[estrop]

Officially, they say it was the hottest day Britain ever had and my 2.3T Griffin had all on to keep cool - 36-37C was reading on the outside temp

All motorway dring today and whilst I did not overheat, using the turbo or going up hills, my temp guage would climb to about 2mm below the red zone but never over.

I don't think I have cooling probs as my new thermostat is an 82 degree and was fitted about 15 months ago. Fan also cuttinng in finer.

anyone experienced the same unusually high running temps with a full turbo this weekend???

 

[Skiddins]

My '97 Aero still hasn't gone above the 9 o'clock point despite me pushing every now and again.

Skiddins

 

[1989agoodvintage]

I didn't drive mine today, but yesterday the gauge was hitting 3/4 on a regular basis. On fast cruise 50mph+ it was around 1/2, but would creep above that once the speed dropped.

 

[donut]

rich. I have a 96 Aero and I had a new thermostat fitted at a Saab Dealer in Turners hill. the readid on a trip from Hastings on the hotest day only produced a reading at 8 o'clock on the guage.

 

[wrighar]

Dont the new 9000's have a pseudo temp guage?

Myt 93 Aero was up to 10 o'clock yesterday at 40 mph, but dropped when I went faster.

The carly was up at 11 o'clock.

I thought that from mid 90's the gauge was pseudo (controlled by TRionic (edit not TCS as I said Doh!) just like the fan is) and wouldn't show temp changes until they become more extreme.

Does you Rad have a fan switch or not?, mine do.

Andrew

 

[billj]

Yes, the '96-on 9000 had the placebo gauge that gives an indication of temerature "severity" rather than temperature directly. I didn't drive the Aero today, but no doubt it wouldn't have gone past the 8 o'clock position (just like on the rolling road at Wilshers Garage). I did drive the CSE (2.0LPT) which is a '95 and has the proper gauge. The gauge on that didn't pass 8 o'clock in town or cross-country driving. I know it can go higher because I saw it at 9 o'clock the day I drove it round Goodwood.

 

[Skiddins]

Dont the new 9000's have a pseudo temp guage[/b]

'96-on 9000 had the placebo gauge that gives an indication of temerature "severity" rather than temperature directly[/b]

Interesting, because in damp weather my dodgy, temp gauge connection shows up.
If i fiddle with the wire it can vary the temp wildly and it reads all over the place.

Surely if it does give a fake reading, it shouldn't effect the gauge so much, and these changes happen almost instantly.

Skiddins

 

[zeke]

Boy, the 5 hour time difference makes it difficult to get in on some of these discussions.....Bill, the '96 up temeperature gauge - I wonder what the rationale is for the use of a logarithmic scale? For the past 45k miles, I have never seen mine go above 9 o'clock. Have you ever seen yours go that much above 9? Has anyone ever seen a '96 or newer get near the red? The thing is, it is a real temperature gauge - well, at least from cold start-up until 'normal' operating temperature.
Andrew:

I thought that from mid 90's the gauge was pseudo (controlled by TRionic (edit not TCS as I said Doh!)[/b]

The coolant temperature sensor goes near, but doesn't enter the Trionic. It's wired directly to the EDU on the instrument cluster. The EDU does all the calculations and feeds the temperature gauge.

For what it's worth, the temperatures here in New England have been averaging about 26 Cº daily, for the last 10 days.
Very humid (75-80%), lots of gray sky, and frequent showers - the mold, mildew, etc is driving me crazy. My temperature gauge remains @ 9, however!

Skiddins, are you moving the wires at the sensor under the manifold or are you moving the wire at the temperature gauge? I have questions both ways! If you're moving the yellow wire at the gauge, then you must be driving around with the dash top removed. If you're noticing instantaneous movement of the gauge while moving the wires at the coolant temperature sensor, then you move really quick!

I just sent a sample of my last oil change (6200 miles on M1 15W-50) to be analyzed. I'm concerned because I've noticed a very small decrease in coolant. Since the last oil change, the coolant level dropped from the "full" mark down to the "minimum mark". Now, it may be that the level dropped before I pseudo-fixed (Bars' leak) my water pump leak. It had a very slight drip from the bottom of the gasket on the cover plate. I'm hoping that explains my coolant drop, but I'm keeping my fingers crossed until I get the oil analysis - should have it tomorrow, I hope. If anyone is interested, I'll post the analysis when it arrives.

 

[Mark B]

Please post the oil analysis. Thanks.

 

[Skiddins]

Skiddins, are you moving the wires at the sensor under the manifold or are you moving the wire at the temperature gauge? I have questions both ways! If you're moving the yellow wire at the gauge, then you must be driving around with the dash top removed. If you're noticing instantaneous movement of the gauge while moving the wires at the coolant temperature sensor, then you move really quick![/b]

I was fiddling around under the manifold, while someone else was watching the gauge.

Are there definitely circuits behind the gauge, not a direct connection?
If this is the case, what sort of 'intelligence' are these circuits providing for the reading?

Skiddins

 

[mark e]

Originally posted by robert lavergne:
[qb]The coolant temperature sensor goes near, but doesn't enter the Trionic.  It's wired directly to the EDU on the instrument cluster.  The EDU does all the calculations and feeds the temperature gauge. [/qb][/b]

And it's a placebo... BillJ and co have had their cars on the track and the temp gauge never moves above 9 O clock... it was done like this in response to customer concerns that the gauge seemed to fluctuate a lot. Basically, unless it's really getting too hot, it won't move from a position round about 8 O clock.

