996 Instrument Cluster Failure - Third Time

2003 Carrera 4 62,000 miles never been wrecked or raced Help! I am going absolutely crazy. I have now replaced the instrument cluster for the third time in less than one month. The clusters have been acquired from reputable dismantlers, programmed and installed by the dealer, no engine/charging faults determined. Then within a few days, the two replacement clusters exhibit the same symptoms as the original unit. Gauges go crazy and then simply freeze up and dash lights go out. The only remaining part of the cluster remaining functional is the digital odometer and outside temperature display. This is becoming an expensive nightmare and so I am seeking any feedback as to what might cause this. Could it possibly be some electrical spike from the alternator or voltage regulator that destroys the circuitry, which does not show up in the diagnosis? I do not want to spend another $1,000 when the incident will probably repeat itself.

Do you not think that you may have a short in a wiring harness on the car and that is what is causing each of the clusters to fail? Pete

Has to be a short somewhere that is frying the clusters.

Wouldn't a durametric help spot errors in power? If you can spot the alternator messing about or diagnose from which regions have the erros reported you should be able to figure out which branch is feeding the problem. I had a 64 mgb in high school and that thing had positive ground. I doubt england still uses positive ground.....wouldn't be a british model I'd figure anyway. I'd buy a durametric before letting it get away with that trick again. Ask them at their website if their stuff will extract info you need to spot the problems source.

Quotes source post: 7783688

Nope. What others have said. There's a short. More accurately a wire has some insulation scraped away or is caught under a bolt head or between a bracket and the dash and some movement of the car results in this wire making some sort of electrical contact (to ground or to a 12 volt hot wire) that results in a fried instrutment cluster. These wiring problems are very hard to solve. I expect for you in order to solve this one you'll have to remove the dash and as much hardware under/behind it and examine all wiring for any signs of any possible problems. Trace any and all wires and wiring harnesses as far away from the dash area as you can. The problem might lie just beyond the are where you think it lies. Sincerely, Macster.

I feel ya man. My odometer and temperature outside are fried. Maybe if we join both of our cars, we will have one complete?

you might be money ahead to get a pro shop to fix this for ya.. someone chime in a shop for him

These can be really hard. As Macster said, tracing each wire is often the only solution. This assumes that: - either something was done wrong during the car's build - there has been unusual wear, or - some mod has caused this -- either by inappropriate wiring or fastener that caught or damaged a wire. Since this is not common, it might be worth it to review any mods that have been done to the car (even if they're non-electrical) and look there first.

I would check fuse sizes on that cluster (make sure the fuse is to spec).... If there is a short, the fuse should blow... not the cluster. Cheaper to replace fuse BTW....

Greetings: Thank you for all your input. A couple more facts about the situation. * As far as I can tell, there is no fuse in the cluster itself. I have had it apart and see no fuses in the instrument cluster. * The clusters went haywire when I hit a rough patch of road or bump, with the first going out after going over some railroad tracks. * I have checked and rechecked the fuses in the fuse box next to the driver side door. All are to specification and none have blown. So, the malady seems to bypass the fuses. * Before installing cluster #3, the dealer's mechanic checked the alternator. He indicated the diodes show some wear, but does not believe it to be extensive enough wear to throw an electrical spike that would fry the cluster. And, no other electricals indicate any problems. So, I suspect I am going to have to bite the bullet and have my independent mechanic start tearing the car apart to figure if there is some loose or frayed wiring. Any other suggestions would be appreciated. Chris Stocker

Quotes source post: 7786622

I'd seek confirmation the cluster circuit not protected by a fuse. This seems wrong to me. Let me check... In the meantime, the info the cluster went out/haywire after car encountered rough road strongly suggests an electrical contact is the source of the problem. I'm not knowledgeable enough to advise you beyond generalities but this suggests if you could hook up a voltage/amp meter to the instrument cluster circuit -- the instrument cluster does not have to be present -- and with the key on to power this circuit moving whatever wires/harness and any thing under the dash you can and observing the meter dial you might be able to match up touching something/wiggling something and seeing the volt meter needle react. You almost certainly will have to remove some hardware once you've located the suspected wire/harness and to gain better access to these potentional sources of the problem you might be better off removing the dash to begin with. As I started out with: generalities; I'm short on specifics. Sincerely, Macster.

