Possible cheap coolant level sensor replacement?
 

Found this part while searching for coolant level sensor prices. Damn cheap! Anyone ordered one of these by chance? Probably to good to be true.

http://www.autopartswarehouse.com/se...0&addfitment=1

In terms of what they are, they should be cheap, as its just a magnetic reed switch in the central casing with a sliding magnetic float. It'd be a interesting to do a side by side comparison as to function, fit and switching depth level. The only problem I see is that it screams non-original as its black instead of the OE white.

Possible cheap coolant level sensor replacement?
 

If the black is the only difference and it would indeed fit/work...For the price i could live with it. :thumbup:

The OEM ones i found online were $160 :confused:

That price doesn't bear any relationship to the manufacturing cost of it ... it's just the Porsche price, which applies to a few other things, for no particular reason that I can understand, other than to escalate the price because they can. They used to be about $50 seven or eight years ago ... which was also too much back then.

MJ,

Interesting find given many of the original items are quite possibly not working by now. Better still, they are intended as a direct 928 replacement without any mods.

Hopefully someone can/will offer some feedback.

Rgds

Fred

I'll go ahead and order one to inspect/test, I'll let everyone know my findings. However, if anyone else has any further thoughts or opinions I'd still like to hear them.


Thx

Ordered...We shall see. :corn:

When you apply the "Chinese Water Pump Formula" and it is that much cheaper than the factory one (or other expensive source) you should know automatically that it is no good.

True to a point.

One issue with the "Cheap Chinese Stuff" is quality control. It isn't that all of the stuff is junk, it's that you can't be certain of the quality. One might be fine. The next junk. The one after that fine.

And coolant level sensors aren't water pumps. From the standpoint of "essential for the car to function" and from the "how much time/money does it cost to replace it?" one.
From a durability/environment question too. What does the WP have to do and where does it have to do it? Ask the same of the coolant level sensor.

I'm curious as to how it does in the real world.

Does anyone have the part number for the original Porsche sensor ?

928 641 503 01

Thanks MJ - awesome find if it works - may be able to use the original cap to keep the white look on the tank. Ordered one as well - we will see.

Roger,

Am I right to understand the Porsche part is no longer available that or there are not many left to be had?

Rgds

Fred

It was NLA when I bought mine a few years ago:

https://rennlist.com/forums/928-foru...el-sensor.html

I think they were put back into production at some point between then and now, as I've found them currently for sale online.. However the prices I've seen are insane!

Porsche restocked them about 6 months ago like they seem to be doing with most 928 parts. Once they have them made again the price always escalates significantly.
Last time I checked they had plenty in stock - problem is the price!!!

Roger what's the price on the Porsche part?

List is $185 - I do not think that is too bad for a part that will be good for 25 years.

Followng that link a day later finds a list of coolant TEMPERATURE sensors, goes to other sensors like oxy and brake pad wear sensors. No levelsesrs. Am I missing something?

Hmm..I just clicked on the link I initially posted and it went right to the coolant level sensor just as its supposed to. I guess try it again?

This is what comes up for me. The headline says "Coolant Level Sensor". The items are not coolant level sensors.

Possible cheap coolant level sensor replacement?
 

That's so strange! This is what i get.


https://cimg8.ibsrv.net/gimg/rennlis...6a901cb0a6.png

I got the same page as MJ.

At $20 a pop well worth a shot to nothing I reckon. Not sure I agree with Roger on the stock item being "reasonable" at $185 a pop - outrageous is a word that more easily springs to mind in this instance.

Rgds

Fred

Note that the top cap is the wrong color on these replacements.

Yes sir, we pointed that out in the beginning of the thred, Roger suggested we may be able to swap the cap out.

Related tangent, but I presume the coolant overflow tank must be out of the car to remove the sensor? I replaced my overflow tank last weekend while the engine was out and could not see an angle to get the sensor in with the tank in place (unless maybe you remove the fender).

Ordered one of these a couple of years ago, the one I received was white even though the pic showed black. Worked great. Should have a pic around somewhere

Great to hear!...Tracking shows mine should arrive tomorrow.

Thank you for sharing Matt.

Guys,
Happy to see it works fine.
Might order one as I'll change my coolant tank soon.

I received mine - but no cap. As I was hoping to use the original white cap not an issue.
I will try and test it over the weekend.

I recieved mine, No cap as well...Which pisses me off because i needed the cap, which I was lead to believe was included. Anybody have a spare cap they would part with?

Well that s****, ...
Who knows what other information is lost with closed threads and blocked members. :corn:

Awesome, at 10% of the Stealer price, it is worth a shot. I tried to make one out of a BMW part, but it was a bust.

Has anyone confirmed this works? Im getting low coolant warning with a full tank of coolant.

Hi guys - looks like I need a coolant level sensor for my '91 S4 - has anyone confirmed this non-Porsche version works yet? Also if you receive it without the cap - does this make it unusable?

I guess if I get to needing one of these, its something to re-wire and ignore.

If we overheat..we still have a pressure sensor, and a temp gauge, and when it leaks..we still have a nose, and when we check oil every few tanks of gas...we still have eyes.

listing has gone?

Well electrically it works but the cap does not fit from the original sensor. It looks like it can be made to fit with a bit of whittling. As long as you have the original seal/gasket you should be good.

Mark - I have one here for you to try.

Nigel - I can get more if needed.

That'd be great Roger. I sent you a couple of emails yesterday with bits I need - can this be added to those?

Sean and I looked at this further today. It will take a lot of "whittling" to make this work and seal on the top of the expansion tank.

The sensor will report a different level of lack of fluid - not a show stopper.
The new sensor will not pass through the hole in the original cap - probably can be overcome.
However the sealing face cap to sensor is none existent/way too small. As you screw the cap down the sensor will fall into the tank.:banghead:

Here are some pictures.

So much for the direct replacement that they advertised. :confused:

MJ - it was great that you found it and it would have been perfect if it worked. I will certainly try and return mine.

Oh well, looks like I'm in for the Porsche version.

^ +1

Can't you find a washer to go under the sensor and over the cap? What is the OD of the sensor under the lip?

Karl - I am not sure it is that easy. The washer would have to be a special with a recess or the electrical connection would not protrude out the top of the cap enough.

OD under the lip is 19mm.

That 'replacement' is an outright misrepresentation. The only correct claim is that it's a level sensor .. the rest is BS.

I'll bet there are many others out there that are similar, and probably some that are a closer fit and function.

Is the washer for the original sensor is still available?

I'm changing the expansion tank and this would be a good time to change this washer.

The sealing washer has never been available as a separate parts as far as I know.
However I have used the washer for the windshield washer fluid tank level sensor with success.
It is 928 628 737 00 but comes as a flat disc and you have to cut out the hole in the middle - go figure!!!

Thanks Roger.

I'll look if I could get something close...
Or use the old one.

Why don't we simply make a plastic adapter ring that fits the new sender and the top of the tank? It could actually be made out of the top ring of the original failed sender; or out of some new plastic round bar stock.

