Air Cleaner Temp Sensor

[wrcsixeight]

LA-318 TBI

This sensor is inside the air cleaner housing, in board of the round air filter.
It has two vaccuum lines attached to it, one vacuum source from the throttle body, one to a vaccuum modulator, which lifts a door inside the snout leading to the aircleaner.
This blend door diverts air from the exhaust manifold mixing it with fresh air captured from in front of the radiator.

Apparently the engine computer is tuned for this device to maintain a 100F intake charge temperature at low throttle levels.

I found that this blend door was always allowing in only manifold heated air at low throttle levels, and removed and plugged the vacuum source, then started driving across the country. My MPGS were about 10 to 15% less than expected on that first tank of fuel.

At that point, I wedged a 1/4inch ID fuel hose under the blend door, lifting it ~5/8" of its ~2 inch travel, and at warmer ambient temps, and with some tailwinds MPGS were ~15% more than expected, but the fuel needle dropped faster when driving at night with lower ambient temperatures.

This was enough evidence for me to insure this heated air system was returned to function, once I reached the other coast.

I found a NOS sensor online, but they wanted 30$ for it and 40$ shipping. Not happening
Found a Carter TC13 that appeared identical, 18$ and free shipping, and ordered that instead.

On arrival it does appear nearly Identical, and took about 15 seconds to swap out the original.

Needing to see how these actually work, I took my flap sander on my angle grinder and sanded the perimeter off, and took it apart.

There is a bi-metal leafspring that pushes a pintle/piston into the snout on the air cleaner side.
When the bi-metal leafspring heats, it allows the pintle to drop, and filtered air enters the sensor, and reduces vacuum to the modulator which controls the blend door, allowing it to drop and mix in cooler ambient air with heated manifold air.

My Pintle was stuck.
There is a good amount of grime under the snout under the pintle
This grime shows shows that the air filter, as we all know, does not capture all particles.

Long Ago, I had a K&N filter, but I found the throttle body all dusty and I could easily see sunlight through its filter media, threw it out, and returned to paper filters.

I have to assume that the K&N filter oil was a contributing factor as to why the pintle in this sensor got stuck, rendering it ineffective at its job.

I've been told this heated air system is ONLY for better drive-ability when the engine is warming up, but if the computer is basing all its spark timing and A/F ratio at light throttle based on a set air temperature of 100F, then hotter or colder inlet air temps will cause imperfection. Disabling this system does certainly cause the engine to run crustier until it heats up.

My experience with the system disabled, no heated air allowed, reducing highway MPGS by ~10% proves to me this system is Not just for improved driveability on initial warm up.

Here are some pics of the sensor and its interior.
Notice the grime on the one side of the interior under the inlet snout.
~ 230K miles on the sensor, which I assume is original.






The carter TC13 appears nearly identical, but the ID of the snout allowing filtered air inside thereby reducing vacuum to the modulator, which then lifts and lowers the blend door, is slightly different, and the tiny holes to the vacuum lines could be different size, and the bi-metal leaf spring could be calibrated differently. I think I will clean up the original and glue it back together and return it to service.

The brass snout can be screwed in or out and change the position of the blend door.
I could attach a Ktype thermocouple and really go a bit nuts playing with this system.

I know these sensors were used on a lot of vehicles, not just Chrysler, up into the MPFI era and back well into the carb'd area.
I have to assume that the failure mode of this sensor is similar, the pintle gets stuck, eventually, and negatively affects efficiency, to some degree.

Highway MPG'S are important to me.
I'll be recrossing the country fairly soon, and am glad I noticed the blend door was always slammed against its stops at light throttle/idle, or I would have remained ignorant of the dis function.

 

[circuitsmith]

Had this same type of system on my 1975 Civic and 1981 Accord.
The sensor simply directs vacuum to the flapper mechanism; no computer involvement.
When I swapped the spaghetti hosed carburetor for a Weber on the Accord I bought an air cleaner housing from the junkyard.
With some cutting and fastening I managed to add a second snorkel to the housing and keep the flap and 100F temp control.
The consistent air temp allowed me to put a leaner jet in the Weber primary barrel and improve cruising economy.
The Weber was a part of modding the engine, also a non CVCC head, performance cam, re-curved distributor and exhaust header.

