[Rich-L79]

I got a new sending unit today. I tested it and got basically the same readings as found on my old unit though I do not see the needle on my multitester bouncing occassionally as with the old unit. The ohms do not measure up to the ranges listed by Dave in my earlier thread. I now have to wonder if my multitester is out of whack. A new battery in the tester made no difference in the readings. I'll stop by Radio Shack tomorrow to compare it against a new one.

Assuming the sending units are okay, and since they both read basically the same I now think they are both okay, I'm left with a bad ground at the dash or a failing gauge. Who can I contact and how can I contact them about repairing the gauge if it is indeed the culprit? If I measure the ohms from the sending unit wire at the engine to a ground it is a very low reading (under 100) so I believe the wiring and ground are okay, again that leaves me with the gauge itself. The full open/full closed test of the gauge seemed to show the extremes on the gauge I would expect and while that confirms the connection is good it does not necessarily confirm the gauge itself is good (right?).

Why can't it ever be the easy solution?!

 

[John_Muha]

Have you considered picking up a potentiometer to use in place of the sensor? It’s the next test I would do after the full/empty test.

 

[65elcamino]

Rich,

The full deflection test of the gauge shows that it works properly. The sender is a variable resistance device that varies the amount of current flowing through the gauge.

Your sending wire should be an open circuit when it is disconnected from the sender and the gauge. The ground for the circuit is through sender body to the intake manifold. If you are measuring the wire to ground at the sender end without it disconnected from the gauge, you are reading the resistance of the gauge.

Verify that the sender wire is not grounded when disconnected from the gauge and the sender unit. Verify that you have a full 12 volts at the supply to the gauge when the ignition switch is in the on position. Other than these two checks, the circuit should be good. You have a new sender and the gauge shows to be good from your testing.

 

[Rich-L79]

I'll check for 12v at the gauge, should be easy enough to do. The one thing I haven't checked is real world operation of the sending unit I just installed. At the very least I'll be driving the car to a club meeting on Saturday so we'll see how it works then. It's supposed to be hot and sunny so I'll certainly know if it's reading low!

I do have a good ground at the sending unit. I get the same ohms reading when measuring between the wire-connection post and the body of the sending unit as I get between the wire-connection post and any other ground in the engine compartment.

This unit is measuring nearly 600 ohms at room temperature so either my multitester is whacko or my new unit is working exactly wrong to the same degree as my old one. Highly unlikely.

I need to do some checking at the gauge. I hate getting under that dash but at least things are easier to reach under there than they are under the dash of my MR2! I did recently fix a 2-year old problem with the radio in my MR2 so perhaps I'm on a roll!

 

[John_Muha]

Originally posted by Rich-L79:
I did recently fix a 2-year old problem with the radio in my MR2 so perhaps I'm on a roll!

Intermittent power on the memory 12 volt line?

 

[Rich-L79]

Originally posted by John_Muha:
</font><blockquote>quote:</font><hr />Originally posted by Rich-L79:
I did recently fix a 2-year old problem with the radio in my MR2 so perhaps I'm on a roll!

Intermittent power on the memory 12 volt line? </font>[/QUOTE]Exactly. I ran a new power feed from a non-switched 12v source and it works perfectly. When I was told that power feed must be intermitantly dropping voltage I couldn't understand:
1. how could it drop voltage, shouldn't it either be on or off?
2. how dropping voltage would kill the radio but still allow the tape deck to play.

What I found out was that the radio tuner is separately powered by this non-switched 12v source while the tape deck was powered by a separate switched 12v source. I had previously thought there was just a single 12v source.

I should probably go tell the repair place I'm an idiot since they kept telling me it was a wiring issue and I kept telling them they were nuts. I guess I also never needed to spend that $185 to have the deck "fixed" in the first place! It was my choice to have the deck "fixed" so it's not like they did work I didn't authorize. I was convinced the deck was at fault so I told them to crack the box and fix things up. They did get the panel lights in the deck working again! Those few little lightbulbs cost me nearly $200!

 

[Rich-L79]

Well, the multitester checked out. Follow up checks on the car found the following:

-the lead from the sending unit does not have a leak to ground (good)
-the gauge has a static ohms reading of around 100 between the terminal (good, I think)
-the sending unit wire has the same ohms at the gauge end of the wire as at the unit itself, about 650 (bad? seems high)
-the hot wire at the gauge reads just over 12v with the key on (good)
-the sending unit wire alone, shorted to ground at the sending unit reads about 10-15 ohms at the gauge end with the power off (bad? this seems a little high for a simple wire) Could this be my source of a low temp reading? How do I determine where the cause of the extra resistance is? And confusing me is the fact that if the ohms reading at the sending unit is the same as the sending unit wire at the gauge end there really doesn't seem to be an extrodinary amount of extra resistance in that circuit. Right?
-another confusing bit is that the gauge reads low when the power is off, but the needle pushes even farther to the low side with a cold sending unit when the power is turned on. The cold reading on the gauge with everything hooked up and the ignition on is way past the cold end of the scale so SOMETHING has to be adding more resistance to the circuit than it should have, right? But what? Could it come down to the gauge being out of adjustment? Can I adjust it myself?