Yes Please! :bowdown:

OK, I will do it. I'll order one of these replacements shortly. Then I have a couple of questions, because I haven't even yet looked at any of them on my cars.

Why are there two male wire connectors on the replacement and only one on the factory one shown? Are there actually two sets of threads in the original cap? It looks like large female threads to put the cap on the neck of the coolant tank, but also some male threads kind of recessed on the top of the cap.

What about the two little "hook-like" clips at the top (or is that the bottom) of the replacement? What are they for?

Guy says it worked great a couple of years ago--so, what's up with that?

What's the difference between the $18 one and the $20 one in the link?

It's broken.

Check! That's one answered. Thanks Sean.

Jerry

Pm me your address and ill send you the one i purchased.

Do you happen to notice it in my avatar? And thanks.

I always view Renn from my phone with the Garage Talk app, which doesn't show it. I brought it up on my browser and i see it now.

Ill get it out Asap.

OK, Mj. The Coolant tank is still in the 928 tub that I am going to start the Radical Custom work on so I'll take it out and remove the sensor/sender and then I'll better know what we have to work with.

Jerry

The sensor has shipped, should be there in a couple days.

Thx

I have one of mine out. It looks like what I thought were threads on the outside of the top is actually just a groove around the smaller radius round top, probably originally intended for some kind of rubber boot or weather seal for the electrical connector.

It also looks like the seal inside the cap for the top of the tank neck might be made with a suitable size O-ring.

I also see that the "cork" at the bottom that floats up and down does not appear to be making any kind of contact that would close a circuit and send a signal. I wonder how it is supposed to work?

I found another thread about this part and it explains about the float--it has some kind of magnet material around its core that changes some kind of resistance inside the core of the sensor and that apparently creates the signal.

Now I will wait for Mj's replacement sensor and do some of my own evaluation. I think I am going to find the answer to the question about how Guy got one to work.

It's a magnetic reed switch in the central section, which is a fairly simple switch often used in burglar alarm door and window switches, proximity sensors, counters, etc etc.

https://encrypted-tbn0.gstatic.com/i...Tw3cPdy3kjak1T

http://thumbs3.ebaystatic.com/d/l225...dt1kYGmqFQ.jpg

Mj's sensor/sender showed up in the mail today. I have made come comparisons and measurements and have the following findings and conclusions:

The hole and cylinder in the top of the original cap is .725 in. in diameter, and inside its .300 inch tall, and .200 inch tall outside. The top of the new sender, the part that might fit inside the top hole and cylinder is .765 inch in diameter, including the width between the outsides of the two barbed prongs on top of it, so it is too big to work in the hole as is.

In order for the new sender to fit inside the original cap the cap needs to have its top hole opened up "whittled" perhaps, as previously suggested, to the new sender top diameter of .765 in., and the cylinder-like projection above the cap top completely removed. That will leave the hole in the top of the cap the correct diameter and about .100 inches thick.

Then the grip range of the two barbed prongs is .125 inch. So it will be forced through the new size hole, and grip outside the top of the cap, leaving about .025 inch of space for a new rubber seal between the inside top of the sender and the inside top of the cap.

Thus there needs to be a new rubber seal made from about .025 to .032 rubber that has the inside hole at .0765 in and the outside large enough to seal around the top of the neck of the collant tank. It can be about 1.300 in od.

So the problem is that either the original cap needs to be carefully modified as described or a new cap fabricated. I might look into someone with a CNC lathe and/or mill to see if a cap can be made cheaply enough out of something suitable, and then find the correct rubber compound and thickness to cut the rubber seals.

I don't see any kind of adapter washer doing anything for us here. Because of the relative sizes, that is going to be kind of like a wheel spacer with a negative thickness.

In the meantime, I'll go out and modify my cap to show what I mean.

Sounds good Jerry...Personally I think the whole new cap idea sounds appealing as the originals are very yellowed and ugly at this point in their life cycle.

I just replaced the actual tank with a new one thats a nice bright white, it would be a shame to put a yellowed cap on it.

First I turned the top of the center hole/cylinder off. Then I opened up the center hole to slightly over the top diameter of the new sender. Then I tried to set the new sender into the hole so that the locking prongs locked outside the top.

Then both of the locking prongs broke off!!!

The last picture shows what I was hoping to achieve, but the next to last one show that both tabs or prongs are broken.

Well, I guess I didn't think this one through very well. I kind of had it in the back of my mind that the prongs were likely to break, but I guess I had to find out for myself.

However, I think this is still the direction to go. What we will need to do is make or modify the cap so that it has two small slots for the prongs and barbs to slip through then twist the sender about 90 degrees, perhaps up a bit of ramps, so that it tightens itself against whatever we find for the sealing washer. That ought to keep the prongs from breaking and yet keep the sender in place.

Jerry without making this too difficult,
simply flip the cap over and insert the sender ,
then find an O ring that will fit into the caps inner circumference,
add some DC111 to the o ring and screw it onto the bottle.
This will provide an adequate seal,
and also leave it possible to orient the sender pins, so no cutout for the locking tabs is necessary. Simply open the hole in the cap so the locking tabs will pass through.
Eazy peazy

Stan, you got me with this idea. If I take your suggestion to "simply flip the cap over and insert the sender" literally that seems to me to put the sender sticking up out of the top outside of the cap and measuring nothing but air, even if it would fit into the bottle under the fender.

However, given your usual brilliance about these sorts of things, when I figure out what you have in mind I'll bet it will be a winner.

On the other hand, my problem in doing these sorts of things is not just how I would get one of these things to fit in one of my cars, but rather how the solution of fitting the replacement sender to the original overflow bottle can be utilized by just about any one of the other 928 guys who might be in need. Even turning the original cap like I have done is not in the cards for just about 99 percent of the needy, even if that would have worked the way I had envisioned. I can fix this one to work fine with some kind of adhesive and something like your o-ring suggestion, but that is not very duplicatable for the rest of the guys.

Here's what I am thinking now. And I am sort of back to something like the original washer idea, but in a different way.

This depends on our ability to have a replacement cap made that is suitable for this conversion, or some program to modify some original caps, if some reasonable source of good used ones might be found. In addition to cutting the center hole flange down and opening up the hole like I have already done I will counter bore the inside of the top of the cap around the hole cutting its thickness to about half, leaving it about .050 or .040 thick in the center and just out far enough for the large diameter of the new sender to fit and maybe a bit more. In that counter bore we will fit a thin flat rubber seal. Then we will break the two hooked prongs off of the new sender and then insert it in the hole after the seal. Then we will fabricate a thin plastic washer that is the id of the body of the sender just under its largest diameter, and with the od equal to the id of the cap, then fit another thin rubber washer, same size as the plastic washer, under that and screw it onto the bottle, orienting the contacts correctly just before it is tight. Tightening the cap will close both seals with the plastic washer and sender in between. The small rubber will seal above the sender and the larger below at the top of the bottle.

The conversion kit will amount to the new cap, the small rubber washer, the larger rubber washer and the plastic washer for inbetween.

