SaabCentral Forums banner

1 - 20 of 90 Posts

·
Registered
Joined
·
2,791 Posts
Discussion Starter #1
Who else has this issue? I have had it on a succession of 9Ks. What is your theory? My fixes seems to have been temporary so far. I'd like to get a fresh opinion of malady without showing my cards yet.
elaboration is car is running and normal, then fuel and coolant gauge drop. Then pop back up.
 

·
Registered
Joined
·
2,942 Posts
Who else has this issue? I have had it on a succession of 9Ks. What is your theory? My fixes seems to have been temporary so far. I'd like to get a fresh opinion of malady without showing my cards yet.
elaboration is car is running and normal, then fuel and coolant gauge drop. Then pop back up.
Scantar's been fighting this for a bit so I'm sure he'll chime in --

Mine has started to do the same thing. Probably the connection to the EDU module mounted on the gauge cluster, or some crusty solder joints lying within the module itself.

The real danger with this is that the module also sends a temp signal to the ECU. With no temp signal input the default is the ECU thinks the engine is cool and won't turn on the fans -- this could cause you to over heat the engine before you know it in city driving with the a/c on...
 

·
Registered
Joined
·
2,791 Posts
Discussion Starter #3
Scantar's been fighting this for a bit so I'm sure he'll chime in --

Mine has started to do the same thing. Probably the connection to the EDU module mounted on the gauge cluster, or some crusty solder joints lying within the module itself.

The real danger with this is that the module also sends a temp signal to the ECU. With no temp signal input the default is the ECU thinks the engine is cool and won't turn on the fans -- this could cause you to over heat the engine before you know it in city driving with the a/c on...
I talk with him outside SC, ohh I know. He thinks it's the ribbon and I think it's the brain. Ive been swapping EDUs and was good for awhile but problem came back. I've had 3 different 9Ks do this same thing. EDU swap fixed it on one '98, I've went through 2 different EDU on another '98 and once Ive caught it on my '94 which I have taken no action on.
On '95 and above it will turn off the stuff you say plus turn off the AC! '94 temp gauge don't talk to the ECU so might effect differently.
 

·
Registered
Joined
·
2,791 Posts
Discussion Starter #5
I swapped my temp/fule gauge and that seams to fixed it for me.
No longtime test yet. I suspect the soldering and nuts behind the gauges.
I tried swapping cluster only, had no effect. What year is your car, not listed in your avatar.
 

·
Registered
Joined
·
4,197 Posts
I talk with him outside SC, ohh I know. He thinks it's the ribbon and I think it's the brain. Ive been swapping EDUs and was good for awhile but problem came back. I've had 3 different 9Ks do this same thing. EDU swap fixed it on one '98, I've went through 2 different EDU on another '98 and once Ive caught it on my '94 which I have taken no action on.
On '95 and above it will turn off the stuff you say plus turn off the AC! '94 temp gauge don't talk to the ECU so might effect differently.

No, I don't think it's the ribbon. I know it's the brain.



As soon as I can clean this sharty heat sink material off the board I'm going to put it under a microscope and reflow the solder
 

·
Registered
Joined
·
4,197 Posts
Here's the bullcarp part of all of this. If we can't come up with a way to either repair or just bypass the edm - or at least the part that sends temp signal to the ecu - the 9K is destined for extinction. This is an issue that is happening with more frequency if you look through the threads over say the last 12 months. That means the edm is aging out...not a good thing!


The edm is essential to the operation of the cooling system. When the right side cluster drops out, the edm tells the ecu that the car is cold and the ecu shuts down the fan. Not terrible if it's winter and you're highway driving. Head Gasket popping if it's summer and you're city driving.


My observation is it does seem to be heat related, as in behind the dash heat related. That tells me that either a solder joint is bad (there are literally hundreds) or it's internal to one of the eeproms or transistors (there are dozens)


Ideally, it would be cool (no pun intended) to be able to bypass the system altogether and send engine temp signal to the ecu a different way, regardless if the gauge drops out...but that takes thinking time, like a weekend camping in the woods thinking time.
 

·
Registered
Joined
·
2,942 Posts
Here's the bullcarp part of all of this. If we can't come up with a way to either repair or just bypass the edm - or at least the part that sends temp signal to the ecu - the 9K is destined for extinction. This is an issue that is happening with more frequency if you look through the threads over say the last 12 months. That means the edm is aging out...not a good thing!