Quotes source post: 7786622

While I don't have the car in front of me, nor the wiring diagram, I'm pretty sure there's a fuse, maybe two, associated with the instrument cluster. You need to locate a fuse diagram for your car. I consulted with one for a 2001 model and found in Fuse Holder B, fuse #S1, B1 "CLUSTER TIP BUTTON PSM PSM DIAGNOSTICS POWER TOP" and in S1, B18 "INSTRUMENT CLUSTER DIAGNOSIS HEADLIGHT VERTICAL AIM-CONTROL". And in Fuse Holder E, fuse #S1, E1 "TERM/86S CU ALARM-CL/RADIO/CLUSTER/INFOSYSTEM/CU SENSOR OVERTURN/DAYTIME RUNNING LIGHT". In this particulare model car one of these fuses protects the instrument cluster and the protection may be split between 2 possibly even 3 fuses. Same in your car. Though the fuse holder and fuse number and groupings may be different. Sincerely, Macster.

According to the 996 Essential Companion Row B Fuse 1 = Instrument cluster (15 Amp fuse) Row B Fuse 10 = Instrument cluster (15 Amp fuse)

As a second update on my ongoing saga of three instrument cluster failures in less than two months, my problems still exist with no solution in sight. I took the car to the Porsche dealer, (Rector Burlingame), who installed and programmed the previous two replacement clusters. They were unable to determine the root cause of the problem. I then took the car to a reputable auto electric shop, whereupon they spent approximately 12 hours utilizing their extensive experience and Porsche diagnostic equipment to diagnose the root cause of the cluster failure, ($1,500 out of pocket). Unfortunately, after an invasive investigation, (dash removal, tracing wiring, sensors, fuse block analysis, etc.), they were unable to identify any issues, (alternator, wiring, fuse block, harness, charging system, battery, cluster connectors, etc.), which would cause repeated failures. I then took my car to my independent shop to see if they could attempt to sort out the issue. Apart from the cluster itself no other fault codes are present. They also contacted two other Porsche dealer service managers, (apart from the one I initially took the car), explaining the issues and as expected, they too have never had any other Porsche of any model experience similar recurring malfunctions. I subsequently wrote to Porsche North America for a second time replaying the history of the issue, which includes nearly $5,000 spent to date and the Porsche dealer, independent auto electric shop and independent Porsche mechanic unable to diagnose the problem. I stressed it hard to believe my car alone among thousands of late model 993/996/997/Boxters is the only one experiencing similar multiple instrument cluster failures, requesting the issue be referred beyond customer service to their technical department to research their database to determine if similar incidents have been reported and how the issue was remedied. Their response was to inform me that the car was no longer under warranty and therefore they bear no financial responsibility to remedy the problem and that I should take the car to my local Porshce dealer. Most irritating, I was not asking them to assume any financial responsibility, but only to help in its diagnosis and identifying a solution to this problem. So, it appears I am resigned to driving the car sans any gauges as I am loathe to purchase and install another instrument cluster as the likelihood of failure is high. Has anyone had any luck in the past getting Porshce to assist in determining a problem their dealers have been unable to resolve? Your feedback would be appreciated. Thank you. Chris Stocker

Typical Porsche response. They are just covering their *** and never want to acknowledge any on going issues. I would check all the grounding points on the car, clean them and re-tighten them. There is an article in the latest issue of Excellence on Porsche electrical systems and common faults. While there is nothing specific to the problem you have been having, they do talk about bad grounds and the problems these cause, and how ECU's get burnt out by bad ground. The fact that it seems to happen on rough roads could indicate that a ground is vibrating and arcing. There are a lot of ground points on these cars. I will try to look and see which ones are closest of more involved with the instrument clusters. Seems to me that if a dealer sold you the cluster and installed it, it should be covered under some sort of warranty.

