3rd Time Trying to Get Temp Gauge to Read Right - No Luck

[1987 GMC Jimmy]

I've replaced the little sender thrice now trying to get a proper coolant temperature reading. It sucks because all my other instruments work correctly now. The first one was from Autozone about two years ago, and it read about three quarters to pegged hot. I replaced it again this summer with a NAPA unit. It worked pretty okay for a little while. It read about fifteen degrees over, which is dead center on the gauge. After a while, it started pegging almost all the way to the right, about where the red line starts, and then it would move back a little bit to about three quarters. I replaced it again tonight with a BWD one from O'Reilly, and it's doing the same exact thing. Are all these just made in the same sweatshop and I'm running around in circles trying to make my gauge work, or am I having a PC issue? The only thing that's fishy with the cluster is that sometimes the bulb to the left of the speedo goes out, and the voltmeter reads to low when that happens so I smack the top of the dash, and the bulb comes back on, and the gauge reads closer to where it's supposed to. I just noticed the voltmeter doing that in conjunction with the bulb because I thought the bulb's connection was just bad before. The truck heats up normally, and it reads 197* warmed up at the cylinder head with an infrared temp gun.

 

[Honky Kong jr]

Could be a bad gauge. Could be a broken wire that is screwing with the values.

 

[74 Shortbed]

Autometer...

 

[yevgenievich]

Check all grounds to the cluster and between engine and body. Check continuity on the sender wire

 

[chengny]

Are all these just made in the same sweatshop and I'm running around in circles trying to make my gauge work, or am I having a PC issue? The only thing that's fishy with the cluster is that sometimes the bulb to the left of the speedo goes out, and the voltmeter reads to low when that happens so I smack the top of the dash, and the bulb comes back on, and the gauge reads closer to where it's supposed to. I just noticed the voltmeter doing that in conjunction with the bulb because I thought the bulb's connection was just bad before. The truck heats up normally, and it reads 197* warmed up at the cylinder head with an infrared temp gun.

For these gauges to work right, there needs to be two separate paths to ground. One is for the control coil and the other one is for the sensing coil. As I am sure you know, the control coil is run straight to the ground bus block - and as such should have essentially no resistance.

On the other hand, the coil for the sensing leg sees a variable resistance to ground. How much resistance the sender generates is dependent on the conditions that the sender is subjected to at the time (i.e. low coolant temperature = high resistance and higher temps result in lower resistance).


The coils both share a common voltage supply.

If the indicator pegs high when the DK GRN 35 sensing leg is grounded that would seem to indicate that the sensing side of the gauge circuitry is good.


But remember, these two coils are constantly fighting each other over where the needle should be positioned. The sensing coil is attempting to move the needle to HOT while the control coil is trying to pull it to COLD.


As noted above, the coils share a common voltage supply – so that isn’t generally an issue (unless the indicator has internal problem).

So what it boils down to is this; whichever coil has the least resistance to ground – and consequently the greater current flow through its windings – will be the coil that exerts the greater pull on the needle. Since you are getting an artificially high indication, it would appear that the sensing coil is winning the battle. And because the control coil is wired straight to ground it should always win - at least in the beginning. But then, as the coolant warms up, the current flow through the sensing coil starts to increase and begins to pull the needle off the low peg.

Bottom line: It looks like there is a problem with the control coil’s grounding circuit. Ignore the comment about the DK GRN 35 lead in the dwg below - it was for someone else:

You must be registered for see images attach

I would pull the cluster and check the connector clip for the control grounding leg. Especially the one in the "G" spot (hahaha sorry) - "B" is for battery and "S" is for sensing.

You must be registered for see images attach

Those clips can get oxidized (especially down on the levee) and loosen up over time.

You must be registered for see images attach

Check for clean tight contact at both the indicator studs and also where the ears interface with the printed circuit. Usually they can be salvaged with just a pencil eraser or some fine emery.

 

[hiltijoe]

Autometer...

yep that's what I did put a mechanical gauge both for temp and oil pressure

 

[74 Shortbed]

yep that's what I did put a mechanical gauge both for temp and oil pressure

 

[chengny]

Some folks think their truck looks good in stock condition and would rather diagnose and repair an issue to keep it that way.

