When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
Hey guys. I just installed a brand new auto meter mechanical water temp gauge into my turbo b16 1988 sedan. I did tjis for multiple reasons which dont really matter, but im getting some funky readings. I researched for weeks as to the proper location to put the probe in the coolant system. I came up with the upper radiator hose area being the best/hottest reading. I understand this will probably read 20-30 degrees hotter than the oem due to being at the exit point of the engine and just prior to hitting the radiator and being cooled. I read that some people were successful in using the stock bleeder screw location for a sending unit. So what I did was buy a new water neck and welded a bung in place for the probe to thread into. I did this toward the bottom at a "negative" angle as to keep the probe submerged at all times. Im typing this from my phone so ill attatch a photo of what im working with. So I installed everything and bled the system. Got lots of air out. Coolant is full. After a while of driving my temps are anywhere from 200-220, I would call that normal considering where the probe is. My oem dash gauge is the normal 1/3 point. Than after some cruising, my new temp gauges just climbs and climbs and pegs at 240+ while movng. Keep nin mind the oem still reads normal, hasnt budged at all. eventually is starts to go down, after slowing down some. Any thoughts? Heres some additional info that may help:
Water pump is OEM honda and maybe 25k on it along with tstat.
Capillary tube did not appear to contact anything when installed into the bung and water neck (I thought maybe heat soaking?
- I am running a cheap ebay slim fan, but today I ordered a SPAL 1700cfm fan and relay kit as a precaution, but this should only effect when sitting still.
I also read about putting a block in thr grill to force all the forward air into the 1/2 radiator.(im running the skunk2 alpha radiator for ef chasis) so I temp made some cardboard pieces to block off grill and force air over to the radiator. I drove for 1/2 hr which wasnt really long enough and ambient temps were substantially different from earlier in the day, so i dont know if this really helped yet.
-my crappy fan is a pusher do to space constraints in my bay, may effect free air travel when moving?
-hard to belive the new gauge is defective. It reads normal for a good while.
Open to thoughts, experiences,ideas...
Not sure if you read my post but I did bleed the system. The new water neck from honda does not have a bleeder screw port, it is cast closed. I didnt have a correct tap for a bleeder I had laying around. I cracked the flared nut fitting for the sensor at the bung on order to bleed. I also allowed to ide for 1hr, jacked up, rad cap off and let it bleed.
good idea on the temp gun. Ill have to wait and bring one home from work and try it.
I can attempt bleeding again.
Im really doubtful that the capilary tube is defective from new. They are filled with argon and if it was bad, chances are it would read inaccuratly all the time not randomly.
so today i took the car for a drive. temp gauge worked perfectly fine for about 40 mins than temps started climbing, while the oem cluster gauge was perfectly normal and my laptop logs read normal. so somethings not right. air would mess up the other ECT readings i would imagine as well as the autometer one. good news though on my air damn thing: while moving my temps remained much cooler(until my gauge started freaking out). so here is what i did:
came home and temp gauge was actually maxed out. let it cool down a bit, popped off rad cap and coolant resevoir cap (running a rywire coolant tank), ran the car and bled the system again at the radiator, kept topping off the fluid and i did get a lot more air out. temps stayed in check per the OEM gauge, and the autometer appeared normal for some time. about 30 min into it the temp maxed on the autometer again. i continued running it and at this point i was not getting anymore bubbles from radiator. oem gauge still didnt move, rad fan came on, ecm log of ECT was normal. only thing was the autometer gauge was maxed. is it possible this thing is heat soaking? im still waiting to borrow the temp gun from work so i couldnt hit it with the temp gun today. i mean chevy guys thread these things into aluminum intake manifolds all the time, i dont see why threading it into my aluminum bung would cause it to heat soak, unless the probe is contacting something inside that i couldnt see. but than it would not draw tight when i tighten the fittings , the probe has a flared fitting on the capillary tube much like a brake line except bigger, 5/8-18 unf to be exact, that is threaded into autometers brass adapter to 3/8npt which is threaded into the bung.
Last edited by 1988dxsedan; Sep 5, 2016 at 01:48 PM.
having the probe AFTER the thermostat is actually a really bad idea (or in your case in the water outlet, which is removed from the engine). You want to monitor engine temp, not radiator inlet temp. the only way to accurately read actual engine coolant temp is to install the bung as close to the ECT sensor as possible. that is the only location that has constant coolant flow from the main engine coolant circulation path.
there's a good chance the probe is also picking up some ambient temps from the turbo system, which can run upwards of 1000 degrees.
Coolant temperature should only be referenced from a known good measurement point. On most engines, this is the cylinder head itself, and not the cooling circuit between the hoses and the radiator. Readings from those areas are heavily skewed by environmental conditions like air temperature. The weld bung location is incorrect for your application. I have tried for a long time to get Ford Mustang owners from placing fan switches in the upper hose. A few have listened, and have engines that don't spout coolant everywhere. The real premise here is to detect an overheat at the ENGINE, not the RADIATOR HOSE. This allows the cooling system time to react to the data sooner.
Thermistors are very reliable in providing temperature feedback readings, and should be used when possible in factory, or near factory applications, as the ECM will use the readings in these spots to determine fuel modifications, and fan control. Not all thermistors and fan switches share the same housing. Be warned.
Okay what your telling me makes sense. Finally someone that knows. Everything I searched and found on this forum and others stated upper radiator hose or water neck where the bleeder screw is. So I put the bung right there, which to my ubderstanding is in the direct coolant path from the oem ect probe in the cylinder head. Flows right from there and out the water neck. Didnt realize that 1/2" from my bung to the head made a 80*+ difference in readings. So with your information:were to I put a 3/8 npt fitting in a honda cylinder head? I havnt found anything near the oem sensor that large.
I did run it with a temp gun. Idle for 55 min total. Bung surface temp never exceeded 180 degrees. I monitored ect with the laptop and it didnt exceed 205 even though the autometer pegged at max hot (260*+)
I removed the sensor and boiled it for 1 hr straight. Sat exactly at 220 the entire time. I even held so the probe touched the bottom of the pot (should have been hot as fu**) and the gauge didnt budge which tells me its not a heat soak issue.
no clue at this point. Even autometer tech support said at that location it may just read a little warmer that factory but it should be fine.
ECT sensors are normally positioned near the engine-side of the thermostat. The placement you have shown will not engage the fans until water reaching the cold side of the thermostat is overheated, and will be too late. Ideally this sensor is placed on the other side of the thermostat. In this scenario, once the engine-side of the cooling system overheats, the fan will operate.
Is the fan sucking in air into the engine compartment?
mechanical water temp gauge funny readings?
