HSV gauges in G8 flashing

[J Wikoff]

This morning, on the way to work, my gauges were flashing, several flashes at a time, then it would pause. It was raining and the headlights turned themselves on with the wipers. If I turned the headlights off, they stopped flashing. It was still raining when I left work. Turned on the wipers, the headlights come on, and the gauges start blinking again.

I hit night mode (so all the dash lights turned off but the speedo and DIC), and the they stopped flashing, but stayed lit. Normally, they turn off with night mode.

I turned the brightness all the way up, and they stopped flashing, but turned off.

I turned the brightness a few clicks down, they started flashing again, but in between flashes, they stayed off. But stayed at the same brightness when lit.

Turned the brightness all the way down, and they stopped flashing, and stayed all lit up.

It happened to get dark enough in the forested area I was driving through for the lights to want to turn themselves on. When that happened, they acted normally no matter what I tried. Lit up/dimmed just like they should.

????

I have a video and possibly other info here:

Flashing HSV gauges - GRRRR8.net - Pontiac G8 and Camaro Forum (http://www.forum.grrrr8.net/showthread.php?t=6632)

I should add: This is new behavior. I've had them installed for a few months, and it just started doing this.

[steve_t]

It looks like they might be flashing in time with the wipers actuating. This could be due to the current draw of the wipers. Who wired them up for you?

[J Wikoff]

Turning the wipers off didn't change the flashing. It seems like if the lights are on (whether I turned them on or the wipers did) without the twilight sentinal telling them to be on, the gauges flash.

I wired them up.

[senatorVIII]

Are you sure you aren't driving a VW Beetle called Herbie??

[J Wikoff]

Has this happened to anyone else?

Here's another video.

Started headlights off and gauges are lit like normal.
Turned headlights on (dimmer set near middle) and gauges start blinking, pause for a moment, the flash again.
Turned the dimmer up and they turn off.
Turned the dimmer down and they stay lit.
Set dimmer near middle and they blink again.
Covered light sensor so headlights would turn themselves on and they stop blinking.
Set dimmer up and down and they dim like normal.
Uncovered light sensor and they start blinking again.

http://i19.photobucket.com/albums/b184/jwikoff99/G8/th_020.jpg (http://s19.photobucket.com/albums/b184/jwikoff99/G8/?action=view&current=020.flv)

[HSVGTS215i]

Has this happened to anyone else?

Here's another video.

Started headlights off and gauges are lit like normal.
Turned headlights on (dimmer set near middle) and gauges start blinking, pause for a moment, the flash again.
Turned the dimmer up and they turn off.
Turned the dimmer down and they stay lit.
Set dimmer near middle and they blink again.
Covered light sensor so headlights would turn themselves on and they stop blinking.
Set dimmer up and down and they dim like normal.
Uncovered light sensor and they start blinking again.

http://i19.photobucket.com/albums/b184/jwikoff99/G8/th_020.jpg (http://s19.photobucket.com/albums/b184/jwikoff99/G8/?action=view&current=020.flv)

Could be faulty LED globes if thats what they use in the guages, Ive had to replace LED globes in the centre console for the Auto number illumination before. They too would flash on and off.

[J Wikoff]

I'm not familiar with "LED globes". Can they work normally under certain conditions and not under others?

[smokey777]

i pulled up to a servo & noticed in the window that my parkers (which are LEDS) one was dead & the other was flashing so i think its dead LEDs

[J Wikoff]

If all three gauges light up evenly across the face and uniformly from one gauge to another, could it still be a dead LED?

[SSV8TE]

Gday mate,
Exactly the same problem that i have had with mine since day 1.
I have found that you have to dim the dash lights to the point where they dont flash and im told there is no fix yet but im sure a few people are looking into it.
When you find the fix let us all know if you do.
Heres a short vid with me adjusting the dimmer switch.
http://i303.photobucket.com/albums/nn121/SSV8TE/th_31102008038.jpg (http://s303.photobucket.com/albums/nn121/SSV8TE/?action=view&current=31102008038.flv)
Cheers,

Andy.

