View Full Version : LED battery light?
Everywhere I read about LED conversions, the one thing there is a consensus on is that the battery light must remain an incandescent or the charging system will not function properly. I've read it has to do with the resistance of the bulb, and LED's have next to zero resistance compared to their incandescent counterparts. As such, they will not work as a drop-in replacement. This, however, raises two questions in my mind:
1) If resistance is the only limiting factor, couldn't a resistor be added to the circuit to give it the extra resistance required for proper function?
And
2) If resistance matters at all, does that mean there is always some current going to the bulb that would need filtering with LED's?
I've seen electronic projects of all kinds coming from the electrical engineers on these forums, surely allowing an LED to be used in the battery light is a walk in the park. Everyone seems to have always advised against it though, and there's no speculation that I could find on it being possible to do. The benefits may not outweigh the work involved, but it's hard to imagine that when completely re-engineering everything else in the car, that is the one item that always remains unmodified. Is this actually doable, even if unnecessary? I'm interested in doing this myself, having an upcoming binnacle replacement and LED upgrade ahead :)
DMCMW Dave
05-02-2013, 12:05 AM
Everywhere I read about LED conversions, the one thing there is a consensus on is that the battery light must remain an incandescent or the charging system will not function properly. I've read it has to do with the resistance of the bulb, and LED's have next to zero resistance compared to their incandescent counterparts. As such, they will not work as a drop-in replacement. This, however, raises two questions in my mind:
1) If resistance is the only limiting factor, couldn't a resistor be added to the circuit to give it the extra resistance required for proper function?
And
2) If resistance matters at all, does that mean there is always some current going to the bulb that would need filtering with LED's?
I've seen electronic projects of all kinds coming from the electrical engineers on these forums, surely allowing an LED to be used in the battery light is a walk in the park. Everyone seems to have always advised against it though, and there's no speculation that I could find on it being possible to do. The benefits may not outweigh the work involved, but it's hard to imagine that when completely re-engineering everything else in the car, that is the one item that always remains unmodified. Is this actually doable, even if unnecessary? I'm interested in doing this myself, having an upcoming binnacle replacement and LED upgrade ahead :)
The circuit needs a bidirectional resistor in it. A diode is not bidirectional. You might be able to bypass the LED with a resistor but then the LED probably would not light up. Why is this such a huge deal? In normal circumstances this light is on for a moment when you start the car, so it's not like it is contributing to the huge (:tongue_stick:) current consumption of the dash lighting system. I've never seen one of those bulbs burn out. When they don't work it's usually the socket that is the problem.
Starglider
05-02-2013, 03:32 AM
The circuit needs a bidirectional resistor in it. A diode is not bidirectional. You might be able to bypass the LED with a resistor but then the LED probably would not light up. Why is this such a huge deal? In normal circumstances this light is on for a moment when you start the car, so it's not like it is contributing to the huge (:tongue_stick:) current consumption of the dash lighting system. I've never seen one of those bulbs burn out. When they don't work it's usually the socket that is the problem.
A bridge rectifier with parallel resistor would probably work, but I didn't think it was worth the hassle.
Tillsy
05-02-2013, 04:03 AM
A bridge rectifier with parallel resistor would probably work, but I didn't think it was worth the hassle.
You would also need to cross-connect the phase inducers to provide a regenerative power source... ;)
Bitsyncmaster
05-02-2013, 04:34 AM
Probably a low value resistor and an LED with a diode would work. One good reason to do an LED there is that you can solder everything in place and not have the lamp socket anymore. But since the bulbs never burn out, you could also just solder it in also.
LordFly
05-02-2013, 06:47 AM
I like Shep's idea (especially since I just filled my binicle with LEDs a few days ago). Sure there is no NEED to have it be an LED, but that doesn't mean some of us don't WANT one in there anyways. Do any of us really NEED a stainless steel car? Of course not, we just WANT one. I'll have to put some research into the whole bulb need for charging thing, pretty sure I can make something work.
Mark D
05-02-2013, 11:40 AM
You would also need to cross-connect the phase inducers to provide a regenerative power source... ;)
haha... I think what you meant to say was that you'd need to employ a drawn reciprocating dingle arm to reduce sinusoidal depleneration.
http://en.wikipedia.org/wiki/Turboencabulator
The circuit needs a bidirectional resistor in it. A diode is not bidirectional. You might be able to bypass the LED with a resistor but then the LED probably would not light up.I was actually thinking more along the lines of cutting one of the wires and sticking a resistor in the middle, rather than using the LED as a resistor itself or bypassing it entirely. An LED will still pass current through regardless of which direction it passes in, but it will only light up in one direction. I'm not sure where the unidirectional part factors in.
I've never seen one of those bulbs burn out. When they don't work it's usually the socket that is the problem.Are you sure about that? The bulbs are 32 years old, there's no way they have a 0% failure rate. There's at least a dozen threads documenting problems owners have had with a blown battery bulb. Steve even talked about this at the NC Tech Day a few weeks ago, the video of which was posted in Post #119 (http://dmctalk.org/showthread.php?6327-North-Carolina-Tech-Event-(Spring-2013)-Saturday-April-20th-Featuring-DPI&p=100709&viewfull=1#post100709) of that thread. I honestly have not heard of the socket being a problem before :hmm:
Probably a low value resistor and an LED with a diode would work. One good reason to do an LED there is that you can solder everything in place and not have the lamp socket anymore. But since the bulbs never burn out, you could also just solder it in also.This is what I was thinking. I'm not sure, but is the diode necessary? A resistor and an LED is essentially the same thing as an incandescent bulb in terms of current, passing current in both directions and having resistance, with the exception being that LED's only light up when current passes through the right way. I'm not sure this would be an issue though considering it shares a common ground with other binnacle lights. I'm really just curious if it's as simple as sticking a resistor in the middle. I could tackle that in a matter of minutes, and the peace of mind knowing that there will always be resistance in the circuit is itself enough of a reason to do this.
