That's what I discovered too. I didn't have the cluster completely removed, so I couldn't see the back side to determine which side the ground was on. For lights which come on every time you start the car, it was easy to test. But there are a few which don't normally come on, and I'll have to trace the contacts to know which way is up.
Once you have the wiring harness connectors removed, can you get the cluster out without removing the steering wheel? My wheel only adjust up and down, it does not extend forward and backwards towards the driver. I didn't remove the connectors, and it looked like the cluster might not fit between the wheel and the dash.
Can you reach the connectors from below the dash and up the backside (or I guess that would actually be the front side)?
From your picture, it looks like the two connectors may have locking features on the ends. Do you just squeeze the ends of the connectors and pull?


Thanks,
Greg

 

Oh yea, just tilt your wheel all the way down and you can ease it out to the right side and those connector wires are fairly long though sometimes you have to kinda pull them up a little to get them up fully.

 

Just to add...I had trouble with my speedometer a few days ago It would seem to have poor connection, worked one minute and didn't the next.
Removed cluster and examined the removed connectors. A black oxidation film were covering multiple copper connectors. Cleaned and trouble is over.
The lost hours of research and testing wiring to and from transmission to find the fix is over.

While you guys have the cluster off check the connectors.

 

Greg & SouthernGypsy,

Have just puled the instrument cluster from my car and checked the warning bulb circuits. As you probably know, it's pretty difficult to get a volltage probe to one side of each bulb because of their canted orientation.

Anyway, the Check Engine, Sport Mode & 1st gearlock bulbs all seem to pull .01v when not being lit. The other bulbs don't seem to pull anything when not being lit.

So, is .01v really sufficient to power an LED dimly? Maybe it is. But I still don't really understand why those 3 circuits are pulling some current when apparently not in use whereas the other warning bulbs circuits don't seem to pull any current until asked for?

I also don't really understand current draw on LEDs. If I changed the bulbs for LED ones that each had 2 LED units would they glow dimmer or brighter with .01v than an LED bulb with one LED unit? My naive logic says they would glow brighter with any voltage compared to a single LED unit. But maybe there's a reason why .01v isn't a sufficient voltage to cause a 2 LED unit to glow at all, whereas it can make a single LED unit glow dimly? Does that make sense?

Any thoughts?

Paul

 

Paul,
What I thought I knew about LEDs turns out to be wrong. I thought they were either on, or off, with no dimming capability. Apparently that isn't correct. My guess is that if the LEDs are connected in parallel, and there is enough current to drive both of them sufficiently to light, then 2 would show brighter than 1.


I wonder where that small voltage is coming from. You say it is on 3 circuits, and I see 3 transistors in the photo from Southerngypsy. The transistors are the black devices with 3 leads coming out of them. Do the circuits to the lights with the small voltage come from one of the transistor leads?


I didn't have time to pull the cluster on my XJS this weekend. Too many other things that I needed to do.


Thanks,
Greg

 

Two glow very dimly when the LED's are in them, the Sport light and the Check Engine light.

The other thing about the sport light is that when an LED is in it but a regular bulb in the /1 light then neither work and vise versa, but an LED in both will make both work (though sport then glows dim which it also does if an LED is in it but regular bulb in /1) and if you put a regular bulb in both then both work with no glow from the sport.

 

After doing a little online research, I'd almost bet money that the transistors are what is causing the slight voltage to some LEDs. It seems that it is a common situation where a transistor does not shut fully off. With bulbs, this situation would probably be inconsequential, and the engineers probably decided to ignore it. It looks like there are ways to get transistors to shut off completely by the addition of resistors, diodes, or both to parts of the circuit, but it might be difficult to determine the correct solution.
But the apparent interaction between the Sport mode circuit and the /1 circuit still is baffling. It is almost as if they are wired in series, and the presence of one LED prevents enough current flow to light the other bulb in series. But if that were the case, the one with the LED should still work. They can't be wired in series anyway. I'm guessing they operate mutually exclusively. One or the other can be on, but not both at the same time.
On my cluster, the Sport mode LED glows dimly all the time, but the Check Engine light does not. And Paul, it sounds like it is only the Sport mode led that glows on yours, and not the /1 or Check Engine led. If that is true, then something is going on that makes the Sport mode led glow, and not the other 2 even though all may have a small voltage continuously.

