Topic: Baffling 944 gauge pod problem

Small problem: Something is putting resistance back into the grounds in my 944's gauge pod. It now lights up at night (good), but it pegs the coolant temperature gauge whenever the car is on (bad!).

I can't figure the stupid thing out, and given this car's previous adventures in overheating, that's not good. The ground measures about 20 ohms of resistance when the car's on, and 50 if you turn on the lights, so something's screwed up somewhere. Any ideas?

The car is based in Austin, and we're heading down to MSR-Houston for the full-24 on Thursday night. Contact info is stef [at] puffalump [dot] de - help!

Re: Baffling 944 gauge pod problem

When you checked the resistance of the ground(s) - where exactly are you measuring? Did you make the measurements with everything still connected? If you simply measure between the ground of one of the  instruments (say, coolant temp) and the vehicle structure, then you are not really measuring just the resistance of that wire, because there are many other parallel paths that exist when everything is connected.

If you think there is a problem with the gauge grounds then you should remove each ground wire (one at a time) and then measure the resistance from the end of the ground wire (where it normally connects to the instruments) to structure. The gauge grounds are likely small gauge wires - on our Audi they are 0.5 square mm (20 AWG) which should have a resistance of about 10 ohms per 1,000 feet. Since your ground wires are likely less than 10 feet long, you should measure something like 100 milli-ohms (0.100 ohms) maximum. If you see higher resistance, then maybe you have a damaged wire or the wire's connection to structure is corroded/damaged.

Your coolant temp sender (or the wire connecting it to the gauge) could also be bad. Your sender is likely a thermistor, so the resistance of the sender goes down as temp goes up. If the sender or the signal wire is shorted to structure, then the gauge sees zero ohms and will be pegged high. Try removing the signal wire from the sender, that will look like infinite resistance and the gauge should drop to its lowest value.

We Audi Be Faster
'85 Audi Coupe G(in &) T(onic)

Re: Baffling 944 gauge pod problem

how accurate is the OEM coolant temp sensor anyway?

Easiest solution might be to just go to the nearest auto parts store, grab an aftermarket mechanical or electrical temp gauge, and use that.

Don't know enough about the stock gauges to tell you what to chase.

20 Time Loser FutilityMotorsport
Turbo Dodge Powered E36 Build
2008 Saab 9-5Aero Wagon
Retired - 1989 Dodge Daytona Shelby 2011-2015 "Lifetime Award for Lack of Achievement" IOE, 3X I got screwed, Organizer's Choice

Re: Baffling 944 gauge pod problem

TheEngineer wrote:

how accurate is the OEM coolant temp sensor anyway?

+1 on this. The gauge in the cluster likely gives you 3 indications: cold, "normal", and oh shit (too late).

After losing 2 motors to undetected overheating problems, we added another coolant gauge and installed the sender in the coolant hose between the engine and the radiator. Required making a custom sender mount out of some steel pipe, but not a big deal. Teams in the NE are likely to hear us on the radio asking the driver to check coolant temp almost every lap!

We Audi Be Faster
'85 Audi Coupe G(in &) T(onic)

Re: Baffling 944 gauge pod problem

ninjacoco wrote:

Small problem: Something is putting resistance back into the grounds in my 944's gauge pod. It now lights up at night (good), but it pegs the coolant temperature gauge whenever the car is on (bad!).

I can't figure the stupid thing out, and given this car's previous adventures in overheating, that's not good. The ground measures about 20 ohms of resistance when the car's on, and 50 if you turn on the lights, so something's screwed up somewhere. Any ideas?

The car is based in Austin, and we're heading down to MSR-Houston for the full-24 on Thursday night. Contact info is stef [at] puffalump [dot] de - help!

it's not running the gauge through the dimmer circuit is it?

Mistake By The Lake Racing (MBTL)
88 Thunderbird "THUNDERBIRDS ARE GO!"
A&D: 2014 Sebrings at Sebring (NSF), 2014 NJMP2 Jurassic Park (SpeedyCop), 2012 Summit Point J30 (PiNuts)

Re: Baffling 944 gauge pod problem

Just start replacing every wire in the car until the problem goes away.

Terribly Awesome Racing
Dozen time loser. You'd think I'd know better by now.

Re: Baffling 944 gauge pod problem

I hate to admit it, but smarter folks than I were the ones who measured it. I believe one end of the little meter (forgot the name) was attached to the body on a metal nub and the other though that wire, as you noted.

I don't think it's running through the dimmer circuit AFAIK. The interior light dimmer and the light switch are their own little connected pod that's below and to the side of the main gauge cluster. The only thing we did to that connection was move it below the rollbar to relocate our dumb wiper switch.