OK with the sender and the cap positioned like your last picture shows
get an O ring and install it into the cap after the sender then screw this onto the bottle.

NOTE when the old seals fail you can just fit an O ring into the cap after the sender is in position,
the O ring will hold the sender in place and the cap will seal the O ring between the inner edge of the sender and the top of the coolant bottle.
Eazy Peazy resealed lots of sensors like this none have leaked

Yes your right, a picture is worth a thousand words

Ok, Stan. Now I understand. But, what you are talking is about resealing the original sender. What we are trying to accomplish here is a way to make the supposed "direct replacement" new sender actually work with the original cap and/or bottle. The new one is smaller and nothing about what you suggest will help.

On the other hand and on second thought, a very thick o-ring, something like a quarter inch cross section, might very well both hold the new sender in the top of the cap and seal both around it and the top of the bottle. If there is such an o-ring, that might merit some further consideration. But we still need a new cap or enough used ones to modify as I have described. The real solution to this new sender is going to be the cap.

And, it appears obvious to me that the cap usually pictured with this new sender will not fit the bottle in the 928.

Just guessing that the threads and seal are some standard that the cap for a heavy-duty Nalgene bottle might fit. I'm a country away from the car so can't measure myself, but someone else might if it's handy. Nalgene caps are metric sized, flat-topped, and can be holed with a gasket punch or similar. And they are the same color and consistency as the reservoir.

What in the world is a heavy duty Nalgene bottle?

I ordered some of these senders to mess further with and to replace Mj's that I broke. I ended up making the order by phone and the guy on the other end, probably in India, told me that they are replacement only for 924 and 944, not 928.

hnniei - unless you are a bot please read the thread 8>)

Friggin Bot...That's word for word my first post that started the thread.

Yep...he/she/it/they got in my thread. No way in hell am I clicking whatever the link is. :grr:

Seth K. Pyle

It's a link bot. trying to get links back to their crap spam site.

Report the posts and help get 'em out of here.

I had previously opened up the hole in the top of my cap so it would fit around the top ring of the replacement sender. Today I recessed the under side of that hole so that the sender would fit a little farther through the cap. Then I turned the top ring of the sender, just above the widest flange, so that the sender would go even further through the top of the cap. I want to have room for a rubber washer/seal that I hope to make to fit around the top of the sender and still have the top all the way through the cap.

I drove around today to a couple of machine shops that I was referred to looking for someone who had a CNC lathe and could make the replacement cap for this conversion. I found one and will be looking into getting him the material to make them out of. I think I will use asetal (delrin) for the cap. Then all I need is to locate some rubber sheet to make some washers out of. I'll make the intermediate washer out of some of the scrap ABS that I have left over from my other formings. I have been saving a lot of that for various washers that I make from time to time.

Next I have to make a couple of dies to cut the washers with. I can do that in my lathe without too much effort. That should go pretty well if I have the material to make them with.

Actually the part is made for BMW and has a BMW number. Comes with a cap but I doubt that it will work with our expansion tank. I have one coming and will share the pictures to help the project.

If this works I can sell the basic sensor for $15 and make money 8>)

Great job Jerry...You never cease to Amaze me! :bowdown:

It looks like it might go somewhere, so I went out and processed some of my scrap and ABS drop material into washer blanks and turned some of them to the outside diameter. I can make about 20 of them at a time in one turning in the lathe, but I will have to open up the center hole in the lathe one at a time. That will be more time consuming. With the material I processed, it looks like I'll have about three hundred of them or more.

Here's the first batch of the partial washers

P.S. Instead of "washer" I think these are going to be our "sealing ring."

Nalgene supplies plastic lab ware including cylindrical bottles. High quality stuff. I'm away from my usual resources for finding this stuff. Rob may have info on sizes and fit.

I measure roughly 8 threads per inch (thread pitch is 0.125", or 3.1 mm) with a neck ID of 28.2 mm on a spare coolant reservoir I have here. It's made of polypropylene.

Looking through the Nalgene catalog (http://www.thermoscientific.com/cont...genecontainers)

I don't see anything off hand that matches, but will look around some more.

https://webfiles.uci.edu/redwards/pu...%204-22-16.jpg

Thanks Bob and Rob for your inputs on this. It would be nice to have an inexpensive source for bottle caps that fit that I could modify for this conversion. I doubt that we are going to be that lucky; and I expect we would have to buy the whole container just to get the caps.

Damn Spam Bot!..These are getting bad lately.

I have spent some more time on the material I'm making the sealing rings out of and have some of them finished. Out of the batch of material that I worked up for these I have started about 400 of them and have finished maybe 125 or so. Fortunately I can finish the larger center hole in these two at a time so that is saving me a lot.

I find this process kind of pleasant, but mindless, kind of like folding newspapers before your paper route. Here are some pictures of some of the process.

The first picture is a duplicate, I think.

The second one shows how several of the sealing ring blanks are screwed onto a mandrel in my lathe for turning the outer circumfrence. The third and fourth pictures, that show all the ribbons of black plastic that has been turned off the blanks, I'm showing because when I get clean shavings like that I save it to recycle into ABS glue. I simply stuff one of the empty or near empty glue containers with these shavings then soak it in MEK and create a full container of home made ABS glue.

The Picture with two clear plastic jugs in it kind of shows where I am with this now. The one on the left is the original turned blanks and the one on the right has the finished rings. Across the middle of the front of that picture is kind of the process from sawn and drilled blanks to finished rings. Then, the last two pictures show where these Sealing Rings will go.

I'll have the rest of the Sealing Rings opened up shortly--I now have about 300 of them finished.

I have located some Delrin round bar stock in either black or natural color in Denver. I think I'll order the natural since that is probably closer to the originals in the 928. The machinist I have found will make about 200 of these out of two 8 foot bars of it.

Then, I have located a rubber supplier on-line and I will either buy the material from them for the seals or may even have them cut the seals for me. I need to contact them to see what the cost is. They can supply rubber in .032 in thickness and with an adhesive back, which I think will be best. Then I can stick these to the cap and sealing ring myself.

I might of missed it but does the level sensor work in the 928 - might be worth trying before doing any more work.

Yeh, I guess you missed it. A couple of guys named ROG100 and Sean tested it and found that it works electrically, but that it will show low level of coolant at a slightly different level, which was no show stopper. Check out posts 38 and 40 above.

Sort of response I expected from you - we tested it electrically but not in a 928 because it would not fit. Just trying to be positive - I would fit one before you make a lot of parts - trial installation and hook it up to a 928 harness and make sure the light works on the dash. Simple stuff. I think it will work but better to prove it first.

The hell sorta response is this? I never plugged it in to a 928, in to a 928 connector and sure didn't test to see if it worked.

Nevermind, Rog addressed it.

Yes, I know. I was just pulling your chains.

I tracked down a better quality part from a more reliable wholesaler. Part comes with the cap but as already defined it does not work however easily removed.
Once we know the part works in anger on a 928 and coupled with Jerry's parts we should have a winner.

New unit (with cap) next to old unit

Is it going to be accurate?

