Forums » General Electronics Discussion

Okay here it is simple but I haven't done any EET since 1983. Can anyone tell me the specs to add into the Led Parameters? All we used in '83 was forward current an min/max voltages. It is also unknown at this time if R6 is needed for Flasher load balancing.

https://www.circuitlab.com/circuit/6z4bxy/turn-signals/

signality

Thanks, one of the links led me to a textbook about the spice calculations. Down loaded it; however. It may not answer all of my questions about the parameters that are listed on the edit led parameters. I will read it and see.

Okay; signality,

Completed I hope, current flowing through each leg of the array is very close to the spec of 120ma @ 3.5 volts. 131ma Total operational R is a bit higher than a 1156 I think a 20 ohm 10watt R6 may be enough.

https://www.circuitlab.com/circuit/t7d4q9/led-turn-signal/

[ cough ] Don’t get me wrong, but any vehicle (to be used on public roads) is the wrong place to play with. You probably play with your life.

R6 with 20Ohm and 10W (+…) in your turn indicator? Which mass? The 12V supply will be up to 14,8V regularly.

See: http://www.superbrightleds.com/cat/tail-brake-turn/

And: http://lmgtfy.com/?q=motorcycle+led+turn+signal

Regards, Sancho

Sancho_P Said

[ cough ] Don’t get me wrong, but any vehicle (to be used on public roads) is the wrong place to play with. You probably play with your life.

R6 with 20Ohm and 10W (+…) in your turn indicator? Which mass? The 12V supply will be up to 14,8V regularly.

No offense taken.

No I do not play with my life. That is why I want it correct before installing. Lighting is the key to visual recognition of a Scoot. R6 may be to high in ohms but not watts. An 1156 turn/marker bulb is 1.5 Ohms.

The voltage supply in the TSSM only releases 12V +/- .15v to the signaling system. My major concern is the current and forward voltages. They need to be balanced for optimum Luminosity! 3.5V 30mA for each Led.

The Cree diode is being used by some makers of your links. In arrays of up to 15 Diodes ( 3 x 5) for brake lights. Extremely Bright!

Would you pay a High Price for something You know you Can Make?

Regards

If you add a ground to the -ve end of V1 in:

https://www.circuitlab.com/circuit/t7d4q9/led-turn-signal/

sweep the voltage of V1 between about 12V and 14.4V and then plot P(R6) in a DC Sweep you'll see that it is quite a high power part.

As I said, depending on the type of flasher unit you are using - i.e. if it is a 2 terminal bimetallic strip unit - R6 has to dissipate the remains of the 21W or so that the original indicator bulb dissipates minus whatever power is dissipated in the total of the LED strings.

Obviously, if the power in the LED strings is relatively small compared to the original 21W then the power in R6 will be high.

Also be aware that strictly there are rules about the exact colour (spectrum) of vehicle lamps so you may have to do some research on acceptable LED colours to meet the vehicle testing requirements in your locality. If it's for a vintage vehicle then you may be exempt but again that may only apply to original - and not aftermarket modified - equipment.

:)

Updated 4/10/13

https://www.circuitlab.com/circuit/xe4ndj/led-turn-signals/

Sorry but now you're being silly.

1) Run the DC Solver

Look at the currents throught the individual LEDs.

Compare against:

ABSOLUTE MAXIMUM RATINGS

in:

http://scn.cree.com/products/pdf/ledlamps/CP41B-WES&WGS.pdf

2) You need to read more about how you feed and control current to LEDs:

https://www.circuitlab.com/circuit/mp673a/two-leds-separate-vs-shared-resistor/

https://www.circuitlab.com/forums/basic-electronics/topic/rt582n4q/need-some-help-on-a-project/

Yes I have that download.

not completed, and The Cree Leds are a four pin Diode. still working on calcs resistors will change I know that. max voltage is 4.4v and reverse breakdown is 5v.

@ 4 volts forward there is about 60 ma forward current. this prog does not allow tested parameters as these are, So this is just a sketch board right now.

Just didn't want you to go off working on a publicised design that was doomed to fail ...

The number of pins on the Cree is not really relevant here: there's only one LED in each package.

Is your 12V really regulated? Or is it like most vehicle supplies: 12V battery up to 14.4V at max generator output?

If it's a stable 12V then you can use lower series resistors than if you have to cope with 12V to 14.4V etc.

If it's can be as low as 12V though you probably can't reliably drive 3 LEDs and a series dropper resistor.

3 * Vf max = 13.2V > 12V

End of.

Say you want the LED current to vary by no more that 10mA (33%!).

Call this delta_Iled.

Let's suppose the 12V supply is 12V to 14.4V.

Let's do the sum for 2 LEDs per string and ignore the Vf variation with forward current and temperature.

Max Vsupply = 14.4 Min Vf_led_string = 6V

Min Vsupply = 14.4 MaxVf_led_string = 8.8V

delta_Iled = 10mA

(14.4V-6V)/Rseries - (12V-8.8V)/Rseries) = 10mA

:.