[6.2L.Club]

It is the calibration of the bcm and it's output voltage to control the brightness of the LEDs, it can be fixed. Your guages have to be recalibrated by Continental (VDO), who are the makers of the guages or one of their authorised repair centres. I have no idea on what it would cost coz l never bothered to go down that path. l just dimmed them.

[HSVGTS215i]

I'm not familiar with "LED globes". Can they work normally under certain conditions and not under others?

I think LEDs are more sensitive to voltage variations than normal globes..

[J Wikoff]

It is the calibration of the bcm and it's output voltage to control the brightness of the LEDs, it can be fixed. Your guages have to be recalibrated by Continental (VDO), who are the makers of the guages or one of their authorised repair centres. I have no idea on what it would cost coz l never bothered to go down that path. l just dimmed them.

So since it just started, is it because they have become uncalibrated?

Is there a VDO repair center in the US?

By "globe", do you mean "bulb"?

Has anyone tried a different illumination circuit? Mine are always lit even if I don't have the headlights on. But I've read if I switch to another wire in the headlight switch, they only light up with the dash.

[J Wikoff]

Talking with a couple coworkers that have been working with can-bus and engine programming. They say the can-high and can-low wires should be twisted around each other.

Thoughts?

[Wonky]

I'd say the best thing to do is try it and see - if it works you'll be a hero :bow: to many (including me who'd eventually like HSV centre gauges) and if it doesn't that's one more thing eliminated!

[SSV8TE]

As food for thought mate,
If you twist the two high/ low can bus wires together which of the two dash harness wires would you connect these high/low can's too?
Any thoughts on that?

Cheers,

Andy.

[J Wikoff]

After I wrap them around eachother, I'll probably hook them back up to the same harness they already are. I'd have to dig through the write up to see which harness that actually is. It's next to the steering column. They told me a specific number of twists per centimeter, but I can't remember what that was right now. Supposed to interupt EM interference or something along those lines.

I also discovered today that they flash when I use the remote start too. They used to just light up. Kinda cool effect when I use the remote start at night... three sinister red gauges glowing in there by themselves.

[gmh308]

Talking with a couple coworkers that have been working with can-bus and engine programming. They say the can-high and can-low wires should be twisted around each other.

Thoughts?

You installed these personally.....was the right connector already in the harness?

They should be twisted to reduce susceptibility to noise. There should also be a CAN bus terminating resistor after the VDO CANCockpit module - it may be there if you connected to existing harness connector. Unlikley that lack of twist is causing any backlighting problems.

6L.Club likely pretty well has the reason with the BCM control of the backlighting. If you search Google for VDO Continental/automotive instruments in your area, you may find someone that has the CANCockpit programming tools and some knowledge on this. Though CANCockpit is more present in truck/bus markets than car.

:)

[J Wikoff]

There was no connector. I had to splice them in to the car's wiring.

Please expand on the terminating resistor. Never heard of it. What's it do? Is it a simple resistor for one or more wires? What resistance?

Is the VDO CANCockpit module the black box behind the gauges?

[gmh308]

There was no connector. I had to splice them in to the car's wiring.

Please expand on the terminating resistor. Never heard of it. What's it do? Is it a simple resistor for one or more wires? What resistance?

Is the VDO CANCockpit module the black box behind the gauges?

Yes the CC module is the black box behind the gauges, normally with some yellow wires hanging out of it. (normally).

It is common, in GM cars at least, that the CAN is terminated with a 120ohm resistor somewhere. Without going into the deep and meaningful reasons for this, in simple terms it is there to ensure electrical integrity of the CAN so that data corruption is minimised or eliminated.

The VE/G8 design calls for this resistor on the CAN after the CC module. If you spliced into the CAN somewhere under the dash, and you now have a "Y" connection, that is probably ok. If the CAN was getting too corrupted, you would have DTC's being raised as modules could not communicate.

Would not be surprised if the CC connector is hidden somewhere on the dash harness anyway. It is a rectangular 6 pin connector, with 6 pins used, and a lock tang.

It has a pink/blue, black, brown & brown/black (CAN), blue and white/yellow wires into it.

What did you connect the "backlighting" wire to?