Just has to tell you when the level drops below a certain point and as long as that point is approx. at the half way point on the expansion tank it will be WAY accurate enough.

If it is accurate enough for the same function in BMW's I think it will be accurate enough for a 928.

I found a source for the compound to make the seals out of. They recommend Nitrile. Then they gave me a quote to make the seals for me, but with the cost of "engineering" it makes them too expensive for the modest quantity needed for this project.

Then, last night, it occurred to me that I can do this with only one seal, the larger one for the top of the coolant bottle. What that requires is to simply glue the sealing ring I have made to the under side of the flange around the top of the new sender; then there is no need to seal the surface between the top of the sender and the cap.

I'm going to order a sheet of nitrile and then make the cutting die in my lathe and make some seals. That is, as soon as I complete the testing of these senders in the 928.

Well, I haven't done the 928 test yet, but I went ahead and made my seal-cutting die in my lathe. Here are some pictures of it with the other stuff so far.

I'm still waiting on the material for the seals. Then, now that it is a new month I can order the delrin for the caps and have that show up on next month's credit card bill. I hope I have the seal cutter sharp enough; but if not I can turn it again in the lathe. It is long enough so that I ought to be able to make several new blades on it if needed before I finish.

The Nitrile finally showed up after over two weeks since I ordered it, but it looks like pretty good material. I tried my seal cutting die on it and made a few seals, but I think my die is too sharp for as low grade of steel as I had to make it out of, so after just a few of these it is dull. I'm going to sharpen it and try some more, but I think I will be able to get plenty out of this material to meet our needs for a long time.

Here is a picture of the few that I have made so far.

I haven't ordered the delrin for the new caps yet but will probably in the next week.

Nice! :thumbup:

Hi Jerry, can you try case-hardening the cutting end of the punch?

Possible, but I haven't done that since my basic welding and forging class, back in the day.

I think I can sharpen it differently than I did the first time so that the cutting edge lasts better. I had the outer edge turned pretty thin and very sharp, but the inner cutter was more (less?) tapered and kind of blunt. The inner cutting edge actually is surviving very well and cuts pretty nicely the way it is. However, in order to sharpen the outer one I need to do both so that they cut on the same plane.

Even if I have to sharpen them a few times it will be less trouble than trying to figure out how to case harden them.

Heat the end up red hot and dump it in a bucket of used engine oil.

Any updates on this? I'd love to try one as I've got my coolant tank out for service.

Cheers

I am only about $1300 and one surgery away from finishing up on this project. The Delrin rod material is about $300 and it is $1000 to have the new caps machined. I have my back surgery in the morning, so I'll be out for a while.

Jerry--

Best wishes for a quick and relatively painless recovery from a successful back surgery.

Thanks Dr Bob.. I had the surgery and have tried twice since I got home to post an update about it and this thread, but my hands are swollen and I can't type for ****, and I keep loosing what I have typed. I'm going to give up and try later.

reviving an old thread.... any progress? This is quickly approaching my to do list.

I picked up the Delrin/Actetal rod material when I went over for my surgery, but still don't have the money to have the caps machined. When cash flow picks up I'll try to get the caps made, then I can do a bit of machining work on the senders and assemble a kit to market through my small business membership.

The Acetal rod I bought is called "natural" and it looks almost white.

I was putting this project together to take my Acetal Rod over to the machinist with the CNC lathe for making the caps and discovered that in my haste before my surgery I bought rod that is too small. I bought 1.5 inch diameter and need 1.750. Im going to order the correct rod monday and hope it is in stock for shipping; then I'll get some of these finalized. I also ordered another large batch of the senders that don't fit and that I am going to modify for use in the 928. They came yesterday so now I have 64 of them. I'll start the minor modification shortly.

To save some cash I have decided that I'll have the machinist turn the caps but leave the tops round and then I'll have to make a jig and mill the hex flats on the tops in my mill. That should go pretty quickly.

This morning I worked up a drawing for the machinist to make the caps from in addition to giving him the original cap as modified by me and the neck off of the failed coolant bottle out of my car. Those ought to give him enough information about how to machine the threads inside the cap. Tomorrow I'll call the plastics place and see if they have in stock what I need for Acetal/Delrin rod.

As can be seen in the earlier post by Roger, the difference in length of the two senders in question is about 7/8 inch. In terms of "accuracy" either sensor is going to say either on or off, so I don't know how much more accurate either could be. The on or off will however occur at a different level in the coolant bottle; and I don't know if there is any real issue about where in the bottle that will occur. You will likely have about a quart or so more coolant in the system when the low level warning comes on with this replacement sensor than with the Factory one.

So this setup will not tell the owner that the level is low but rather that they are loosing coolant?

Here are some pictures of where this project is now. The cap shown is the original one I modified, and the new ones will be a bit different, with more "meat" in the top. The ring around the new sender is in fact both the plastic ring and the rubber seal ring both of which I have made. The plastic ring will be glued to the under side of the slightly modified top of the new sender. That way it will not need an additional seal. You can see the different sending level by comparing the two senders, as previously indicated by Roger. I haven't called the plastics place yet to order the correct Delrin, but I'll find out about that shortly. These pictures are taken next to my drawing for the new cap.

looking good Jerry.

When do you think these will be ready for prime time?

I ordered the acetal rod today and it is in stock in Denver so it ought to be here this week. I'll take it over the the machinist and see what his timing is like. When he has the caps done I can ship you one, Mj, to try.

Sounds good my friend.

The Rod came today. I may not have time tomorrow to take it to the machinist, but probably Friday. Then I'll see when he might be able to get to the new Caps.

I took this project to the machinist today and left it for the caps. It looks like it will be two or three weeks before they get to it.

In the meantime I still have to figure out just how best to affix the sealing ring to the underside of the mounting flange ring on the replacement sender. I consulted with my plastics expert about it today and we cannot figure out just what plastic or nylon the sender is made of. It looks like I will have to glue them together with epoxy. I might even consider some sort of liquid rubber compound since whatever I put them together with only needs to make a seal since they will be held together by the cap and the top of the bottle neck.

I stopped by the machinist's on tuesday to check on any possible progress on the new caps, even tho it was before the two weeks predicted, and they said possibly this thursday or friday--today. I am expecting a call from them to come over and see the first one run and to give it my blessing for the full run, which I hope will be done early next week. If I can remember to do so I'll take my camera and take a picture of what they have produced.

Close, but no Cigar, yet. I stopped by the machine shop this afternoon and they were working on converting my drawing to one in their computer for use in the CNC Lathe. It looks like monday before they will have one turned out for me to approve. I was, however, able to explain a couple of things about my drawing.

Here are the first two newly designed Coolant Sensor Caps. They still need to be machined to hex around the top, and I am going to do that myself. I have to make a jig to go in my rotary table in the mill then machine the flats. I might work on that this evening.

This is looking good, coming along nicely.

Looking good so far Jerry.

I have the sealing ring glued to the underside of the top ring on the sender, so this item would function just fine the way it is. As to the flats around the top of the cap, last night it occured to me that I doubt that anyone has ever installed one of the original senders with a huge spanner and even that no one has likely had to use such a tool to remove one of these.