 

Sorry, I meant that to be referring to the symptoms on SouthernGypsy's cluster, not Paul's.

 

Sorry again. I'm going to fast and not reading everything carefully. I looks like SouthernGypsy does have two LEDs glowing all the time; the Sport and Check Engine. That makes some sense if slight current flow through the transistors is the problem. It doesn't explain why the same thing isn't happening on the /1 led (although the transistor on that one might be shut off more fully), and it doesn't explain the interaction between /1 and Sport.

 

Tks guys for the replies.

Although it doesn't necessarily alleviate the problem, I wonder if the reason for the /1 bulb not working with an LED in the Sport mode is for the following:

- the 2 circuits are probably linked as the gearbox can not be in /1 if it is in Sport, and vice versa.
- If an LED is in Sport, then the circuit is always complete because the LED needs less current draw and glows dimly with even the residual voltage
- If the Sport circuit is "seen" as being complete because the LED has completed the circuit even when the button is not pressed, then maybe the voltage is then reduced on the /1 circuit when the /1 button is pressed.
- However, when an LED is also in the /1 circuit, then the reduced voltage is still sufficient to make it light when the button is pressed.

Does that make sense?

I don't really want to start messing around the wiring to pull the transistors down to zero. So, I might try a 2-unit or even 4-unit LED just to see if that stops it from glowing with .01v.

Or (he said hopefully), maybe these circuits are a bit variable. And with only .01v on the circuits in my car, maybe they won't light an LED at all? Whereas other cars are pulling more than .01v and they do get the problem. Maybe that's wishful thinking?

Paul

 

Paul,


.01v does seem to be extremely low. Perhaps your LEDs won't glow.


Regarding your theory on what might be happening, I don't think an LED would "complete" a circuit any more than a bulb filament. If there is voltage present, the bulb will pass a small current too.


In fact, LEDs typically pass less current than bulbs (which means they have a higher total resistance), and that sometimes causes problems with things like the Bulb failure warning system when LEDs are used in the circuits that have that feature.


From the wiring diagrams that I have, both the /1 and the Sport mode bulbs get their signal from the transmission control unit. The switch on the center console does not appear to be directly in the circuit for either light. The switch is connected to the ACU, and the ACU has output wires that go to the instrument cluster for the /1 and Sport lights.


Since the typical problem when using an LED rather than a bulb is the fact that the LED draws less current, the interaction between those two lights is very strange. It's like they are wired in series, and there isn't enough current to glow a bulb filament if there is a LED in series with it. If both are LED's, the current is sufficient to light both. If both are bulbs, the current is sufficient. But it doesn't make sense that they would be in series because both would light if either was selected.


Strange

 

Of all the threads (including the one I started some time ago) I've seen on fixing the annoying dim-warning-lights-on-XJSs this is by far the most informative and best!!!

I have a question about the B8.5DLEDs you choose: The SuperbrightLEDs' website say they are 6 lumen. I this weekend tried their T1.5-xHP (Search Results For "pc74 led bulb - high power instrument panel led " | LED Products | Super Bright LEDs | Super Bright LEDs) bulb which is rated at 10 lumen and saw no difference from the original Osram 12/1.2 bulbs (I tried both a red (parking break) and a blue (high beam)) Am I incorrect in believeing that the lumen-rating is a measure of the brightness?

I am also looking forward to reading if a solution is found for the mysterious glowing Sport bulb. Like Greg Hastings stated above, that was the only one of my original bulbs that was burned out, so clearly not a coincidence.