Cheers man. :)

[J Wikoff]

It has been hooked up to the yellow wire in the headlight switch harness. With PeterG22000's advice, I've switched it to the white wire with silver bands. It doesn't flash anymore and dims like it should. But it doesn't light up with the remote start anymore either.

I'll drive with it like that for a while, then decide if I liked the lit remote start gauges and keeping the dimmer out of a certain range, or if I like no flashing no matter where the dimmer is.

[gmh308]

It has been hooked up to the yellow wire in the headlight switch harness. With PeterG22000's advice, I've switched it to the white wire with silver bands. It doesn't flash anymore and dims like it should. But it doesn't light up with the remote start anymore either.

I'll drive with it like that for a while, then decide if I liked the lit remote start gauges and keeping the dimmer out of a certain range, or if I like no flashing no matter where the dimmer is.

The yellow wire to the head light switch does not carry any load. None of the wires to most switches do, they interface to the BCM which then controls other items. It was flashing because the BCM thought it had a short/too much load.

The wire that controls the LED Backlighting should be White with a Yellow stripe.

The white wire you are talking about is most likely the headlight control wire which is why it does not dim.

Pretty well all the BCM and relatd wiring has silver paint markings on them.

Cheers! :cheers:

[J Wikoff]

They dimmed with everything else before when it was dark enough for the headlights to turn on, and they still do with the white/silver banded wire. If I turned the headlights on without it being dark out, the gauges would flash in the middle range of the dimmer. Sorry if I wasn't clear about that before.

When I switched wires from the yellow to the white with silver banding the behavior changed like this:
No more flashing
Does not light up with remote start
The gauges only light up when the headlights come one. They used light up without the headlights.


There is no wire in my headlight switch harness that is white with a yellow stripe.

[cashie]

They dimmed with everything else before when it was dark enough for the headlights to turn on, and they still do with the white/silver banded wire. If I turned the headlights on without it being dark out, the gauges would flash in the middle range of the dimmer. Sorry if I wasn't clear about that before.

When I switched wires from the yellow to the white with silver banding the behavior changed like this:
No more flashing
Does not light up with remote start
The gauges only light up when the headlights come one. They used light up without the headlights.


There is no wire in my headlight switch harness that is white with a yellow stripe.

I would guess that this is now normal operation...?
I would have thought the gauges would only light when the headlights are on.
Do they light with your remote start when the headlights are on "auto"?
Maybe just leave your headlights on "on"?

[J Wikoff]

I would guess that this is now normal operation...? Seems to be, but I liked that they lit up with the remote start.
I would have thought the gauges would only light when the headlights are on. I agree.
Do they light with your remote start when the headlights are on "auto"? Not anymore.
Maybe just leave your headlights on "on"? If dings incessantly if I do that. Not worth it even if it works.

[gmh308]

They dimmed with everything else before when it was dark enough for the headlights to turn on, and they still do with the white/silver banded wire. If I turned the headlights on without it being dark out, the gauges would flash in the middle range of the dimmer. Sorry if I wasn't clear about that before.

When I switched wires from the yellow to the white with silver banding the behavior changed like this:
No more flashing
Does not light up with remote start
The gauges only light up when the headlights come one. They used light up without the headlights.


There is no wire in my headlight switch harness that is white with a yellow stripe.

Ok got it now. Looks like you were on the switch illuminator wire. If thats like the ign switch, it has a normal bulb.

The White/Yellow is not in the switch harness, it is an output from the BCM and doesnt go near the switch. It is for LED illumination. Probably hard to find buried up under the dash somewhere. But if it is there, it is available in the region of the binnacle gauges.

Cheers.

[J Wikoff]

If I were to find the white/yellow wire, how would it change the behavior? Is that not the wire the other guy that had the same flashing was using?

[SSV8TE]

Other guy might be me john, :lol:
I am going to try these tricks in mine once the ute is back on the road.
Sounds like alot of great info here towards fixing this mysterious blinking problem.
Good info lads :cheers:

Andy.

[bigdogdazza]

Gday mate,
Exactly the same problem that i have had with mine since day 1.
I have found that you have to dim the dash lights to the point where they dont flash and im told there is no fix yet but im sure a few people are looking into it.
When you find the fix let us all know if you do.
Heres a short vid with me adjusting the dimmer switch.
http://i303.photobucket.com/albums/nn121/SSV8TE/th_31102008038.jpg (http://s303.photobucket.com/albums/nn121/SSV8TE/?action=view&current=31102008038.flv)
Cheers,

Andy.