A thought I have previously had is that it might be better with knurling around the top rather than being hex shaped. However knurling on this hard plastic must be cut in rather than rolled in as with softer plastics, or aluminum. Then I woke up last night thinking about cutting the flats on these with the jig that I made yesterday and was dividing the circle into sixes flats, again, and was thinking what it would be like to divide the circle into seven or 8 flats, just for the fun of it.

Then it came to me that the better way to finish the tops of these caps was to divide the circle into several more segments and simply mill rounded slots (verticle grooves) several places around the top to form it in a finger grip fashion rather than hex. I think that is what I am going to do, unless anyone else can convince me that a huge spanner is likely to be needed.

Here are some pictures of the current status. I need to drill some holes in the flat bar to ut it together and mount it in the rotary table and then I can do the milling. I also need to find the correct washer for the top to hold the caps to the jig. I'll have the finished batch of caps tomorrow, I think.

Why not just cut some vertical grooves into the rim of the cap and call it a day?

That's what I tried to say I was going to do; but thanks, Shawn. (Edited it)

Oh! Sorry, I see that now. That last paragraph didn't register on my brain the first time I read it.

Disregard. Carry on! :bigbye:

Or just leave them as is.

The fear is that "as is" the smooth plastic doesn't offer enough grip to hand tighten the unit sufficiently for it to seal reliably.

Then send a surgical glove with it! :)

I drilled the flat bar and put the jig together then lined it up in the center of my rotary table and cut the flutes in one of these first two caps. I used a 3/8 inch round nose mill cutter and cut the flutes at 30 degree intervals. That resulted in 12 of them, but I think that is a bit too busy. I'm going to cut the next one at 40 degree intervals and see if I like that better. The only problem in that this acetal material cuts leaving plastic fuzz around all of the cuts that has to be trimmed away with a knife.

Jerry -- Re: the plastic fuzz, would it be possible to start the flute cuts from the top of the cal and then complete flute by cutting from the bottom of the cap? I don't know with your equipment if that would be quicker than trimming off the fuzz by hand, but it should get rid of the problem.

Here is the other of the first two caps cut with flute spacing of 40 degrees. I like this the best because the flutes are just about as wide as the space between them. Also there are 9 of them and I like the odd number. This picture also shows the fuzz just after cutting. I think cutting the fuzz off with a knife is going to be less trouble than trying to cut them twice or some other method of cutting the flutes.

Here I just put the cap on the jig and bolt it down, then set the rotary table at 0 degrees and cut to 150 on the Y table axis with the cutter set to cut just below the top ring. That gives a tiny bit of the flute, but rounded, cut into the bottom ring of the cap. Then rotate the table 40 degrees and cut again. Otherwise I think there are just too many maneuvers to deal with.

Still no thoughts about the finger flutes versus the hex top?

I like it with the fuzz still attached.

2nd vote for the fuzz.

Sorry, Hacker; here it is with a shave.

I picked up the box of nearly finished caps this afternoon. They look great. I'll start working on them, but The machinist suggested that the problem with the fuzz may be cured, somewhat, with a sharper mill cutter. I'll look thru my tooling and see if I happen to have a new one or I'll order one and then start to finish these out.

I still wonder if there is going to be any interest in having any of these finished out with the hex top, such as for concours cars?

Jerry -- I'm sure someone would prefer the hex cap. However, for all the reasons you mentioned, I think the version you came up with is more practical and is the one I'd prefer.

Go with the one that's easier to produce. The hex cap version will finger tighten fine. It might not be as comfortable as your version, but I it's not something that'll be removed and put back on all that often.

Thanks David (X2) for your feedback about these.

Definitely like the 9 flute one on the left. Awesome job Jerry!

Thanks, Chris. I favor that one too.

I think it turns out that the machinist was correct about the sharpness of the cutting tool. I looked thru my tool drawer and found a sharper bit, but not new, and tried it. It still fuzzed up the cuts, but the fuzz could be removed with my thumb nail and a bit of knife clean up. So, I ordered some new bits. I think that will almost totally remove the fuzz issue.

As it turns out, with my jig in the rotary table, I guess I could do a whole lot of different things with the top ring of the caps. By using a much smaller diameter bit I could put a higher multiple of flutes in it and get closer to the original knurling I kind of thought about early on. However, each flute, however wide, takes about the same amount of time and care to cut, so I am going to stay with the nine. I may still do some with the six flats, but even that will take a little more time than the nine flutes. I think.

I still can't figure out why the original caps are hex shaped. There is simply not enough pressure to need to torque these caps down with a wrench. And, hand tight, even over time, they are not likely to need a wrench to be unscrewed.

There is some merit to just leaving them plain as now finished, but I think the only drawback to that is the puzzlement for anyone later on to wonder how it was installed or should be removed. The hex says "with a big wrench" and the flutes say "with your hand." Plain may even say it can't be removed.

Test the tightness before doing anything to the caps. If they can be properly tightenned then doing more just adds to the cost. The fuzz will always be there, even if the cutter is brand new and will need tk be trimmed. That's just how plastic behaves when machined. I've had to do this with plastic gears that were cut from blank disks. Trimming them by the hundreds became a real PITA.

Can the flutes be the same pattern/size of cam gears?

Doesn't everyone have an extra timing belt laying around that could be used like a strap wrench for grip?

I'm pretty sure that enough torque can be put on this cap with your fingers to destroy the rubber seal inside it, so what's the point?

I bought my first '64 Porsche 356C in Bellingham, WA. Still have it.

Mj, the one you bought to "inspect/test," and that you inspected and then sent to me, is the one I have modified a bit on the top to remove the two barbs and turned the top a bit smaller and the top ring a bit thinner for the sealing ring, is the one I have been using in this development and in the pictures. I have glued the new sealing ring and seal to it and added the new 9-flute cap and put it in the mail back to you. Now I think you can test it. Let me know what you think.

+1

I'm not sure why anyone is concerned with getting this tight. The round one will be fine, and hex would be cool simply because it looks original.

Think about it....your municipal water pressure is 60-90psi. How big of a wrench to you use when connecting up a hose?

The issue is when you are working on a 30 year old part that has never been removed, that may require some persuasion and a good old set of channel lock's will make quick work of any shape.

I vote hex just so Rob doesn't loose points at the next concourse, and leave on the fuzz. Let the end-user do the trimming.

Shadow work???

Jerry

That Sounds Great!

Ill keep you posted once received.

Thanks Again for doing all of this. :cheers:

Mj, its going to be interesting to see just what you come up with for testing this later style sensor. The first test will be fairly simple in just determining that it can be installed in the 928. I'm not concerned about that one, but you are going to have to turn the connector over for it to work.