I have not yet checked the fuses for a weak fuse or a poor connection at the fuse but if I'm getting the 12v I should be getting what's the point? Time to look over the wiring schematics I guess. I'm more confused than ever now.

 

[65elcamino]

Rich,

The wire resistance of 10-15 ohms is most likely resistance through the body. If I understand the method of your resistance measurement. You grounded the one end and measured the resistance to ground at a point near the gauge. Measure the wire and only the wire (from one end at the sender to the other at the gauge) with the bulkhead connector connected. You should see very minimal resistance, less than 0.5 ohms.

The gauge uses a magnetic wire coil to cause the needle to deflect based on the amount of current flowing through it. The needle action you described sounds normal. The needle is not free to spin easily so that it doesn't bounce. This makes the needle return to a slightly different position when the ignition is off. When you turn on the 12 Volts, the current causes the needle to pull to the full cold position. It should not go past this point.

It sounds like the needle may be bent and indicating incorrectly. Use Dave's resistance values to adjust the gauge needle for the correct position. Here are Dave's values from your first post.

Low Mark (Cold) = 300 to 320 ohms
1/4 = 145 to 150 ohms
1/2 = 103 to 115 ohms
3/4 = 79 to 88 ohms
Full Scale (Hot) = 58 to 64 ohms

Get a resistance decade box and setup the circuit to check the gauge needle positions at the cold and hot resistance values. If both are off the same amount (indicating lower in your case), the needle is bent to the side in error. These needles are very easily bent if the faceplate is removed such as having the gauge package removed from the dash bezel.

 

[Rich-L79]

Originally posted by 65elcamino:
If both are off the same amount (indicating lower in your case), the needle is bent to the side in error. These needles are very easily bent if the faceplate is removed such as having the gauge package removed from the dash bezel.

I understand that's possible but the gauge was working fine a few weeks ago. With nothing at all being changed it just stopped working properly. I haven't been inside the gauge cluster since I rebuilt the car in 2000 and the gauge cluster itself has never had the face plate off for as long as I can remember because I've always been afraid of breaking a needle or scratching a gauge plate.

I'll measure the resistance in the sending unit wire from end to end. I will check the fuse panel for a bad fuse or bad fuse connection. Where can I get a resistance decade box? Radio Shack perhaps? Expensive? And if the gauge readings are off how can it be adjusted (other than bending the needle which I don't believe is the issue in this case)?

 

[65elcamino]

Other than the needle getting bent, I guess the magnet wire could have gotten fried if the wire got shorted. Inspect the wire for any visible damage or burned lacquer. Do you remember any time when the needle was pegged at hot (indicating a shorted wire and high current flow)?

A bad fuse would cause all of the gauges to stop function since it would interrupt the 12 volt supply. You verified that you have 12 volt with the ignition on, so the fuse is good.

An electronics shop will have the decade box, not sure on price, I just borrow one from work if I need one. You can buy a couple of precision resistors at the values for proper operation at Radio Shack. Just connect each resistor, one at a time to the sending wire and to the battery negative post to test the gauge.

 

[Rich-L79]

I wonder if the bulkhead connector has gotten some corrosion in it that is causing more resistance in the sending unit wire? The car got rained on again on the way to the Mid America Regionals and that's when the problem began. Perhaps some water got in there or perhaps I nudged it loose while cleaning the firewall area. Worth a look-see. I'm now very eager to get home and measure the resistance of the sending unit wire from end to end.

 

[John_Muha]

Rich
A small potentiometer will do the same thing as a decade box. You adjust it to Dave's numbers using your meter and then hook it up. A muliturn is easier to adjust. A linear taper pot works better than an audio taper. Maybe a 500 ohm pot. Nothing too large because it would be too hard the dial in the numbers.
PS
If you don't have a couple of aligater clip leads buy a small pack. Easier to attach a pot or resistor.

 

[AZCamino]

I wonder if the bulkhead connector has gotten some corrosion in it that is causing more resistance in the sending unit wire? The car got rained on again on the way to the Mid America Regionals and that's when the problem began. Perhaps some water got in there or perhaps I nudged it loose while cleaning the firewall area. Worth a look-see. I'm now very eager to get home and measure the resistance of the sending unit wire from end to end.

When you restored your car did you use dielectric grease to seal the harness connection at the bulk head connector? Dielectric grease will keep moisture out of the connections so they don't corrode.

 

[Rich-L79]

No, but that's a darn fine idea to grease up those connections even if they aren't related to my temp gauge problems. This car is basically a trailer queen but it is definitely a rain magnet everytime I travel to a distant show. I suppose the day I buy an enclosed trailer that will all stop. I'm sure I could cure a drought if someone would arrange a car show hundreds of miles away from me and then invite me to the show.

 

[AZCamino]

Rich,

We have a lot of low reservoirs in the West. Perhaps there would be some financial opportunity arranging car shows (one car show) all over the western U.S.