FredR

I posted a similar thread a while ago and never found a resolution but there is a bit more info to add so will try again.

I posted a pic a few days ago about my dash panel in my PSD flush thread. For some reason when the engine is switched off the needle will swing to full scale displacement and stay there [see below]. If the engine has been switched off for a while and I knock the display with my knuckle the nedle will sometimes drop back down to zero. Switch the ignition back on and the needle goes to full scale and stays there and will not reset no matter what. It seems to do this every time and even more baffling, if I disconnect the battery the needle still stays full scale.

My take on this is that these scales work on the principle of induced magnetism and if I did not know better, I would say that somehow it seems as though the magnetism is being retained semi permanently for some weird reason- whether such is possible or not I do not know.

Once I start the engine the gauge more or less behaves normally. I consider it to be working normally when the reading goes up and down with the revs. Hot idle usually shows 2 barg and in a departure from my late S4 that my motor came from, it seems to max out at 4.5 barg that I attribute to a possibly different gauge calibration [or some signal interference].

When this problem first surfaced I attributed it to the 3 pin sender failing- purchased a new sender unit and had it fitted but when I got the car back it was better but still far from perfect. Did a bit more investigation and found the engine bay wiring from the 14 pin connector was a dog's breakfast. Ended up replacing about half of the cores with new wire myself. That seemed to improve things but even so the display behaviour catalogued above was present and occasionally the display would seem to freeze at say 3 barg and not move up or down irrespective or revs. Just as the problem would manifest itself it would mysteriously dissappear again.

I removed my display- checked the terminal rails and everything appeared fine. Fitted the display from my late S4 and it did the very same thing.

The only other pertinent piece of info I can add is that I do not have the stock 3 pin connector plug- that dissappeared before I got the car but no idea why. It seems the agents used cable crimps to fit onto each pin of the sender- crimped to the wire and then used as a push fit on the other end of the crimp and to be fair, they are well seated as it takes a fair bit of effort to remove them. Whether or not there is any signal leakage- no idea but I doubt it. I did consider fitting some heat shrink over the pins but as we have little to no rain did not bother.

Took the car for a spin this afternoon and nearing home the gauge seemed to freeze at 3 barg. When I shutdown the engine the display reading very slowy dropped and as it reached zero the display suddenly shot up to full scale deflection.

Any suggestions?

FredR

Obviously this one has you boys as perplexed as I am.

Did a bit more internet trolling and came across a thread on Pelican parts talaking about earthing and possibly a partially bad sender unit- bummer if the latter is the case given this was a new sender unit not long ago.

Anyway took the thing for a spin this afternoon and again after about 20 minutes or so of driving the pressure indication seemed to peg at 3 barg- no variation with rpm's. After I parked up the needle slowly dropped and then suddenly sprang to full scale [again]. About an hour later took alook at the thing, thumped the display with knuckle and the needle dropped back down to zero where it should be. This suggests to me a bad contact somewhere in the display and yet when I ran with a different display unit the same problem occurred- does this put the problem somewhere in one of the four connector strips?

Also forgot to mention that my tachometer signal also breaks down at about 5k rpm with the needle swinging around around full scale but somewhat erratic- could there be some kind of common earth for the display unit? however if such were the case how come other instruments are not seemingly afffected?

Urrgh-I hate electrical problems!

Tomkat80222

I'm sorry I have no answers for you, but this one belongs in the strangest things my 928 does thread.

FredR

Did some continuity testing today at the 14 pin connector terminals. For terminals 2, 4 and 7 going to the sender unit I could not detect any continuity to earth or any cross connection- good I believe?

Terminals 2 and 4 back to the display no earth continuity. Terminal 7 to the display had about 1.5 ohms resistance- given that is the earth connection to the display I take that is acceptable.

Mrmerlin

you may not get any ground indication as its a resistance thats being measured

FredR

Stan,

At the moment I am not finding anything obviously wrong. Whether or not there are any more tests I can carry out I do not know. Tomorrow I hope to find my previous sender unit and do some testing on it and if the tests prove positive I might try putting it back in now that I beleive the wiring is sound.

Still struggling to understand why the gauge pegs at full scale when switched off and after a while a tap on the display gets it to drop down to zero- clearly that cannot be caused by the sender unit.

The issue with the display freezing may well be the sender but I am always a bit suspicious when it seems there is more than one problem running concurrently even though it is quite possible.