The next issue, which ROG100 says is not a show stopper, and which I have not accomplished, is to see just how it may function in a 928. That has to do with the shorter length of the new sensor compared to the original. One thing for sure is that it is going to need to function with the coolant level in the expansion tank/bottle at right at 1 inch higher level. I too thought that was not going to be any significant issue, but now a thread pops up that suggests that the coolant level in the tank is going to always be at one or another particular level, and may not be subject to simply filling to a level one inch higher. If that is the case, most of this effort may in fact be wasted, because of the difference of one inch in the on/off range of float of the cork on the respective sensors.

First I think we need to determine if the coolant can be kept at a one inch higher level from that suggested in the manual. If that is too high and will cause a loss of that much coolant, we are likely to end up with the low level warning on all the time. Then I think the only solution to that is going to be a different tank with the sensor mounted in a recess on the top that is one inch lower than the rest of the top. However, even that, with this sensor, is likely to be only a few dollars more that the 180 to 230 cost of the factory sensor.

And then, the mounting neck on the new tank could be one to accomodate the cap that should come with the new sensor rather than the replacement one I have designed. That's of course assuming that one could buy one of the new sensors with a cap, which I have been unable to do except for two of those that I received.

So, I think the test is going to need to be simply filling the coolant tank an inch higher than stated and then driving it to see if there is any coolant loss after reaching full operating temperature. If there is no loss then I am sure this sensor conversion will work fine. If the level drops back to the recommended level, with the extra inch being dumped overboard, then - - - - -

There is about a 1" difference between the two sensors in question.
Taking the normal level to be the center point (joint of both section halves) of the expansion tank and about 2.25" of coolant.
The original indicates a level of less than 1" when the float reached the bottom of travel.
The new sensor about 1.75".
I have not tried with water in the tank as yet so indication points may be slightly different.

Either way the sensor will tell you when the level has dropped about 1" or more from the standard fill mark.

Once the sensor is tested in the application you will know if it does the job or not.

I tested the floats on both of the sensors, new and original, and find that the original one floats to the top only when it is 3/4 submerged in water and the new one at 7/8 submersion. Based on that, what Roger says about how much coolant will be in the tank when either float is at the bottom of its travel is going to be pretty accurate.

The owner's manual (81 anyway) says that, after draining the system, one should fill the tank to the bottom of the filler neck and then run the engine to operating temp. At that point the coolant should be at the mid point in the tank. My first thought about that is that the coolant level drops because the coolant in the top half of the tank is filling up some voids or air bubbles throughout.

However, I went out and looked at one of my S4s and could not see any level markings on the tank except for a sticker flat on the top of it. The botton line says: "COLD WATER LEVEL." Now I am wondering how the coolant expands as it cools off.

In any case, if the starting point with a cold engine is to the top off the tank, there is no opportunity to add more to try to accomodate this shorter new sensor. However, I'm not sure about that; and based on Roger's calculations with 1 3/4 inch of coolant in the tank when the new float is at the bottom, the new sensor is simply going to tell us that the coolant level is low when it drops about half an inch to 3/4 inch below the middle of the tank rather than about and inch and a half. Then there is another eighth of an inch in our favor since the original sensor float travel is 7/16 inch and the new one is only 5/16 inch. At least I think that is in our favor.

Therefore, when you get a coolant level low indication, which should be rare anyway, it is still an indication that you are loosing coolant. Maybe that is all it is supposed to be in any case.

I have this project essentially completed, at least so far as the development goes, and including much of the fabrication of components. I have found that in the process of sharpening the nitrile cutting die for the seals that the outer diameter tended to get smaller and I ended up cutting about a hundred of them slightly smaller than I want. I then re-sharpened the cutter and now it is cutting just where I want it and I have it sharpened at the right angle to stay sharp now thru another hundred seals, and likely many more.

I have also received and tried the new HSS mill bit and it is cutting the flutes with a bit of fuzz, but which can be simply rubbed off. I am however still deburring the flutes with my pocket knife, but that takes only about a minute or less each.

Before I list these for distribution in my small business membership I think we need to know how high in the coolant expansion tank the level of coolant can be held above the cold fill mark without loosing coolant. Hopefully the level can be held at about an inch higher than the mark without the coolant leaving thru the fill cap; or that the sensor can still function at the original coolant level and still float the cork on the new sensor without sending the warning.

Hey Jerry

I recieved mine on Tuesday but haven't had a chance to get up to the shop to install it yet. I plan to get up there on Monday, So i will hopefully have feedback for everyone by Monday evening.

Thanks Again.

Went this far and still don't know what level the sensor is getting triggered at?

Roger and Sean have actually tested one of these sensors electrically and since all of the ones I have bought have corks that move freely up and down, the only other test that is really needed is to determine just how high in the expansion tank the coolant can be maintained above the cold fill mark without loosing coolant. Anyone with a 928 can determine that without the need to remove the tank to install one of these replacement sensors.

At this point we have only one data point and that is the one posted in the other thread by F4GIB where one of you state that your coolant level likes to be at 1/2 inch above the middle tank seam. Even at that level this sensor will work just fine because the coolant will have to drop nearly a quart from there to drop the cork enough to give the low level warning. That would be a significant drop/loss that should not occur in the normal fluid expansion and contraction.

So, the only test that is really needed is several more data points showing just how high in the expansion tank the coolant can be maintained. When I am comfortable in knowing what the highest level of coolant is that can be maintained without loss I will either put these on the market or put them on the shelf. So far half inch above the middle says put them on the market.

Is anyone else willing to add about a quart of coolant to their system and run or drive it to operating temp then send me or post the respective coolant levels?

I wonder if Greg Brown would be willing to torture test it for you (us). When he's doing a dyno run on one of his engines, he could swap out the existing sensor of one of these. In a matter of weeks, he (or one of the other professional 928 mechanics) could document how it works on several different cars under a wide variety of conditions. If the new sensor works, I assume it would be a benefit to Greg's customers -- unless he's got a private stash of sensors to draw upon.

As for me, I'm not volunteering to test the unit because my sensor is working and I don't want to risk damaging it by removing it for the test. I'm not suggesting that the risk of breaking the sensor for someone competent with a wrench is all that high. I just know my limits. If it's possible to break a difficult to find part, I'll find the way!

This is a low coolant sensor so as long as it sets the light off when the sensor drops to a low level in the expansion tank then it works.
When the coolant gets hot it expands and the level rises which is not an issue as you only want to detect low coolant levels.

The factory set the minimum level triger point for a reason. If the replacement sensor is innacurate then it isn't really doing what it supposed to (warn the driver that the coolant is at the minimum). It would be just a coolant level loss sensor.
Relative to the overflow tank seam, how far does the coolant need to drop to trigger the OE sensor?

There is likely no particular "reason" for the cold coolant level to be set where it is and I suspect it is set there very arbitrarily. Too, there is probably some "minimum" level of coolant for the system, but it is not likely at the point that the original sensor gives the low coolant warning.

For many decades automobiles all over the world ran with the coolant expansion space simply in the top of the radiator. There were no expansion tanks then. That suggests that the minimum amount of coolant in any such system is in fact when there is some coolant missing not only from the expansion tank, but also from the radiator.

