RazMan
|
| posted on 9/3/07 at 11:13 PM |
|
|
A question for the electronic guys
I have converted all of my lights (except headlights) to leds and they work very well, consuming very little power and are very bright after a little
experimenting.
The only problem I have is the Evodash warning light for the indicators. Although the indicators flash as they should, the warning light (also led)
just stays on and doesn't flash. I assume that the nature of leds means that the circuit is staying 'high' and needs to be
discharged in some way to allow it to flash. My question is what value resistor should I insert between +ve & earth to make this happen?
Any guidance will be appreciated 
Cheers,
Raz
When thinking outside the box doesn't work any more, it's time to build a new box
|
|
|
|
|
stevebubs
|
| posted on 10/3/07 at 12:17 AM |
|
|
12V / 21W Bulb -> 1.75 Ohms
Assuming that you have 2 bulbs in parallel (Front & Rear) then
1.75 / 2 -> 0.875Ohm
I think Premier Wiring do off the shelf resistor packs for this application...
|
|
|
RazMan
|
| posted on 10/3/07 at 08:08 AM |
|
|
No Steve, I think you misunderstood. I have converted the whole system to leds. The flasher 'relay' is now an electronic flasher
controller (supplied by Premier) so it is not a case of putting a ballast resistor in parallel with the indicators to make them flash at the correct
speed.
Everything works fine except for the panel warning light which stays on constantly as long as the indicators are activated. I assume the reason is
that, because of the low power consumption of leds, the voltage is not dropping low enough between flashes, so it needs to be discharged to the
'low' condition somehow.
Cheers,
Raz
When thinking outside the box doesn't work any more, it's time to build a new box
|
|
|
martyn_16v
|
| posted on 10/3/07 at 08:41 AM |
|
|
You could try a normal relay between the flasher controller and the warning lamp?
|
|
|
RazMan
|
| posted on 10/3/07 at 08:59 AM |
|
|
Yes, I suppose that would do the trick but extra load of the relay coil might overload the electronic flasher unit. iirc it is rated at 9 watts and I
must be pretty close to that limit with my setup - each 'side' consists of 6 x 1 watt luxeon star leds + 48 conventional leds.
Cheers,
Raz
When thinking outside the box doesn't work any more, it's time to build a new box
|
|
|
stig mills
|
| posted on 10/3/07 at 09:40 AM |
|
|
As you may have dicovered a light emitting diode is directionally sensitive and pos must go to the long leg. A 1K resister would give the led about
10milli amps at 12v but you would need one led for left and one for right as the polarity changes when you switch left to right if your trying to put
one across the 2 positives.
Hope this helps, regards Stig
|
|
|
RazMan
|
| posted on 10/3/07 at 11:31 AM |
|
|
Just to clarify, here's a rough circuit diagram. Everything is working fine, so I don't need to change the circuit - its just the dash
lights don't flash on & off.
 
Rescued attachment Indicator Circuit.jpg
Cheers,
Raz
When thinking outside the box doesn't work any more, it's time to build a new box
|
|
|
dave r
|
| posted on 10/3/07 at 12:14 PM |
|
|
cant you just put the dash warning lights on the same connection as the main indicators ?
|
|
|
RazMan
|
| posted on 10/3/07 at 02:18 PM |
|
|
As far as I can tell they are on the same connection, only internally in the Evodash. Actually the repeaters are also not quite right - they
don't quite flash off, more like they go very dim.
Cheers,
Raz
When thinking outside the box doesn't work any more, it's time to build a new box
|
|
|
Confused but excited.
|
| posted on 10/3/07 at 02:36 PM |
|
|
I don't know if you just left them out of your diagram for clarity, but there should be a separate resistor in series with each individual LED,
the value of which is determined by the current required to give the required evel of illumination of the associated LED (see device data sheet).
Going dim rather than off may indicate a bad earth.
[Edited on 10/3/07 by Confused but excited.]
