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.
UPDATE: You can read the thread, but most of us have come to the conclusion that the magnetic float looses buoyancy over time. Cleaning the bottle may help for some period of time, but the bottom line is.........long term, you either buy a new bottle, or just unplug it and check your coolant level when you check your oil.
Actually, I just found a service bulletin that confirms the float!
See the pdf attached at post #134(if that number doesn't change) I posted it on 1/22/2011, or just follow this link..........
Special Warning:Have seen a lot of people breaking the plastic nipple of the radiator when changing out the coolant bottle, thereby causing them to have to buy a new radiator. One way to avoid this is to cut the hose at the nipple and avoid this situation!
Maz I just included a pic on the 'Nipple' situation in your first post, I hope you don't mind ASH8
Now..............on to the thread and you'll see how the conclusions came about if you are truly bored out of your skull!
So.............after dealing with the dreaded coolant sensor and light for a few months going off and on(usually after spirited driving) like everyone else, I finally decided to go ahead and get a new coolant bottle.
After taking the old bottle out and installing the new one, I played around with the old bottle/sensor and it will in fact pop right out of the bottom of the bottle after taking the entire bottle out! See pic 5 for the clip locations. I hacked my way into this bottle as you will read, but you can get the sensor itself out, but there isn't really a need to do so..........read on!
I butchered my old bottle for this experiment. I wanted to get at the float and the float shaft, that's why you will see this piece of the bottle cut out! Pics 1 and 2 are just of the sensor itself and this was after I decided to go further and hack the bottle apart.
All this sensor itself really consists of is two wires running from the plug to the sensor through the plastic plug that connects at the bottom of the bottle. You can see some of that plug(so to speak) in pic 1.
The sensor itself is a small reed switch enclosed inside a small glass tube.
Again............pics 1 and 2, notice the reed switch is not closed at this time.
This sensor(reed switch) slides into a plastic tube when you install the plug/sensor assembly into the bottle(pics 3, 4, and 5 have it installed in the plug and therefore in the shaft/tube). There is a round magnetic float that surrounds the plastic shaft in which the sensor sits and this floats up and down with the coolant level(same pics.....3 through 5). When the coolant level is to low, the magnetic float rides lower on the shaft(pics 3 and 5) and if low enough it causes the reed switch inside the shaft to close, thereby illuminating the idiot light. BTW, the resistance of this switch is about 10 ohms when closed.....keep this in mind.
So...............I assume that the PCM/ECU sends out a voltage on one leg and reads the other leg every so often to determine if the switch is closed or not. This could happen all the time or it might query it every so often(not sure), but either way.............if the coolant is low enough, the magnet falls and the switch is closed.
Having said all that, and remembering how this usually start to fail......getting on it hard usually causes the light to come on and then as you back off, the light will go off. This along with the fact that this goes on for awhile until eventually the light stays on all the time, no matter if you have enough coolant or not, or how you are driving........something eventually fails. (See pics 4 and 5)
Correct me if I'm wrong here, but the spirited driving effect must mean that more coolant is being sucked out of the bottle when getting on it(or possibly a big increase in heat inside the bottle?....see below) and once you back off, the coolant level returns to the bottle(or heat goes down...see below) and the light goes off.................this is of course, until it sticks or fails for good!
I see a couple of possibilities, but think I can eliminate a few real quick.
1.) It's possible that the contact resistance of the reed switch itself could increase over time enough to where the PCM/ECU doesn't like what it sees for the voltage being returned and no matter what the float does, it's stuck or it thinks it's closed. <--- Nope not the issue
The reed switch is enclosed in a glass tube and assumed to be air tight, so I doubt this switch could be sticking or building up much contact resistance. I did throw a meter on the one I took out and slid the magnet up and down and I could see(and measure) the switch open and close. It went from approx. >2meg to 10.0 ohms when it closed.
Worth noting............if it was metal fatigue on the switch contact arms, you think it would be more prone to staying open anyway as opposed to not being able to be closed by the magnet float. <--- Nope not the issue
2.) The plastic shaft that the sensor sits in and that the magnetic float rides up and down on could swell with heat and keep the magnet from sliding freely even when enough coolant? <--- Nope not the issue
Examining the plastic shaft and the distance between the shaft and magnetic doughnut float, the gap is big and the plastic is pretty thick. It would have to expand a whole bunch and cool down real quick(shrink) in the time I've noticed it takes it to go back off.
3.) Look at my pics on the next post and while I didn't show pics inside of the bottle itself, you can see I had crap on where the magnetic float comes to rest(with low coolant level) and it was also on the inside bottom of the tank. I think it's the crap that builds up in the bottle and at the base of the float shaft as coolant gets older that might possibly cause the float to stick. This is why banging on the bottle might dislodge the stuck magnet float and cure it for some period of time. <--- Still possible as cleaning helps but doesn't cure
If I didn't miss anything(and I probably did)..........I think the real solution here is kind of the obvious one.
