So, I've been thinking about moving my temperature gauge sending unit to the thermostat housing for a more accurate read vs the stock way in drivers head. I think it has too much of a swing in temp especially after WOT. So moving it over to the housing seems like a better way of doing it and I will find out. So since I didn't want any chrome housing I saw this one from Dorman and I can polish it up since it's aluminum. I turned it over and saw how it was machined and thought that this was yet another machined piece of crap by EMS. WOW! not even a round hole, but egg shaped. The pic doesn't show it good enough, but it is noticeably egg shaped. I guess I'll touch it up, but this kind of crap machining even on the simple stuff is unacceptable. It was made in China?

For all the people that have sent me PM suggestions on where to have my manifold pours done and machined, this is a perfect example why I refuse to have it done in China or any other place outside of the USA. I would have ZERO control on the machining again and nothing but issues that I don't want to deal with. It's hard enough being here and having to deal with poor quality machining. Ugh!

 

you realize GM made hundreds of thousands of t-stat housing with a sensor hole in them for OEM uses.

 

Yes I remember throwing one of them away. Pretty ugly.

 

What does GM making tons of these have to do with anything, but I thought for $11...what the heck. Just didn't think the quality would suck like that, but I can fix it. - FIXED already!

 

Personally, I think you would get more temperature swings there due to the open/closing of the thermostat. Mine is in the stock location for the gauge, plus I have a sensor in the intake manifold crossover for the EFI. They match up pretty closely on readings.

 

I' m with Steve. Seems like the "simulated" temps would be iffy at best at the water-neck. That housing should be the hottest reading of the entire cooling system because that coolant is on its way to the Rad.
I would think the most consistent and accurate temp readings would be in the head. Those GM engineers were not stupid.

As far as the egg-shape in the housing? Doesn't matter. Could be square, as long as it works. The thermostat doesn't care about the shape.

 

It is an experiment on this project to see if it is any better, maybe not. I use my IR gun and check temps at the housing and compare that to my laptop temps and the gauge. The gauge is off high, so I got the variable resistor to tweak in the gauge more inline with what the housing temps are.

 

The one that came on my car actually had a crack like hole in it so I ordered the one you have or one very similar.. The only issue was getting those plugs to seal. you could screw them all the way in and they never got near tight, Combination of RTV and Teflon tape finally did it...

60

 

I would highly suggest in the future that you use Permatex 56521 High Performance Thread Sealant, it works wonders. No tape required.

I was also just looking at this housing after polishing it up and looking good that I will have problems with the temp sensor sticking up maybe a little high and will have to remake my hard fuel lines. Not sure if I want to do that right now, so I may wait a little on this swap and go with the adjustable resistor on the water temp gauge first and see if that works. If so, great. If not, I will start bending tube again.l. It works like a champ every time.

 

Perhaps, before any retooling for your manifold, you create a change to include a threaded port with temp sensor in mind.

 

Just curious... all those cars that used the "041" and similar type heads that did not have provisions for temp senders in the heads.... where did those smart GM engineers put the temp sender??
I would think the water neck would be the ideal location, if hot there it's hot at the head.

 

I don't know. But the waterneck is last place where the coolant is before it makes its journey to be cooled. The coolant has picked up every ounce of heat from the block, heads and Intake. Then exits. So logically, the waterneck is the hottest coolant spot. So I would think tapping a sender into that area would give inaccurate / too high of readings.
But, better your gauge reads too high, than lower than actual temps. Just my theory.

 

So, this is gotten a tad out of scope, I think. I intended to just show what shoty machine work was done on a very simple piece and then explain what I was going to do. Yes, it doesn't affect the operation of the stat or cooling system that it was machined egg shape on a round hole, but it just looks damn bad on their part. Kind of like the shoty machine work on my manifold that EMS did on a lot of the T-stat holes, it was functional, but it looked like **** and I would not have put that out if I saw that first.

So, that brings us back to everyone's opinion on where the temp sensor should really be for a more accurate read for a gauge. Everyone has their right to their opinion and there will be many I'm sure, but who is right? I do not see any reason why the sensor can not go in the housing (since it was used there as well by GM) vs going into the head. Like I mentioned before, this is an experiment to see which works best for me and not me trying to reinvent the wheel. As far as GM engineers back then being pretty smart... You couldn't prove that by me on some things they did on the 82, one example... I would not think that putting an ECM (which is electronics) into a boxed in compartment with liquid acid battery, and the fumes in that environment would be a great idea or an engineering marvel in 82. People wonder why their ECMs go bad, I wonder why? Corrosion comes to mind and seen a lot that were very bad, components and ECM connectors.

The temp gauge in an 82 is not linear and I do not think that any year C3 gauge is either, but I can not speak on that question. I'm willing to bet that Wilcox will say the same thing about the gauge that it is not linear, but their adjustable resistor makes it somewhat linear and may work just fine for me and what I'm trying to achieve here. I don't think putting the sensor in the housing would be anymore susceptible to swings than the head placement and may in fact be better, but we shall see.

At WOT, the sensor starts reaching higher temps much quicker because of where it is placed which is also because of the aluminum heads I would think. By putting the sensor in the housing, it should see a more true reading of the actual water temp just before it goes into the radiator which is fine for me and if it's the hottest temp, I want to know what that temp really is at a stable reading. Also like mentioned before, I run a 180* stat and it opens at 180* according to all my readings. The motor seems to stay steady at 190-195* by my laptop readings and IR gun which are very close and only a couple degrees different. However, my gauge starts off by staying about 195-200* and then starts climbing for no apparent reason and looking at the laptop it is cruising along at 190-195* I actually thought the gauge was bad at one point, but I think it is just more stable and accurate at the lower temps, hence the not linear thing coming into play.

The reason I say that is because for another experiment that I was doing for another unrelated issue, I removed the center out of an old stat and installed it in my motor. So, effectively, it was just acting like a restrictor plate and no longer as a stat whereby it was just letting full water flow all the time and the motor would run at whatever the cooling system temp was that it was going to run. My motor ran fine at around 165/168*, IR said about 165*, gauge was reading just fine at around 165/170* and steady and the laptop was reading very close to that reading. Huh? Is that gauge actually bad? I didn't think so, but I do have a brand new NOS AC Delco temp gauge still in the box if I wanted to swap it out. My motor did run just fine at those temps, but since it was just another experiment, I didn't leave it that way to run at that temp which is a bit cold for my liking. The DeWitts radiator works just fine.

Anyway, thank you to all that posted your thoughts and I will see what happens and maybe post up the data for member's to mull over.

 

you know a IR gun and a few seconds of your time will answer a lot of your guess

 

Dunno if there's a "best" location.
IIRC, all scratch built race-only cars I've been around have the thermocouple/sender in the intake. That's certainly much more handy location than in head.

 

Yes, the T-stat housing is theoretically the hottest location as it's the last place before the radiator, that's only AFTER the t-stat opens. Any time the t-stat is closed, it's not getting an accurate temp reading of the coolant. So a temp probe in the neck works fine on race-only cars that don't run a t-stat, but on a street car it's less than ideal.

You could put it in the water crossover of the intake manifold so it gets the hot water before the t-stat (thus, doesn't have to wait for the t-stat to open), but I don't see the point. The front of the driver's side head is pretty much the last place before the t-stat so it's not a significant difference. Besides, gauge indicated water temp doesn't need to be a laboratory spec reading as it's just an estimate of overall engine temp.