Thanks MacGyver. You have been a great asset to this forum and your inventions/wit/advice go a LONG WAY!

Looking forward to your final design.

 

[QUOTE]Originally posted by MacGyver
[B]Hopefully the parts will arrive in the next couple of days.

 

Ding! Ding! Ding! We have a winner modifry, if he so chooses to purchase a lighting kit, will receive 10% off his order. Although I need to make sure it will work properly when several millimeters of plastic and air are between them (shouldn't be an issue), I have mounted a magnet in the base of the screen and a reed switch in the folding mechanism. When the mechanism is up, the magnet closes the reed switch and everything operates as normal. Fold the windscreen down, the switch is pulled away from the magnet in the base thereby opening, top no longer functions.

Technically, I do have to run wires someplace, so I might as well include them in the small bundle that runs into the folding mechanism. I have two choices which I haven't made up my mine on. I can either use a single wire to connect to the reed switch (with the other end of the switch connected to ground) which would require the use of another relay on the circuit board to deactivate the top, OR I could use two wires on the switch and merely have the reed switch inline with the e-brake switch instead of the relay being inline with the e-brake. The first option allows me to use two wires for ground return (a nice thing) and prevents this looooong wire being added into the e-brake switch wiring, but the second option is cheaper since it doesn't require another relay. I'll probably go with option 1.

Bob, no problem at all sharing my circuit design...it's not like someone couldn't open up the box and reverse engineer the damn thing in 2 minutes In a nutshell, though...3 transistors to provide current regulation, 1 for each color. All LEDs for a particular color are in a parallel cluster which is inline with the collector. A 2W resistor is inline with the emitter, set to give 120mA for an emitter voltage of 7.3V. The base is biased with a resistor from Vcc to base, and a trimmer pot from base to ground. The pot's arm is tied to ground, so turning it will bring the base from a max of 8V (@ 10kohm) down to 0V(@ 0ohm).

The brake signal trick is cute, but nothing revolutionary. Two DPDT relays are controlled by the brake signal wire. When activated, the pot's wiper arm on the red LEDs is opened, thereby popping the base up to the full 8V and turning on the LEDs to full intensity. The transistor base of the green and blue LEDs is connected to ground, thereby popping it to 0V and turning off the LEDs.

So total parts count for the control circuit at the moment:
3 transistors
3 potentiometers
6 resistors
2 DPDT relays (brakes)
1 SPDT relay (soft top)

Various other parts:
Multiple LEDs
1 switch (on/off)
1 reed switch (soft top)
1 magnet (soft top)


I would say parts count is pretty slim. Granted, it's still not tested, but in theory it shouldn't be an issue.

 

I'm rereading a few things, and your examples got me wondering...

For example 1, they probably tried some sort of driver circuit. I would find two points...one point that's powered by the headlights, and a seconds that's powered when the car is running or key is in the ignition. Connect the pieze between those two points...when the car is turned off, power flows from headlights to grounded ignition point making the buzzer sound. Seem about right?

Example 2 would take me a bit more time and some paper/pencil, but I would expect some sort of RC network for a simple timer to turn on each SCR. First light starts the first timer, cap is charged ( or discharged) which turns on the first SCR and lights the second light. This starts the second RC timer, time is up, turns on the second SCR which turns on the thrid light. Am I close?

Example 3...I've always been interested in learning how to build high-quality power supplies, but I usually end up buying some premade component off the shelf for simplicity. I suppose if I thought long and hard enough I could come up with a few designs, but without actually putting them down on paper I have absolutely NO idea on component count. You'll have to share on this one.

 

I'll see if I can come up with a way to eliminate some of the relays. As for my examples:

[QUOTE]Originally posted by MacGyver
[B]I'm rereading a few things, and your examples got me wondering...

For example 1, they probably tried some sort of driver circuit.

 

OK, tried out the circuit on a smaller scale...and it WORKS! (3 LEDs at 40mA instead of 6 at 120mA...that little transistor gets warm, but it's supposedly good up to 800mA)

modifry has made some suggestions that I will look into, but after that it looks like things are a GO.

 

Sounds great. I'm looking forward to seeing the finshed mod -- on my car!

 

I have faxed modifry the latest and greatest version of the circuit for a sanity check. It looks good, the circuit seems to work without issue, but having a second pair of eyes is always a good thing.

 

So, Mac, now that Rick's is shipping the newly modified screens that "LED Ready" to work with your design, do you have a rough estimate of when you will be able to start taking orders ?

 

modifry and I went over the schematic night before last night, and I'm glad we did it. I was still in MOSFET design mode but using BJTs (anyone speak Greek? ). Anyway, the design has been finalized, we both agree on it, so I'm happy.

I will begin laying out the PC board sometime tonight or tomorrow (hopefully tonight)...should only take me a few hours from the start. Once I find a board house I'm happy with, the order should take anywhere from 5-15 days to make its way back into my hands.

Once THAT is done, I'll start accepting windscreens.