Tell them about the bent treacle edges!
|
|
|
stevebubs
|
| posted on 10/3/07 at 07:18 PM |
|
|
quote:
Originally posted by RazMan
No Steve, I think you misunderstood. I have converted the whole system to leds. The flasher 'relay' is now an electronic flasher
controller (supplied by Premier) so it is not a case of putting a ballast resistor in parallel with the indicators to make them flash at the correct
speed.
Everything works fine except for the panel warning light which stays on constantly as long as the indicators are activated. I assume the reason is
that, because of the low power consumption of leds, the voltage is not dropping low enough between flashes, so it needs to be discharged to the
'low' condition somehow.
Quite right - had misunderstood...
Quick question: If you connect a dash bulb instead of an LED does it all work ok?
If you're short of a bulb to test with, I can bung you a holder in the post...
|
|
|
stevebubs
|
| posted on 10/3/07 at 07:28 PM |
|
|
Second question wrt dash LED - is this an off the shelf LED or one of the dash bulb replacement types?
Off the shelf don't have a resistor in (see above) but the dash bulb replacement ones do...
|
|
|
RazMan
|
| posted on 10/3/07 at 07:41 PM |
|
|
CBE - Yes you are quite right, my diagram just shows the leds and not the internal resistors which are part of the assembly. The Evodash also has
internal resistors to allow them to operate from 12V.
Steve - The Evodash is a sealed unit so I cant retrofit a filament bulb but if I connect a temporary low wattage bulb it flashes correctly, therefore
proving that it is just the warning leds that don't flash for some reason - they stay on until the indicator is switched off.
Cheers,
Raz
When thinking outside the box doesn't work any more, it's time to build a new box
|
|
|
Bob C
|
| posted on 11/3/07 at 01:56 PM |
|
|
Raz - get yourself a 1kohm potentiometer & stick it in parallel with the indicators.
Start flashing & wind the resistance down 'till it works. Then measure the resistance & permanently attatch a fixed resistor with
slightly fewer ohms.
I'm assuming you'll be careful to avoid blowing up your flasher circuit (by taking the resistance to zero) & can work out what wattage
resistor you'll need (W = V^2/R)
cheers
Bob
[Edited on 11/3/07 by Bob C]
|
|
|
RazMan
|
| posted on 11/3/07 at 03:22 PM |
|
|
Excellent idea - so 1K would be a good starting point? Now why didn't I think of that? 
Cheers,
Raz
When thinking outside the box doesn't work any more, it's time to build a new box
|
|
|
02GF74
|
| posted on 11/3/07 at 08:41 PM |
|
|
something is not right with the circuit digram.
firstly the box to the left of the one named "dash" is the flasher unit?
I assume there is 12 V feed on the left, then when the indictaor stalk is flipped, the contact inside this box makes contact to the rail going to the
LEDs so they are fed with 12 V? This would light the dash bulb as per your circuit.
but when the stalk is still flipped and the LED lamps go off, this is when your dash lamp stays on?
and you say that putting a low power bulb in parallel with your LED lamps causes the dash lamp to go off?
I can't see how that can happen with your circuit as the bulb you put in is connected across 12 V feed.
I
[Edited on 11/3/07 by 02GF74]
|
|
|
RazMan
|
| posted on 11/3/07 at 10:33 PM |
|
|
The box is actually the indicator switch which is fed by the electronic flasher unit and starts flashing whenever the switch is flipped.
I am going to play with resistors during the week and see what happens (or explodes  )
Cheers,
Raz
When thinking outside the box doesn't work any more, it's time to build a new box
|
|
|
02GF74
|
| posted on 15/3/07 at 11:09 AM |
|
|
so what happened? fixed it or not?
|
|
|
RazMan
|
| posted on 15/3/07 at 03:16 PM |
|
|
Ah yes, I forgot to post my conclusion
I connected a small dashboard lamp (probably about 1 watt) in parallel with the Evodash warning lights and PRESTO everything flashes exactly as it
should now. I wrapped the dash light in heatshrink and stuffed it out of the way - less hassle than finding an equivalent resistor.
