Wonder if that scoring indicates that the water pump is somehow binding and not spinning as fast as designed?

With the new thermostat installed, have you re-checked the inlet/outlet temps on the radiator?

If that delta still exists, you have flow issue somewhere.

I honestly don't think this is the case. Before I removed the pump I checked for binding and it spun freely. I also dry fitted the pump to the housing without a gasket and even then it didn't bind. Also, I didn't notice any belt squeal. I bet that scoring happened long ago!

The delta does still exist! After I swapped in the new OE thermostat I took it for a spin and remeasured all the temps. It was ~75-80C on the top and 45C on the bottom. Since everything seemed to be flowing freely during my hose flush, I theorize it is a water pump issue or an extremely stubborn air-lock/bleeding issue.

At this point, I consider the housing scoring to be suspect and I will replace the housing. I went ahead and grabbed an alleged good used housing online, in addition to a water pump balance banjo tube since it was not present on late '85's.

Should have added, weak heat is another symptom of poor flow. The temp dropping while the heater is running is normal. Think of it as another mini radiator.

What’s the state of your hoses? Is it possible you have a kinked hose somewhere?
At temp, the top radiator hose should be full of hot coolant. If you’re getting no resistance when you squeeze it, you have problem.

Check your hoses and make sure they’re not collapsing/collapsed during operation.
It’s also worth doing a very thorough air bleed. Get the nose up as high as you can —ramps or equivalent should be sufficient. I use the top radiator hose to refill the system. When it’s full, reattach it and top off through the radiator if needed. Keep the rad cap off. Turn the heat dials to full. Run the engine until 82-85 C. Should take 20mins or so. There should be zero air left in the system after that.

I'm late to the party. I believe I read that you replaced the red cap. Did you pressure test the cooling system? While less likely than other items, it is possible that you might have a head gasket issue. Either way, a system pressure test (over a few hours) will help determine if you have a leak somewhere.

Good luck.

The scoring in the pump housing almost certainly is old; the tan discoloration is the tell.
The scoring, per se, is unlikely to be the culprit. Excess clearance between the impeller and the housing cone is more likely. How could there be excess clearance? 1) The cone of the housing was machined too deep during initial production. 2) The housing was re-machined at some time after production. 3) The impeller is located too close to the pump gasket surface.

A comparison of the present housing with the replacement one should provide some insight as to the depth of the conical surface.

The head gasket/head failing is a possibility that I’ve purposefully not put much thought into for my own sanity. :) The car shows no symptoms of a failed head gasket other than running slightly hot. This combined with the big gradient across the radiator, no heat, and air in the top tube point me towards a less serious coolant flow issue.

I did pressure test the system, although my test kit was a bit wonky and didn’t seal great to the expansion tank. Here’s my coolant system pressure test results from page 1:

• Holds 15psi for a few minutes
• Holds 8psi for an hour
• Holds 5psi indefinitely (overnight)

Does that seem normal?

I will not discount the possibility of a head gasket failure however! Is there anything I can do to check for headgasket leaks while I have the cooling system apart? I suppose I could start it for a minute or two and try to observe combustion gasses exiting the cylinder head coolant passage.

That's good that you tested the pressures. The readings are hard to comment on because pressure is a function of temperature. Typically once you are at operating temp, the system should hold pressure as rated on the radiator cap for at least 30 min. It will decrease over time as the system cools. So in hot summer weather, the pressure drop will be less and over a longer period o time than in the winter. If a combustion leak, it will get really high, causing the overflow tank to fill, and after reaching operating temp, shutting off the engine, if you were to open the rad cap, you would have a large amount of escaping coolant/pressure. Also, you upper/lower rad hoses would be quite firm after a few minutes post shut down.

Unfortunately, none of this can be tested with the system open as you have it now.

The only other suggestions I can offer is to thoroughly clean your radiator, making sure no leaves or debris are in between the condensor if equipped and radiator. I would also recommend a citric acid flush of the radiator itself to make sure it is not coated with oil internally restricting heat transfer.

The odd thing is your temperature differential in/out of the rad. Seems like it's functioning well, which would also indicate the pump is flowing. Possible restriction between the pump and head, maybe reverse flush between the two if disassembled.

I'm out of ideas on this one. Best of luck

Alright, I am back with more. Novel incoming, so here's a TL;DR: the water pump and housing was not the issue, nor is the "issue" solved. However, I'm back to suspecting my temp gauge is significantly wrong after re-checking temps again by making direct contact measurements with a thermocouple.