The edm is essential to the operation of the cooling system. When the right side cluster drops out, the edm tells the ecu that the car is cold and the ecu shuts down the fan. Not terrible if it's winter and you're highway driving. Head Gasket popping if it's summer and you're city driving.


My observation is it does seem to be heat related, as in behind the dash heat related. That tells me that either a solder joint is bad (there are literally hundreds) or it's internal to one of the eeproms or transistors (there are dozens)


Ideally, it would be cool (no pun intended) to be able to bypass the system altogether and send engine temp signal to the ecu a different way, regardless if the gauge drops out...but that takes thinking time, like a weekend camping in the woods thinking time.
I've got a pdf of the ECU pin-out somewhere...
The T-5 suite project shows how to add T 5.5 to a C900 so the pinouts exist and a bypass could easily be achieved...

I have to imagine it'd be more accurate at this point too -- having the eng temp reading come into the dash and going back out is using a lot of wiring and bulk head & dash connectors -- any of which can be affecting the resistance / ohms which is translated to eng temp for the ecu to determine best fuel mixture...
 

·
Registered
Joined
·
2,942 Posts
Related symptoms?
Just completed another CO to Ohio drive (1300 miles) and intermittently had the following glitch:

The temp and fuel gauge would start reading high -- this would be accompanied by a drop in the voltage gauge readout as well as an actual dimming of the edu (digital numbers of the voltage readout). Very hard to tell during the day but at night it was easy to correlate that the lower the voltage read (went as low as 10.2v) the higher the temp & fuel gauges read.
The headlights did NOT dim when this occurred -- I don't believe there is any issue with the actual charging system but can't confirm...

The car didn't drive any differently this trip although I did notice a slight loss of mpg overall average this trip.

Regrettably I'll have to drive it back to CO in a couple weeks with this problem probably unsolved...
Without a really good wiring diagram I don't know how to even begin with this one...

Once back in CO I'll probably replace the dash cluster since I have a "low mileage" one in stock and hope that solves the gauge reading and occasional drop-out symptoms...
 

·
Registered
Joined
·
2,791 Posts
Discussion Starter #12
Related symptoms?
Just completed another CO to Ohio drive (1300 miles) and intermittently had the following glitch:

The temp and fuel gauge would start reading high -- this would be accompanied by a drop in the voltage gauge readout as well as an actual dimming of the edu (digital numbers of the voltage readout). Very hard to tell during the day but at night it was easy to correlate that the lower the voltage read (went as low as 10.2v) the higher the temp & fuel gauges read.
The headlights did NOT dim when this occurred -- I don't believe there is any issue with the actual charging system but can't confirm...

The car didn't drive any differently this trip although I did notice a slight loss of mpg overall average this trip.

Regrettably I'll have to drive it back to CO in a couple weeks with this problem probably unsolved...
Without a really good wiring diagram I don't know how to even begin with this one...

Once back in CO I'll probably replace the dash cluster since I have a "low mileage" one in stock and hope that solves the gauge reading and occasional drop-out symptoms...
In regards to high readouts. My 1994 reads like its about 2/3 temp up. I will not bore you with the story of how I tried to diagnose. The national guard needed to be phoned to find me though. In the end I just trust the IR meter is right and not the gauge. Wonky temp gauge.
I hoard clusters now.
I once went and looked at a near death CL 1998 where dude had a rubber hammer he used to beat the dash board in "just the right spot" to get the T an G gauge to work again. Believe ive told the story. Wanted 1,500 dollars for it. HAHAHHAA. I remember what an idiot he was well. He called me a few minutes after when I left and said "then what's your best offer?" ..
"I was afraid you'd take $400 so I didn't say anything"
Click.
 

·
Registered
Joined
·
2,942 Posts
In regards to high readouts. My 1994 reads like its about 2/3 temp up. I will not bore you with the story of how I tried to diagnose. The national guard needed to be phoned to find me though. In the end I just trust the IR meter is right and not the gauge. Wonky temp gauge.
I hoard clusters now.
I once went and looked at a near death CL 1998 where dude had a rubber hammer he used to beat the dash board in "just the right spot" to get the T an G gauge to work again. Believe ive told the story. Wanted 1,500 dollars for it. HAHAHHAA. I remember what an idiot he was well. He called me a few minutes after when I left and said "then what's your best offer?" ..
"I was afraid you'd take $400 so I didn't say anything"
Click.
:lol:

I have no fear of the engine overheating in the winter on the hwy...
My temp gauge normally reads 1/3 of the way up the scale and the gas gauge is normally consistent/accurate...
I should add that when the voltage readout dimmed and the gas & temp needles shoot up; no other lights on the dash dimmed.
So I'm assuming the problem is restricted to the cluster wiring/edu wigit on the back of the cluster/or wiring between the two...