Change/Flush out your coolant level every year.
Take a look inside the bottle and if you see crap down there, go ahead and take the bottle out and clean that crap out.
BTW........It was pouring down rain this morning, so I did experiment with the piece I cut out and you'll see that in the pictures below.
I sprayed contact cleaner on it to hopefully get rid of the crap, then flushed with water.........this didn't remove it.
I didn't want to rub obviously because if it was still in the bottle, you wouldn't be able to get to it!
I let it dry and then hit it with WD-40(again no rubbing) and rinsed with water and you can see the result in the last pic.............it did take it off.
Of course you can't do this with it in the car (and wouldn't want to!!!!!!).........and WD-40 might not be the best to use(remember I wasn't going to reuse this bottle or piece!). If you remove the bottle from the car, and find something that will dislodge that crap and get it loose(even a good stream of water) you'd be in business. The piece I cut out(the magnet/sensor shaft) is located on the left side of the bottle when looking inside from the front of the car and behind a couple of channel slot openings. It might be hard to get to it with any real stream of water. That's why I'm thinking of finding something that will loosen it up on contact, but not hurt the plastic and then flush it out really well with water, let it dry, re-install, and add NEW coolant!
Of course, you could always just
1.) Buy a new bottle. (I did after only ~20k miles)
2.) Unplug the connector. Just check your coolant level often.
3.) Take it out and clean it the **** out of the bottle!
Why doesn't someone sell just the sensor instead of having to buy a whole new bottle?
Great stuff. Thanks Mazurfer for this very much needed thread.
I have the coolant light issue. On Second bottle and was going to get a third. Now I'm going to try this and see what happens. I'll let you know.
There's a million and one threads on this subject, but I don't think there is one with as much detail.
I hope I was clear.................don't remove the sensor, remove the entire bottle. Even if you are going to attempt to clean and flush, there is no real need to remove the sensor itself. Flush it out really good and get all that crap outta there, however you decide to do that????
What breaks up old crusty coolant? WD40 works, but not sure you should use that!
Concentrate on the area where the float is.......it will be hard to get to from the filler cap opening, but is obviously more important in that area than the rest of the tank.
I figure next time, I will use something to break the crap/crud up, then start flushing with water while shaking the **** out of the bottle until I get all the crap out.....either that or just friggin buy another one!
Here's the piece of the puzzle I am struggling with:
If the switch is open(magnet float riding high indicating good coolant level and no light) and magnet falls with low coolant(closing the switch and illuminating the light), then why would unplugging the sensor cause the light to be on(it would be an open circuit)? Hmmmmm.........this kind of tells me that the PCM is looking for a given range of voltage on the return(not completely an open circuit.......ie....infinite ohms), but expects some range of resistance(say 10 ohms to ~2m or so)..................and will illuminate the light if outside that range, either low(magnet) or high(unplugged). Remember when I measure it without the magnet being there, I still read about 2m(if I remember right?)........and it was dropping. I don't remember where it stabilized cause I didn't care, as long as it didn't drop
all the way to 10 ohms and it didn't.
So knowing this info, it should therefore be possible to find the right value resistor and just plug it in to where you disconnect the sensor and you'd have no light. You would however have to check your level and not depend on that light ever coming on. I'll see if sometime in the coming days I can figure out what that value should be, but anywhere greater than 10 ohms and below ~2m might do the trick.
UPDATE: I was going by what some had told me in that unplugging the sensor caused the coolant light to come on. That is not the case in my car(an 05) and if yours does.....I think you have another issue. This is a simple switch. If the float magnet gets too close to the reed switch(low coolant) or is stuck, then you will ge the light. Unplugging the sensor should NOT give you a light as far as I know.
Excellent question! Wow........that could suck!
Well, I didn't see anything that would hold it on, but I'll have to go examine the bottle where I cut it out to make sure. I will assume that the round part at the top of the shaft would fit into something or be close enough to not allow the magnet float to come off, but I'll go look!
Nope..........it can't come off.
For lack of better terms, the shaft that the magnetic float goes around(and rides up and down on), fit's into a receptacle when the assembly is plugged into the bottle, therefore the magnet can't come off.
Well, looks like I got some more research to do because I was just poking around on the net and it's possible that this magnetic float is actually Engineered in terms of buoyancy to this specific application. This could throw into the fray that it might be less buoyant in old coolant and possibly different using water, water wetter, or anything else besides what it was Engineered for. It's also possible for the buoyancy of the material to change over time. I just read that depending on the material used, it's possible for these floats to absorb fluid over time and that would change it for sure and also change the temperature coefficients which could/would effect buoyancy.