You guys havent been feeding the hsv party drugs have you?:woohoo::party:

[gmh308]

If I were to find the white/yellow wire, how would it change the behavior? Is that not the wire the other guy that had the same flashing was using?

Holden connects the "LED Backlighting" circuit on a White/Yellow wire to the CANCockpit unit in VE HSV's. They also connect a circuit "Comm Enable" via a Blue wire. Comm Enable is used to power CAN based modules in a passive mode so they can talk to each other if needed.

We just try and do what the factory does, and then things generally work. :)

[J Wikoff]

Well, no more flashing since I change the wire. And the behavior mimics the dash gauges spot on.

[Grant]

You installed these personally.....was the right connector already in the harness?

They should be twisted to reduce susceptibility to noise. There should also be a CAN bus terminating resistor after the VDO CANCockpit module - it may be there if you connected to existing harness connector. Unlikley that lack of twist is causing any backlighting problems.

6L.Club likely pretty well has the reason with the BCM control of the backlighting. If you search Google for VDO Continental/automotive instruments in your area, you may find someone that has the CANCockpit programming tools and some knowledge on this. Though CANCockpit is more present in truck/bus markets than car.

:)

To be pedantic, dual wire CAN high/low are run in parallel to ensure they are subjected to the same noise, not so they are subject to less noise. CAN is voltage-differential encoded; the relative difference between the levels of the two lines dictates whether the current bit is a zero or one.

For example, if both lines are subject to some induced noise that makes both lines increase by 1v, the relative difference between the two lines is still the same, so the data remains valid.

[6.2L.Club]

this is the schematic for the guage pod for HSV's

http://i180.photobucket.com/albums/x83/sixlitreclub/wiringdiagramhsvguagepod.jpg

Led dim is the far left "yellow" wire

if it wasn't such a bitch to get to the bcm l would have connected it to this point. :lol:

[gmh308]

To be pedantic, dual wire CAN high/low are run in parallel to ensure they are subjected to the same noise, not so they are subject to less noise. CAN is voltage-differential encoded; the relative difference between the levels of the two lines dictates whether the current bit is a zero or one.

For example, if both lines are subject to some induced noise that makes both lines increase by 1v, the relative difference between the two lines is still the same, so the data remains valid.

G'day mate, That explanation in "twisted pair" seems very Wiki like... If you read the post again, it doesn't mention "they are subject to less noise" it mentions that the CAN/GMLAN is less "...susceptible to noise...".

Which is recognition of the design that embodies standard techniques to minimise normal mode noise (noise relative to ground) by ensuring that noise is common mode. Common mode noise being more easily seperable from the data signal during decoding at layer 1 and above.

What the Wiki doesnt seem to go into is the reason for the highly necessary twist, which is to ensure a reasonably constant capacitance spec, which is part of the 120 ohm impedance spec of the CAN circuitry, which takes into account capacitance, reluctance and resistance, and uses a 120 ohm resistor to ensure that termination of the cable is compliant with the spec and that energy reflections at impedance changes are minimised/negligable which ensures that re-transmits due to packet deformation/corruption are minimised and the bit rate on the CAN is maximised to ensure timely inter-module response.

:teach:

They are actually run in parallel, because they go to all the same places :). The twist helps them do it in the most effective way! :goodjob:

Enjoy.


this is the schematic for the guage pod for HSV's

http://i180.photobucket.com/albums/x83/sixlitreclub/wiringdiagramhsvguagepod.jpg

Led dim is the far left "yellow" wire

if it wasn't such a bitch to get to the bcm l would have connected it to this point. :lol:

Yes the BCM is tough to get to, with its dozens and dozens of connections across the 7 connectors. :confused:

Great post. Had not seen the detail schematic of that before. That will help anyone else doing this install!

Interesting in that (and this is not uncommon) that the schematic is a little different to the actual wiring and shows the SAS - Steering Angle Sensor wiring differently.

