Red coolant light slow blinking after tank replacement

My 99 996 was leaking coolant from the overflow tank recently. I replaced it with a new Mahle tank, buttoned everything back up and low and behold the coolant level light is slow blinking at me. Double checked the coolant level and added a tiny bit to ensure it was between the lines. Still blinking. Figured maybe I broke the sensor during install, so I ordered a new coolant level sensor and pop it in yesterday. Light still slow blinking. What should I check next here and in what order would you suggest? - engine compartment fan (how do I diagnose this) - engine temp sensor - wiring harness for the coolant level sensor (I drove it for weeks with the leak dripping down the bottom and kept it topped off until I could get to the replacement) - other thoughts? (could the magnetic float in the replacement tank be causing this)

Quotes source post: 19606091

How did you refill the coolant in the system?

Just reconnected the hoses to new tank and refilled from the top/tank. Are you thinking it could be air in the coolant system? If so, what the recommended steps?

Quotes source post: 19606173

At this point, it is moot. It's highly probable that you created air pockets in the system. Either purge the whole system and start over by refilling the coolant using air-lift tool or open the purge valve and hope for the best. Now, did you examined the oil for any intermix? I assume you inspected the coolant color before swapping out expansion tanks, correct? Lastly, is the blinking light 1 per every 2 seconds or once per second?

Quotes source post: 19606184

I have opened the purge valve now and will monitor the level and see if that helps. No intermix noticed in oil. Correct coolant color was/is Pink. Light is blinking once every 2 seconds.

Check your reservoir level and add more fluid when it's cold. Leave the bail open /up. Only need to drive it a block or two then check again. Probably have to do it three or four times. When It no longer needs fluid then you can close the Bale. Check periodically for the next 5 or 6 drives to confirm it does not need more fluid.

Just disconnect the low coolant sensor wiring and leave it unconnected " just as a test temporarily".... If the slow flash continues then the problem is NOT in the low coolant warning circuit .. There are several other things that can cause the "slow flash" , no need to go into disscusing them untill you confirm the need..

Quotes source post: 19606701

I disconnected the harness from the coolant level sensor last week and the flashing stopped (it was driving me crazy after replacing the tank and I was sure the coolant level was fine and no longer leaking). Assumed it was most likely a broken/failed sensor from the tank install process and ordered/installed a new one. Slow flash came back once I connected it.

Quotes source post: 19607133

Ok , then the coolant level is low, or the float is stuck , or the sensor is shorted..... Does the slow flash start up after 19 seconds with the coolant level above full , or does it not start untill driven for a while..?

Quotes source post: 19607247

Coolant level is just above the MIN line when cold, should I top it up to the MAX line? I usually keep it between the two marks. Haven't had the coolant level above full, however the slow flash comes on as soon as I turn the key in the ignition. Seems unlikely that the old and new sensors both were bad, but not impossible I suppose. Could the wiring harness to the sensor be causing the short? Or perhaps the float is stuck as you mentioned. Can that be corrected with a magnet or tank needs to be removed?

The low coolant waning is a very simple circuit... The wiring from the electronic warning system has 1 wire that goes to the low coolant sensor/switch...when the switch/sensor closes the other wire is just connected to chassis ground... The sensor/switch is just a magnetic reed switch that opens and closes if a magnet is near by... The float contains a magnet that moves up and down dependant on the coolant level.. When the coolant level is full, the float/magnet should be floating high away from the reed switch and redd switch will be in the OPEN state and not grounded...( no slow flash) ...when the coolant level goes low the float/magnet drops to near the reed switch and the reed switch CLOSES and grounds the wire from the warning system to chassis and the slow flash begins 19 seconds later.. This can be bench tested with a simple OHM meter, or can be tested in-car using an OHM meter and controlllijng the float level and using some basic logic...

I bench tested the sensor and it is working fine. I assume the warning light flashing is part of the closed circuit loop and not a status/signal stored in the ecu right? I ask because you mentioned it comes on 19 seconds after the circuit closes, however mine is coming on the second I turn the key in the ignition. If this sounds reasonable based on the circuit being closed, then the only conclusion I can think of is that the float is stuck on the new tank. Any chance of getting it unstuck without removing the tank?

PT3 is the best, but I would add from my experience, that the only surefire way to fill these cars and their ridiculous number of coolant hoses is to invest in an air compressor and a vacuum fill kit. My indy knows the 996 well, but it never gets filled all the way without the vacuum fill despite his best efforts. You don't need to drain the whole things before you vacuum fill. I don't. I suspect that not being filled all the way, despite having the right level at the coolant tank, is a source of valve sleeve and other heat related issues.

