Dash Dimmer Circuit Project For LEDs

[spoke]

I've been working on a circuit to allow the use of LEDs for dashboard instrument lighting and give complete dimming from 100% down to 1-5%. The circuit will provide a PWM output and will work with LEDs and incandescent bulbs.

Circuit simulations show very good functionality and linearity over the full range of dimmer adjustment.

Since the light switch potentiometer is relatively low resistance, when changing to LED dash lights, the LEDs tend to dim just a bit then turn off. This circuit would provide a ballast resistor to sense the change in resistance of the pot.

One thing I need though is the range of resistance of the potentiometer of the light switch. If anyone has a light switch laying around, I would appreciate a measurement of the resistance from bright to off. It looks like the pins for the pot are 30 (12V power) and 58a (dimmed output to the bulbs). The pins are from the '86 930 schematic. YMMV.

[Canada Kev]

I took an old light switch, probably from the SC era where the knob unscrews from the switch shaft - don't know if that'd make a difference. Measuring with a Fluke DMM and a reasonably fresh battery, I opened and closed the rheostat a bunch of times, measuring each to get some consistency and try to eliminate errors. My measurements are the same regardless of polarity of the meter.

Across pins 58a and 30 with the rheostat fully closed (counter clockwise until the knob just starts to unscrew) I measure 4.5 ohms. Open until the rotation meets physical resistance just before turning on the interior lights, but dash lights still should be full bright, I get 1.1 ohm. After the switch is rotated fully open, I measure 0.1 ohm.

Leet me know if you need anything more.

[spoke]

Originally Posted by Canada Kev

I took an old light switch, probably from the SC era where the knob unscrews from the switch shaft - don't know if that'd make a difference. Measuring with a Fluke DMM and a reasonably fresh battery, I opened and closed the rheostat a bunch of times, measuring each to get some consistency and try to eliminate errors. My measurements are the same regardless of polarity of the meter.

Across pins 58a and 30 with the rheostat fully closed (counter clockwise until the knob just starts to unscrew) I measure 4.5 ohms. Open until the rotation meets physical resistance just before turning on the interior lights, but dash lights still should be full bright, I get 1.1 ohm. After the switch is rotated fully open, I measure 0.1 ohm.

Leet me know if you need anything more.

Thanks Kevin. The largest measurement of 4.5 ohm seems low. Someone measured the pot from a 914 headlight switch and came up with 35 ohm. 4.5 ohm doesn't seem like it would dim incandescent bulbs that much. However depending on how many bulbs are connected, the pot might be that low in resistance.

One question: when you turn the pot to maximum resistance, does it eventually go to infinity? That is, at the high resistance end, the pot should open up which is equivalent to turning off the dash lights totally.

[Canada Kev]

Just checked again, this time using another meter.

The results differed, but mostly fell in line with similar numbers. Switch full counter clockwise, fully dimmed: 5.5 ohms. Fully open, bright: 0.7 ohms.

My switch never goes to infinity, so doesn't completely open the connection. Perhaps there's a design reason for that?

If I were you, I think I'd get some more data points and not rely on just my readings. Again, this is an old switch that I had in my used spares box. I got it from someone years ago because I had a faulty one that needed to be replaced, which I did, but with a new switch that arrived the same time as this one. I've never used this one in my car and really don't know for sure if the dimming works to spec.

[FrankM_]

I will try to measure mine (85 model) switch, but indeed, it never cuts of the dash lights entirely (that would even be illegal and fail MOT here in BE). Will you share your schematic fi it all works ?
It will take some time as I have to get the switch partially out to reach the terminals.

I have the LEDs from carmagic in place and they dim using the standard switch though not 'fully' as you mentioned.

[famoroso]

Interested in what the LED dimmer would look like and how it would mount / install.

Thanks for doing this. I have Carmagic instrument LEDs in both of my cars (just swapped the Bright Whites in my '87 out to the Classic Whites) and I love 'em!

