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.
My concern with that location is that once the thermostat opens the flow through that particular hose is closed off. From your data it sounds like the opposite is being observed however. The sensor providing the signal that the ultragauge is displaying is in the intake mani where the coolant should be the hottest due to proximity to the point of heat generation (orange dot in the pics). It then exits the block and flows into that hose where flow is restricted by an open thermostat. That sounds counter intuitive but the thermostat opening closes off that particular hose where the new sensor lives (green dot). I'm not too surprised that you're getting a good reading off that location because of the heat transfer properties of the coolant, and some flow has to be occurring since the thermostat will allow some blowby. I tried to mimic a drawing I saw here, here's my (crudely drawn) understanding of the flow characteristics. When the engine is cold (thermo closed), the coolant flows out of the engine via that J pipe in the center (item 8), straight through the T and past the new sensor, then down into the top of the thermostat where it makes a 180 and goes up into the hose that feeds it back into water pump and through the engine again. The crux of the situation is that the thermo being closed is what forces the coolant straight through the T, and because it's closed the coolant can't exit the radiator through hose 15 so therefore it can't enter the radiator through hose 12 either.
However, when the coolant is hot and the thermostat is open, coolant can no longer enter the top of the thermostat because it's a dead end and it's instead forced to make a 90° turn when it enters the T, which sends it through the radiator instead. It flows through the radiator and out the bottom, then into the bottom of the thermostat, which is now open, and the coolant is diverted out the other hose where it previously made the 180 and does not flow past the extra gauge's sensor.
Obviously your gauge is doing its job so I don't think this is necessarily a problem or even a flaw in the design, but it seems not to be ideal due to the restricted flow when the thermo is open. Placing the sensor in the T itself, or in front of it (difficult), or in between the thermostat and the water pump (also difficult), should ensure that there's flow past it under all conditions.
Well this just got infinitely easier. I'm not sure why it took me so long to come around to it, I guess I was hung up on using the old sensor. Thanks for the inspiration, Extinct. The thread in the bleed tee is 1/8-27 ...
I wonder why I never measured that thread myself. Instead of installing the sensor directly into the bleed hole, could I use a tee like this?
Space would be probably enough, but I wonder if there is enough heat conduction to the temp sensor, so it can measure temperature accurately. The advantage of the tee is to keep the bleed screw and use a sender with a plug (I found some cheap 1/8-27 senders on ebay, TX71 - Jeep/Chrysler/Dodge around $7).
I measured the thread in the tee again with a caliper and is actually M10x1. Test with a nut says it is M10x1 definitely and not 1/8"NPT. A sender that would fit is, e.g.:
That's an interesting approach. M10X1 is a straight thread and 1/8"-27 is a taper, but the pitch crosses over nicely and you get the benefit of the taper to help with sealing. You still need the o-ring in there, I wouldn't try and seal it with threads alone but otherwise it worked fine for me using the VDO sender above. As with any sender of this type ground is required and for now that's a little schlocky, I used a ring terminal slid down over the post and held captive with a couple extra o-rings under the signal terminal. I need to employ a better approach there, but the hood closes without interference as well.
Pitch is different (1mm vs. 0.94mm). I tried with a M10x1 nut on a 1/8"NPT thread. It doesn't fit, not because of the taper, but the pitch. I can get it in a bit, but not to the full extend and I might damage the thread in the plastic tee. Not worth the hack, since M10x1 is easily available.
Not worth arguing about, you are right that there’s a difference in the pitch however metal on metal is different than screwing a metal sender into plastic threads. I was willing to sacrifice the slight thread mismatch for the better sealing the NPT offered, enhanced by the plastic stretching to accommodate the pitch variance of six hundredths of a millimeter times 3 or 4 turns. Not only does it sound inconsequential, it is evidenced by the picture of my application.
I’d encourage you to use the M10 in yours and post an update. For $30 I’d certainly throw this sender in a drawer and get the M10, but I question the ability to torque the M10 in a plastic thread to get it to seal without risking stripping the T. There’s a reason the M10 comes with a crush washer and despite not having to use it, I don’t think you’re going to get the same results with a metal on plastic straight thread, and this was the motivation for my decision to use the hack approach. In my view it’s still superior. Metal on metal is a different discussion.
The original M10 plug comes with a rubber O-ring. It will work the same way with a different M10 piece, given sufficient contact surface. I modified my brass plug, soldered some stranded leads to the NTC, isolated the leads with heat shrink hose. Now going to seal it with JB weld and need to wait for a pair of automotive plugs I just ordered to equip my Frankensensor with that. Will post an update as soon it is completed.
