Honda CRX Forum banner
1 - 20 of 35 Posts

·
Registered
Joined
·
1,492 Posts
Discussion Starter · #1 ·
Ok .. so, i'm going to attempt glowing gauge needles :) ..

I orders some SMD ( surface mount diodes ) basically tiny LED's ( light emitting diode )
I ordered them without really knowing just how small they are and now is see ..
:shock: they are TINY ! .. so .. the good news is .. they will definitely "fit" under the needle .. haha
the bad news is .. ( or may be ) .. will it be enough light ?

From what i understand the forward voltage has to match and connecting several LEDs in parallel with a sum near or equal to the load voltage ( 12-14 volts ).
So .. from the data on the graph at this link ..
http://www.ebay.com/itm/10pcs-Red-1206- ... 713wt_1165

2.3 x 5 = 11.5 .. nice right ? so .. 5 of these should in one circuit should be close to the load of one typical incandescent lamp .. right ? 6 would be 13.8 , would that be better ?
are either of those so close that a resistor wouldn't be needed ?

I don't think i ordered enough ..
I'm thinking 5-6 for Tach & Speedo and 2-3 each on the Fuel Temp ..
( i can get away with only 2 on the temp because .. eh .. it's never cold right ? .. lol )

anyways .. i have some time to figures it out .. the SMD's wont arrive from Hong Kong for another week and the gauge needles are coming from Bulgaria .. lolololol .. so .. yeah .. got a little time to work it out ..

All help sincerely appreciated .. :)b
 

·
Registered
Joined
·
565 Posts
Jeff,

Here is the datasheet for the 1206: http://media.digikey.com/pdf/Data%20She ... 150CKT.pdf

Just putting them in series and adding the forward voltage drops together is not a very good design technique because you can't really guarantee the amount of current that will go through the LEDs. Here is a better way to design what you want to do:

The max DC current allowed is 40mA, but the specs are all centered around 20mA, so that is where the LEDs are really intended to operate. This will also give you a safety margin in case of voltage variations that are common in automotive applications.

While a car's battery is nominally 12 volts, it falls well below that when the starter is cranking, and goes well above that when the alternator is charging the battery. The voltage can typically run up to 15V while the engine is running. For design purposes, I design automotive circuits to survive at least 18V. In fact, for circuits that go into production, I add a schottky diode such as a 1N4148 to cause the circuit to survive someone putting jumper cables on backwards, and I add an 18V TVS (transient voltage suppressor) to suppress voltage spikes generated from things like the battery getting unhooked while the alternator is spinning.

For a 1-off circuit such as the one you want to make, you don't really have to add those protection devices, unless you can't afford to make another circuit should one of these events happen. TVSs are kind of hard to find, and the 1n4148s aren't really needed, since your circuit is made from LEDs (diodes). I just include them FYI.

For the basic circuit you want to make, I would design for a 15V supply, with 20mA running through each diode. The forward bias voltage at 20mA is 2.3 max. So for a circuit with one diode, you would subtract 2.3V from 15V to get 12.7V. From Ohm's law, V=I*R, or R=V/I we know that the current limiting resistor you will need is R=12.7/0.020=635 Ohms. So you would want to pick a current limiting resistor that is the next size up from 635 ohms. To check the wattage you will need, you use the power equation P=V*I=12.7*0.020=0.254W. So to be safe, you should use a 1/2W resistor.

If the voltage is lower than 15V, you are safe. If it goes up to 18V, you will get 18-2.3=15.7 across that 635 Ohm resistor, which gives you 25mA, which is well below the 40mA max.

If the light from one diode is not enough, you can run more than one in series. For example, if you run two in series, the calculations would be as follows:

Subtract 2*2.3=4.6V from 15V to get 10.4V. To get 20mA through the LEDs, divide 10.4V by 0.020A to get 520 ohms.

For 3 LEDs in series, subtract 3*2.3=6.9V from 15V to get 8.1V. To get 20mA, divide 8.1 by 0.020 to get 405 Ohms.

For 4 LEDs in series, subtract 4*2.3=9.2V from 15 to get 5.8V. To get 20mA, divide 5.8 by 0.020 to get 290 Ohms.

For 5 LEDs in series, subtract 5*2.3=11.5V from 15V to get 3.5V. To get 20mA, divide 3.5 by 0.020 to get 175 Ohms.