 

[1987 GMC Jimmy]

For these gauges to work right, there needs to be two separate paths to ground. One is for the control coil and the other one is for the sensing coil. As I am sure you know, the control coil is run straight to the ground bus block - and as such should have essentially no resistance.

On the other hand, the coil for the sensing leg sees a variable resistance to ground. How much resistance the sender generates is dependent on the conditions that the sender is subjected to at the time (i.e. low coolant temperature = high resistance and higher temps result in lower resistance).


The coils both share a common voltage supply.

If the indicator pegs high when the DK GRN 35 sensing leg is grounded that would seem to indicate that the sensing side of the gauge circuitry is good.


But remember, these two coils are constantly fighting each other over where the needle should be positioned. The sensing coil is attempting to move the needle to HOT while the control coil is trying to pull it to COLD.


As noted above, the coils share a common voltage supply – so that isn’t generally an issue (unless the indicator has internal problem).

So what it boils down to is this; whichever coil has the least resistance to ground – and consequently the greater current flow through its windings – will be the coil that exerts the greater pull on the needle. Since you are getting an artificially high indication, it would appear that the sensing coil is winning the battle. And because the control coil is wired straight to ground it should always win - at least in the beginning. But then, as the coolant warms up, the current flow through the sensing coil starts to increase and begins to pull the needle off the low peg.

Bottom line: It looks like there is a problem with the control coil’s grounding circuit. Ignore the comment about the DK GRN 35 lead in the dwg below - it was for someone else:

You must be registered for see images attach

I would pull the cluster and check the connector clip for the control grounding leg. Especially the one in the "G" spot (hahaha sorry) - "B" is for battery and "S" is for sensing.

You must be registered for see images attach

Those clips can get oxidized (especially down on the levee) and loosen up over time.

You must be registered for see images attach

Check for clean tight contact at both the indicator studs and also where the ears interface with the printed circuit. Usually they can be salvaged with just a pencil eraser or some fine emery.

Check all grounds to the cluster and between engine and body. Check continuity on the sender wire

I'll do just this. Of course, my break is over so it'll have to wait a few weeks, but I need to check on this and the other small aforementioned issues, lube my speedo cable, and I've got another, supposedly working, Delco 2700 radio to swap in so I'll kill all these at once. I did ground the sensing side, and it pegged way right at the three o'clock position so that's good. I will also say that early in my exploits, when the speedometer was the only thing that worked, I swapped in another cluster, and the coolant temp gauge did the same thing. I need to check everything behind the cluster, but I should probably check the cluster ground, huh? Where would that be? By the way, all my grounds under the hood are good, and this is the only electrical concern I have. Everything else is tip top.

 

[crazy4offroad]

IIRC there are year breaks for the temp gauges. IE, the resistance will be different for certain year breaks. You may have an older temp gauge, using a newer sender will give a false reading. Mine does the same thing yours is doing, I just ignore it. I know what the temp is based on an autometer gauge I have mounted near the shifter. And, I know my instrument cluster is a cobbled together Johnny Cash Cadillac pile of mismatched parts.

 

[Honky Kong jr]

I have some 89 gauges when you get to that point. I was gonna keep it complete but I robbed the fuel gauge out to fix mine so bets are off so to speak now it’s parts

 

[yevgenievich]

There is a colored resistor on the back that indicates what sender to use.

 

[chengny]

IIRC there are year breaks for the temp gauges. IE, the resistance will be different for certain year breaks. You may have an older temp gauge, using a newer sender will give a false reading. Mine does the same thing yours is doing, I just ignore it. I know what the temp is based on an autometer gauge I have mounted near the shifter. And, I know my instrument cluster is a cobbled together Johnny Cash Cadillac pile of mismatched parts.

c4or is right about the necessity of having a compatible sender/indicator pair. I was going to ask you about that possibility - of a mismatch - but didn't want to confuse the issue. Your truck is way newer than when the last change in sender/indicator specs were made - in 1979. So you are probably okay on that issue. I don't even think you can jam a 1978 cluster into a 1987 dash - or mount the indicators in the bucket


For Truck Years: 1967 - 1973

The Temp Gauge Needle Should Point to:
Left Line (Cold) when sender resistance = 350 Ω
Middle Line when sender resistance = 76 Ω
Right Line (Hot) when sender resistance = 51 Ω