As to your hording clusters -- does that mean you think that's where my problem lies -- the cluster/edu and not in the wiring in, and around them?

I hate that there's no way of testing a cluster/edu other than to install it on a car and see if it works... This is only (sorta) guaranteed if the car was working correctly before swapping parts...
 

·
Registered
Joined
·
4,197 Posts
As to your hording clusters -- does that mean you think that's where my problem lies -- the cluster/edu and not in the wiring in, and around them?

I hate that there's no way of testing a cluster/edu other than to install it on a car and see if it works... This is only (sorta) guaranteed if the car was working correctly before swapping parts...

My .02 is it's the edm itself not the wiring around in and out of it. And yeah, sucks that the only up or down test for it is to swap the part. When you do though, do yourself a favor and drive it for a week without buttoning down the dash top until you know it's ok.


One thought I did have was finding the pin in/out on the ecu that gets the info and splicing in an analog temp gauge. Might not help with the dash signal but it may prevent the ecu from shutting off the cooling fan
 

·
Registered
Joined
·
2,012 Posts
My .02 is it's the edm itself not the wiring around in and out of it. And yeah, sucks that the only up or down test for it is to swap the part. When you do though, do yourself a favor and drive it for a week without buttoning down the dash top until you know it's ok.


One thought I did have was finding the pin in/out on the ecu that gets the info and splicing in an analog temp gauge. Might not help with the dash signal but it may prevent the ecu from shutting off the cooling fan
I'm fairly certain that the coolant temperature indicator has no effect on the operation of the radiator fan. It is a shunt circuit that only allows the operator to monitor the conditions. You could remove it completely from the cluster and the EDU would continue to control the fan relays normally.


Anyway, I started writing this a few weeks ago. But for some reason, it was forgotten.

It seems all interested parties are experiencing the same issues – which are confined to a very specific area of the instrument cluster. Specifically, the common problems are:

1. Intermittent (but very temporary) erroneous readings provided by both the coolant temperature and fuel tank level indicators.

2. Bill B noted that his Low Fuel Level warning lamp was illuminated as well.

I assume he is referring to the one in the RH section of the cluster – an oval under the gas pump icon (indicates that there is less than 10 liters of gas remaining).

Not the one in the center section - a little arrow/dot combo next to the voltmeter - that when flashing, indicates the driving range has fallen below 30 miles. That range is calculated by the trip computer - based on current fuel consumption and quantity of fuel remaining.

3. No one reported a loss of either the instrument backlighting lamps (2) or the RH turn signal indicator lamp. But that may well be due to the fact it was daytime and the directional was not in use.

So we are looking at this printed circuit:





Fuel Gauge 47a



When the ignition switch is in the Drive position, 12 V is supplied to pin 2 of the temperature gauge from pin 22 of EDU 210. Fuel level sensor 46 is connected to the EDU unit which performs a calculation and sends a 1-9 V signal to pin 1 of the tank/temperature unit.

1 V = tank empty
9 V = tank full


Indicator Lamp 47b
Low fuel warning lamp 47b is controlled by EDU 210. When the fuel level in the tank drops below 10 litres , the lamp is grounded via pin 3.

Coolant Temperature Gauge 47c



The temperature gauge displays the temperature of the engine coolant.

When the ignition switch is in the Drive position, 12 V is applied to pin 2 of the temperature gauge from pin 22 of EDU 210. Coolant temperature sensor 45 is connected to the EDU unit which performs a calculation and sends a 1-9 V signal to pin 7 of the tank/ temperature unit.

1 V = min. temp
9 V = max. temp


Instrument Lighting 18
Backlighting of the instruments is regulated by rheostat 16 (lighting for instruments and controls) via pin 6.


Indicator Lamp 47i

The indicator lamp for the right-hand direction indicators is controlled by flasher relay 23 via pin 4. The lamp can also be activated directly from the anti-theft alarm control module.

The temperature gauge 47c, fuel gauge 47a, instrument lighting lamps 18, and indicator lamp 47i are connected to ground via pin 5. As noted above, the low fuel level lamp is grounded independently of the other 4 components.