I replaced the water pump and housing with a good-used assembly. This pump housing seems to have been lightly used and had significantly less scoring than the original. The new is the first image, the original is next image.
https://i.postimg.cc/nCY4xpdg/IMG-6438.jpghttps://i.postimg.cc/hX2VJMR0/IMG-6441.jpg

The pump seems to be a NOS or rebuilt "GK" unit, and listed "west germany" as its location of manufacture. So it's at least old. :) It spun smooth and the seals felt that they had resistance so I went ahead and used this "new" one. When comparing the impellers, they're slightly differently angled and the length from the gasket surface to the end of the impeller differed within my measurement capabilities error.
https://i.postimg.cc/30jv3FZm/IMG-6443.jpghttps://i.postimg.cc/7Cc7966G/IMG-6442.jpghttps://i.postimg.cc/fSQd4rYM/IMG-6444.jpg

I also installed a balance tube since my car was missing this between the water pump housing and head. Seems to be a '85 thing. Also changed the short tube and drive belts while I was in there.
https://i.postimg.cc/ygcJTytt/IMG-6452.jpg

I pressure tested the cooling system to 10PSI, filled it up thru the reservoir, and then did some extra "burping".

• I first jacked up the front end way higher than I had previously. No burps.
• Then I pulled the top-tube off of the thermostat housing and filled the pump assembly until it overflowed.
• I then filled the radiator using the top hose. Once the tube overflowed, I pinched it off and quickly plugged it onto the thermostat housing. This resulted in a higher coolant level in the expansion tank, and otherwise had no obvious effect on the amount of air in the top tube.
• I turned on the key to the glow position and turned the heat on max. I heard the aux-pump running and it sounded like it had air in it. I popped off it's output tube (that connects to the water pump housing) and let the coolant flow out a bit to try to clear any air.
• I then started the car and observed no coolant level drop in the expansion tank. I let the car come up to temp (per the gauge) at idle, revved it a bit to clear any air. Again, no coolant level drop was noted.
• I drove the car for 10 minutes and observed no coolant drop and the same gauge behavior as before, except now I had cabin heat! Woo!

Still, the gauge climbed to mid-90's*C, and close to 100*C with the AC on. Cabin heat on reduced the engine temp slightly as expected.

Frustrated, I did a sanity check. Instead of my IR thermometer, I used a thermocouple reader. I taped the thermocouple wire to a pencil and used the eraser to push the thermocouple bead onto different surfaces.

https://i.postimg.cc/t7dZZDhV/IMG-6457.jpg

Interestingly, I measured 81.1*C at the base of the temperature sender! Check this out:

Gauge reading: 98*C-ish
https://i.postimg.cc/kRyB0Xny/IMG-6456.jpg

Temp Sender base: 81.1*C
https://i.postimg.cc/bs3ZLJVy/IMG-6458.jpghttps://i.postimg.cc/tYX1hP5v/IMG-6459.jpg

Cylinder Head: 81.1*C
https://i.postimg.cc/k2HB3HdL/IMG-6460.jpghttps://i.postimg.cc/ftcLdT0h/IMG-6461.jpg

Thermostat Housing Inlet:80.6*C
https://i.postimg.cc/5YdyjFy3/IMG-6462.jpghttps://i.postimg.cc/v16DQvYT/IMG-6464.jpg

Thermostat housing outlet (this seemed a little low honestly):47*C
https://i.postimg.cc/BtfMHqV7/IMG-6467.jpghttps://i.postimg.cc/w7624wHh/IMG-6468.jpg

And as a double-triple-sanity check, I checked this thermocouple vs a meat thermometer vs my electric kettle display at tap-water boiling point (100C). Each read within a few *C of each other.
https://i.postimg.cc/fVyKDkVy/IMG-6479.png

I think I may have been lead astray by a combination of a faulty dash gauge and an IR thermometer reading high. This also backs my observation that no-matter what I change the needle acts EXACTLY the same. I think everything was mechanically working fine from the start.

What do you folks think? At least I got my drive belts done. :)

I continued down my engine temperature rabbit hole today.


First I tested the temp sender with my electric kettle. I set the kettle to 85*C and suspended the sender in the water such that I could access the pin on top with my DMM. I checked the resistance at 85*C and it was around 54ohms. This was within the VDO chart tolerance, so this sender checks out.