For Truck Years: 1974 - 1978

The Temp Gauge Needle Should Point to:
Left Line (Cold) when sender resistance = 350 Ω
Middle Line when sender resistance = 68 Ω
Right Line (Hot) when sender resistance = 46 Ω

For Truck Years: 1979 - 1990

The Temp Gauge Needle Should Point to:
Left Line (Cold) when sender resistance = 1,365 Ω
Middle Line when sender resistance = 96 Ω
Right Line (Hot) when sender resistance = 55 Ω

I will also say that early in my exploits, when the speedometer was the only thing that worked, I swapped in another cluster, and the coolant temp gauge did the same thing. I need to check everything behind the cluster, but I should probably check the cluster ground, huh? Where would that be? By the way, all my grounds under the hood are good, and this is the only electrical concern I have. Everything else is tip top.

This is good info I wish you had shared it earlier.

1. I have a 1987 cluster in my lap right now. I traced the etching for the coolant temp indicator all around the board and back to the harness socket - that pin grounds a lot of stuff on the PC. If there were a problem with continuity to ground on that common leg, in addition to the coolant gauge problem, you'd have problems with about half the board's components.

2. Because you changed the cluster and the gauge behaved the same, it is most likely not a grounding problem internal to the cluster. So forget my suggestion about it being a bad connection at the stud/clip/PC interface. We kind of have to assume that the cluster is good.


3. Also, if the rest of the components that share that ground leg are okay, that would pretty much eliminate a continuity issue between the control coil/cluster and the grounding bus block. BTW - the final grounding point for everything inside the cab is at the ground bus block. That block is located up above the E-brake on the outside wall. It is grounded to the sheet metal and the sheet metal is grounded to the frame..... Ultimately all grounds end up back at the battery neg post.

4. Keep in mind that you are having a problem with the indicator reading high - not low. As described above (and as you now know) there are basically only two conditions that will cause that; a short to ground (or a ground fault) in the sensing leg of the gauge - between the gauge and sender - or a poor connection to ground in the control coil's leg.

To summarize:

You seem to have confirmed a good cluster PC and coolant gauge (by replacing with a known good).

And because everything except the coolant gauge is working (and it shares the same etched ground circuit with lots of other stuff), the cluster is properly grounded externally.

Battery voltage supply to the various components in the cluster is similar to the ground loop - it services a number of things. So, if that were a problem - again, like a bad ground it would cause you a lot more trouble than only the coolant gauge.

So it follows that, if the gauge has battery voltage to the coils, the gauge is good, and the control coil is properly grounded...the high indication at the gauge is due to a wiring issue somewhere in the lead between the sensing coil and the sender.

Have you ever traced that lead (DK GRN 35) and inspected it for insulation failure. After it leaves the loom behind the distributor, it travels through some tough neighborhoods - for about 3 feet. It is designed to be secured to the top of the LH valve cover with these nice clips:

You must be registered for see images attach

but, by the time you or I get the truck, they are usually long gone. The DK GRN 35 usually gets shoved onto the intake manifold and sits in hot grease/oil for years. Even worse, a lot of times it is allowed to hang down near the exhaust manifold.

I would take close look at that lead all the way from the point where it comes out of the loom to the sender button. Wipe it down down clean with a paper towel soaked in gasoline, then twist/bend it always looking for signs of insulation breakdown. It doesn't like you will find any actual conductors poking through (you might). More likely the insulation will be chafed/pinched/squished/cracked allowing a ground fault to develop at operating temperature.

Another option would be to cut the DK GRN back by the firewall and splice in a temporary lead. If it doesn't help, just splice the OEM back together. This is when meggers come in handy - but I'm sure you don't have right?

 

[1987 GMC Jimmy]

c4or is right about the necessity of having a compatible sender/indicator pair. I was going to ask you about that possibility - of a mismatch - but didn't want to confuse the issue. Your truck is way newer than when the last change in sender/indicator specs were made - in 1979. So you are probably okay on that issue. I don't even think you can jam a 1978 cluster into a 1987 dash - or mount the indicators in the bucket


For Truck Years: 1967 - 1973

The Temp Gauge Needle Should Point to:
Left Line (Cold) when sender resistance = 350 Ω
Middle Line when sender resistance = 76 Ω
Right Line (Hot) when sender resistance = 51 Ω

For Truck Years: 1974 - 1978

The Temp Gauge Needle Should Point to:
Left Line (Cold) when sender resistance = 350 Ω
Middle Line when sender resistance = 68 Ω
Right Line (Hot) when sender resistance = 46 Ω

For Truck Years: 1979 - 1990

The Temp Gauge Needle Should Point to:
Left Line (Cold) when sender resistance = 1,365 Ω
Middle Line when sender resistance = 96 Ω
Right Line (Hot) when sender resistance = 55 Ω





This is good info I wish you had shared it earlier.