 

[wrcsixeight]

I wonder if most manufacturers that employed similar systems, if all were for an attempt at a 100F steady temperature, for economy, at light throttle.

Something magical about that 100F temperature?

The 18$ Carter TC13 I bought, cross references to Ford, not Dodge.

 

[Lubener]

If the incoming air is very cold, your heated air door will remain open all the time. This helps cold weather drivability and also helps prevent icing of the carb or throttle body. Fords had that setup going back to the 70's.

 

[wrcsixeight]

I've made a new collector which is larger and encompasses more.of the exhaust manifold than the original and subsequent coffee can versions after original rotted away

My New manifold heal collector also acts like a heatshield, blocking the back of the alternator which houses the rectifiers.

I previously measured 5 more.amps at hot idle when the heatshield was first employed. But lots of influencing variables were not closely controlled at that time.

 

[wrcsixeight]

I cleaned out the sensor with some isopropyl Alcohol, and rtv'd it back together.

Many Q tip heads were turned black, and kind of sticky, and this pintle/piston interface was the worst.

I intend to install a K type thermocouple nearby inside the aircleaner near this sensor, and see what temperatures the aircleaner sees inside at light throttle with the carter TC13 controlling vaccuum to the blend door, and then the original, now cleaned and reassembled.

 

[wrcsixeight]

Grommett installed and thermocouple slid into place.

 

[wrcsixeight]

Drove 20 some miles today watching temperature display regularly with this carter TC13 air cleaner temperature sensor installed.

The temperature inside the air cleaner, as read by this thermocouple, once the engine was at full temperature, ranged from 89f to as high as 129f, though it only hit 129f once and very briefly before dropping quickly.

The usual range was 94.5 to 109f at light throttle, and it always seemed to be rising or falling while moving even at the same throttle and speed.

The only time it seemed to be stable was when parked with transmission in park or in neutral. Stopped in drive at a redlight, foot on the brake, and it still rose and fall regularly.
I will estimate the overall average temperature that I noticed would be about 102.5 degrees F.

When accelerating with ~ 1/2 available throttle, temperature would quickly drop to about 19f over ambient( 70f), and then once returned to light throttle once at speed, would quickly rise to 112f then lower and bounce around the 102 f area +/-5 degrees f.

I'll return the cleaned original sensor to this location, and capture some data from it as well, one of these days.

 

[ZeeOSix]

^^^ Sounds how I'd expect the temerature to behave based on the location of the thermocouple.

 

[wrcsixeight]

The sensor's simple bimetal leafspring dropping a pintle limiting the amount vacuum to the actuator lifting the blend door, is hardly precise.

I was expecting a bimetal coil spring opening an aperture inside the sensor for more precise metering of vacuum.

The blend door likely leaves hot and cool spots not made uniform just passing through filter.

Need one of those intake tornado gimmicks after the blend door.

If the ecm bases everything on a 100f intake charge. Seems best efficiency could be attained if it were steady 100f entering throttle body, not 100f +/- ~6f....
If that is what is occurring, and not just measurement error.

Where would be a better spot for the thermocouple in your opinion?
I can easily relocate it.

I thought nearby the sensor at same height as the sensor's intake snout, was not just easiest, but a good starting point.

 

[ZeeOSix]

Where would be a better spot for the thermocouple in your opinion?
I can easily relocate it.

I thought nearby the sensor at same height as the sensor's intake snout, was not just easiest, but a good starting point.

Yeah, probably the best you could do is locate the bead of the thermocouple in the air right next to the inlet pipe on the sensor. The thermocouple is low mass, so it doesn't take much temperature change to make it react. Whatever heats/cools inside the bi-metal sensor is probably higher mass and won't react as fast as you see the temperature change on the thermocouple. But if the bead of the thermocouple was near the intake tube, at least you'd know what the air temp is at that location if you think it might vary inside the filter area.

 

[wrcsixeight]

Thanks for your input, Zee0Six.