Rseries = ( (14.4V-6V) - (12V-8.8V) )/10mA

= 520 Ohms

See:

https://www.circuitlab.com/circuit/ga8gz3/delta_iled_01/

But then the max current is only about 16mA. Try to increase the LED current and then delta_ILed increases.

This is why LEDs are usually driven by constant current sources or sinks which are often themselves based on switch mode supplies:

https://www.circuitlab.com/circuit/f646vh/2-transistor-constant-current-sink-01/

https://www.circuitlab.com/circuit/7aavyg/constant-current-sink-led-driver-linear-01/

https://www.circuitlab.com/circuit/63wt5z/constant-current-source-led-driver-linear-01/

https://www.circuitlab.com/circuit/872x9p/constant-current-smps-led-driver-01/

Even so, 10 off LED strings @ 30mA each string is only 300mA so you will need that parallel resistor if the flasher unit depends on seeing bulb load.

I've also just remembered another potential gotcha with LEDs.

An incandescent bulb has a low cold filament resistance so the flasher unit sees an initial very high current as it is connected to the cold bulbs. as they light up the filament resistance rises so the current through the flasher unit drops. Some flasher units may not actually flash if they only even see a roughly constant hot bulb current as defined by the nominal (i.e. hot) bulb power rating of say,

21W/12V

:)

https://www.circuitlab.com/circuit/bwwf8m/diagrm-of-circuit/

This is the problem I'm Working on!

I am considering an lm317(?) Triac(?) Circuit.

Sorry: typo.

Min Vsupply = 14.4 MaxVf_led_string = 8.8V

should be:

Min Vsupply = 12 MaxVf_led_string = 8.8V

"This is the problem I'm Working on!"

Forget implementation details.

Two things you need to establish before you can go further.

i) how does your flasher unit behave with different simple resistive loads;

ii) what does your 12V supply really do? Is it really regulated to 12V? Or is it like most vehicle supplies: from around 12V battery then limited at 14.4V max at higher generator outputs?

Also you need to define what an"1156" bulb is as an electrical load. 12V 21W?

You don't need to go as detailed as "A Tungsten Lamp Model" in:

http://www.intusoft.com/nlpdf/nl11.pdf

but it does give you an idea of how complex a load a filament bulb really is.

signality

Will this work as diagrammed (with low heat) or, is there a different/better way to obtain a constant voltage with a constant current? kasman57

oops 3V will operate/energize the Cree Led

https://www.circuitlab.com/circuit/yjreqe/controlled-voltage-and-current/

You should be starting with something like this:

https://www.circuitlab.com/circuit/7aavyg/constant-current-sink-led-driver-linear-01/

or:

https://www.circuitlab.com/circuit/63wt5z/constant-current-source-led-driver-linear-01/

or simpler still:

https://www.circuitlab.com/circuit/f646vh/2-transistor-constant-current-sink-01/

Start with one of these for a single 2 LED string.

Play and get the hang of it.

"This is a nice simple little circuit:"

Okay got it to run in dc sim but not dc sweep.

Can it be used for 3 Leds in a string (leg) that energize at 3.0v-3.2v. [Used Two (old 1.554V and 1.457V) AA batteries will energize the cree with legs in parallel wire mode.]

Also just add another string for a larger array I presume!

I've edited the example slightly to clarify.

It does run in Time Domain but shows current spikes at startup which are artefacts of CLs solver getting it wrong.

"Can it be used for 3 Leds in a string (leg) that energize at 3.0v-3.2v."

As long as:

V(supply) - V(Q2_C) > n*Vf_led

where n = number of leds

Vf_led = max forward voltage of one led

then Q2 (and it's little friends Q4, Q5 & Q6) won't saturate and so it'll work as a constant current sink.

Note that V(base) ~= 0.6V

(where ~ means aprroximately)

and V(Q2_C)-V(base) >~0.25V for Q2 to not be saturated.

"Also just add another string for a larger array I presume!"

Correct.

" ... will energize the cree with legs in parallel wire mode."

Sorry, don't understand this. As far as I can see from the datasheet:

http://scn.cree.com/products/pdf/ledlamps/CP41B-WES&WGS.pdf

the LEDs have two anode and two cathode pins. The anode pins are connected together and the cathode pins are connected together.

So there's only one way to connect the LED.

I've also reduced R7 to 1M and it now runs OK in DC Sweep.

I tried it with 1M last night and it didn't run then ...

Doesn't inspire confidence.

:(

However .....

You still haven't answered the questions posed in:

https://www.circuitlab.com/forums/optoelectronics/topic/t92e42e5/turn-signals-for-motorcycle/#comment_4905

Until you do you have no idea if your circuit will work in practice.

I have researched and asked for an answer from the Manufacturer, No answer to date.. I also need to take a break for other reasons. Will continue in about 1 week.

See you then.

When you come back, could you not make some measurements yourself?

A voltmeter and a few high power resistors?

P = V^2/R

21W @ 12V = 6R8 21W

15R @ 9.6W

33R @ 5W

68R @ 5W (actually dissipates about 2.1W)

See what the flasher unit does with those loads and sling a LED with a 470R in series across too to see if it actually flashes properly.

Measure the battery voltage engine off and at mid to high revs.

Plot some graphs?

:)

Okay will do!