For 6 LEDs in series, subtract 6*2.3=13.8V from 15V to get 1.2V. To get 20mA, divide 1.2 by 0.020 to get 60 Ohms.

I would not go above 6 LEDs in series, since your required current limiting resistors will fall below 0 Ohms, which is not possible. If you need more than 6, you will need to run them in parallel.

To run them in parallel, you multiply the amount of current used by one LED by the number of LEDs you will use. For example, for two diodes, you would do the following calculations:

Multiply 20mA by 2 to get 40mA or 0.040A. You will get the same 2.3V drop across all LEDs in parallel, so to get the current limiting resistor, you will divide 12.7V by 0.040 to get 317.5 Ohms.

For 3 LEDs in parallel, you would divide 12.7V by 0.060 to get 211 Ohms, and so on. To calculate the resistor power dissipation, you multiply the voltage by the current. For 3 LEDs, you would calculate 12.7*0.060=0.720W, so you should round up to a 1W resistor. While the wattage of the resistors in the circuit where the LEDs are run in series goes down, the wattage in circuits where the LEDs are run in parallel goes up, so you need to recalculate the wattage for the current limiting resistor for every case.

So you will need to run some tests to see how much light you want in order to see how many diodes you will need.

You will also need a substrate upon which to mount those SMD devices. They are really tiny. I'd recommend that you hack up a prototype such as this one from Radio Shack: http://www.radioshack.com/product/index ... Id=2104052

Or this one: http://www.radioshack.com/product/index ... Id=3173937

For SMD semiconductors, you will want a steady hand and a very small soldering iron. I use this Radio shack unit: http://www.radioshack.com/product/index ... Id=2062728

Be sure to use rosin core solder, not acid core solder.

HTH
 

·
Registered
Joined
·
1,492 Posts
Discussion Starter · #3 ·
Thanks for the examples Mark.
That's just what my ADHD Dyslexic eyes needed .. hehe ;)

I hope running them in parallel works out better.
It would be much easier in regard to mounting.

2 rings of copper wire would do the trick , tack glue the copper wire rings to some thin plastic after soldering on the SMDs ,
done :)
I could do pretty much the same thing for series mounting but the copper wire would have to be
"woven" in and out to make 2 circles of offset copper dash marks ..

I'd have to find something to take the heat of the soldering but still be thin and non-conductive.
I was thinking of melting silver solder and "painting" traces on some plastic similar to a dinner place mat for kids.
Ya know .. kinda like the plastic circuits on back of the cluster ;)

I just have one hope ...
I hope the best way and the easy way are the same thing .. lolol
 

·
Registered
Joined
·
1,492 Posts
Discussion Starter · #5 ·
CrxMusix13 said:
markolson said exactly what I was going to say! :rolleyes1:

No, but this is really good info...

This should be interesting to see Jeff! Keep us updated!
haha ... yeah everybody was just about to say the same thing as Mark ..
mark is just faster :rolleyes1: :rolleyes1: :rolleyes1: LoL :mrgreen:

Yeah some guy on ebay has these "innovative" new rings for Lexus but he obviously doesn't speak english as his first language. He's not getting that even though my CRX doesn't have the "required" needles .. I'm adding the "required" needles .. :doh: ..
He says it's some carbon fibre ring with a special coating .. no pics .. :( ..

anyways .. i'm pretty excited about it .. I'd also like to take the clear plastic lens out of the cluster and replace it with painted glass , like the one in the photo in my thread about changing the needles ..

ok .. well stay tuned .. :biggrin:
 

·
Registered
Joined
·
565 Posts
I just re-read the data sheet, and they say that hooking up the LEDs in series as I recommended in the first half of my post above may give inconsistent brightness from the different LEDs. So hooking them up in parallel may give you more consistent light output from LED to LED.

If you could run an experiment where you first hook them in series and then hook them in parallel, at least I would be interested in hearing what you find. Might be hard to do without soldering, though...
 

·
Registered
Joined
·
3,879 Posts
markolson's analysis of resistor calculation is correct but there are additional considerations.

With this part the number 1206 is the package size not a part number. 1206 is a surface mount package that measures 0.120 x 0.060 inch. There are probably hundreds of different red LEDs available in a 1206 package. I don't see anything to indicate the referenced data sheet is for the same LED listed in the ebay link.

