude drawing above), there will be a voltage drop through the gauge's resistor. Then there will be a second drop across the sending unit's resistor. Thus, when the sending unit resistance reacts to temp changess, its resistance will change, and the voltage drop across both the sending unit's resistor and the gauge's resistor will change. This would mean a potentiometer as the heart of the gauge would have some variable potential to measure, which would serve as an analog to temp.

Anyhow, not sure if all that is right, but that's why I was thinking there had to be a gauge resistor in series, not parallel, with the sending unit resistor. Am I nuts? (I mean, other than in general. :) )

Mike

[txtaz]

Yes you are nuts...:wavey:

Total resistance in parallel does change with a varistor/thermistor and a fixed resistor but it is smaller change than series. The folmula is

RT = 1 / ((1 / R1) + (1 / R2) + (1 / R3))...

So say you want a fine tune RT. Place a variable potentiometer in parallel with the thermistor and the meter movements armature windings. Now you can fine tune the gauge reading. You can also do this in series. It depends on voltage and current requirements/limits on the circuit.

General accepted design principles (in this application) are to design in parallel for better stabilization during fluctuations.

Make sense?

Da Taz

[DickB]

The heart of most electrical gauges is a moving coil galvanometer, which measures current.
http://hyperphysics.phy-astr.gsu.edu/HBASE/magnetic/galvan.html#c1 (http://hyperphysics.phy-astr.gsu.edu/HBASE/magnetic/galvan.html#c1)

Ammeters, voltmeters, and ohmmeters are built around this basic galvanometer. Because the moving coil meter responds to very low currents, placing a shunt resistance in parallel with the sensitive meter makes a practical ammeter that can measure most any size current. Placing a resistance in series with the meter to limit current makes a voltmeter. Adding a voltage source with a calibration resistor makes an ohmmeter.
http://hyperphysics.phy-astr.gsu.edu/HBASE/magnetic/movcoil.html#c1 (http://hyperphysics.phy-astr.gsu.edu/HBASE/magnetic/movcoil.html#c1)

Auto Meter describes their short sweep, three terminal water temperature gauge as “nothing more than a fine tuned & calibrated ohm meter”
http://www.autometer.com/tech_faq_answer.aspx?sid=1&qid=57 (http://www.autometer.com/tech_faq_answer.aspx?sid=1&qid=57)

[BigGrizzly]

Just to be fair, they are both similar. The point is the meter is calibrated to the sender. So any sealer or Teflon will int erupt the signal. Vette your are going to have to run the engine longer.

[Conquistador_del_mar]

The heart of most electrical gauges is a moving coil galvanometer, which measures current.
http://hyperphysics.phy-astr.gsu.edu/HBASE/magnetic/galvan.html#c1 (http://hyperphysics.phy-astr.gsu.edu/HBASE/magnetic/galvan.html#c1)

Ammeters, voltmeters, and ohmmeters are built around this basic galvanometer. Because the moving coil meter responds to very low currents, placing a shunt resistance in parallel with the sensitive meter makes a practical ammeter that can measure most any size current. Placing a resistance in series with the meter to limit current makes a voltmeter. Adding a voltage source with a calibration resistor makes an ohmmeter.
http://hyperphysics.phy-astr.gsu.edu/HBASE/magnetic/movcoil.html#c1 (http://hyperphysics.phy-astr.gsu.edu/HBASE/magnetic/movcoil.html#c1)

Auto Meter describes their short sweep, three terminal water temperature gauge as “nothing more than a fine tuned & calibrated ohm meter”
http://www.autometer.com/tech_faq_answer.aspx?sid=1&qid=57 (http://www.autometer.com/tech_faq_answer.aspx?sid=1&qid=57)

Excellent explanation. Bill

[txtaz]

Good stuff Dick,

I forgot to add the current limiting of series resisitors.

Grizz is right, run it longer or get a pan of boiling water and dunk the sensor.

Da Taz

[Ghost]

Ding!! And the light shineth!