I used to think this whole system was only to help drivability/efficiency on warmup, but on a recent drive across country, with the system disabled and only ambient temperature air fed into aircleaner I got ~15% worse fuel economy than I am used to seeing, and I wedged the blend door open about 5/8" to 3/4" some 600 miles in, then got the expected MPGS for the remainder of the trip, the minimum expected though, and about 1mpg less than I could realistically hope for. based on previous experinece driving this route 2x a year for the last 3 years with the same tires, and driving this same vehicle for 19 years across country no less than 25 times.

This last trip wtith the sensor disabled and blend door wedged open a set amount, the colder it got out, the faster the gas needle dropped. Warm days i was happy with miles accumulated per hashmark, at night when colder, I was dismayed that that day's gains were being eaten up.

I installed the thermocouple in free air, not to measure the temperature of the sensor itself, but to see what the air temperature allowed inside the filter is while the sensor, adjusting vacuum to the hot/ambient blend door is doing its thing.

So the way i see it, either this Carter TC13, which is for a Ford, is constantly varying the blend door as it cannot slowly and precisely control the vacuum to the actuator which lifts the blend door, and/or the air inside is not thoroughly mixed where the thermocouple resides, and different speeds will have different pressure on either the manifold stove/collector or the cold air intake out in front of the radiator cause the ~ 10 to 20f swings I was noticing.

Say the latter is occurring, then perhaps the air in the throat of the throttle body is not an average 100f across the full aperture, but today, perhaps 70f or ambient on one side and 130f on the other. Perhaps in the intake it mixes thoroughly and an equal temperature charge reaches each cylinder, or perhaps cylinders 5678 see a different aircharge temp than 123 and 4.

I know those intake tornadoes marketed years back just add restriction, and likely reduce max HP and torque, but I'm looking for max fuel economy, and lowest emissions.

When the blend door is half open, hot manifold air is on the bottom and ambient air is on top when it gets split and forced around the airfilter.

Perhaps if there was a vortex generator on either side of the splitter blending hot and ambient, perhaps there would be a more homogeneous temperature air charge, not only reaching the sensor, which might then not be lifting and lowering the blend door rapidly, but perhaps instead able to more properly keep a homogenous air charge right at 100f always, instead of averaging 100f, and allow for more complete combustion at light throttle, increasing MPGS and reducing hydrocarbons.

Picking Flypoop from the pepper I know, I could just slow down and get better MPG's bla bla bla.

Seems to me this system stands a chance of being improved upon and it might not take much effort to do so.

Or the minimal efforts expended might have no effect whatsoever.

But changing variables and capturing actual data to compare, appeals to me, sometimes.

I need to change some variables to test some hypothesis.

Seeing if the blend door is opening and closing rapidly when my thermocouple display rises and falls would be telling.

The splitter inside the aircleaner facing the snorkel where blend door is, perhaps some of those stick on vortex generators would present minimal restriction yet allow the hot and cold air to mix more thoroughly.

 

[circuitsmith]

As I remember it the system on my 1981 Accord had a check valve and small reservoir on the vacuum supply line.
So when manifold vacuum dropped due to throttle opening the door retained its position.

 

[wrcsixeight]

Earlier models of my vehicle, with carb's, had such a system. The vacuum hose diagram under my hood shows nothing of the sort, though earlier models do.

I am wondering how well the hot and ambient temperature air are mixed outside the air filter, and whether the rather fast swings in temperature I observed were from the blend door opening and closing in response to the bimetal spring inside the sensor manipulating vacuum to the actuator, or because the thermocouple itself was perhaps on the border of temperature stratified air, or perhaps some combination of both.

I'd have to think if the ECM is basing all air fuel ratio and spark timing based on this system maintaining a 100f aircharge at light throttle, then it would be most efficient if the intake charge temperature was steady 100f, not bouncing +/- 6 degrees f around that mark, but perhaps 6 degrees makes no difference at all.

There is a temp sensor on the throttle body, but apparently it is only consulted by the engine computer on startup, and a quick startup hot or cold, has never been an issue.