[Edited on 15-3-07 by RazMan]
Cheers,
Raz
When thinking outside the box doesn't work any more, it's time to build a new box
|
|
|
tks
|
| posted on 15/3/07 at 07:16 PM |
|
|
basicly
i guessed that the inputs from the evo dash are transistors (you can check this with yout multimeter)
if they don't see a clear grond (lower as 0,6volts) it won't shutdown your led..
you say that the others lights dimmed
thats because they aren't grounded enough also. if you have it sorted you can choose to let it be this way...
saying that i don't like possible breakdown / more work then doing things right now (if you forget about the bulb and it breaks down it will/can
give you someday allot of work)
anyway the input on the evo dash isn't grounded enough.
your side leds to and thats the reason the reason they light up a small bit...
the reason wy they don't ground enough i cant tell you.
if you want to raise the trigger voltage of the evodash inputs add a diode in series
then you raise the limit with +/- 0,6volts..
regards,
Tks
The above comments are always meant to be from the above persons perspective.
|
|
|
RazMan
|
| posted on 15/3/07 at 10:51 PM |
|
|
Thanks for the suggestions tks - I had not thought of a diode to solve the problem. The bulbs are in a place that I can access easily so I will leave
them as they are for the time being. The important thing is that they now work perfectly and I want to spend some time on other parts of the car (its
never finished is it )
[Edited on 15-3-07 by RazMan]
Cheers,
Raz
When thinking outside the box doesn't work any more, it's time to build a new box
|
|
|
02GF74
|
| posted on 16/3/07 at 11:56 AM |
|
|
quote:
Originally posted by tks
i guessed that the inputs from the evo dash are transistors (you can check this with yout multimeter)
if they don't see a clear grond (lower as 0,6volts) it won't shutdown your led..
......
but doesn't it all depned on what is inside the flasher unit?
If the unit witches between 12 V and open circuit, the open circuit condition should be the same as if the indictaor stalks are not in an indicating
position?
anyway, it has been sorted ... but I;d still be interested to see the full circuit diagram.
If I were dsiging the dash, I'd have LED connected to ground via resistor and have power to it that comes fro,m the same point the power goes to
the flasher lamps themselves, no need for transistors or anyting else, seem very strange.....
|
|
|
MikeRJ
|
| posted on 16/3/07 at 04:31 PM |
|
|
quote:
Originally posted by tks
i guessed that the inputs from the evo dash are transistors (you can check this with yout multimeter)
if they don't see a clear grond (lower as 0,6volts) it won't shutdown your led..
I think it's more likely that they are just LED's in the dash, but that the indicator assemblies have several LED's in series.
Any leakage through the flasher circuit only has to raise the voltage by one LED Vf (about 2-2.5volts) to illuminate the dashboard LED, but at this
voltage the indicators won't turn on.
|
|
|
RazMan
|
| posted on 16/3/07 at 05:39 PM |
|
|
quote:
Originally posted by MikeRJ
I think it's more likely that they are just LED's in the dash, but that the indicator assemblies have several LED's in series.
Any leakage through the flasher circuit only has to raise the voltage by one LED Vf (about 2-2.5volts) to illuminate the dashboard LED, but at this
voltage the indicators won't turn on.
Thats my understanding too. With so many leds connected together (each side has 6 Luxeon Stars & 48 conventional) the dash led wasn't seeing
a complete 0V and it was just enough to prevent it turning off between the 12V pulses. The filament bulb simply discharges the high condition and
allows everything to do its job.
Cheers,
Raz
When thinking outside the box doesn't work any more, it's time to build a new box
|
|
|
tks
|
| posted on 16/3/07 at 07:05 PM |
|
|
mhh its simple,
i gues if you get a 600Ohm resitor 1/4 watt and ditch the bulb that everything is just fine. think its a problem of minimum charge in the flasher
unit.
also you could try disconnecting the dash input (maybe voltage is comeing from there...)
Its clear that if you pull down that line everything works fine..
i gues the charge/voltage leak is comeing from the dashboard.
Tks
The above comments are always meant to be from the above persons perspective.
|
|
|