Then I borrowed a block tester from a buddy to see if I could see combustion gasses in the coolant. I took the car from a spin, and got the temp gauge up to the usual 95-98*C. While idling I used the block tester bulb to suck up air from the expansion tank. No color change of the fluid was observed. I then used the radiator top tube to 'squish' more water into the expansion tank and force air thru the tester. Again, no color change was observed. While this is probably not a conclusive test, I feel that I would have seen the fluid change color if there was enough of a leak to bubble coolant out of the head.
https://i.postimg.cc/0zD9bmMk/IMG-6480.jpg


With the car still at idle, air conditioning on max, and temp gauge reading around 98C, I measured the resistance at the sender to be 55.7ohms. This tells me that the car is actually running between 84*C and 90*C based on the tolerance of the sender. I feel this generally aligns with my thermocouple measurements, since a loss of a few *C through the body of the sender is expected.
https://i.postimg.cc/DWcFcTPR/IMG-6486.jpghttps://i.postimg.cc/gXkpSTMj/IMG-6483.jpg

I believe this confirms that my gauge is reading about 10*C too high. Also, while plugging and unplugging the temp sender connector, the gauge sometimes read as high as 110*C. I would then re-plug the connector and the temp would return to around 100*C. Next steps will be pulling the cluster to investigate grounds and burnt traces.

Great,
make sure you check the old forum posts about iffy or jumpy temp gauges , like:

https://www.peachparts.com/shopforum/diesel-discussion/383571-erratic-temp-gauge.html

You may want to use google advanced search for better results.

P.S. IR thermometers can be tricky, even those with adjustable emissvity (and yours isn't. It is pre-adjusted @ 0.95 emissvity factor). Always use the same surface when comparing differences (electrical or masking tape being the most convenient).And even then one can get wrong readings.

It turns out TCouples can be wrong too (I did not know that...some alloys can rust). It may be necessary to sacrifice/lose one or two during a test,lol...aspecially wthen the ways to attach them to the surface is limited (JB weld ,Gorilla tape etc.).


Happy Fourth of July !

Bingo! Thanks Vox. Problem (mostly) solved. Maybe even two birds with one stone...

I tore into my instrument cluster to see what I could find. A few initial observations:

• My oil pressure gauge was leaking at some point and coated most of the multi-gauge innards with a film of oil.
• No obvious burnt traces were found.
• The main grounding pad was oxidized and stained with oil.
• The little wire spool spindle on my fuel gauge was cracked and the spool was slightly loose.
• There was a bit of oil between the ground pads and the temp/fuel gauge actuator grounding mating surface.

https://i.postimg.cc/Ff4Ty7ZV/IMG-6492.jpghttps://i.postimg.cc/5HSnnbKd/IMG-6493-1.jpg

I started by ringing out all of the ground traces, they seemed mostly ok. No smoking guns or cracked joints found. Some of the pads seemed to give 'cleaner' readings than others.

I disassembled the multi-gauge instrument assembly from its PCB.

I cleaned the PCB with some de-greaser and lint-free cloths and swabs.

I re-flowed the main grounding tab solder joint and reinforced the joint with a small piece of wire. I then polished the pad on both sides of the PCB with a fine scouring pad to remove the corrosion and staining.
https://i.postimg.cc/9DTxwGGf/IMG-6493.jpghttps://i.postimg.cc/k6jwPwdT/IMG-6495.jpg

I re-flowed all of the other solder joints on the PCB.
https://i.postimg.cc/CZS4Z9hv/IMG-6502.jpg

I added a small jumper wire from the ground point on the temperature gauge PCB to the shell of the gauge actuator as shown in the thread that Vox posted. I sanded away any coating that may have been on the shell before soldering.
https://i.postimg.cc/dh8BVJYk/IMG-6501.jpg

I added a dedicated ground wire to the metal back panel of the cluster at the main grounding pad screw. I attached the other end of the wire to the big stack of grounds behind the cluster. I used a single-wire connector that I had on hand so that I can easily remove the cluster without an additional tool.
https://i.postimg.cc/5QY5dVgt/IMG-6504.jpghttps://i.postimg.cc/6y4cFdbh/IMG-6499.jpg

These mods seems to have made the gauge match my sender resistance measurements more accurately! Instead of the usual 98*C-110*C, the gauge now shows 85*C-90*C. For reference today was ~90*F in Maryland, and I had the AC on max.
https://i.postimg.cc/V5pB9Js0/IMG-6508.jpg


Even though we're still a few degrees above 82*C, I feel much more confident in the health of the cooling system and the engine in general.

As a bonus, my fuel gauge seems to be reflecting a more accurate fuel level than before. I wouldn't see any fuel gauge movement for around ~150miles. Now I see gauge movement at <100miles.

I won't say /thread yet. Don't want to jinx it. But this seems promising. Thank you all!!!

Interesting, and thanks for the update. Fingers crossed!