Tweet
Sorry about adding more discussion on Temp senders, but I've got an interesting problem. I purchased a temp sender from Corvette Central - of course I didn't read all the posts about how inaccurate they are before purchase. However, I ran a little experiment that someone recommended on the board. I placed the unit in the oven with a thermometer and took resistance readings every 20 degrees F. I also bought a variable resistor at radio shack and tested the temp gauge itself. Interestingly, the two curves were not that far off. To keep this brief, to measure 180 degrees, the gauge needs to see approx. 124 ohms from the sender. My "study" showed that the sender, when exposed to 180 degrees, did read 124 ohms. Here's the wild part. The darn system still reads 240 degrees on the gauge when the engine is running at 180. I measured the resistance of the sender, installed in the engine after warm up, and it read 124 ohms. Hooked it up to temp gauge and it pegged above 220. Next I connected the variable resistor again and used the temp sender case for my ground connection. At 124 ohms, temp gauge read correctly. Went back and forth several times with the same result. What in the heck am I missing?
-
-
Re: C1 Temperature Senders
Boy you have me stumped with that one, your diagnosis process sounds right on, and the only thing I can think of is a lossy ground circuit to the guage itself, creating impedance in the ground path as well as the hot path. But, increasing the overall resistance would lower the reading wouldn't it? As I recall, the resistance of these sending units drop with elevating temperature, so I am not sure my though process is valid. Maybe someone else has some thoughts on this one, what you did sounds dead on to me...Craig -
Re: C1 Temperature Senders
Stephen,
I, too, purchased the Corvette Central sender and had the exact same problem when I fired up my newly rebuilt engine the week before last. The motor was running for about 10 minutes, I poked my head in to have a look at the gauges, and my temp was pegged at 240 deg. I posted the same question on this board, and tested the sending unit in the oven, and on the stove in water. The ohm readings were pretty close to what I was supposed to be looking for, but when it is installed in the car, it goes to 240 when the coolant temp is 175-180. I ended up purchasing the Wells TU-5 at Autozone, and although it reads very close to the Corvette Central sender at the different temps in the oven and on the stove, it reads right around 180 on the gauge when by thermometer is reading 180 in the expansion tank. This whole thing with the senders has me pretty baffled....It shouldn't be this complicated, but it seems to be a common problemComment
-
Re: C1 Temperature Senders
The only thing I can fathom here is when the sender is in a loaded state with current flowing, the actual impedance is changing from the static ohmmeter measurement....CraigComment
-
Re: C1 Temperature Senders
Just a little side-note. I read earlier regarding Junk-Yard for old AC temp. sender. I went out for a hoot and found a '69 Chevy truck. Removed sender cleaned it up (looks brand new) and for the first time, my '62 gauge reads right on at 180+/-. PS... man at yard gave me the old sender.Comment
-
Re: C1 Temperature Senders
Stephen-----
Did you apply thread sealer or Teflon tape when you installed the temp sender? If so, remove the sender, THOROUGHLY clean off the sealer/tape, and re-install. Then, see if it reads correctly.In Appreciation of John HinckleyComment
-
Re: C1 Temperature Senders
High resistance = low temp. Thought about the ground path, but just didn't seem to make sense. I had a similar thought to your other note. If I remember correctly, the ohmmeter uses extremely low voltage or current to measure impedence. Under vehicle power conditions, maybe the resistance values of the sender change. I think I'll try that autozone sender and see what happens. It still bugs me that I can't solve this sender issue.Comment
-
Re: C1 Temperature Senders
Don - I had originally used some thread sealer, but have cleaned it off. Just to be sure, I tried hard wiring the sender case to ground with no change in results. I had a similar thought to Craig's response below that the resistance value changes under vehicle power conditions and the variable resistor does not.Comment
-
Re: C1 Temperature Senders
With the gauge in the circuit, you might try measuring the voltage drop across the sender when it is at 180 degrees and compare that to the voltage drop across the pot. at the same temperature. If they are different, which I assume they will be, the sender resistance is changing with current flow.
I would be surprised if the resistance element was heating up with the current flow, but that is a possibility.