If any or some of this is the case, then there is no way to get at the float. You can get the sensor out, but you're not getting that float out without hacking into the bottle. Now thinking if it's not the crud build-up causing sticking then cleaning may or may not help! Argh!
Damn those Engineers...........and I is one!
This is probably why we are up to revision "M' on the bottle. They probably made it far more complex than it needed to be and can't figure out(or be bothered) to discover what it is that's failing Guess engines are more important than a $160.00 bottle!
BTW................when I ran downstairs and into the garage to answer Nuke's question above, I grabbed the float, put it in a pan of water(not coolant) on the stove and it sank to the bottom. The buoyancy does change as it heats up though and it started to become more buoyant and almost floated, once I saw that I just turned off the stove. Probably would have been even more buoyant in hot coolant! Probably Engineered specifically for FL-22
Here's the piece of the puzzle I am struggling with:
If the switch is open(magnet float riding high indicating good coolant level and no light) and magnet falls with low coolant(closing the switch and illuminating the light), then why would unplugging the sensor cause the light to be on(it would be an open circuit)? ...
Uh....It doesn't ... for me at least. I started getting the light and just unplugged the sensor. Haven't had the light since. Mine is a 2005.
Here's the piece of the puzzle I am struggling with:
If the switch is open(magnet float riding high indicating good coolant level and no light) and magnet falls with low coolant(closing the switch and illuminating the light), then why would unplugging the sensor cause the light to be on(it would be an open circuit)?
Very nice write-up Mazurfer!
Does the glass bulb appear to have fluid in it?
If so, you might have a small amount of current flowing through the liquid (think back to science class with the salt water) that will give you a large resistance.
When the float drops and the contacts touch each other, you'll then have little resistance.
This could explain why the light stays on when you disconnect the harness.
You need to have some kind of "feedback" or resistance to the PCM.
Easy way for you to try it out is to measure resistance with no float installed.
If your meter reads its an open circuit, then I'm wrong with my theory.
If you do measure resistance, then you have current flow even through the contact are not touching each other.
Uh....It doesn't ... for me at least. I started getting the light and just unplugged the sensor. Haven't had the light since. Mine is a 2005.
That weird........and to be honest, I haven't tried unplugging it myself. I was merely using others peoples observations that unplugging the sensor caused the light to come on.
EDIT: Neither does mine! I was just going by what people said........that it came on when they unplugged.
Does the glass bulb appear to have fluid in it?
If so, you might have a small amount of current flowing through the liquid (think back to science class with the salt water) that will give you a large resistance.
When the float drops and the contacts touch each other, you'll then have little resistance.
This could explain why the light stays on when you disconnect the harness.
You need to have some kind of "feedback" or resistance to the PCM.
Easy way for you to try it out is to measure resistance with no float installed.
If your meter reads its an open circuit, then I'm wrong with my theory.
If you do measure resistance, then you have current flow even through the contact are not touching each other.
Yeah, I thought about all of this. Right now, It would appear as the glass bulb doesn't contain fluid, but it might. I need to take a closer look at it. I did measure it without the float anywhere near the shaft/sensor(it was completely off) and I still read about 2m........so there may be fluid in there. If there is, it is completely filled and there is no air what-so-ever.
I'll examine more closely and double-check readings.
To me the read switch looks like the standard glass vacuumed ones used in 12V house alarms on doors and windows (I have a few spare).
The pics of the plastic tube surrounding the reed switch looks slightly bent?, I am wondering if over time the "tube" bends with Hot/Cold heat which might catch or snag the black round float magnet?
Good observation........it is infact bent!
I'm not sure it could catch as there is about 1/16 of an inch gap.........but I did realize it could be a possibility. I have to assume the shaft is straight when new, and nothing I did in removing it caused that bend, so...........................
for some reason, I never got any problems with coolant level sensor. I replaced my last one cuz the tab is broken.
Yeah, but from what I know, you're pretty **** about changing fluids, so tell me on that old bottle, how often did you change/flush out your coolant? I wondering if "old" fluid effects the bouyancy of the float in some regard and then other things happen.
