Sounds like you need a TIG welder.

It's been on my list for years.
But this last year, it's been rapidly climbing to the top of my list.

Just trying to help!!! You also need a mill to go with that lathe....

And, perhaps, a bigger garage.

The bigger garage is at the top of the list.
We're simultaneously shopping for a new place and shopping estimates to build a garage on our current property. It's not at all stressful.

I went through this a LONG time ago.

The fin mounted sensors do not work because of the fluctuation in fin temperature that is ambient temperature, load and airflow dependent. Yes, the heads can warmup quickly so that combustion is stable. But there is more to warmup than just head temperature and resulting mixture.

The reason the factory location for CHT (3.2) works well is that it is not attached to a fin. It is a segment in the volume of the head below the fins. So the thermal mass is orders of magnitude greater than a fin and gives a more constant reading. This is similar to how a spark plug mounted system would be.

If you wanted a head mounted sensor then I would center drill a bolt, epoxy a thermistor in the bolt, and drill/tap a corresponding hole in the head to mount the sensor. The cylinder 3 segment behind the tin (forward side of engine) would be a good place to drill the hole.

I install the engine temp sensor in the vent cover. I drill and tap NPT on the driver's side flat section for the GM Coolant sensors. Works well and gives you an overall view of the engine and oil temperature. It is also exceptionally stable from fluctuations.

I also find the vent cover location easy to wire and access. I typically run EFI wiring looms across the rear shock member and it is easy to bring a pigtail forward for that sensor.

Other locations can be the chain cover like Al showed previously. You will need to make your own temperature curve for the calibration of the Bosch/VDO sensor. I think it is slightly different than the GM default values.

The software coolant warmup curves are adjusted to match any of the installed location types. I typically watch the engine oil temp gauge and when it reaches the first mark, match that temperature with the coolant reading in the EFI and that is when warmup cycle is terminated. Then you can adjust the enrichment factors below that to match your target AFRs during warmup.

Thank you, Jamie for sharing your experience and insight.

Another data point using the 964 CHT sensor -
I have mine attached to the first fin on Cyl 3 (researched that the rear cylinders experience the least amount of air cooing).
yesterday (75F) while exiting the highway and taking a residential road for a mile i noticed 198F on my gauges. It came back down to 170ish once RPMS were back up above 3k.
My intake temp was ~95-100F i seem to remember.

How is your sensor mounted to the engine?

A little square piece of aluminum that is threaded so the cht sensor can screw into it and clamp against the cooling fin on the head. It has a slot cut into it so the whole thing can be slipped over the fin.
Similar to the pic in post #2 but instead of a bolt clamping the contraption to the head, the cht threads in and physically touches the cyl head fin.

Got it.
I wonder if your fixture is the one I found on here (somewhere) that I modeled mine after.

Good morning.
Thursday morning.

Kelly had plans last night, so I was left to my own devices. Once I got home from work, I changed clothes and headed out to the garage to change the CHT for an oil/engine temp sensor.

The ever helpful, if not wicked ugly, Jimtweet recommended I install a Bosch oil temp sensor in the breather cover. This seemed like it wouldn't work too well because it's not immersed in oil despite being in the crankcase. So I started this whole thread to ask the brain trust what they thought and the general consensus was that old Tweet was right.

So last night, I pulled off the breather cover and got to work.

Here you can see where I drilled and tapped the big boss at the rear right of the cover.
http://forums.pelicanparts.com/uploa...1649938824.jpg

And here you can see the sensor in its new home. (with aluminum crush washer to prevent seeps/leaks)
http://forums.pelicanparts.com/uploa...1649938824.jpg

From the inside
http://forums.pelicanparts.com/uploa...1649940514.jpg

I couldn't see inside the crankcase, but I jammed my fat finger in there and there felt like plenty of room. I 'measured' the stick-out of the sensor beyond the mounting surface of the cover. It's just enough to catch the straightedge, but not enough to measure without the use of a dial indicator. I anticipated no problems.
http://forums.pelicanparts.com/uploa...1649940632.jpg

After this was done, I reinstalled the cover & hose. I then extended the sensor wires and jammed an EV1 connector on there, went inside, turned on the oven, washed my hands, and ate some veggie buffalo wings. (Spicy false-chicken nuggets is a more appropriate description. But our friends at Morningstar Farms are fancy like that.)


***AND NOW FOR THE THRILLING CONCLUSION***

I drove the car to work this morning. It was clear and 34*F on my way in. With the CHT setup, the ECU would be reading engine temps that never got to 50*C. Likely in the low to mid 40s the whole way to work. With the new setup, and me doing C->F->C in my head, the engine temp tracked right along with the oil temp from ambient to 180*F/68*C. Pretty excited with how that turned out. We'll see how it does this evening when ambients are in the mid-50s. (F)

Well, nice work.

Except... I don't want to come over as Debbie Downer, but it seems to me you've now got a perfectly good oil temp reading. Which can also be achieved with a sensor in the TTS hole in the left chaincase, as Al suggested...

Your motor almost certainly responds differently to mine, of course - and I very much doubt that our warmup comp is the same (I've ended up with slightly richer than stock Motec, with an extra time-based (10 seconds or so) trim after start for 0C and colder).

I found that, when using chaincase sensor for warmup, my motor would be spitting, crackling & popping through the headers on overrun (and sometimes on steady throttle in warmer weather) - and throttle response was decidedly woolly & ragged. Better than my CIS warmup was, certainly (which was a bear), but not what I'd hoped for from EFI...

And this at a point where I now know that the CHT sensor would read 60C (with chaincase sensor reading 16C, or 6C over ambient).

