SharkSkin

And if that doesn't work, you'll come on over and grab a wrench, right?

AO

Of course!

SharkSkin

Alan

Andrew (and Dave),
I was waiting in a Dr's office today killing time (!) and was thinking about this one. The first posted 2 wire version you linked - Andrew - I don't know how they can do it... I cannot design a circuit that works reliably that way - at least not one that works acceptably well. Hence maybe their comment about it dimming noticeably when the door initially closes I think!

A 4 wire version (Bulbs, Relay O/P, 30, Gnd) immediate dimmer that has to be spliced in after the relay and feeds the bulb seperately is very doable with good performance - but requires circuit fabrication. It could even fade up too! - I did do some searches (yes a Drs office with free WiFi!) but I found no good canned solutions.

So I think you will have to build your own if you really want one... I've added this idea to my back burner list - its got a bunch of things ahead of it though!
Alan

SharkSkin

So, Alan -- you don't thing the one at radiolocman.com would work at all? It looks like the easiest one to implement since it seems that all I would need to do is tie J1-1 on the diagram to pin U2 on my panel, then tie pin J1-2 to ground. The one at mp3car.com looks doable too, and seems it would be equally easy except for the additional ACC wire. I have to think about it some and decide whether I care about it dimming right away. I could slap this on a protoboard pretty quick and give it a shot, but I don't work with this kind of circuit every day and I can't just visualize the circuit dynamics in my head. Oh, I get that both circuits are charging a cap whose voltage output makes a transistor's path from the light to ground increase resistance gradually to open circuit -- but beyond that I'd have to sit down with pencil & paper & scratch my head a while over the specifics.

AO

You guys are doing a great job. Anyone need a beer yet?

Alan

The issues with these 2 wire types is two-fold:

1) They consume some aditional (charging & bias) current even when off - I don't like this...

2) Since they run off stored power in a capacitor - they will run out of juice pretty quickly...

They can only get their own supply when the interior lights are off (via the bulb) they have to store
this to operate for as long as the bulb is on - hence the big capacitor....

From a detection perspective they can only know the lights switch is either open or closed. When the switch is open it has to charge itself - its the only time it can.

When the switch first opens is when the delay and dimming must happen.

So it either needs to monitor the switch - detect the first opening (bulb going off) and immediately turn its O/P transistor 'on' (risking a flickert) and then delay and dim. But what if the light was on (e.g. car door open) for >5 minutes... the capacitor will discharge and no delay or dim...

The other option can avoid the flicker by having the circuit drive its output transistor 'on' the whole time the switch is closed - but then its very hard for it to detect that the switch went open - it has to look for a small difference between switch and transistor resistance (and probably deliberately make this big enough to see easily) - then must initiate the delay and dimming. This type has an even harder time maintaining its temporary capacitor supply...

In either case there is the problem that it won't work if the car door stays open too long, it also will either flicker when the switch opens or dim noticeably when the switch opens - unacceptable I'd think.
If you had LED lights the flicker would be worse but the dimming may be less noticeable.

I don't see a path to a good performance 2 wire version. I haven't built one yet but I would recommend a 4 wire version w/ 30 & Ground as well as relay O/P as a trigger input and take over the bulb control completely with the module. This can be built to have no flicker/dim and to consume no continuous current when inactive. It would also be easy to build a dim only type - with a ramp up quickly and ramp down slowly... I jotted down an outline circuit to try out on a rainly day.

Alan

AO

Should be done today or tomorrow!

SharkSkin

OK, I see what you're saying as relates to the following circuit... this will discharge as you say if the door is open too long, then do essentially nothing when you close the door:



Hmmm... with I=E/R, the circuit below would appear to draw .06mA through the two 100K resistors and the collector cutoff for the main transistor is also in the micro-amp range. I could be goofing up my interpretation of the spec, of course.... but the circuit below looks to me as though it would draw insignificant current after the cap is charged. Also, it should charge just fine with the door closed.

This would also be pretty easy to hook up I think, as long as I only want it activated by the doors. Allowing the hatch to activate it would require tying in somewhere away from the CE panel, wherever the connectors are that are shown on Page V of my circuit diagram, circuit track 18. I don't know just where that connector is, and if it's too far from the CE panel I may just skip it. I was thinking of interrupting the CE wire at U2(just the doors) and picking off B+ and ACC/IGN from my neutral safety relay socket.

SharkSkin

justin

You guys got it done yet?

SharkSkin

Justin, I have a solution ready just for you. Before you close the door, look at the illuminated dome light. Gradually close your eyes until you don't see the light anymore. Then close the door. Done.

Alan

Dave - the second one IS a 4 wire circuit... the capacitor there is just the timer cap...

Unless you want the delay too (assume you already have one...) its a bit over complex - you can prob. get by with much less..

maybe something like this - needs some work still...

Alan

SharkSkin

Nope, no delay on mine. It's on or it's off, no in-between. Also, I'm interested in doing the least possible re-wiring to make it work. I thought that after post number 20, you were still saying you hadn't seen anything that would work. I suppose that if all of the correct lines are there on later cars it could be made to fit in/on a replacement relay can, but I really have very little to work with.

Alan

Dave,
I'm skeptical any 2 wire would woirk well enough to be fitting for the car. To me - if it only works if I close the door quickly - its just not good enough (in fact its a pretty crappy design).

The second one you posted probably works OK - but you have to cut wires - right? for mine I would add this after the relay but since you don't have a relay - your input would be the pin switches as in the diagram.

Another option - do you have a spare relay slot on CE? I'm sure there are standard bosch footprint relays that do exactly this - plug & play - just need to wire to it... I know you know how to do that!

I think new Audi's have it (and I'd expect new VW's too) - my new 2006 Ford has it... but I can't find any info - just hope its not all CAN bus controled now...

Alan