 

I forgot to post it but the night before last while driving on driving on a dark, cloudy night out in the country (away from all city lights) that the Seatbelt light also VERY dimly glows all the time but it's extremely dim, I have driven at night since putting in the LED's but never noticed it until I drove a friend home on a totally black overcast night and we got out in the country away from any light sources and only then did I notice it. So yesterday I switched it back to a regular bulb in that one and then tested last night by taking a late night drive on another dark cloudy night out in the country and now I have no bulbs glowing at all. So that's the Sport, /1, Check Engine, and Seat Belt lights that need to be left regular bulbs. Oddly all of those seem a bit brighter since putting the LED's in all the others but maybe just because I wiped off the bulbs before putting them back in and with a selection of bulbs I'd removed I choose the best looking ones to use for those four.

 

Anders,


Yes, lumens is a measurement of visible light. It's hard to say why the leds you tried didn't appear brighter than the original bulbs. I would think there would be sufficient voltage and amps to fully drive those leds.


Concerning the Sport bulbs being the only ones that are burnt out on our vehicles. I'm wondering if that may be more from previous owners keeping the car in Sport mode most of the time. That might be more likely to burn the bulbs than a small trickle current which isn't sufficient to even glow the filaments. And that would be the one bulb likely to be on more than any of the others.

I've been sidetracked with a woodwork project in my garage, so the XJS has been under it's dust cover mostly. So I haven't had a chance to pull my cluster and look at the transistor circuits. I hope to be able to do it this weekend. But we are coming to the end of our summer her in Northern Idaho, so I'm trying to take advantage of the nice driving weather while it lasts. Soon enough, winter weather will set in and I won't be driving the car at all for few months. That will give me the opportunity to work on a few minor problems.

 

I have read this thread from start to finish. I have decided to replace some of the original bulbs in the cluster but I am not able to locate which bulbs to purchase.
Can someone offer me where I am able to purchase a few original bulbs for a 1996 XJS.?
Thanks

 

eBay for originals or for LED's there is a link in the opening post of this thread.

 

All,
I finally got around to pulling my instrument cluster again, and got a pretty good look at the circuit traces. On my unit, it appears that the 3 transistors are used to drive the Sport mode, Check Engine, and Brake (low fluid) lights.
Two of those match with SoutherGypsy's glowing LEDs. The Brake light is the odd one that doesn't match his symptoms.
But, mine is a 1993, and that year seems to have a few unique configurations. So it is possible these circuits are different.
And, only the Sport mode light glows all the time on mine. The other two do not. I just checked it in a dark garage at night. Furthermore, the Sport mode glows pretty brightly. I can see it in the daylight with the top down. If this is caused by transistors not shutting off, perhaps there is varying amounts of leakage?
I have damaged copper foil at the socket for my low fuel light. So I'm going to order a new flexible circuit to fix that. I may be able to use the old one to play around to see if there is a way to get the glowing lights to extinguish.

 

Transistors are analog devices, and all have some leakage. Chances are these are cheap ones, so leakage will vary from car to car. Maybe replacing it/them with a small relay would do the trick?

 

My guess is that the signal current is insufficient to operate a small relay if they didn't want to drive those tiny bulbs directly with it.
Maybe a higher quality transistor would do the trick? Surprisingly, the flexible circuit boards seem to be readily available, so I'm ordering a new one today. The foil pulled loose on both contacts of the low fuel light, so I need to fix that anyway. When I get the new board, I may be able to read some markings on the transistors. I'll also be able to diagram the existing circuits accurately and completely. It's hard to do on the installed unit because some of the traces are on the back side and are not always clearly visible.
There are also two diodes, and what appears to be a tiny capacitor involved in each transistor circuit. Perhaps one of those devices is the culprit?

 

Maybe the signal current on those circuits is sufficient to drive the LEDs directly?!
The bulbs are 1.2 watts = .1 amps. The LEDs are 10 milliamps = .01 amps = 10 times less current draw.
That would be very easy for me to test on the old circuit board. I can just remove a transistor and jumper the input signal to where the collector lead on the transistor was.
I wonder if that risks burning out whatever is generating the signal current? 10 milliamps is a pretty low current.