86 951 Driver

I am having LED bulb issues again. This will be my second set where 1 or more LED bulbs start to flicker. I'm not sure if they can't handle the duty or what. Anyone else have this problem with the LED bulbs from superbrightleds.com?

Ryan_U

Are you using straight leds plugged into the sockets? I've seen a couple instructions here that just use leds with a resistor and that is not the best. Using a lm317 chip that regulates the amperage (~10-20 mA) would be better since a car voltage can vary between 12.x and 15.x and cause the leds to receive to much voltage. The only problem with a LM317 is that it won't work with the dimmer. I made a test circuit with a two position switch that allows two different intensities. I could also draw up something with a potentiometer that would give a greater range of intensity but it might have to replace the original dimmer switch. I don't have a 944 yet so I can't test this on a car. Maybe the original dimmer switch could be used but you would have to rewire some things. Also, depending on the colour used, you can only put a certain number of leds on the same chip. There is an automotive version of LM317 that works at lower temperatures and should be used in this application but it is a little more expensive. As soon as I get my 944 (or if someone wants to send me a cluster and a switch to work on) I will make a circuit.

86 951 Driver

http://www.superbrightleds.com/cgi-b...%2FWLED-x5.htm

Those are the LED's I'm using. I am not using anything but the LED bulb in the socket. Whats weird is that usually one or 2 will fine. This last time 2 are flickering.

VWaddict

Yep. I posted a complete how-to some time ago covering an LED upgrade which does away with the sockets.

The issue with the sockets (and possibly your LEDs) is that the LEDs don't draw as much current as the incandescent bulbs do. -Contacts will oxidise over time, but the bulbs actually draw enough current; called "Wetting current". http://en.wikipedia.org/wiki/Wetting_current -The typical definition is used for switch contacts and relay contacts, but is also equally applicable to socket contacts.

For that reason, LEDs are best soldered in.

The circuit described by Ryan is known as a CCS (Constant-current source) and works VERY well at avoiding the fluctuation caused by battery/alternator voltage variation (engine running versus battery operation, for example) although incandescent bulbs also exhibit this. LM317's can easily be configured to vary the current if a dimmer option is required or -alternately- can be driven by a duty-ratio driver which still allows constant-current on-state operation while allowing the driver to dim the display... but that requires a little modification and ingenuity.

LM317's can be configured as voltage regulators OR constant current sources... either offers a way to dim the LEDs if required.

VWaddict

...actually Ryan, I was considering a duty-ratio pulse dimming circuit for the dash, and I bought a spare rheostat to play with, but the resistance of the rheostat is (naturally) very low, which means that you need a bunch of current to make things happen... or you use it in a precision bridge... Meh.

I was thinking about re-purposing a pot from a 'wiper delay' control, if it will fit in that same slot. -I'm pretty sure that the resistance will be much better suited for low-current designs... -Could be used to swing the mark-to-space ratio of a simple 555V chip or similar... then that whole thing could become a module which just 'drops-in' in place of the present dash dimmer, and makes the whole thing less load-dependant than a rheostat. -What think you?

-Dug up my thread from a few years ago:
https://rennlist.com/forums/944-turb...e-writeup.html

Keith

86 951 Driver

I get nervous about soldering LED's to the gauge cluster. Thats why I like the socket type. Is there anything i could to keep using the socket type and get decent life out of them?

Ryan_U

I was wondering what resistance the dimmer switch had and if it was possible to use it as a variable resistance for the LM317. A duty-ratio pulse dimming circuit is better efficiency wise but harder and more expensive to build. I don't mind the voltage drop caused by the LM317 because You can still use three LEDS with a forward voltage of 3.5 and still be within operating range of the circuit. The milliamps of lost current don't make a big difference on a car battery and the chip doesn't generate very much heat. Nice project though. What I am wondering is if you could take apart the dimmer switch and convert it to a 3 step pot.

Ryan_U

I wonder if you could use acid paste to clean the connectors and then use some electrive conductive grease to prevent oxidation.

mytrplseven

TMI!! I think I'll just clean up the OEM plastic and glue some silver tape behind like I did on the last two 944's I had.

VWaddict

The resistance is only a few ohms, and it's a 2-terminal variable resistance, so I was thinking that the simplest way to do a variable linear voltage regulation might be to do a simple 'Wheatstone'-type bridge with another resistance of the same value as the rheostat's maximum, and a gain circuit after, to 'magnify' the small voltage (after the drop from the larger resistances in the bridge), if you follow me.

Then a simple emitter-follower should work acceptably, though you'd always be losing 0.6V across the base-emitter junction.

A 555 timer setup to run in an astable configuration should work as a PWM-based dimmer, and could saturate an open-collector drive transistor for full voltage, and drive LEDs and incandescent bulbs alike fairly easily... Mind you, I say all of this without actually having ever actually BUILT one yet!

Ryan_U

I guess that if you want to use the stock dimmer, you would have to use a circuit that compares the drop and then boosts the difference to a workable level. It's starting to be a complex circuit for my level of expertise. I also wonder how the dimmer works to dim standard bulbs with only few ohms. I'd probably fashion a two position switch that would vary between two resistors to get a hi/lo setting or else build a rheostat with maybe five stops. I've run tests on a sample circuit and after a while you don't see much of a difference in brightness and light dispersion becomes a problem to some degree at lower settings. Also, all of these circuits have to be adapted to whatever leds you are using because some are more efficient than others and will put out more light for the same current. What I would really like to do is build a rheostat that would switch between coloured or white leds. Something like Pos1=White HI Pos2=White LO Pos3=Red HI Pos4=Red LO ect...

I might buy a used dimmer to see if it could be modded. I would prefer a stock look. Maybe there is another switch that can be used that comes from a compatible source.

pettybird

same issue here. mine worked fine for ~5 months and now like two of them are intermittent.

funny because I use others in a couple of my other cars with no issues...

VWaddict

The only other thing that might fit is the intermittent wiper delay control... I know of no pot/switch can easily be adapted to fit in there.

I might fiddle with a bridge circuit... If the circuit works, it could be built as a 'drop-in' where the incoming power and ground is used to power the circuit, and the output is then sent to the bulbs in the same way.

It'd definitely be more efficient and comparable in terms of brightness to use PWM, also less load-sensitive, but I'd have to make it work first.

86 951 Driver

What have you done about it? Its irritating for this to happen. There has to be a solution.

944meister

Had no idea such LED nerditry existed, Srry pls continue.....