I also have full Spoke LED setups in both cars and feel very strongly that this is a no brainier mod that seriously enlarges the visual footprint of these tiny cars. Safer, more reliable AND less current draw? Win / win / win.

As a matter of fact, I just had Timmy2 whip up an instrument cluster wiring harness for me...

[spoke]

Originally Posted by FrankM_

I will try to measure mine (85 model) switch, but indeed, it never cuts of the dash lights entirely (that would even be illegal and fail MOT here in BE). Will you share your schematic fi it all works ?
It will take some time as I have to get the switch partially out to reach the terminals.

I have the LEDs from carmagic in place and they dim using the standard switch though not 'fully' as you mentioned.

Thanks for the info about not going dark. I thought they did this but not sure about it. I'd gladly share the schematic once it works. It's just a simple triangle wave generator and a comparator to sense the setting of the potentiometer in the headlight switch. It gets complicated when factoring in the pot in the switch may only be 5-50 ohms which is why LED dash lights don't dim much. Sensing 0-5 or 0-50 ohm is not trivial.

[spoke]

Originally Posted by FrankM_

Interested in what the LED dimmer would look like and how it would mount / install.

The dimmer would have 4 wires. Ground and 12V power is needed to drive everything. Then 2 wires would go between pin 48a on the headlight switch and the dash lights. The circuit would just be in a small plastic box tucked in behind the dash.

BTW, that's a nice looking harness. What is the main plug to? I'm not familiar with that part of the 911 wiring.

[Iciclehead]

I tested 3 switches I have on my table.

Switch 1 - 911 612 029 01 resistance ranged from 0.1 to 3.6 ohms
Switch 2 - 911 612 029 01 resistance ranged from 0.2 to 5.2 ohms
Switch 3 - 911 612 029 03 resistance ranged from 0.1 to 215 ohms

Measured using my Fluke, low measurement was just past the first stop (fully on) to the last stop (fully off)

They were a bit unstable and jiggled a fair bit, not much rotation for an increase in resistance.

Hope this helps

Dennis

[spuggy]

Originally Posted by spoke

I've been working on a circuit to allow the use of LEDs for dashboard instrument lighting and give complete dimming from 100% down to 1-5%. The circuit will provide a PWM output and will work with LEDs and incandescent bulbs.

You know you can buy all kinds of PWM modules on Amazon, right?

I bought a boxed one that converts a 0-12V control signal into a 10 or 15A PWM output for about $12. Figured it'd be close enough (I doubt the rheostat output would ever get down to 0V, but I don't need dimmer than a watch dial), but never got around to fiddling with it...

Edit: might even have been this; looks familiar:

[timmy2]

Originally Posted by spoke

BTW, that's a nice looking harness. What is the main plug to? I'm not familiar with that part of the 911 wiring.

Plugs into the single dash lighting harness wire for the dash lights coming from the 6 pin female connector behind the gauges.

[spoke]

Originally Posted by spuggy

You know you can buy all kinds of PWM modules on Amazon, right?

I bought a boxed one that converts a 0-12V control signal into a 10 or 15A PWM output for about $12. Figured it'd be close enough (I doubt the rheostat output would ever get down to 0V, but I don't need dimmer than a watch dial), but never got around to fiddling with it...

Edit: might even have been this; looks familiar:

Yes, I have a few here that I've used on a model railroad club that I belong to in Ambridge, PA. The one you show likely uses an isolated dc/dc converter and uses the industry standard 0-10V (0-100%) control signal.

What I'm trying to do is interface the potentiometer (rheostat) in the OEM light switch. This is the issue. The pot only varies from 0-5 ohm or 0-35 ohm, or 0-200 ohm and is attached to 12V. Such a low impedance is difficult to sense and difficult to know when it has reached its minimum brightness (maximum resistance).

So to sense a 0-5 ohm resistor, say if I want to drop 1V across it, then I'd have to drain 1V/5ohm = 0.2A. Across 14V that resistor would be 14V/0.2A = 70 ohm. The power dissipation would approach 3W (14^2/70). 3W is massive and would require a heat sink or a large board.