FredR

For anyone following this I did a bit more testing today and found something that seemed interesting. I re-tested my connections at the 14 pin connector to verify what I reported previously in post No4 and got the same results [good].

I then pulled out my previous sender unit [beleived to be defunct]. When I tested between pins m [earth] and mk [pressure switch] there was continuity with no pressure- this contradicts what I found on my installed sender. I then jury rigged a pressure test on my old sender unit [see pic below] using my tyre inflator. fortunately the rig has a small volume and leaked a bit which made testing easier as I could see what happens as the pressure was falling. On my old sensor with the leads connected across terminals M & WK and the pressure dropping at 6 psig the switch resistance flipped from infinity to zero [i.e. continuity]. Given the installed sender shows infinite resistance [open circuit] at zero pressure does that imply the installed sender is goosed?

On the other hand, with my old sender tested on the other contacts I could not find any variation to resistance as pressure varied suggesting that indeed pressure indication on that unit is knackered.

So, to summarise, it seems to me as though there is something wrong with my installed sender given M and WK are open circuit at zero pressure and my old sender is also goosed because there is no variable resistance as the pressure changes.

Is my reasoning logical? Either way a new older type sender is on its way to me.

FredR

Bumping this thread a little-

I have ordered a new two pin sender that is winging its way to me. Looking at the installation I get the impression that to remove the current sender I would need to either remove the oil filter or remove the alternator- any comments/tips?

I would also like to test my [relatively new] current sender given it is behaving a little different to my old one [see above details]. I am assuming that the pressure indicator will have variable resistance in the range of a few hundred ohms- does anyone know what resistances to expect relative to pressure?

Rgds

Fred

Mrmerlin

you will have to remove the alternator to gain a wrench hold I used my 12inch knipix flat jaw pliers to hold the mounting.
Then a 24mm to unscrew the sender.
Use a coat hanger to hold the alternator

FredR

Thanks for that- getting the alternator off is simple enough getting the damm thing back on is a pain in the *** working off jack stands. I usually end up using a scissor jack to help push it into position. With the cables off it is easy enough but when the cables are in situ I find it difficult.

Does removal of the switch lose much oil [I am quite good a getting under car showers]?

Rgds

Fred

Mrmerlin

about 2 oz of oil,
I used a small bowl to catch it on my 88.
I did move the fan shroud and cut the hose tie to move the shroud a bit,
I have Hans MMs as well,
the alternator was pretty easy to put back into place.

FredR

Stan,

So I can look forward to at least 1oz of oil shampoo!

I always remove the fan shroud - I have the Spal 12 inch dual fan setup with my custom made mounting brackets [made from one of my wife's aluminium kebab skewers!]. Removing the fan shroud is a 60 second job.

I like to have plenty of room to work in!

kmascotto

I wish I could help Fred...my answer to most things is replace everything...not great advice though

FredR

Kent,

On these things more often than not that is pretty sound advice- trust me! The oil pressure indication is probably no more than a nuisance but I do like to see the temperature gauge and the oil pressure gauge working correctly.

Engines with real oil pressure issues generally speaking do not last too long. What I am seeing is inconsistent readings suggesting signal corruption to some extent or other. It irks me just as my tach signal breakdown irks me but with the tach signal I know that no matter what I do the high rev fuel cut off is going to protect the engine come what may - the rest is irrelevant.

Rgds

Fred

FredR

Update- just received one of the earlier two pin type senders with the spade connectors rather than the three pin R2D2 interface that no connectors are available for.

Did a bit of testing to verify it actually works and got the following results if of interest to anyone:
5 barg 185 ohms
4 barg 160 ohms
3 barg 130 ohms
2 barg 95 ohms
1.5 barg 75 ohms
1 barg 55 ohms
Low pressure switch- changeover point about 6 psig.

The above readings should be taken with a little lattitude [+/- 10%] as my test rig was slowly losing pressure - this at least made the task at hand easy to complete.

Hopefully I will get round to changing out the switch in the next day or two. I somehow doubt it will solve my problem but it should at least eliminate a variable.

If this does not work then the problem has to be in the wiring from the 14 pin connector to the dash panel given everything else will have been eliminated. I would like to do another test. If I undo the 14 pin connector and switch the ignition to the on position to test what is happening in the cores used by the sender will that cause any problems? Cannot see why it should and I dare say the thing will not start but then I have no intention that it should.

Rgds

Fred