The ebay listing gives a forward voltage and light output at 20mA confirming 20mA as a safe operating current.

Due to variations in manufacturing LEDs do not share currently equally when connected directly in parallel (sharing the same resistor). This is what the referenced data sheet says when you sort out the grammar. Unequal sharing of the current results in unequal light output.

The more LEDs you string in series the greater variation you will see in light output with slight variations in voltage. With each LED having it's own resistor there would be minimal variation in light output over the range of 12 to 15 volts that would be expected during typical automotive operation. With 6 LEDs in series the LEDs may not even light until the engine is running and would vary intensity greatly with alternator load and engine RPM. Another advantage of keeping the LED strings short is for better intensity control with dash light dimmer.

For your application I would suggest series strings of 2 or 3 LEDs with a resistor. Connect as many of those strings together in parallel as you need to reach your desired number of LEDs.

Test a string in the car with the dimmer control to confirm the number of LEDs per string and resistor value before you commit to wiring all your gauges.

Mike
 

·
Registered
Joined
·
1,492 Posts
Discussion Starter · #9 ·
Thanks guys .. :)b :)b

I wonder if we can predict the light output within some degree of accuracy?

The data for the light output is available in the ebay chart.
Do we know the typical output of an incandescent lamp?

I know it would be an estimate but
it would be nice to have some goal and/or realistic expectations going in to this.

The suspense is killing me waiting on the needles to arrive. :cry:
when the needles arrive I just LED's & battery to test light output .. :)
I really don't think it's going to take very much to light the actual needle.
Creating an even "wash" around the bezel will probably be the hard part ..
 

·
Registered
Joined
·
565 Posts
If you can, make a jig and experiment with it. Sometimes you get lucky and they all are close enough that you don't notice differences. Modern semiconductor processes are pretty tight. and if they all came from the same manufacturing lot, they are probably very close.
 

·
Registered
Joined
·
1,492 Posts
Discussion Starter · #11 ·
markolson said:
If you can, make a jig and experiment with it. Sometimes you get lucky and they all are close enough that you don't notice differences. Modern semiconductor processes are pretty tight. and if they all came from the same manufacturing lot, they are probably very close.
yeah and these lights are 120º beams so, all of the "wash" light will be reflective.
Heck, maybe i won't get any wash .. haha .. that would be great :)
like you say .. you just gotta wire it up and "see what it does" ..
some wise man once said .. "hey, what's this button do" .. lol

EDIT: lights arrived
This is a pic next to a few grains of rice :shock:


100_6595 by jeffwilliamsusa, on Flickr
 

·
Registered
Joined
·
1,492 Posts
Discussion Starter · #12 ·
well i put this off for a while but i'm back on it now ..
I found some great info at the xtuners.com website.
They have a step by step CRX2 install .. pics only .. no text .. :(
anyways .. i also found a nice pic of the unit they are using ..
Seems they are using groups of 2 LED's
in the following photo , you can see they use a inner and outer ring.
We'll call the inner ring Negative for now but it could be either way depending on the direction of the LEDs in circuit .. ok .. so
The outer ring has a Positive connection with 2 LED's to the left & 2 LED's to the right.
The ends of these 2 LED chains are followed with a resistor which leads to the inner ring and thus back to ground.

I did wire up 4 SMD tonight and i think i blew them up immediately .. lol
first off my dang power supply is putting out 24v :shock: :oops: ..

I know the 24v deal was part of the problem and no resistor .. :oops: ..
I thought i had 12-14v and didn't think that would burn the LEDs with just a momentary contact .. ( but i was wrong .. lol ) .. i did see them light up for a second .. haha

Now that i have gotten a closer look at these needle light units , i need to make a new assessment ..

so .. question ..
** Do you think all 12 LEDs are in one circuit ? ( 4 tach 4 speedo 2 fuel 2 temp )

As mentioned above, i think these groupings of 2 w/ resistors will have less light variatio with different voltages ..

Ok well anyways .. here is the pic of the unit .. need to figure a resistor size
 

·
Registered
Joined
·
1,492 Posts
Discussion Starter · #13 ·
HaH ! check it out .. a nice calculator ..
http://led.linear1.org/led.wiz

I'm not great with electronics ... so i could use a little help here ..
If each group of 2 LED's is operating as one typical lamp .. then ..
the calculator says i need 1/2w 470ohm resistors

Does this sound right ?