[FrankM_]

You could build a wheatstone bridge with 100 Ohm 2W resistors and the rheostat in the lightswitch. It should be able to limit the current somewhat and still give a usable output of about half a volt. Then feed the output into an opamp to get the range you want.

However, the voltage is spikey with a lot of transients when the engine is on so you will have to deal with those too (resistor and tvs diode before your circuit).


Sent from my iPad using Tapatalk

[famoroso]

So, this is a standalone additional dimmer...
Carmagic, LED Lights for Classic Porsches

In contrast to that, you are talking about developing a box that wires in-line between the light switch and the gauge lights to "fix" the factory light switch's inability to adjust the brightness when using LED bulbs?

Can this box add a fuse for the instrument panel light circuit?

Thanks for working your LED magic for these cars. I really appreciate it.

[Tremelune]

I think/hope the goal is to allow LED bulbs to be dimmed down as low as incandescents. I hate how bright LED dash lights are, but it also seems silly to be using incandescent bulbs in 2020, particularly in cars with expensive alternators and a slight predisposition towards fire...

I have yet to see a dimmer on the market that will bring LED bulbs down to the lowest incandescent level, which is where I like them to be in rural areas. They usually bring them down to ~20% (which is bright to me), and then off after that.

[spoke]

Originally Posted by FrankM_

You could build a wheatstone bridge with 100 Ohm 2W resistors and the rheostat in the lightswitch. It should be able to limit the current somewhat and still give a usable output of about half a volt. Then feed the output into an opamp to get the range you want.

However, the voltage is spikey with a lot of transients when the engine is on so you will have to deal with those too (resistor and tvs diode before your circuit).


Sent from my iPad using Tapatalk

Frank,
I'm curious how the bridge would be connected to provide the necessary output. Could you draw something up and either post it here or email it to me? The circuit would have a TVS in front. I do that with all automotive circuits with ICs.
TIA.

[spoke]

Originally Posted by famoroso

So, this is a standalone additional dimmer...
Carmagic, LED Lights for Classic Porsches

In contrast to that, you are talking about developing a box that wires in-line between the light switch and the gauge lights to "fix" the factory light switch's inability to adjust the brightness when using LED bulbs?

Can this box add a fuse for the instrument panel light circuit?

Thanks for working your LED magic for these cars. I really appreciate it.

Yes, trying to work with the OEM dimmer in the headlight switch. The Carmagic dimmer likely has done away with the OEM dimmer and provided it's own pot to control brightness. If the dimmer in your OEM headlight switch is not functional then the external dimmer is a good choice. If the OEM dimmer is in good shape then I'm trying to find a solution for that.

I could add a fuse for the the instrument panel light circuit or just current limit the output.

[Iciclehead]

Originally Posted by spoke

Yes, trying to work with the OEM dimmer in the headlight switch. The Carmagic dimmer likely has done away with the OEM dimmer and provided it's own pot to control brightness. If the dimmer in your OEM headlight switch is not functional then the external dimmer is a good choice. If the OEM dimmer is in good shape then I'm trying to find a solution for that.

I could add a fuse for the the instrument panel light circuit or just current limit the output.

I like the idea of not adding a separate knob for dimming.....and in addition adding the fuse is a really good thing, but please do not have an inline fuse, if you are adding a little box for all the wizardry then having the fuse built into that is a better solution as there are enough wires floating around in that area.

Very interested in this....sounds like a wonderful solution to a problem....

Dennis

[FrankM_]

Quick and without testing anything, I was thinking of this (powered by a TVS protected DC/DC converter you can get off Amazon cheaper than what you can build yourself) :

This is by no means meant as a final design, and yes the circuit with the light switch and the 47 Ohm resistor will use some current and produce 4,2 Watts of heat at 14V approx, so it could be optimised. (This is where the light bulbs would normally be in the circuit).



You can play around with the resistors around the opamp to get other gain.