Quotes source post: 7834020

In his original post he states that he acquired the clusters from dismantlers and had the dealer install them. I have not used a Porsche dealer for repairs yet, but in my experience most franchised dealers offer warranties on parts and labor purchased through them. At this point, I'd be very tempted to take the car to the dealer, bite the bullet for a new cluster from said dealer and have them fix the car. If it fails as quickly as the used clusters did, the responsibility for fixing it should then fall on the dealer.

Quotes source post: 7834157

That would be my suggestion as well. I believe all parts installed by the dealer have a 2 year warranty. I would confirm that before moving ahead and get the answer in writing if you can. If/when the new cluster fails they'll install another one, etc. or actually diagnose where the trouble lies. Failing that, you could always trade in the car on another one

this one is going to SUCK. -What else has been fixed right before they started to fail? -If you can tell what is fried exactly.. You may be able to track down where to start to look (I am a electronic engineer, I would be willing to look at one if you want to send it) -Any idea when it fries? Just driving? Hitting a bump, you get in it,, turn the key and zilch... -Are you 100% sure the old ones are in fact dead and the short is simple causing it to fail but not really ruining the cluster? Many Many people end up trading in cars with electrical issues.. Best of luck.. and I mean that. Pm me if you want to take me up on my offer.

Can you rent time on a shock-dyno "shaker rig"? I agree with the others - this "bump" factor is too coincidental to be overlooked. And, it might not be a short behind the dash, but somewhere else on the car that eventually leads to the dash wiring. If Kyle can figure out which part of the instrument cluster blew, that might help you figure out where to look on the car. And, as he suggests, some other recent work might have caused the problem. Let's say some suspension work was done and a wheel speed sensor wire got damaged (which leads to the dash for the MPH)... maybe something like that is causing it.

as mentioned, maybe there is nothing wrong with the prev instrument cluster.... one way to tell is to mount it on a second car.... sometimes all it takes is one bad wire to pull down a whole circuit board... but it doesnt necesarilly mean that the board is bad. It just doesnt work with your wiring because of external problem....

Not certain if the shop manual has the electrical read outs for the car but if there is such data a laborish testing would help. Have you ever tried a durametric? Seems a problem might show some place. I'd also check your plugs and sockets for being secured plus the state the pins are in like if they are recessed too much or possibly extending too much giving rise to a short. Might just check the ECU modules connection.

How about one light (bulb?) out for the alternator gauge. The needle lights up, but backgroud is dark. My independent tech had the cluster out, but couldn't find a bulb or any other visible issues. THere have been no other problems with the cluster. Even the dealer couldn't find any replacement bulb in the catalogue/system. Any ideas? Thanks, Mike

Quotes source post: 7833861

Man oh man. Let me think about this... Wiring problems are very difficult to trouble shoot. Part of the problem is some automotive techs think more like mechanics -- understandably -- than electricians. Automotive electrics particularly tricky cause so much of the wiring is buried in the car. I've seen Cayennes gutted of their interior down to sheet metal by techs looking for a bad wiring harness. I think it is going to come down to you sitting in the car testing each wire of the instrument cluster for continuity and shorts/grounds. While I offered the idea the problem was a bare wire, or one caught under a bolt head or pinched between some metal it might be a wire inside the harness broken and upon a sharp bump jabbing into another wire through the insulation. For a new car the solution would be to replace the harness and send the old one back to the factory. In your car's case a new harness -- and which one? -- is expensive. It could cost in parts and labor what you've already spent. I can't help but think there's some shop somebody out there who's encountered this though. Trouble is finding him. Sorry not of more help. I'll think some more on this and ask around and see what I can learn. More when I know more. Sincerely, Macster.