The SAS sits across the CAN after the BCM and before the VDO module, and there is a terminating resistor after the VDO module.

As your connector dwg shows, the actual wire used in the harness to BCM X3-18 is white with yellow strip vs yellow. (that's being nit picking, but for the person searching under the dash to install one of these, it is essential info and is what the actual harness is built with.)

Such is the challenge with GM schematics, if you dont have them all, they do can be a little ambiguous. :confused:


Well, no more flashing since I change the wire. And the behavior mimics the dash gauges spot on.

Excellent result! :)

[Grant]

G'day mate, That explanation in "twisted pair" seems very Wiki like... If you read the post again, it doesn't mention "they are subject to less noise" it mentions that the CAN/GMLAN is less "...susceptible to noise...".


I'll take your obvious implication my text is plagiarised as a compliment. The CAN bus is still as succeptible to noise, but it is very robust in the presence of noise. CAN cant prevent the induced noise, it just deals with it in a robust manner.



Which is recognition of the design that embodies standard techniques to minimise normal mode noise (noise relative to ground) by ensuring that noise is common mode. Common mode noise being more easily seperable from the data signal during decoding at layer 1 and above.


This is a fundamental property of differential encoding. In this context, it can only be common-mode noise because thats how differential encoding (and CAN) works. There is no common gound; the value on the bus is determined by measuring the difference between the two bus lines.

However, the point is that the noise becomes part of the signal at layer 1, but it doesn't change the value of the data. After the effects of the common mode noise have been cancelled out at the receiving device, ideally the data is received as it was transmitted. There is no "noise" above layer 1. Bit errors introduced at layer 1 will cause the layer 2 frame CRC check to fail, and therefore the packet will be discarded.



What the Wiki doesnt seem to go into is the reason for the highly necessary twist, which is to ensure a reasonably constant capacitance spec, which is part of the 120 ohm impedance spec of the CAN circuitry, which takes into account capacitance, reluctance and resistance, and uses a 120 ohm resistor to ensure that termination of the cable is compliant with the spec and that energy reflections at impedance changes are minimised/negligable which ensures that re-transmits due to packet deformation/corruption are minimised and the bit rate on the CAN is maximised to ensure timely inter-module response.


The electrical characteristics of the wire (capacitance, resistance...) are unrelated to the fact the wires need to be twisted together.

The wires are twisted together because thats how differential pairs work. They must be subject to the same noise for the protocol to work effectively.

The terminating resistor is unrelated to the above, but necessary to prevent reflections, as you said.




:teach:

They are actually run in parallel, because they go to all the same places :). The twist helps them do it in the most effective way! :goodjob:

Enjoy.


They are run in parallel because thats how differential pairs work. If they didn't run to the same place, they wouldn't be a differential pair, and it wouldn't be a 2-wire CAN implementation.

[gmh308]

Love you work mate! :goodjob: And your current avatar too! :bow:

Does the impedance spec of the CAN bus contribute to the overall performance?




I'll take your obvious implication my text is plagiarised as a compliment. The CAN bus is still as succeptible to noise, but it is very robust in the presence of noise. CAN cant prevent the induced noise, it just deals with it in a robust manner.



This is a fundamental property of differential encoding. In this context, it can only be common-mode noise because thats how differential encoding (and CAN) works. There is no common gound; the value on the bus is determined by measuring the difference between the two bus lines.

However, the point is that the noise becomes part of the signal at layer 1, but it doesn't change the value of the data. After the effects of the common mode noise have been cancelled out at the receiving device, ideally the data is received as it was transmitted. There is no "noise" above layer 1. Bit errors introduced at layer 1 will cause the layer 2 frame CRC check to fail, and therefore the packet will be discarded.



The electrical characteristics of the wire (capacitance, resistance...) are unrelated to the fact the wires need to be twisted together.

The wires are twisted together because thats how differential pairs work. They must be subject to the same noise for the protocol to work effectively.

The terminating resistor is unrelated to the above, but necessary to prevent reflections, as you said.



They are run in parallel because thats how differential pairs work. If they didn't run to the same place, they wouldn't be a differential pair, and it wouldn't be a 2-wire CAN implementation.