In effect, am i creating 6 groups of 2 that will draw the same current as 6 typical lamps ?

anyways,
I found this cool circuit board material in my electrical junk box that worked perfect :)
I just need to order some more SMDs & the proper resistors and i'm in business :)

EDIT: eh .. oops .. mark said this already
For 2 LED's in Series Subtract 2*2.3=4.6V from 15V to get 10.4V. To get 20mA through the LEDs, divide 10.4V by 0.020A to get 520 ohms.
So .. the calculator suggests using 13.4v for automotive and i calculated at 2.1 instead of 2.3max which explains the different answer i got above.
the calculator said 560ohm at 15v and 470ohm at 13.4 .. so .. hmm
what should i do ? split the difference ? calculate at 13.4 ? 15 ?

EDIT: i zoomed in on a few of the pics from the xtuners forum and it appears as though he's using 620 ohm resistors ( ofc he could also have a diff forward voltage ) .. i found one pic that clearly had 21 as the last numbers. Using the following link , looking at the last part of 4th row , you can see there are only 2 values with 21 at the end 821 & 621 .. 820 ohms would be way too much so methinks it IS a 621 / 620ohm resistor ..

NOTE: i didn't see ANY of these go beyond 1/4 watt .. the calculator said 1/2 watt so .. eh .. ???

ALSO .. geez :oops: sorry for the spreading the info over 2 replies and kinda out of order :oops:
my dyslexia is worse when i'm tired .. hehe :rolleyes1:
ANYways .. please read back through the last few posts to get all my info and questions ..
 

·
Registered
Joined
·
565 Posts
Those rings have the resistor integrated into them, so you don't need to add a resistor externally. If you can read the number on the resistors, I can tell you their value.

Putting 24 volts on the circuit would burn them out even with the resistors. You'd have to add another resistor to the circuit if you wanted to run 24V.
 

·
Registered
Joined
·
1,492 Posts
Discussion Starter · #15 ·
markolson said:
Those rings have the resistor integrated into them, so you don't need to add a resistor externally. If you can read the number on the resistors, I can tell you their value.
erm .. yeah kinda , did you read above where i talked about being able to read the last 2 numbers on the chip ? and i post a link to a chart where i discern / conclude there are only 2 choice 821 or 621.
So , i'm almost certain he's using 620ohm resistors.

Putting 24 volts on the circuit would burn them out even with the resistors. You'd have to add another resistor to the circuit if you wanted to run 24V.
hahahahaha YeS ! i know that ;) my regulated power supply must have a fault. I've had it for years and it's always put out a steady 14v. I'm not sure why it's suddenly giving me 24v ???

Anyways .. the big question here is:
why are our calculations showing smaller resistor values ?
Is this designer putting this higher value for some benefit gain ? more control of dimming ? better bell curve on the dim control ? More protection to the LEDs ?
OR .. is it JUST because he's using a different SMD with a different forward voltage ?

Reading above, you'll see there is also some question of the supply voltage starting point for use in the calculation.
As i change the starting voltage the wizard calculates 60% load for the resistor and changes the wattage value accordingly. Well at about 10v the resistor value changes to 1/2w ( the 60% point )
So .. I think i could get away with a 1/4 watt resistor .. right ???

I don't know why the other guy used 620ohm resistors, i could only get that in the calc by raising the voltage WAY up ( 15+ )

So between 470 , 560 & 620 ohm ..
i'm leaning toward the 560's - 1/4watt ... think i'll be ok w/ that ?
 

·
Registered
Joined
·
565 Posts
Smaller resistor values will give you more current through the LEDs. Larger resistors will give you less current.

Different LED have different power ratings, so you really need to know the specs of the LEDs to know for sure. My guess is that the LEDs in the ring are not 20mA LEDs, so he put in a larger current limiting resistor.

Depending on the material the LED is made of, the forward bias voltage drop can also be different, as much as 1.6V or more or possibly a lot less.

Assuming the drop is 0.6V, and there are two of them, so the drop is 1.2V, and assuming a worst case battery voltage of 16V, then the drop across the current limiting resistor is14.8V. Assuming the resistor is 620 ohms, then the current will be 14.8/620=24mA. Sounds about right to me.