How all components are wired to the EDU - and the associated pins.



Component key:

45 & 46 are the engine coolant and fuel tank level sensors respectively.

47 is the area of the cluster we are talking about but Saab refers to the entire cluster as component 47.

155 is the low fan speed relay

81 is the high fan speed relay



Note: All of the above is just provided as information to aid in diagnosing the problems. The symptoms, causes and remedies will likely vary from individual to individual.
 

·
Registered
Joined
·
4,197 Posts
I'm fairly certain that the coolant temperature indicator has no effect on the operation of the radiator fan.

Oh, but it does chengny. When signal is lost from the temp gauge the ecu shuts the fan off completely...not assumption, this is something I've observed.


It also will affect the acc in that if you have the a/c running it will shut that off as well
 

·
Registered
Joined
·
2,942 Posts
I'm fairly certain that the coolant temperature indicator has no effect on the operation of the radiator fan. It is a shunt circuit that only allows the operator to monitor the conditions. You could remove it completely from the cluster and the EDU would continue to control the fan relays normally.


Anyway, I started writing this a few weeks ago. But for some reason, it was forgotten.

It seems all interested parties are experiencing the same issues – which are confined to a very specific area of the instrument cluster. Specifically, the common problems are:

1. Intermittent (but very temporary) erroneous readings provided by both the coolant temperature and fuel tank level indicators.

2. Bill B noted that his Low Fuel Level warning lamp was illuminated as well.

I assume he is referring to the one in the RH section of the cluster – an oval under the gas pump icon (indicates that there is less than 10 liters of gas remaining).

Not the one in the center section - a little arrow/dot combo next to the voltmeter - that when flashing, indicates the driving range has fallen below 30 miles. That range is calculated by the trip computer - based on current fuel consumption and quantity of fuel remaining.

3. No one reported a loss of either the instrument backlighting lamps (2) or the RH turn signal indicator lamp. But that may well be due to the fact it was daytime and the directional was not in use.

So we are looking at this printed circuit:




Fuel Gauge 47a



When the ignition switch is in the Drive position, 12 V is supplied to pin 2 of the temperature gauge from pin 22 of EDU 210. Fuel level sensor 46 is connected to the EDU unit which performs a calculation and sends a 1-9 V signal to pin 1 of the tank/temperature unit.

1 V = tank empty
9 V = tank full


Indicator Lamp 47b
Low fuel warning lamp 47b is controlled by EDU 210. When the fuel level in the tank drops below 10 litres , the lamp is grounded via pin 3.

Coolant Temperature Gauge 47c



The temperature gauge displays the temperature of the engine coolant.

When the ignition switch is in the Drive position, 12 V is applied to pin 2 of the temperature gauge from pin 22 of EDU 210. Coolant temperature sensor 45 is connected to the EDU unit which performs a calculation and sends a 1-9 V signal to pin 7 of the tank/ temperature unit.

1 V = min. temp
9 V = max. temp


Instrument Lighting 18
Backlighting of the instruments is regulated by rheostat 16 (lighting for instruments and controls) via pin 6.


Indicator Lamp 47i

The indicator lamp for the right-hand direction indicators is controlled by flasher relay 23 via pin 4. The lamp can also be activated directly from the anti-theft alarm control module.

The temperature gauge 47c, fuel gauge 47a, instrument lighting lamps 18, and indicator lamp 47i are connected to ground via pin 5. As noted above, the low fuel level lamp is grounded independently of the other 4 components.


How all components are wired to the EDU - and the associated pins.



Component key:

45 & 46 are the engine coolant and fuel tank level sensors respectively.

47 is the area of the cluster we are talking about but Saab refers to the entire cluster as component 47.

155 is the low fan speed relay

81 is the high fan speed relay



Note: All of the above is just provided as information to aid in diagnosing the problems. The symptoms, causes and remedies will likely vary from individual to individual.
Beautiful, and thank you Sir!
Now I (think I) get it!

1) When the temp/fuel gauges completely drop out, ie: they are receiving NO voltage input.
what we failed to mention also was
that when this intermittent failure occurs it is then followed by the "check radiator fluid" on the edu screen. I surmise then that what has happened is a loss of INPUT power to the edu 210 (when the gauges reactivate, the check rad fluid & "low fuel light" is part of the "start-up sequence" for the edu receiving power. IF the edu is not receiving 12v, THEN it can't send a signal to the fan relays. That's a problem...