1. I have a 1987 cluster in my lap right now. I traced the etching for the coolant temp indicator all around the board and back to the harness socket - that pin grounds a lot of stuff on the PC. If there were a problem with continuity to ground on that common leg, in addition to the coolant gauge problem, you'd have problems with about half the board's components.

2. Because you changed the cluster and the gauge behaved the same, it is most likely not a grounding problem internal to the cluster. So forget my suggestion about it being a bad connection at the stud/clip/PC interface. We kind of have to assume that the cluster is good.


3. Also, if the rest of the components that share that ground leg are okay, that would pretty much eliminate a continuity issue between the control coil/cluster and the grounding bus block. BTW - the final grounding point for everything inside the cab is at the ground bus block. That block is located up above the E-brake on the outside wall. It is grounded to the sheet metal and the sheet metal is grounded to the frame..... Ultimately all grounds end up back at the battery neg post.

4. Keep in mind that you are having a problem with the indicator reading high - not low. As described above (and as you now know) there are basically only two conditions that will cause that; a short to ground (or a ground fault) in the sensing leg of the gauge - between the gauge and sender - or a poor connection to ground in the control coil's leg.

To summarize:

You seem to have confirmed a good cluster PC and coolant gauge (by replacing with a known good).

And because everything except the coolant gauge is working (and it shares the same etched ground circuit with lots of other stuff), the cluster is properly grounded externally.

Battery voltage supply to the various components in the cluster is similar to the ground loop - it services a number of things. So, if that were a problem - again, like a bad ground it would cause you a lot more trouble than only the coolant gauge.

So it follows that, if the gauge has battery voltage to the coils, the gauge is good, and the control coil is properly grounded...the high indication at the gauge is due to a wiring issue somewhere in the lead between the sensing coil and the sender.

Have you ever traced that lead (DK GRN 35) and inspected it for insulation failure. After it leaves the loom behind the distributor, it travels through some tough neighborhoods - for about 3 feet. It is designed to be secured to the top of the LH valve cover with these nice clips:

You must be registered for see images attach

but, by the time you or I get the truck, they are usually long gone. The DK GRN 35 usually gets shoved onto the intake manifold and sits in hot grease/oil for years. Even worse, a lot of times it is allowed to hang down near the exhaust manifold.

I would take close look at that lead all the way from the point where it comes out of the loom to the sender button. Wipe it down down clean with a paper towel soaked in gasoline, then twist/bend it always looking for signs of insulation breakdown. It doesn't like you will find any actual conductors poking through (you might). More likely the insulation will be chafed/pinched/squished/cracked allowing a ground fault to develop at operating temperature.

Another option would be to cut the DK GRN back by the firewall and splice in a temporary lead. If it doesn't help, just splice the OEM back together. This is when meggers come in handy - but I'm sure you don't have right?

No, no megger. I'll take a look at the insulation more thoroughly than I have. At some point, probably a year and a half ago, a big ole rat came, gnawed on the wire, and damaged the insulation about six inches or so from the connector. I repaired it a while ago hoping it would improve or fix the issue but it didn't. I had this issue or a slight variation of it since day one. My sensing wire doesn't neatly travel like the one in the above picture, and I have no clips on the valve cover. It just dangles around near the steering shaft. I don't think it's an earlier gauge, but even if it was, it's in the 1981 and later style gauge font so it'd still be using the same resistance parameters per your above chart.

 

[CRM]

When I first started restoring my 77 my temp gauge didn't work at all. I eventually went to two different name brand stores and bought temp senders but the reading from both was at the 3/4 mark or higher. I finally went into the reference library on this sight and downloaded the service/parts manuals for my truck. I looked up the part number and went to the local AC Delco store. Now it goes up to just under 3/4 and when the tstat opens it drops to 1/2.