We disconnected and reconnected a bunch of wires that we thought might be sender-related, and no dice. Gauge still pegs all the way. I don't think it's the temperature sensor itself—for one, it's tucked way under the intake manifold and a lot of other stuff, so we haven't fiddled with it at all. Two, the darn thing works. The fan comes on when it needs to and turns off when it doesn't. The car hasn't barfed its water out lately, either, so I'm pretty sure the cooling system itself is working fine. It's just a gauge issue.

The gauges are something we messed with to get lights behind them for night use, so whatever gremlin we unsettled has to be up there somewhere. Leading culprit is a ground issue. I'm gonna go around cleaning some grounds tonight, ugh.

tl;dr—watercooling sucks and I'm done with it. Please bring me my 411, which has an oil temp gauge that always reads hot (because the car is running too hot).

Re: Baffling 944 gauge pod problem

The problem is that even if the fans are working as you expect, the gauge may still lie to you. OEM gauges are notorious for not moving past "normal" until you're already in trouble. You can't tell what the fans are doing while you're driving.

A water temp gauge, or even better an idiot light set to come on at some temp like 210-215, is a very smart idea if you're worried about melting down a motor.

20 Time Loser FutilityMotorsport
Turbo Dodge Powered E36 Build
2008 Saab 9-5Aero Wagon
Retired - 1989 Dodge Daytona Shelby 2011-2015 "Lifetime Award for Lack of Achievement" IOE, 3X I got screwed, Organizer's Choice

Re: Baffling 944 gauge pod problem

Agree with TheEngineer on that one. The dash gauge is likely on its own sender and shouldn't be used for primary coolant temp status. There is probably another sender that tells the ECU what the engine temp is, and a third one (temp switch) that lives in the radiator that is used to turn the rad fan on/off.

We installed an aftermarket water temp sensor that feeds a gauge installed in front of the driver. I look at that gauge A LOT. Its got actual numbers so when I get asked the temperature I can give the guys in the pits a value, not just "a little above the green".

Our car also came with a coolant over-temp switch that is installed in the head that was used to shut down the A/C system when the car got too hot. We re-purposed that for a BIG idiot light to indicate when overheating was imminent. Already saved the engine - we came in hot off the track and the wires to the fan switch had come off so the rad fan did not turn on while driving slowly in the paddock. The driver saw the idiot light and turned it off before disaster struck.

We also tied the idiot light to a low coolant-level switch that sits in the bottom of the overflow tank. This gives us some advance warning that we are loosing coolant.

We Audi Be Faster
'85 Audi Coupe G(in &) T(onic)

Re: Baffling 944 gauge pod problem

RSB wrote:

and a third one (temp switch) that lives in the radiator that is used to turn the rad fan on/off.

Although I have seen these and owned a car that had one, it is the least common for fuel injected cars in my experience. 2-3 different senders somewhere in the engine is totally common but the radiator sender seems to have been kind of and interim thing as cars moved from the first to second tier of federal emissions regulations (say 1978-1995).  Seeing one in the thermostat housing is probably more common post 1995.

No idea how Porsche did their gauges but if it is like Volvo, BMW, some VW of an era, you may have that sender and gauge tied to 10V regulator that also drives the fuel gauge.  These things getting flakey or shorted can create all types of fun.  They did it for a valid reason it is just a troubleshooting nightmare if you are not aware of it. regulator can reside in the ECU or on the IP board.

11

Re: Baffling 944 gauge pod problem

OnkelUdo wrote:

No idea how Porsche did their gauges but if it is like Volvo, BMW, some VW of an era, you may have that sender and gauge tied to 10V regulator that also drives the fuel gauge.

Good point - is the temp gauge the only one acting up? (Fuel gauge OK) If so, its likely not the voltage regulator.

The 944 is a product of the 80's so I was assuming that it has a thermofan switch in the rad (like our 85 Audi does).

This website seems to be a treasure trove of info on the 944:

http://www.clarks-garage.com/shop-manual/elect-19.htm

Seems like you need to revisit all the 'improvements' you made to make the lights work.

We Audi Be Faster
'85 Audi Coupe G(in &) T(onic)

Re: Baffling 944 gauge pod problem

Clark's Garage kicks ass. But yeah, we've got a thermoswitch in the radiator, and another sender in the block, under a bunch of stuff.

I opted not to mess with disassembling the top of the engine to get to it. Car survived! But uh, yeah. That one gauge never worked, and I ate through two alternators. I think I was asking too much of the alternator with the lights, and also I think this is probably a ground issue. The driver's side back taillight also goes out if the wire jiggles in just the wrong way. There's a lot of stupid wiring things in this car that GRRRRRRRRRR.