Okay -- this is progress (unless I'm misunderstanding the wiring diagrams) in that now we know it's an intermittent power input to the EDU 12v
Although Scantar and I both drive CSE models we both prefer the analog clocks of the CD so would be unable to see if other edu outputs also all failed at the same time.
My intermittent temp/fuel power loss failure started after I'd done other work behind the dash cluster so I assumed it was my ribbon connection -- that was several years ago. And then it stopped happening so I figured I had a good connection again... but this past summer it started up again.

It would now seem to me that my OTHER intermittent temp/gas gauge failure where both gauges read high (accompanied by my voltage GAUGE reading low -- actual alternator output is fine) is an entirely separate failure issue -- although obviously still linked to the edu 210...

Chengny -- what I couldn't determine from your (alldata?) diagrams was from where the edu receives its 12v input? I'm guessing from the ECU but that's a dangerous assumption when dealing with SAAB...

So I guess while the gauge cluster/edu is out (or in my case -- before I bother to install my spare one) I can play with voltage inputs to the edu and measure what I'm getting on the gauge cluster. I can certainly check continuity between the connections but w/out a sillyscope (oscilloscope) one is really reduced to "swap-shooting" and hoping for the best...
I certainly have little faith in ordering a cluster/edu off ebay... One thing IS for sure -- the solder and connecting ribbons are fragile with age -- I would not really expect such a part to survive shipping...

Scantar (and everybody else) There is an easy fix for the power to the fans though. At this point we all have replacement radiators -- they all have the plug for the early MY 9K's that took a thermostatic switch to power the fans. Simply wire your fan through that as a back up.
 

·
Registered
Joined
·
2,942 Posts
Oh, but it does chengny. When signal is lost from the temp gauge the ecu shuts the fan off completely...not assumption, this is something I've observed.


It also will affect the acc in that if you have the a/c running it will shut that off as well
And in case one finds my above post too long...

(I think!) your statement has the cause and effect backwards:
So it's not when signal is lost from the temp gauge BUT when power input is lost to the edu (or perception of input power by the edu -- eprom glitch?) the following outputs can't happen ie: no voltage signal to the temp/fuel gauge AND no signal to the fan relays...
 

·
Registered
Joined
·
4,197 Posts
And in case one finds my above post too long...

(I think!) your statement has the cause and effect backwards:
So it's not when signal is lost from the temp gauge BUT when power input is lost to the edu (or perception of input power by the edu -- eprom glitch?) the following outputs can't happen ie: no voltage signal to the temp/fuel gauge AND no signal to the fan relays...

I suppose it could be power to the edu...but I don't have the old analog clock/sid. I still have the digital so I do know that all readouts aren't lost. And you're right that the check radiator, low fuel indicator light up so the edu still knows what to do for display when the signal is lost to those gauges, however the signal gets to or from them.


And while who starts the conversation is speculative, we know who ends it...the ecu. Other than the fans, there may be other things that the ecu is doing as well to compensate for what it thinks is a cold run situation, ie fuel enrichment. I can already tell you that if it's hot out and you're running the air conditioner, it will definitely shut that off
 

·
Registered
Joined
·
2,942 Posts
I suppose it could be power to the edu...but I don't have the old analog clock/sid. I still have the digital so I do know that all readouts aren't lost. And you're right that the check radiator, low fuel indicator light up so the edu still knows what to do for display when the signal is lost to those gauges, however the signal gets to or from them.
Then that means it's not as simple as the power input to the edu.
Means it's (probably) the power outputs from the edu...


And while who starts the conversation is speculative, we know who ends it...the ecu.By this do you mean the module on the back of the gauge cluster or the fuel control? Other than the fans, there may be other things that the ecu is doing as well to compensate for what it thinks is a cold run situation, ie fuel enrichment. I can already tell you that if it's hot out and you're running the air conditioner, it will definitely shut that off
Should probably make sure we're all using the same terms...
I think of the EDU as the electronic display unit system (Chengny's diagram part # 210) that attaches to the back of the gauge cluster assembly. But realize this can also only refer to the readouts on the dash and the clock module...

For ECU I was thinking the T5.5 fuel/timing electronic control unit...
BUT I realize ecu can also mean any of the 57 electronically controlled systems found on the average 9K... for instance the "door lock ecu" would be a correct way to refer to the electronic module that controls the door locking function...

It would be really hard to impossible to know if the problem lies in the eproms within the EDU.
 
1 - 20 of 90 Posts
Top