You can measure the LED voltage drop with a voltmeter and then calculate what the actual curent flow is. The larger the current flow, the brighter the LED, until it burns out.

Hopefully, I made that clear.
 

·
Registered
Joined
·
1,492 Posts
Discussion Starter · #17 ·
markolson said:
Smaller resistor values will give you more current through the LEDs. Larger resistors will give you less current.

Different LED have different power ratings, so you really need to know the specs of the LEDs to know for sure. My guess is that the LEDs in the ring are not 20mA LEDs, so he put in a larger current limiting resistor.

Depending on the material the LED is made of, the forward bias voltage drop can also be different, as much as 1.6V or more or possibly a lot less.

Assuming the drop is 0.6V, and there are two of them, so the drop is 1.2V, and assuming a worst case battery voltage of 16V, then the drop across the current limiting resistor is14.8V. Assuming the resistor is 620 ohms, then the current will be 14.8/620=24mA. Sounds about right to me.

You can measure the LED voltage drop with a voltmeter and then calculate what the actual curent flow is. The larger the current flow, the brighter the LED, until it burns out.

Hopefully, I made that clear.
YeS ! Mark .. that helped very much ..

Ok .. so i think i'll do something like this ..
470ohm on board and a 200ohm variable resistor at the termination point on a small board that i'll mount in a place that it can be adjusted without pulling the cluster.
I'll have a total of 3 variable resistors ( tach, speedo, fuel-temp )
This will also allow the 2 resistors to share the heat and allow me to use 1/4watt resitors.
** BUT: does that mean the 470 will be running closer to 100% or will the 2 resistors simply share that load equally ?

ALSO: dang , measuring the voltage drop was a GREAT idea, i'll definitely do that .. :)b

** Finally .. one Last Question .. ( i promise .. lol )
Could we include a capacitor in the + supply to prevent any voltage spiking ?
 

·
Registered
Joined
·
565 Posts
Sorry, I don't have time to study this thread. Are you trying to be able to trim the brightness of each of the modules (tach, speedo & guel/temp)?

When you say, "470ohm on board", does that mean you are planning on replacing the 620 ohm SMT resistor on the board?

I would not replace the on-board resistor. I would just add a small pot in series with each module. The on-board 620 ohm resistor will limit the current to the max allowed for the LED and the pot would reduce that further to dim the module.

1/4 W is probably fine, but it would be nice to know the current that the 620 ohm resistor allows to flow through the LEDs, and the voltage drop across the LEDs before you can calculate the power dissipation.

A cap wouldn't hurt anything, but you probably wouldn't notice variations in brightness caused by noise on the power lines. If you are worried about a power glitch burning out the LEDs, you are better off putting an 18V transient voltage suppressor across the power and ground for the cluster illumination.
 

·
Registered
Joined
·
1,492 Posts
Discussion Starter · #19 ·
markolson said:
Sorry, I don't have time to study this thread. Are you trying to be able to trim the brightness of each of the modules (tach, speedo & guel/temp)?

When you say, "470ohm on board", does that mean you are planning on replacing the 620 ohm SMT resistor on the board?

I would not replace the on-board resistor. I would just add a small pot in series with each module. The on-board 620 ohm resistor will limit the current to the max allowed for the LED and the pot would reduce that further to dim the module.

1/4 W is probably fine, but it would be nice to know the current that the 620 ohm resistor allows to flow through the LEDs, and the voltage drop across the LEDs before you can calculate the power dissipation.

A cap wouldn't hurt anything, but you probably wouldn't notice variations in brightness caused by noise on the power lines. If you are worried about a power glitch burning out the LEDs, you are better off putting an 18V transient voltage suppressor across the power and ground for the cluster illumination.
Thanks ..
When i say "on board" i mean an SMD on the circular wafer that will be installed under the gauge needles. They have to be really thin to fit so i can only use SMD resistors "on board"
NOTE: I do not have the item in the photo. I'm not removing the resistor, i'm building one "like it" from scratch.
I will have 4 SMD LED's on the speedo / Tach , in 2 circuits w/ limiting resistor at ends of circuits
the fuel-temp are 2 SMD - 1 Resistor - 1 circuit ( tied together at the terminal block where the variable resistors will reside )

Ok so .. maybe i'll go a bit bigger on the variable resistor so i can adjust
 
1 - 20 of 35 Posts
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top