Quotes source post: 7834364

After thinking about this a bit I like your idea. Looking at a "fried" instrument cluster to see what's fried, exactly. A postmortum on the cluster might very well pinpoint the wire or circuit that is causing the problem. That no fuses have blown suggests -- I write software/firmware for a living and don't know electronics/electrics like an electronics engineer -- the spike, if it is that, is coming in via a ground line. Sincerely, Macster.

Greetings Again: I greatly appreciate all the feedback received in the past day regarding my ongoing woes to remedy the instrument cluster failures. In particular, the kind offer to analyze the failed cluster. I have decided to bite the bullet and bring the car back to the dealer along with the three fried clusters. In addition, I will hand over to them all your suggestions, the thirty pages from the 996 Workshop manual on instrument cluster diagnosis, reports and receipts from my ongoing efforts to repair the problem and copies of correspondence to/from Porshce Cars North America. My hope is that with some forceful, but respectful cajoling, they will escalate the issue with Porsche Cars North America to identify other instances of repeated failures. And, if necessary they will be authorized to conduct other diagnostic tests, which have not already been conducted. Failing this strategy, in lieu of installing another cluster, I will leave the car by the river near our ranch and wait for the car to disappear and be rid of it. Best wishes, Chris Stocker

Greetings Again: My car has now been at the dealer for the past two days as they seek to determine the cause of the instrument cluster failure. They have already burned through the $500 diagnostic estimate and will let me know tomorrow how much more I will need to spend solely on diagnostics. As far as their findings, the service rep indicated: * They believe the root cause of the problem is a fuel sending unit, which apparently sits in the gas tank. Over time the plastic covering the wires can disintegrate exposing the bare wire to the fuel, (boy this is certainly a wonderful engineering mistake as if I am not mistaken this is what caused the TWA 747 to blow up). * They plan to conduct some additional investigations, but so far have not come up with any other issues. Now for their initial recommended repairs, which are: * Install new fuel sending unit. * Install new alternator; * Install new instrument cluster; and, * Road test car for 300 miles to see if new cluster fails. If it fails again, he indicated they would provide a no-cost replacement cluster, but would need to charge for additional diagnostics. When I asked about the cost, he could not give me an estimate yet, but noted it will likely be quite expensive due to the cost of replacement parts, labor for installation/programming and the time required for road testing. I suspect it will be northward of $5,000. My initial thoughts are to say thank you for your efforts to diagnose the issue and have my independent install the fuel sending unit, either skip the alternator or have the current one rebuilt and purchase another cluster from the dismantler. I believe this would be far less expensive than their recommendations. If the cluster fails again, the car gets left with the keys in it by the river near our ranch. Any comments or thoughts on this would be appreciated. Chris Stocker

Chris, I think your plan is sound. The dealer is guessing, they have no clue. I'd start with just the sender. Just my humble opinion.

Sorry to hear your saga. This is a tough one for sure. Any work done right before the very first failure? There are repair shops specilizing in intrument clusters repair/rebuild if you google. If you send them the units to look at, they may be able to narrow down the cause and give you other useful info.

Based on the fact that the last failures happened due to vibration, I doubt it's the fuel sending unit giving you grief as the failure would duplicate itself much sooner. Failure most likely caused by improper seated wire/exposed wire causing the cluster to ground out. Remove the cluster and hook up a meter and try to duplicate scenario. You may end up driving over railroad tracks for a day, but hopefuly should be able to isolate the wire causing you grief. good luck

One more idea if you are trying to diagnose it by yourself. 1) Send one of the better clusters to a specialist to get it repaired and ask them what's the most likely cause ~$500, I think. 2) From the circuit diagram (of a '99), there are only three plugs going into the cluster (blue, white, and black). The black is the power supply (and other things) so you need to connect that one at least to power the cluster. WIth teh new repaired cluster, you can leave the white and blue disconnected and drive for a couple of weeks and see if problem reappears. If it does, at least you reduce the problem to one third. If problem does not reappear, get the blue or white connected and drive for another few wks and so on until the problem reappears. Hopefully you can narrow down your problem to one third at the end. Combined with what the cluster specialist tells you, hopefully you can narrow it down even further. Also, you may want to tell the specialist the situation and maybe he will cut you a good deal if you need another repair. BTW, I also don't buy the fuel sender theory since from the diagram, the cluster should be able to take the full 12v from the sender even if it's shorted to 12V. That alone should not fry the cluster.