Quotes source post: 19619419

Did you test the sensor with it connected to the tank? If the circuit is closed then that would indicate that the float is at its low warning position and probably not moving.

Quotes source post: 19619419

Simply leave the wiring connector unplugged from the low coolant sensor/switch.....this will for sure have the open circuit and no slow flashing light... The low coolant sensor/switch and the float/magnet and coolant level all simply serve to GROUND the wire from the cluster to chassis ... I have found that the instrument cluster has a 19 second delay in starting the slow flash, I know in 1999 c2 cars, this delay is to keep from having the slow flash during slosh or very high G maneuvers when the coolant level is not actually low...I don't know for sure if the 19 second delay is on ALL models since I don't have all models at my disposal and haven't tested for this on all of them .. All 996 models have the same sensor/switch .. wiring , float/magnet , but some "could" have a different logic built into the instrument cluster where there is no 19 second delay or a need to "set" the slow flash.....Ie.. only the original designer/builder knows for sure ... I haven't tested or experimented with this on ALL models.. But once the slow flash has started, the key has to be turned off to "reset" the system ( on the 1999 c2 for sure) ... This info is not in any Porsche Shop manual or literature... this all comes from my direct experience and testing...

Quotes source post: 19620101

After the coolant tank install and verifying that the coolant was no longer leaking, I did unplug the sensor harness for a few days to stop the blinking. So I can confirm that if I unplug the harness from the coolant level sensor, the blinking does stop. I could also try unplugging the harness, and plugging in a spare sensor without installing it under the tank, then starting the car and using a magnet to close the circuit and confirm that the light stops. If this works, I think it would prove that the issue is with the position of the magnetic float in the tank.

Quotes source post: 19620243

Yes exactly !! Now you understand how the simple system works ...well almost .....the sensor/switch is normally OPEN, when you put a magnet close by the circuit CLOSES ( Grounds the circuit to chassis ). A little logical common sense along with some experience goes a looooong way ...

Quotes source post: 19620302

Observations: Started the car with harness disconnected, no warning light. Plugged in the spare coolant level sensor (loose in hand, not installed under the tank), no warning light. Held a magnet up to the sensor, waited 19 seconds and the warning light began to slow flash. Also, for future folks who may read this, moving the magnet away from the sensor, the flashing continues. Unplugging the spare sensor from the harness, flashing continues. Seems the low coolant flashing, once triggered, will not stop until the car is turned off. Turned the car off. Tried tapping on the side the tank (without removing it) to see if that might dislodge the float. Plugged the harness back into the sensor installed under the tank and started the car, warning light flashing. Do you recommend the next step be to remove the tank and test the float/find some way to verify if it's lodged and then dislodge it? If so, I will likely wait for cooler weather this fall and just monitor the coolant level manually until then.

Is this an OEM Porsche coolant tank? Its extremely rare to have a float failure.. The float is the only moving part in the tank, and it is not accessible. It is built into the tank during manufacturing. I suppose you could remove the tank and turn it up-side down and hit it and maybe jar the float loose, but that wouldn't give me piece of mind that it will not stick again...

Quotes source post: 19625915

It's an aftermarket Mahle tank from FCP Euro. The price for the Porsche tank was exorbitant. I have read all the feedback on the Porsche tank, however the FCP Euro Mahle tank came with a lifetime warranty and they are a reputable parts distributor.

Quotes source post: 19628091

The Porsche tanks IS exorbitant, but having to repeat the job every year or two in order to save money with a 3rd party tank is not the best approach. I did that once and that tank lasted less than two years. I had been assured by the seller that it was exactly the same as the Porsche tank. It wasn't. That's not a mistake I'm willing to repeat.