Dave ChristensenComment
-
Re: C1 Temperature Senders
Stephen,
Sealant would add resistance and would make the gauge read low rather than high.
There are 2 possibilities. First possibillity is as others have pointed out, resistance could be changing with current flow and can be checked by doing a voltage drop test on both the sensor and the pot. I would take a guess that the pot you are using is under rated and can be increasing it's resistance under load, but it could be a sensor problem. Second issue is the supplied voltage. If you are testing the gauge with a pot with the key on but the system not running then you have a probable difference at least 2 volts between the sensor and pot testing conditions and this is enough to make a substantial error.
I did my testing with a battery charger hooked up to the car putting out a constant voltage.
Regards, John McGrawComment
-
Only a couple of possibilities for explanation....
(1) The readings you took are incorrect.
(2) There is a defect internal to the replacement temp sender.
(3) There's a 'third party' agent that's not under control.
Number one is self-explanatory. As a former design EE, I will tell you I've fooled myself making/taking readings in my career....we're all only human.
On number two, the way these senders are constructed is with a 'pellet' (that's the variable resistor) compressed into position against the bottom 'well' wall of the sender's brass probe tip. The pellet is electrically connected to the sender's contact terminal by way of a spring. The inner wall of the brass probe tube is electrically insulated via a dielectric liner (fische paper, Etc.).
The way original senders age and 'go bad' is by coolant entering the inner probe tube. If the insulation on the probe tube wall remains intact, the contact between the spring and pellet can corrode introducing a spurious resistance (overall resistance climbs and temp gauge reads low). If the insulation breaks down, the contact spring can begin to make contact with the brass tube wall effectively 'short circuiting' the pellet. Overall resistance falls and the temp gauge reads artifically high....
Now, one thing that's different between the oven verification test and the in-car installation of the temp sender is the TORQUE/STRESS on the temp sender's case. It's possible (not very likely, BUT possible), that internal geometry of the sender is changing when it's installed via torque/stress and resulting stain.
Issue three isn't all that obvious, but it's worth yaking about. When you measure the resistance of the temp sender out of circuit, you're probably using an analog or digital multi-meter. They execute resistance measurments this way....
There's an internal battery in the meter, typically 1.5-9 VDC that supplies power for the resistance measurement. The control electronics of the meter engage a solid-state voltage regulation device to CLAMP the battery voltage to a precise/tight known voltage level. When you connect the meter across an unknown load, the meter knows EXACTLY what it's supply voltage is, measures the resulting current flow into the target load and computes resistance based on the current flow it sees.
There was a prior thread on this subject where the poster was measuring the resistance of his temp sender with it connected to the car's battery!!!!! This is a GREAT way to blow a meter as well as generate false readings--there are two sources of current (the meter and the car).... I don't know of a multi-meter that doesn't come with an instruction book that clearly tells the owner to take resistance measurements with the target component OUT of circuit.
Next, some things change impedance based on the exact voltage level applied (that's the definition of a semi-conductor, it's also how flourscent lights work). When you take ohmic measurements with a multi-meter, you're looking at the part's resistance characteristics based on the meter's internal battery (typically 1.5 V). When the sender is in-circuit, it's being driven by the car's battery (typically 14 V with alternator active).
There should NOT be a material difference.... But, if there's stray contamination (oil, etc) across the bakelite insulator of the temp sender, you might see a parasitic 'short' at one probe voltage that's not there or of a different degree of severity at another voltage.
Last, when doing these 'before/after' tests, there's another variable--the temp wiring harness branch to the temp sender. When you connect an external variable resistor to exercise the in-dash gauge, there's a high liklihood you've MOVED the temp sender's connecting wire. You just might be fighting a small cut/break in the wire's insulating sheath that comes and goes like Jeckly and Hyde based on the physical position of the interconnect wire....Comment
-
Re: C1 Temperature Senders
Yes, the ohmmeter is usually 11Megohm or greater input, therefore, nil current draw in a circuit. Loaded circuits obviously have a higher current draw, and this is the only thing I can think of....CraigComment
Comment