After 15-20 minutes running, it ran great. Everywhere. So the basic map was fine (plenty of work on that with quick lambda/blending), just the warmup was off.

Seemed to me the motor was saying that it really didn't need warmup trim at that point. So fitted the TurboKraft CHT and used that for warmup cycle instead. Now all my trim is gone by 60C CHT. Which, for me, is before the lambda sensor is even up to temp and producing readings (so closed-loop not an option either).

So my hypothesis was that once there's some/enough heat in the head/piston/valves, you just don't need warmup trim. And that this happens far quicker than the oil (or the rest of the engine block) getting up to temp. Exactly what I'd do with a manual choke.

Works for me, as the car warms up great. I start it, check oil pressure, manually fast-idle it @ 1800 for 5-10 seconds, check throttle responds OK (yup), put it in gear, back it out and drive it like a turbo with cold oil. First couple of stop lights I have to ride the throttle a touch to keep idle up (no ICV) - but it drives fine, responds well (doesn't bog or fluff), runs cleanly.

I don't know why the 993 sensor and/or location doesn't work well for you - although I note that air is a pretty darned good thermal insulator (hence why double glazing works); I'd still expect it to stabilize at some point if trapped in a closed space - but that's above my pay grade....

TK definitely say to fill the airspace in their adapter plate with a fluid. Just can't find the instructions anymore- now I think of it, pretty sure I used a few drops of the left over AT fluid that came with my QuickJack...

Fitting a heat sink to a CPU or other chip without thermal grease - to fill the air gap - does not work well. The cursed sticky thermal pads the manufacturers love/all use (and which I immediately remove on anything I pay for) are for their convenience/cost, not efficiency. Good quality thermal grease works well. All this does is fill the air gap between 2 completely flat shiny metal surfaces clamped together with a big old spring. To ensure good thermal transfer. On an Xeon pulling 85W, you can see the difference in core temps.

I seem to recall Mr. Clewett suggests tapping a hole in the cam tower and fitting a sensor there for CHT. Yup, exactly: https://www.clewett.com/instrux/head-temp-ins-S.pdf

Hey, good luck - EFI is fun, isn't it?

Ask me about rich misfires reading lean on a lambda sensor sometime - that had me going in circles and looking confused for a couple of days at least :D

The MS3X & ITBs I ran the last two years taught me that I don't need to reach 'full temp' to dial out warmup enrichments. I have, and plan to continue to, used the lambda and engine behaviours to tailor the warmup to my specific engine's wants/needs. I'm certain the air gap inside my CHT adapter is a big part of my problem. I'll make a new one once I decide I'm interested in redoing my wiring harness. Right now, that's simply not the case. Now if I switch to Al's suggested single-pole temp sensor in the TTS location, that'd free up wiring for CHT as I'd only need the one wire and I have a spare in my harness already...just no spare sensor ground.

I can tell you that my car drives a-ok 30 seconds after fire-up. It's been a trifle too rich while the ECU thinks the engine is colder than it really is, but I'm going to dial that out this week. No pops or farts or other unruly behaviour.

EFI is fun for sure. Been steadily oscillating between learning and banging my head on walls/desks/etc for a bit over two years now. :D
That whole misfires on lambda thing is good fun. Lean pops with on accel are another fun one that you really just have to learn the hard way.

I have replaced the eye ring on the plug style thermocouple to fit under the cht sender for head temp, just have to make sure its soldered and shrink wrapped. I would hate to have to get that under a sparkplug.

Oh, this also acts as a sanity check as to how warm (or not) the motor actually is. With no ICV and stone cold, car wants to idle at 800 RPM (yeh, nope, not for long).

As it warms, idle comes up. By the time it hits 1050 (set when good and hot), I consider the motor is hot and begin to relax self-imposed throttle/RPM limits somewhat - while being mindful that oil temps are still kind of low. But a little more throttle and RPMs is the quickest way to get heat into it.

And just now it belatedly occurs to me that perhaps the extra oil capacity/cooling from the extended/baffled oil tank and the full-width front cooler results in my oil temps coming up slower than other folks...

I don't know that motor. Do you already have one of these installed? I inserted one on my 2.2 and it was an easy job.

https://uploads.tapatalk-cdn.com/202...f56be51249.jpg

Sent from my SM-G988B using Tapatalk

Thank you all for sharing your solutions. This has been very interesting to me as I was looking for a good, non-invasive, way to sense CHT. I started by using the oil temps but it was very sluggish and ultimately not very representative, as compared with sensing the cylinder heads directly. I'll share my solution.

I machined a very small brass button, filled the bottom with thermal paste, inserted thermistor, and potted the wires in. This is not unlike your typical coolant temp sensor but with much less total surface area (to be cooled by the fan) and with a larger contact surface.

Along with a dab of thermal paste on the fin, I formed a stiff spring which would provide good contact pressure.

1. Formed spring with SS wire:
http://forums.pelicanparts.com/uploa...1651019415.JPG

2. Trial fit on a spare head:
http://forums.pelicanparts.com/uploa...1651019415.JPG

3. Soldered thermistor, heat-shrink:
http://forums.pelicanparts.com/uploa...1651019415.JPG

4. Thermal Paste at bottom of cavity:
http://forums.pelicanparts.com/uploa...1651019415.JPG

5. Potted with epoxy:
http://forums.pelicanparts.com/uploa...1651019415.JPG

6. Installed at #1 cylinder:
http://forums.pelicanparts.com/uploa...1651019415.JPG

Tim K

That is very clever, Tim!
Well done.

Please share your results once you start getting data.