Ok, subject described effect, but not the whole story. #1 response I would give anyone with this problem: Bad/incorrect wiring. Not a problem with my installation.

Car: Acura Integra GSR Coupe (1995)
Player: MK2a (problem duplicated with 2.0b11 and 2.0b3 - timing details of this report specific to 2.0b3)

The empeg's illumination sense is tied directly to the same wire as the sense line for my Clarion head unit. The Clarion does not suffer from the problem I am about to describe.

I have the empeg set to sense +12v on the illumination line. While the engine is running, with my lights OFF (0v), depressing the brake pedal for any length of time will cause the empeg display to dim for a little less than one second. The time the display stays dim is a constant. It is never variable, regardless of how long I hold down the brake pedal.

If I depress the brake again, the empeg will NOT dim. UNLESS, I wait at least 7 seconds (I have timed this many times to get to this number). With the ignition not running, it seems to be at least 12 seconds, but sometimes much longer (beyond 20 or 30).

Again, the Clarion never dims its display in these instances. And when actually turning on the headlights, the Clarion reacts to the +12v faster than the empeg. Slightly faster when applying +12v and dimming, and MUCH faster when dropping back to 0v and restoring original brightness.

I am guessing that applying the brake must somehow apply a small voltage to the illumination line, by no fault of my own installation. This small voltage is enough to get the empeg to trigger its dimmer - but for only a split second for some reason. I have no clue about the delays between replication (as mentioned above, the 7 second timing was observed with b3 - it is similar for b11, but I can't say for sure it's also 7 seconds). I don't feel like ripping open the dash to connect my multimeter to actually read the values. I may try probing the illumination pin on the dock on the weekend if I can manage to squeeze my hand and the probe pin in there.

I hope we can see some comments by Hugo or anyone else familiar enough with the way the sense logic works on the player. I rolled back to 2.0b3 tonight to test (as mentioned) but didn't have time to go all the way back to 1.03. I have no idea why I didn't notice this problem when I first started using the player. I did notice it a while back, but only last week did I trace it down to the brake (I believe Hugo might have mentioned Turn Signals a while back and I thought that might have originally been the problem).

Apart from this player interaction weirdness, there is something out of the ordinary with my rear lighting right now. I have done no electrical work on my car since the installation back in November. I noticed a week ago that my two rear markers were blown (they illuminate when the headlights are on and aren't affected by depressing the brake). The other strange lighting behaviour I've had is as follows:

Lights off: Lighs off in rear. Normal.
Lights off & press brake: left tail light comes on full illumination (both elements in bulb), right side lights only single element.

Lights on: both tail lights illuminate normally (1 element per bulb)
Lights on & press brake: left tail light comes on full illumination (both elements in bulb), right side turns completely off.

This has perplexed me a bit. It was after nboticing this that I found my corners were out. I have to replace my corner bulbs this week to see if they have any effect on the over-all circuit. As mentioned, I didn't touch any wiring recently. I replaced rear rotors, pads and calipres a few weeks ago, but that just seems like an odd coincidence (there wasn't any wiring to be seen right next to those mechanical parts).

I can't stand oddities like this. The thought of having to sort through all the wiring in the vehicle doesn't leave me too excited. There's no way I want to rip up my interior again... I still have easy access to some of the wiring at the rear (and don't see anything wrong with it...)

Bruno

My immediate guess would be that you have a ground fault in the lighting circuit,
probably in the brake lights, probably caused by corrosion.

Since all the lights are generally grounded either directly to the car's metalwork,
or to the light cluster which is then connected to the body, if the ground connection
fails or goes high-resistance odd things happen. For instance, indicators on makes
one brake light flash dimly, or something like that.

Check that all the bulb sockets have the ground (outer) connected to the
bodywork with a multimeter. This should ideally be done with any bulb-holders
in place, rather than removed from the car.

Another possibility I just thought of, although I don't think it's likely: Does your car
have a brake-pad wear indicator lamp? Quite a lot of cars have sensors in the
brake pads, basically just a brass stud that is exposed when the pad wears down
enough, and connects to ground via the brake disc. I have seen some funny things
happen if this sensor wire is intermittently shorting, as can happen if it rubs on the
rotating components.

Pca

The lights-sensing code in empeg_power.c does not have full debounce logic -- a momentary spike will trigger the dimmer.

I just recently examined this code while adding "dimmer" support for the buttonLED's, and ended up coding my own debounce logic for them, since dimming the buttonLEDs is a very time-consuming operation.

I suppose I could move my debouce logic into the main sense in empeg_power.c -- maybe spin off a test kernel for you to try.

I'll put out v237 shortly with this incorporated. It'll get removed again if anyone screams.

Cheers

Okay, v237 is out there now. Give it a try, and let me know what you think.

I'll play with the new debounce logic in 237 tonight, thanks. Installing now...

I was thinking along the same lines with regards to ground. The only problem is going to be checking everything out. The lights are all moduler. They're using plastic sockets that detach from the car body on the interior. Ground is through a ground wire, not through the outside of the socket.

I don't have a brake wear indicator, otherwise that would surely be suspect. The reason I had to replace my brakes to begin with was because of a seizure on the right rear. Unfortunately I had to drive for the entire week with the caliper seized before I could fix it. I still have to clean all the rusted brake dust off that rim...

I'll give Mark's new kernel a try for the problem with the empeg flicker and I'll try to make some time to check out the grounds for all my rear lighting this weekend. Not looking forward to it.

Bruno

Hugo mentioned something similar about the dimming logic a little while back. Thanks for the patch. I'm downloading 239 as I write this and will verify it in the morning.

Do you think you can make a patch to fix the problem with my tail light?

Bruno

The reason I had to replace my brakes to begin with was because of a seizure on the right rear. Unfortunately I had to drive for the entire week with the caliper seized before I could fix it. I still have to clean all the rusted brake dust off that rim...

I had to do the same thing to a friend's car recently. In his case, the caliper seized in action, so thoroughly that when he got home the brake disk was glowing and the plastic wheel trim had melted. How the tire survived I'll never know.

pca

I'll play with the new debounce logic in 237 tonight, thanks. Installing now...

Not working, at least not for me in 239. See my post in the Programming forum.

Damn. I'll still test it in the morning to see if the debounce portion works.

Bruno

Hmm, I should probably actually flash the player before going to bed, or I won't be testing anything...

I began to draught a reply here based upon a bad earth scenario (which altered de-dounce timing may fix but still not completely disprove IMO) and in the end gave up (too many potential causes). Personally I'm with the higher than 0v earth potential at the stop light (bad earth) explanation.

The reason I gave up was I because I needed to make too many assumptions about the empeg's sensing and electrical characteristics such as:
1. Does the empeg utilise a potential divider to sense
2. Is that smoothed by a capacitor
3. What is the input impedance
4. What is the sense threshold voltage

It goes without saying that understanding the electrical theory on gounding problems is beneficial but in itself is not a cure practically speaking. Personally and admittedly OT, I'd benefit from an answer to q.4 (sense voltage threshold) to help cure a sense 'problem' I have with partially powered headlights in sidelight+engine running 'mode'.

The debounce portion of the fix didn't work either. It would still dim when I depressed the brake pedal. I didn't do any of the timing tests mentioned above nor did I look at it more than a couple of times (too busy looking at the road

Bruno

Oh well, ya win some, ya lose some.

Cheers