On really hot days previously, my 93 has managed to get within a few mm of red going up hill, foot down and aircon on.

 

[zeke]

The gauge is run by the EDU. There is a yellow wire that runs from the EDU (pin 30) to the temperature gauge (pin 7). The temperature coolant signal goes through the main wiring harnes, thru a 70-pin connector (H70-1) and into the car, and EDU. The coolant temp' sensor is an NTC (negative temperature correlation? BillJ, I need help here!) resistor linked to the EDU, then the temperature gauge. The EDU calculates the temperature from the resistance, which IIUC, goes down as the temperature goes up. I probably have it backwards, but someone'll know for sure. I've never taken one apart (again, I bet BillJ has ) but I'd wager the hardware is a pretty common embedded controller, of some sort.
Skiddins, have you ever taken the instrument cluster out and looked at the back? There are two round parts, that are on the ends, and a kind-of rectangular module between them. The ends are the gauges and lighting diagrams, the middle rectangular module is the EDU. The EDU is the kind of 'black box' of the vehicle display functions. They're all connected with each other and with other things (ACC, relays, etc.) As to why your gauge is acting funny, there's only three connectors to check - pins 57 (yellow) & 58 (orange) on the main wiring harness, H70-1, the coolant temperature sensor itself, and pins 34 & 27, yellow and orange, respectively, as they go into the EDU. The EDU sends the 'message' to the gauge out pin 30, a yellow wire, to pin 7 on the gauge. That's exactly the path from the coolant temperature sensor to the gauge....I think it was a brilliant refinement of what preceded it -that switch should have never been so close to the turbo. Still, I haven't heard anyone with a '96 or newer reply as to whether they've seen their gauge go much above 9.

 

[mark e]

Yes, Robert, it'll be an NTC- just as the air intake temp sensor is. Most temp sensors are NTC 'cos they're easier and cheaper to manufacture. Resistance drops as temp rises, normally pretty linearly within a set range for each sensor.

The part # for the gauge changed in MY94, and then again in MY96, although they both were superceded by another unit of a common part #, suggesting that MY94 onwards cars will have this sort of gauge.

 

[billj]

Originally posted by robert lavergne:
[qb]The coolant temp' sensor is an NTC (negative temperature correlation? BillJ, I need help here!)[/qb][/b]

"Negative temperature coefficient" is the term you're looking for, but you've described its operation correctly.

I've never taken an instrument cluster apart - I'd be too worried about not getting it back together properly and it looking a bu66er.

I've never seen my Aero's gauge as high as the 9 o'clock position. Not even during a session on a static rolling road where it was held for around 5 seconds at full throttle at every speed from 2500-5000 rpm at 500 rpm intervals (5 seconds of putting out 300hp at 5000rpm

). The other 9000 there that day (Gassy's '91) whacked its gauge right up near the red early in the session and held it there until the end, with the operator keeping a close eye on it and letting it idle for a while between runs.

The gauge will go up to around the 8:30 position under track conditions. I've only ever actually noticed it move up from the 8 o'clock position a couple of times, and it moves in little "jumps", like an electronic clock. I assume this is down to the resolution of the EDU that drives it.

 

[billj]

Originally posted by Mark E:
[qb]suggesting that MY94 onwards cars will have this sort of gauge. [/qb][/b]

Perhaps this type of gauge can be fitted to a '94, but Cissie's gauge (MY95) does the up-down thing just like your '93.

 

[mark e]

Hmm... I'd forgotten that...

 

[arawson]

Just for the record, my '95 Aero seems to have one of these placebo gauges because it never ventures about the 8 o'clock position even when driving in +30C temps. I even changed the stat back in April as I thought the car was running a little cold. Not sure which temp version I fitted, just used the one that Bond St supplied.

 

[mutungi]

Phew....with the AC running at full on Sunday, my needle was just below the RED!

 

[markcanderson]

I've had my 2001 9-3's temp gauge rise.. at the ring last week..

At the end of the 20KM course, just before booting it out of the last curve to the long straight at start-finish.. my gauge hit the 10-o'clock position.. ambient temp was 31c.. that's the first time I've seen it do that. Usually it never budges from the 9-o'clock.. Marky's TiD did the self same thing.. and Maarten's NG900 did it too in my hands.

My 1992 9000's gauge is usually sitting between the 9 and 10 o-clock positions.. but when pushing on in high ambient temps.. i.e. at the ring in May during 27c temps.. I saw millimeters from the red.... halfway into the course.. backing off brought the temps right down... i've since removed the aircon radiator from the front.. ducted air from the front fog light covers into the engine bay.. and drilled holes in the wheel arch liner for hot air to escape.. this seems to have lowered highspeed/highload operating temps.. we'll see in a few weeks at the ring.

 

[aeropilot]

Well in these recent high temps (including out all day on the hottest day on Sunday) my MY96 Aero has been running at dead on the 9 o'clock position for most of the time, but never higher, even in static London traffic with the outside temp showing 45C as a result of the heat soak.....