Yeah, but from what I know, you're pretty **** about changing fluids, so tell me on that old bottle, how often did you change/flush out your coolant? I wondering if "old" fluid effects the bouyancy of the float in some regard and then other things happen.
yeah ... im pretty crazy about changing fluids ... I change my coolant 3 times this year ... LOL
Coolant was on sale man ~~~ it was like 7 bux for Prestone 100% coolant, went to the supermarket to get a gallon of distilled @ 99 cents each ... lol
Anyway, there's always some "crap" sink to the bottom of the coolant bottle,
everytime I flush my coolant, I will take the bottle out, put some dish washing soap & water in there and leave it overnight(or couple hours, depends on when I start the flushing) then I will flush it like 20 times to make sure its clean (no more soap)
yeah ... im pretty crazy about changing fluids ... I change my coolant 3 times this year ... LOL
Coolant was on sale man ~~~ it was like 7 bux for Prestone 100% coolant, went to the supermarket to get a gallon of distilled @ 99 cents each ... lol
Anyway, there's always some "crap" sink to the bottom of the coolant bottle,
everytime I flush my coolant, I will take the bottle out, put some dish washing soap & water in there and leave it overnight(or couple hours, depends on when I start the flushing) then I will flush it like 20 times to make sure its clean (no more soap)
Yeah, I thought about all of this. Right now, It would appear as the glass bulb doesn't contain fluid, but it might. I need to take a closer look at it. I did measure it without the float anywhere near the shaft/sensor(it was completely off) and I still read about 2m........so there may be fluid in there. If there is, it is completely filled and there is no air what-so-ever.
I'll examine momre closely and double-check readings.
If the bulb doesn't have fluid in it, it could also be using the same principles as a smoke detector.
A small amount of radiation is used to allow current to flow through the air.
If you're measuring resistance and the two tabs aren't touching, something is in the bulb to allow current to flow.
yeah ... im pretty crazy about changing fluids ... I change my coolant 3 times this year ... LOL
Coolant was on sale man ~~~ it was like 7 bux for Prestone 100% coolant, went to the supermarket to get a gallon of distilled @ 99 cents each ... lol
Anyway, there's always some "crap" sink to the bottom of the coolant bottle,
everytime I flush my coolant, I will take the bottle out, put some dish washing soap & water in there and leave it overnight(or couple hours, depends on when I start the flushing) then I will flush it like 20 times to make sure its clean (no more soap)
Okay, finally got around to doing a little more with this coolant bottle and sensor.
Ash8 may not like this, but this reed switch is a normally open(NO) switch. When the magnet comes in close proximity to the reed switch......it closes!
It's nothing fancy as I was thinking in a previous post, it's just a simple open or closed switch after all.
Here's what I did today:
1.) Start car and verify no coolant light on the dash. I have a new bottle and proper coolant level, so this was of course expected.
2.) Car still running and unplugged the coolant sensor...........this produced no light. This is contrary to what I was lead to believe by several people, but I had never tried it myself. Again...........the unplugged sensor gave no light! Yours could be different, but I don't know why it would be?
3.) With it unplugged, I measured the resistance on the "good" coolant bottle end of the connector and it read infinite ohms. The switch was open meaning I had adequate coolant. This is the same reading I got on my butchered one with the magnet away from reed switch.
4.) With it still unplugged, I then measured the PCM side of the connector and got 14.2v to GND on one leg and nothing on the other. Remember.............my car was still running, so that's why 14.2v and not 12.0v.
5.) Next, I took my old sensor from my butchered up coolant bottle, slid the magnet off the shaft(meaning it would be floating.......as if in the bottle), plugged it in and there was no coolant light on the dash.
6.) Walked around under the hood, took the magnet and slid it down to where it would be sitting if low coolant or possibly stuck. This of course produced and instant low coolant light.
7.) For the sake of argument, I repeated #5 and #6 about three times just to make sure.
So......................I believe you can get a low coolant light for the following reasons.
a.) Your coolant is low.
b.) The shaft that the magnet rides up and down on has been bent by heat and therefore the magnet tends to stick. <----- Still unlikely IMHO
c.) Crud buildup in the bottom of the coolant tank and gums up the magnet and shaft so that it won't float freely and sticks.
d.) a faulty reed switch in the sensor that has stuck closed. <----- Not sure how likely this would really be.
Now.............as some have reported getting a coolant light even with the sensor unplugged. I would have to say that you have some other issue like a short or something in that wiring somewhere.
If you have the low coolant light and don't want to replace the bottle, then you can just unplug the sensor and check your coolant often. I think I would wrap the PCM side of the connector that will be dangling with electrical tape or something in order to keep it dry(and keep crud out) because if the two terminals or wiring short on that side, you are going to get a coolant light.
09-19-2009, 11:28 AM
See pic 5 for the clip locations. I hacked my way into this bottle as you will read, but you can get the sensor itself out, but there isn't really a need to do so..........read on!
Guess engines are more important than a $160.00 bottle! 
Probably Engineered specifically for FL-22 
Mine is a 2005.
This is the same reading I got on my butchered one with the magnet away from reed switch.