Thanks VERY MUCH Bill, Wes, Dick, and Randy. All of this helped, and these last few posts got me the rest of the way over the bar. For anyone who's curious, you may get a laugh out of knowing the first major hurdle was that I blockheadedly was thinking a potentiometer was synonymous with a voltmeter, rather than being a variable resistor. Knowing what it actually IS changed a LOT. ("Oral, rectal, it's just a thermometer, right?")

Also (and Dick's post was huge on this), the succinct info on galvanometers and the relationship of the 3 major meters to one another, commonly using a galvanometer at their core, really got at the heart of what I was missing.

My whole notion of a fixed-resistance "gauge resistor" coupled with a device that purely measures voltage, or a different device that purely measures resistance, or a third device that purely measures current, sorta melts away in the real world.

When I made that textbooky assumption about the gauge being one of three types of pure measuring devices, I think my primitive analysis made sense. A guy picking his multimeter setting and slapping the leads down onto my circuit would need to look at it that way. Put the fixed and variable resistors in series and I can use the voltmeter setting to figure out temperature. Put 'em in parallel and the voltage drop across both is the same, and I need to use the ohmmeter or ammeter instead.

But understanding a bit more about the guts of the gauge, and that all three gauge types are working pretty similarly, using shunt resistances and other tweaks, makes the whole pure voltmeter OR pure ammeter OR pure ohmmeter assumption kinda fade away, and as Wes pointed out, you can then set stuff up in parallel and it can be even better in practical use.

I really appreciate all of your patience. I hope there is some consolation that unlike trying to teach a pig to sing, while this may have frustrated you all, it didn't annoy the pig--it did quite the reverse. It made my day by restoring some stuff that has been rotting in my head for FAR too long, and adding to it as well.

Thanks much,

Mike

P.S. All of this raises a quick and easy question: do some/most gauges have an adjustment the layman can set? (To calibrate them by dialing in the potentiometer resistance?) Or is that considered factory-only, inside-the-case tinkering that you could try yourself but they don't encourage by putting an easily accessible screwhead or something on the back?

[txtaz]

AND now we are up to date for the 40's era technology...

Today everything is digital. TI makes a digital temperature chip that is IP addressable. I have 8 running in Bugs Bunny now (max the IP bus allows). So imagine having a few hanging out on your engine connected to an addressable controller in a cool digital gauge with a push button and LCD or VFD read out. Since mine are in a computer, I wrote a device driver for them connected to a internal USB header and display temps on a front panel VFP (vacuum florescent display).

Being a purist, I don't know if I'd want something computer looking or even digital in a boat. Maybe a depth sounder but that's it.

Da Taz

[BUIZILLA]

("Oral, rectal, it's just a thermometer, right?")
now there's a new sig line :biggrin.:

[Ghost]

AND now we are up to date for the 40's era technology...

Wes,

LOL. Maybe that's why I like Big Band, too.

The digital stuff sounds pretty cool, though like you, I'm a fan of analog gauges for all but maybe depth. Well, and this (attached), but it's a bit bulky for a Donzi classic.

Really appreciate your patience in going through the gauge stuff. (Been 20 years since I saw the formula for resistances in parallel--that alone was worth the admission price to me. :) )

Regards,

Mike

[mrfixxall]

I have a brand new Gaffrig water temp gauge and I can't get it to work.
All the gauges move just a bit when power is turned on except for this and the oil temp gauge. The oil temp gauge isn't hooked up so I don't expect it to work (no sender yet). But, this one is hooked up and I get nothing. I haven't run the boat for too long either, just 2 minutes max, so that could be why it hasn't moved.
I'd like to test it... can I test using regular electrical tester, or some other quickie hack to see if the needle moves? This is also the only gauge that has water in it so I'm thinking it could be a brand new dud.


Vette, me being a gaffrig dealer you must use their senders to get the accurate readings..They use a differant ohme range sender so if you need them let me know and ill hook you up..:beer:

[VetteLT193]

Opened up a whole can of worms with this one...

I picked up a T-stat, once that goes in it will at least warm up. It's running too cold to even tune the new carb.

I'll get the new sensor(s) from fixxall and should be good to go. I actually have a handful of new gauges that don't have sensors so I need more than just this one. good learning on this thread.