All the coolant low sensor is telling is that the coolant level has dropped, not that it has reached some "minimum" level. It drops only because it is being lost and not "used up." The engine in a 928 does not use coolant like fuel and to some extent oil; so the only purpose of the sensor is to tell that you are loosing coolant. You should then not only fill up the system but you need to figure out where or how it is being lost.

Therefore, we still need only to know at what higher level, if any, the coolant can be maintained in the expansion tank. If it can be held, hot or cold, at about an inch above the indicated levels, then this new sensor will function exactly as the original. It will tell you that you have lost about an inch to inch and a half of coolant in the tank which you should not be loosing.

Nothing Porsche did arbitrarily. It was tested, re tested and tested more before limits were established. As with anything engineered there is always some amount of safety margin built into these limits. However, to say that these limits are arbitrary is simply isn't so. The systems that don't have an external expansion tanks have them in the radiator so they all had expansion thanks or some sort. The factory sensor tell you more than that you've lost coolant, it warns you that coolant has reached a minimum level and you need to top it off. The shorter sensor that you have eliminates this part of the sensor's function. This sensor will only tell the driver that the coolant level has dropped, nothing more. Why not just redesign your cap and lower the sensor to the factory height?

It might be interesting if there was any authority for your speculations.

As to the redesign--why haven't you already done it?

Actually I am not 100% with you on this - most sensors perhaps I would agree but the level of water in the coolant tank plus or minus maybe an 1" is not going to make much difference in the grand scheme of things. This will not make any difference to the performance of the cooling system providing the warning is heeded.
In the case of the BMW sensor it is going to tell you that you have low coolant about 0.75" sooner than the Porsche sensor so not a bad thing.
Will it work on the 928 and not give premature warnings is more of interest.

I found the COLD WATER LEVEL indication on the side of the coolant tank in my 88 S4. It consists of an arrow from the sticker on top down to just about half inch above the tank's middle seam. Then, with the aluminum replacement tank I have been doing some measurements with I find that the new sensor will just barely float the cork when the coolant is at the low level mark. Just about anything below that will drop the cork and turn the low level warning on.

Compared to the original sensor, as stated numerous times in this thread by me and Roger, that is about an inch above where the original will give the cold level indication.

Nothing in the information about this sensor and its indications even suggests that the level of coolant when the low indication is given is also some kind of "minimum!" The only purpose of the sensor it to tell when there has been a coolant loss of something less than a quart. With the level at the COLD mark, just about any loss of coolant is going to give the Low Level indication. What needs to happen to justify use of the new shorter sensor is to determine if the coolant level can be maintained at about half inch to inch above the original cold level mark.

Speculation? What part is speculation?

You really believe that the trigger point of the OE sensor doesn't mean minimum? Is it logical to expect that this point is a minimum established by the factory during severe weather and severe condition testing (then they added a bit more for safety)?

I agree, in most conditions a 1" higher level of coolant wouldn't matter but still the new sensor will not be accurate so it's purpose will not be as designed. It's like having an inaccurate oil dipstick. It will tell you the oil is dropping but that's it.

Imo -- Jerry is referring to your statement that Porsche precisely determined the minimum amount of coolant for the 928 and labeled the coolant reservoir accordingly. That is, your assertion that a sensor that sends a "too low" signal earlier or later than the original sensor would have is providing false information, and your hinting that using such a sensor might be worse than having no sensor at all (because it might give the driver a false sense of security or cause the driver to ignore warnings). This, lumped together, is speculation on your part. Unless you have personal knowledge of what the Porsche team did or did not do in this regard, you are speculating as to what they did and what they knew. Your speculations may be based on reasonable assumptions, but they are still speculations. The size and shape of the reservoir bottle may have more to do with the space available than it did with an extensive study. Similarly, where they set the fill mark may not have been determined with any precision. It's possible there was a range they were comfortable and simply picked somewhere within that range for the fill mark based on other considerations, such ease of manufacture.

The real question is why you're so determined to tell everyone it won't work? Jerry, after significant time and effort, is tying to test the sensor to verify that it works. If it doesn't, Jerry will either come up with a fix or he won't sell it. So, in the interim, what is the harm in waiting on the test results? It's not as if Jerry's trying to foist an unproven product on the public.

Mj reports that he installed the new sensor in his 928 and that it functions just as the original one, with no low warning either hot or cold. He doesn't, however, report what the coolant levels are in his tank, either cold or hot. I have asked him for those readings.

I found one other data point online in a lengthy report by someone who showed his way thru replacing his expansion tank in a 928. The most interesting thing about it is at the end he shows maintaining the coolant level at about an inch or little more below the top of the tank, rather than just above the middle seam.

I need more input about the ability to commonly keep the cold coolant about an inch or inch and a half above the mid seam. If I have to do all of this myself it will be some time before I am comfortable trying to market these new sensors.

Again, I know exactly where the new and the old sensors will trigger the low indication; and that is right at about the middle seam for the new one and about an inch and a quarter below for the original sensor. The cork on the new sensor does not float until it is fully submerged, at least in plain water; the original one at about 3/4 submerged. That means that if Mj's is working correctly and he is starting at the Cold Level Mark, just about any loss of coolant is going to send the low warning, rather than nearly a quart of loss for the original.

I'm not sure just what all circumstance are that might lower the coolant level below the Cold fill level, except loss, and it may be that the only one is coolant loss, but I am not comfortable thinking that any minute loss is what we want for the warning to come on about.

I am still hoping that some of you will go out and add either a pint or a quart of coolant to your tank and tell me what happens to it hot and cold.

Reviving this thread to find out what the outcome was. After reading the whole thing and sitting at the edge of my chair the discussion abruptly dies... What happened?! Is someone selling the cap and seals to make the BMW sensor work?

https://cimg1.ibsrv.net/gimg/rennlis...99a81f5454.jpg
https://cimg2.ibsrv.net/gimg/rennlis...ff47ea7429.jpg
https://cimg3.ibsrv.net/gimg/rennlis...d40921662c.jpg
I completed the development of the alternate Coolant Level Sensor/Sender and left it pretty much just where the last post by me says. I did recently sell one of them to a new young member with his order for some of my other items, but have not started a thread about the availability of these generally because of the lack of any additional testing or feedback about keeping the level of coolant in the overflow bottle a bit higher than the level mark on it.

If anyone else is interested in trying one of these in their 928 they are available for $45 plus $5 for S&H.

Thanks for the reply Jerry. I am going to pull mine out and fully characterize whether it's functioning or not as the "coolant" indicator seems flaky right now. And rule out pressure sensor malfunction.

Do you have a website describing your offerings? Can't find it on The Google...

I don't have a web site, but my offerings are listed in my signature. You can probably do a search and find a thread about most of them here on Rennlist 928, which is just about the only place I do any marketing.

I didn't push the sale of these replacement sensors mainly because of all of the missinformation about the different level that the sensor will need to function at. I think the onlly one I sent for use is the original one I modified for Mj. He reported that it works whether hot or cold. No one else gave me any feedback about trying to keep the coolant level in the overflow tank at about an inch higher than the level mark on it. I think Mj's sensor is simply floating in thin air above the coolant. At least he has a nice white cap and the tank is sealed.