Quotes source post: 19631353

The research I did showed the same results that you mention, however the feedback seemed to be centered around the URO tank and similar OEM tanks. I decided to take a chance with FCP Euro seller and the Mahle tank after my research showed FCP Euro seems to be a trusted parts supplier and they offer a lifetime warranty and Mahle seems to be a supplier to BMW for factory installed parts/coolant tanks similar to the one they make aftermarket for Porsche applications. Honestly, after having gone through the process once, it was not an overly difficult or time consuming job, once I got over my stubbornness and bit the bullet and lowered the engine down to allow for space. I figure if the Mahle tank does crack and leak within a couple years and the seller will replace the part, I don't mind going through the install process again if it saves me $700. After all, one of the things I enjoy about wrenching on this car is the money I save in labor costs.

https://rennlist.com/forums/996-foru...996-1-a-2.html the 996.2 tank (much cheaper OEM) can be made to fit in the 996.1

FCP Euro agreed to send me another tank to try. Wanted to run my thinking by this group and see if it makes sense. If I hold the new tank upside down, I assume gravity will cause the float mechanism to slide to the top position simulating a full tank. If I then install a second sensor in the new tank and plug the harness into it, I could then start the car and see if the warning light blinks or not. If it doesn't blink, then I could turn the tank right side up, causing the float to slide down to empty/low coolant position and see if the warning light starts to blink. If that works, then I would be confident it's the old tank and proceed with installing the new replacement tank. If the warning light still blinks with the tank upside down (simulating full tank), then I know the issue is likely not related to the float in the current tank. Am I overlooking anything here?

I don't have a wiring diagram to guide me, but you can see if the float is doing the right thing using a voltmeter. If you don't have one, now is the time. Set the voltmeter to the lowest OHMS setting and touch the ends of the probe together. The meter needle should move to the opposite side from where it was, indicating a closed circuit. Now hook those leads to the tank. Tip the tank over to simulate full and the needle should move to the other side. The needle will either read closed one way and open when tipped the other direction. It may not be that simple, but that is usually the way a float works. It is either a closed or open circuit.

Quotes source post: 19644654

If I am following you correctly, you are suggesting a similar test except removing the wiring harness from the equation. Is that correct? Your method would simply test the float and the sensor's ability to read/respond accordingly to the float (magnet). As this method is an easy first step, I will do it to confirm the new tank float setup before moving to the second test (more elaborate method I mentioned, involving the harness and starting the car to check the warning light blink).

Quotes source post: 19644728

Yes. The process of diagnosing a problem is to isolate the components. Using a voltmeter to test the float before it connects to the rest of the circuitry will allow you to determine if it is the float or some other part.

Quotes source post: 19631397

Hi Way2fast, is this regarding Expansion Tank 996 106 147 56 ? We'd love to hear your feedback, please send us a PM with the details so we can investigate. Tags: 99610614756, 996 106 157 04, 99610615704

Quotes source post: 19644626

if you�re gonna do that, I would use rubber gloves and rubber bands to seal up the holes and fill it with water so you�re actually simulating the float moving with the level of the water.

Quotes source post: 19644985

Thank you so much for this tip. Finally some decent temps today locally, so I tested the new replacement tank. First, I plugged the new tank (uninstalled) with sensor into the wiring harness and held it upside down to simulate full coolant condition. No blinking occurred. I then turned the tank right side up and could here the float slide down into to the bottom of the tank simulating empty/low coolant condition and after 19 seconds the low coolant light started to flash. I was convinced it was time to start removal and install of the new tank, however based on the comment above, I decided to take the time to plug and fill the new tank with water, hold it level, plug in the harness and start the car and much to my surprise after 19 seconds the slow flashing light occurred again. The one thing to note here that has me scratching my head, why did it take 19 seconds for the light to come on in this last test with the new tank filled with water? With the currently installed tank, the light comes on the second I start the car. In the first test of the new tank I ran without fluid in it (holding it upside down), if I hold it right side up (empty condition) and start the car the light came on the second I started the car. Why then, when I filled it with fluid and held it right side up, did it take 19 seconds for the light to come on?

Quotes source post: 19669111

Because that is the way the programmer designed the algorithm in the Instrument Cluster, read post #15 again.....

Quotes source post: 19669322

Let me clarify, I am curious why the last test caused the light to start after 19 second delay and the previous test did not but rather it mirrored the currently installed tanks issue of the slow blinking light starting the second the car is started. The tests both lead to the low coolant light coming on, but why is one immediately upon starting the car and the other after a 19 second delay?

deleted duplicate

Quotes source post: 19669399

Let me repeat again, that is the way the programmer designed the algorithm. It is not just 1 set of rules, there are quite few "IF's ", "and's" and "also's" in the numerous lines of code in the program.... It is not explicitly explained ANYWHERE, this is just my Empirical evidence and logic from having tested quite a few..

On the advice of FCP Euro support, I plugged and filled the replacement tank again and observed visually that the float is in fact moved up with the fluid level. What other conditions would cause the slow flash to start after 19 seconds?