Greetings Again: To answer recent comments/questions: * I am finished trying to diagnose this myself. The dealer is now the fifth Porsche specialist, (self imposed), that has spent their time and my money seeking to determine the cause of the cluster failures. * I am actually hopeful that the problem was caused by the wiring in the fuel level sending unit so that I can stop the burn rate of now close to $5,000 already spent trying to fix this situation, (although I am furious that frayed wiring with a live current in the gas tank is like driving around a bomb). * The dealer is only diagnosing the issue which causes the cluster to fail. They are forbidden by PNA to identify the specific transistor/component/circuitry in the cluster itself to seek a third party repair. PNA only allows them to sell/program new clusters at $2,500. * I have attempted to locate a service that can diagnose the transistor/component/circuitry and repair the cluster at a cost of $850, which is higher than simply purchasing a used cluster from a dismantler. * All other cluster repair services I have spoken with indicate that they are unable to acquire circuitry diagrams from VDO due to PNA restrictions and thus are unable to repair. * I would very much appreciate any recommended repair services that have the capability of repairing late model Porsche's would be expected. Thank you.

Ok, for repairs look for a good aircraft mechanic specialized in avionics. Your local airport might have a shop or ask around (maybe friend of a friend). I've seen these guys go through what seems miles and miles of wiring looking for a grounding issue on helos. Avionics shops charge good money, so if you can offer a guy some extra bucks on his time off, it would be the way to go. I would bet good money that a capable AME will find your problem. Good luck!

Greetings Again: Sorry for my delayed response, however my instrument cluster situation has been resolved. Diagnosis by dealer of faulty fuel level sending unit was determined to be the cause of the cluster self destructing. Apparently the insulation on the wiring can dissolve over time due to contact with fuel. So, with a full tank, fuel contacts exposed wiring, causing electrical spike at sending unit. No fuse exists between sending unit and instrument cluster, thus sending unit circuitry receives electrical spike, destroying cluster. Total cost of multiple hours of shop labor and three replacement instrument clusters ran just over $5,000. Finally, I sent the failed clusters to BBA Remanufacturing for repair. Unfortunately, none were repairable as company unable to diagnose specific circuitry malfunction and Porsche refused to divulge information. So, problem solved. Now if I can only find the lone key I had for the car, which is another long and sorry tale, not to mention an expensive, time consuming fix. Chris Stocker

a wire with coating thats in a gas tank and disolves when in contact with gas should be reported to the Feds.

Quotes source post: 8312282

+1 spike = spark

Glad to hear that it's been resolved, but sorry to hear about the cost. I hope you have better luck with the key.

I would suspect that part of the insulation dissolving problem is the corn based gasohol subsidy. I think the US and maybe Brazil are the only countries that use alcohol infused fuel. I think Brazil uses sugar as a base. I don't know if that would make a difference?

Also living in the SF Bay Area with oxygenated fuel....I'm considering changing the sender as a precaution....thoughts?