I lately fiound that I mistakenly relied too much on information about the BMW sensor from others on this forum and from the bad information that the BMW sensor was a direct repalcement for the 928 sensor, even though we all knew that it was not, at least in terms of fit. What I have found out is that that is not the only difference. I happen to have an original sensor laying around near my computer so it recently occurred to me that I ought to check to see if they function in the same manner notwithstanding the difference in length. What I found was that all of my efforts in this project were wasted. The 928 sensor makes an electronic connection when the float drops to the bottom, which is of course intuitive, but the BMW sensor, that I have nearly a hundred of, makes the electonic connection when the float is at the top. DUH! I wonder if the opposite configuration of the contacts on the top of the sensor might be a hint to the opposite manner of functioning.

Anyway, that pretty much kills this project except for the outside chance that there is actually a similar alternative sensor that functions like the 928 and might adapt easily to the 200 caps that I had made, or that the caps I had made might be useful to some minor extent to repace an original. There is a post in this thread way above where someone says that they purchased some kind of the repalcement sensor that had a white cap and that it works perfectly. I still wonder what that really was. I actually seriously doubt that, but I may form time to time search for such an animal.

P.S. That was GUY in post 27 above.

I've been learning a bit about relays in connection to a headlight harness that I want to make for my 944. It's a well-known project and problem for those cars. Anyway, as I was reading this thread and what the outcome was due to the function of the new sender vs. the old, I wonder if a small relay wouldn't be the answer to making this work. I know, it's an extra layer of complexity, and I'm not actually suggesting that something like that be incorporated, but it struck me as a way to try to reverse the signals that the BMW sender is providing. I'm not an EE, so there are probably other ways to do that but this one is in line with other things that are currently living in my head.

Bummer about what happened with this project, but I love the pictures of the design and machining work.

Cheers

Thanks Zirco. That is not a bad idea and probably would work. In fact that is probably how the BMW sensor itelf works in real life. The problem with it is what you suggest, that it would make the use of it too coplicated. It really needs to be simply a plug and play situation.

One of my thoughts is that maybe there is some other sensor somewhere in the world that could be adapted that functions the way ours does in the 928. I have searched eBay which lists over 8 thousand collant sensors and have found one that looks like it has possiblities, but I cannot tell from its pictures and listings how it actually works. I have ordered one to see just how it works and if like ours, just how it too might be adapted to the modified BMW sensor top which will still be needed for the tank seal and cap and for the connection. That will mean that it takes two different sensors to adapt into one with my new cap, so the cost will be much more than I had planned, but still much less than a Porsche Factory one.

I don't have great hopes for this new idea because I have a hunch that the 928 coolant sensor is much like the 928 gas cap; that is, there is no other like it in the world.

Porsche, there is no substitute.

Sorry, couldn't resist.

Interesting point, Joe. However, when you think about it that slogan leaves a lot to be desired. Since in most, if not all contexts, a substitute suggests something that works but of lesser quality, whether in sports, automobile racing, clothing and just about everything else. For Porsche that slogan results in suggesting that there is noting worse than Porsche to substitute. In fact with the fabrications I have created for the 928 that holds true since all of my creations are better than Porsche and most have been highly successful as replacements for the stuff Porsche came out with.

I think my slogan is going to be "Jerry - Has a better replacement!"

I received the sensor I found on eBay and the first point of encoouragement is that it works the same as the 928 sensor, that is, it completes the circuit when the float drops to the bottom of its travel rather than breaking the circuit as with the hundred or so "substitutes" I modified with my new caps. But, that is the only bit of encouragement since the float on the new sensor will not fit thru the neck of the overflow tank in our cars. Otherwise I think the switch in the new sensor can probably be adapted to the ones I have put together.

With the new sensor I have found the float is at the bottom but is connected to a kind of ladder-like extension up to the top of the sensor where the switch is actually located. And, the float assembly is quite a bit longer than the original 928 sensor. What that would have meant is that to adapt the float and switch part of the new sensor to the top of what I have put together would require shortening the float assembly/ladder to match the length of the 928 sensor. Again, that is if the float would go thru the neck of our overflow tank.

Now I think that I can probably adapt the switch part of the new sensor to what I have made using the top of my replacements for the connection part and my new caps, but cutting off the bottom of the float and its extension and replacing the float bulb on its bottom at a different location and with something small enough to go thru the neck of our tank. I don't have a good idea about what that float might be made of, but one thought is that I could make it out of wood and cover it with epoxy so that it does not become saturated over time; or, maybe even cork, which is actualluy a kind of wood; or even bulsa. Maybe i could machine a float out of some kind of plastic and glue it onto the ladder, but that seems like too much work.

I'll just have to think about it some. I think the whole thing would come out to about $75 to make it worth while, and for how kind of cobbled together it might end up it probably will not be merchantable, even to save over a hundred bucks over a factory sensor.

The brain is a wonderful thing. I was driving home early today from the office in one of my cars from my other collection, a 99 Eldorado, which is frequently displaying a signal that the coolant level is low. Then I was wodering where the coolant level sensor is located since I have not seen it on the top of the overflow, so I was thinking that it must be somewhere under the plastic tank. Then it came to me that the solution to the otherwise failed sensor project might very well be to figure out a way to adapt my replacement sensor/sender to the under side of the 928 tank. Volla! There is another thread on here about a smaller metal tank that I think Greg B. might have posted about. I think it deserves some further consideration in terms of putting a suitable sight gauge on one side of it and then welding on a new bung for the sensor/sender under it, plus a new bung for some kind of pressure cap on top.

My current thought is that the smaller tanks we see in the other thread, and I have about 3 or 4 of them, are just too small for the 928 application, notwithstanding the lack of a sight gauge. I think the 928 overflow tank is carefully sized to allow for just about twice the amount of expansion of coolant with heat. The minimum level for coolant at cool is about half of the tank volume and the expansion above that is just about the same so that any more than that is overflowed onto to ground. The smaller tanks that we looked at with Greg are about half the size of the originals and I think are just about that much too small. To make my sensors work with any new tank is going to take a compete new tank with about the same volume as the originals but with my sensor and bung on the bottom. I guess I need to look into the design of such a tank. I should also consider the inclusion of a sight gauge as suggested by Greg B since I doubt that I can come up with any kind of tank that anyone can see through.

Jerry, thanks for continuing to think about this. I know we all wish the project had worked. I'm needing to order a sensor soon. I had a bad tank cap that led to a dry reservoir and could have overheated the engine. $200 is hard to pay for such a simple liquid level switch.

I don't know if $200 is too much to pay for one. Mine is still running strong...I have Carl's aluminum expansion tank....only issues I've had is getting the sensor tight enough so it doesn't leak at all when the system is pressurized and the hinky connector won't stay on well enough to keep from occasionally getting false alarms...

If/when mine goes...I plan on paying $200 for a new OEM one...this one is 35 years old and counting...that's a good life for a sensor that operates in a pretty harsh environment...my humble opinion. Worth $200.