Quotes source post: 19672071

Nothing .....it is a simple switch......either open or closed.

So with the sensor unpluged the light stays off? Maybe the sensor itself is bad? Have u checked it with a meter? Im guessing should have continuity if a magnet passes over its and show open as a normal condition.

Quotes source post: 19672289

Yes, with the sensor unplugged, the light stays off. I have checked both the installed sensor and the spare that I ordered with a volt meter and both are working as expected.

I haven't seen it asked/mentioned...do you have the UAOS system installed? If unaware of what it is - it is the Porschetech3 solution for the frequent AOS failures. It can cause a slow flash as well. -Tom

Quotes source post: 19674339

I do not believe any UAOS is installed. Previous owner was not much of a hobbyist. Car appears to have remained bone stock.

Quotes source post: 19687085

Yeah, I didn't think you had the UAOS, you would have noticed it by now......But just to confirm if you will just take a pic of the engine I can tell right away... You have just overlooked a step in the diagnostics..

After a quick search, car does not have the UAOS installed.

Quotes source post: 19688326

I didn't think so... Almost 2 months on this blinking warning light...I'm also sure you don't work on Flat-Rate ...

Quotes source post: 19690247

Texas is finally set to cool off from the upper 90's temps next week, so hopefully more time spent in the garage soon. During the last round of testing I did leave it unplugged, so it's not blinking currently. I figured the warning blink wasn't doing me any good anyway. Trying to determine my next steps so that when the heat breaks, I can get busy. Thinking I'll rerun the test of filling the spare tank with water and plugging it in again to see if the flash starts again after 19 seconds. If it does, i'm not sure where to go next. PT3, I know you said it's a simple circuit, but from my reading (dangerous on the internet, I know) I've seen references to slow blink possibly caused by coolant temp sensor. Which is why I asked before if it could be caused by anything other than the tank/coolant level sensor. Does the fact that unplugging the coolant level sensor causes the blink to stop prove anything?

Quotes source post: 19691196

I haven't chimed back in on this thread, but having done this job and even producing a video on the subject, I understand what the experts say over and over. That is, when it comes to the coolant expansion tank, never use an aftermarket tank... go with a genuine Porsche tank and genuine sensor purchased through Porsche. So many knockoffs and even counterfeits. The only one that is proven to work consistently and without problems is the genuine part. I'm sure it's been mentioned on this thread, but also make sure you have a new Porsche coolant tank cap.

Quotes source post: 19691196

You are simply not taking what I am telling you literally....Explaining to you how the system works, answering specific questions, and troubleshooting for you step by step, are three separate concepts and I think your are conflating them or not following .. From where you are now it is very simple...the wire from the dash when not grounded there is no flash....when the wire is grounded there is flashing....the wire grounds via the switch/float ..you just have to use simple deductive reasoning to figure it out for your self. Trust me...your new tank is faulty

Is the coolant level sensor easily removable from the expansion tank? If so, you could remove it from the tank, connect it to the harness, and use a magnet to simulate the position of the tank�s float. Have another person nearby to confirm whether the light still blinks. If it goes away, it�s got to be an issue with the float.

Quotes source post: 19691508

Yes, I ran that test earlier and the sensor worked with the continuity/resistance test on the meter and plugged into the harness (no blinking light, until I applied the magnet), both tests with a magnet to simulate the float. The weather in Texas dropped below the 90's finally, so I swapped the replacement tank in today with the original sensor and ran the car up to hot and took a short drive. So far the light has not come back on. I will check the coolant level and wait a few more days of driving before I declare victory officially, but so far so good.

Resolution: swapped out the tank for a replacement under warranty and installed. Thank you everyone and big thank you to PT3!

Good !! That didn't take long ...( sarcastic remark) Most people don't realize that a lot of Porsche Engineering is 3 demensional, or 3 layered and can make things confusing to a lot of people... The first dimension or first layer is seemly simple and can be viewed or even tested as such for all practical purposes.. The second dimension or second layer is highly sophisticated and complex and takes a lot of study or training to comprehend all the nuances and purposes of it all .. The third dimension or layer of Engineering takes the sophistication and complexity even deeper and is so deep that if you understand the second dimension completely it becomes simple again... and quite user/DIY friendly......because the extreme deep level of complexity and sophistication doesn't negatively effect the basic function or test-ability, just adds depth and nice features....

Quotes source post: 19716395

Swapped out for another MAHLE coolant expansion tank? Or genuine OE Porsche part?