Man what a pain. As far as the exposed wire in the tank I gotta laugh because the metal connectors contacts to pump and sending units sitting in the gas are always soaking in the gas so an extra exposed portion of wire isn;t going to give rise to an outcry. I don't like the arrangement but its the way it goes after many fuel pump repairs. Just doesn;t feel right sticking exposed wires in the tank. I have a feeling if it was the sending unit it was a case of the wire contacting something other than the fuel or being an intermitent deal where the variations causes spikes trashing the cluster. Or if the fuel pump was replaced a contact wasn't good some place allowing voltage from the pump to enter the readout circuit frying your cluster. A bad ground for the cluster might do that. Where the bad ground was doesn't have to of been in the tank. Notice I didn't say you be getting mislead by the repair guys on what was the cause although......I hope it doesn't pop up again and that they never did any fuel pump or gauge repair before this problem.

Slightly OT, but I have one extra US instrument cluster, 2002-->, manual, used. $500,-

Quotes source post: 8312695

I agree the insulation should not be affected by anything in the gas but the fuel tank atmosphere contains an overly rich mixture of gas vapor to air and is not explosive. This is true of our cars -- they have the sender in the tank but also the fuel pump which runs most of the time (but not all of the time, for instance when the tank level is low) submerged in gasoline -- and this is the principle upon which truck/rail car and shipping tankers which carry thousands of gallons to unimaginable numbers of barrels of refined or crude oil rely upon. The shipping tankers do sometimes explode but it is during the storage tank cleaning operation which does permit outside air to enter/mingle with the residual oil in the tanks that can and does form an explosive mixture. The very act of the cleaning -- using jets of high pressure water -- generates static electricity (like lightning in storm/rain clouds) that sparks the explosion. (Newer tankers use filtered/cooled engine exhaust gases routed to the tank that of course results in a non-flammable mixture because there is not enough oxygen in the exhaust gases to support any combustion.) Sincerely, Macster.

Quotes source post: 8313660

My 02 Boxster has been in the bay area since 04 and has covered approx. 160K miles in that time (total miles is 239K currently and counting) and all on oxgenated (ethanol mainly) gasoline and its sender is just fine. In fact not too long ago (at over 220K miles) the original fuel pump quit and the tech replaced the pump and upon my request inspected the tank and in-tank hardware and found no reason to recommend anything other than the non-functioning fuel pump be replaced. Might add the old fuel pump's wiring, connectors, and fuel hoses looked just fine after all that time in the tank. (Just looked them over -- the pump's in a cardboard box under my office desk.) Based on my experience, Porsche gas tank level sensor/senders do not appear to be falling prey to gasoline attacking them in numbers to worry about. Sincerely, Macster.

This exact situation has happened with my 2003 996 4s Coupe. I went through the same process - instrument cluster, then another instrument cluster, then pin-testing each pin in the harness, then checking the cluster, then time and money. I wish I had read this blog earlier. When I took the car in after the second cluster failed, the expert (Mike Callas @ Rennsport in Sealy, TX) said the OBD showed a fault cose for the fuel sending unit. There is a guy in Houston @ Austin Radio who repairs the clusters. We sent the cluster for repair and commenced pin-testing and attempting to diagnose where the short was. Meanwhile we have ordered a new fuel sending unit. I will tell the forum whether that solved the problem, but I am cautiously optimistic. I just thought you should know that other people have had the exact same problem. Thanks!

Don't mean to revive a dead thread, but... I have to say thank you. On my way home tonight, the dash lights on my 2003 C2 cabriolet instrument cluster went out. I tried turning the lights off and on to remedy it. The dash lights came on briefly after one of the tries, but immediately went back out. For some reason, when I parked the car they came back on again, but I couldn't help but wonder what kind of costly problem I had on my hands. I immediately searched rennlist for the problem, and reading this thread I got very concerned that it would be very hard and expensive to figure out. I had basically relegated myself to driving at night without a dash. Then I read the solution, and thought, holy crap, I literally just filled up my gas tank... Coincidence? I'm not sure, but it makes sense to me to replace the fuel level sending unit and see where that gets me. I'm thinking before my cluster shorts for good (it appears to still be working at the moment). So thank you for helping me avoid a potential nightmare. Anyone know the part number for the sending unit? How hard it is to replace? I'd like to do it myself if possible.