Sure there is no NEED to have it be an LED, but that doesn't mean some of us don't WANT one in there anyways.Exactly! There's really no need to do the Lambda light either (without repurposing it, of course), yet making that one an LED is practically a given when converting to LED's. This is partially where my perfectionism comes into play: all the lights in the binnacle being converted to LED's except for one doesn't seem like I'm going far enough! :)
haha... I think what you meant to say was that you'd need to employ a drawn reciprocating dingle arm to reduce sinusoidal depleneration.
http://en.wikipedia.org/wiki/Turboencabulator:hihi2:
Bitsyncmaster
05-03-2013, 04:48 AM
I was taking of using one LED and not an LED bulb replacement. Each LED has a maximum reverse voltage that you can not exceed or it will burn up the LED. That is the reason for using a diode so you don't exceed that reverse voltage. Current does not flow both ways in an LED (it is a diode).
I think what the alternator needs is a current (maybe a 100 ma.) flowing to the alternator to get the field started. Then the alternator can keep the field running with it's own power. If the voltage is to low from the alternator I think it drives current back to the bulb to indicate a problem. If this is what happens, you would need a diode going to the alternator and the LED with a current limiting resistor would be attached over that diode in the reverse direction. You may also need to limit the current through that diode (not sure about that).
dmc6960
05-03-2013, 12:04 PM
In other words, just use an incandescent.
Even on the new instrument cluster I'm building (yes, I'm still working on that) with digital electronic controls for everything and LED lights on everything, will STILL have a plain old bulb for the Alternator.
In other words, just use an incandescent.
Even on the new instrument cluster I'm building (yes, I'm still working on that) with digital electronic controls for everything and LED lights on everything, will STILL have a plain old bulb for the Alternator.
+1
..and as far as it looking different for a split-second or two, it's not a big problem substituting a different rated incandescent bulb in to match the appearance, the resistance/load isn't "that" tight and regulator don't care once 12V hits it...all you have to worry about is not to go bright enough to melt plastic.
:smokelot:
In other words, just use an incandescent.
Even on the new instrument cluster I'm building (yes, I'm still working on that) with digital electronic controls for everything and LED lights on everything, will STILL have a plain old bulb for the Alternator.This post got me thinking. With all that modern technology you're using, surely there's a way to pull it off. It can't be that much harder than a digital speedometer, odometer, etc.
Then I read another thread on running new speaker wiring, and had an epiphany of a far simpler solution to this problem than what we've all been discussing. Instead of tapping into the existing circuit and trying to combine 1981 technology with 2013 fixes, I simply patch it and create a new one. The wires coming from the alternator are rerouted to a bidirectional resistor of identical ohms to the stock bulb. This keeps the charging circuit functional, but (obviously) bypasses the bulb. The bulb is then replaced with an LED version and rewired to a modern voltmeter. There are voltmeters available that will illuminate a light only when the voltage is within a specific range, and one of these will be used. Then it's just a matter of hooking up the voltmeter to the alternator, and viola, problem solved! Simple, reliable, and easy to do.
Maybe once I get this done, I can sell plug-in kits :)
dmc6960
05-05-2013, 11:33 PM
KISS
'nuff said.
KISS
'nuff said.It is simple. You replace the bulb with the resistor and run the LED to a new voltmeter. What's so complicated about that?
Elvis
05-06-2013, 04:00 PM
coool, so easy. show me voltmeter please :goatee:
dmc6960
05-06-2013, 04:32 PM
http://imgs.xkcd.com/comics/the_general_problem.png
I've already stated how ridiculously simple it is. Instead of providing actual input, you chose to be an ass. Trolling much? :rolleyes1:
Particularly ironic coming from the guy who's completely redoing every other part of his instrument cluster and getting zero actual benefits whatsoever :hihi2:
coool, so easy. show me voltmeter please :goatee:I don't have any part numbers yet, but I know all the parts exist. Just have to figure out the details ;)
Bam:
http://www.pac-audio.com/productDetails.aspx?ProductId=573&CategoryID=31
$40.95 sensor that turns an LED red at 11.5 volts or below. That's 90% of my work cut out for me.
And you guys said it was too complicated :hysterical:
dmc6960
05-19-2013, 12:38 AM
That still is complicated. It also bypasses the original purpose of the light which informs you of a charging problem. This solution turns it into a low-voltage dummy light. Not the same thing. It could light up on you every time your driving in the rain on a humid day. Headlights, wipers, cooling fans, blower motor, all running at idle could bring the legitimate voltage down this low at idle. Also turns it into a damn expensive LED.
Tillsy
05-19-2013, 01:43 AM
Just create a smaller version of this circuit...
http://dmctalk.org/attachment.php?attachmentid=19145&d=1368942178
Tomcio
05-19-2013, 07:59 AM
My car is not available for testing so I cannot verify this but as an idea...
What would you say about using a circuit like this:
19147
It is a simple circuit that replaces the bulb. A resistor provides the necessary path to energize the regulator and would have a resistance similar to the bulb (a bit higher actually). Then a LED would be installed with a current regulating diode - it's that square element next to the LED. This component provides constant current and protects the LED at voltages from 1.8V over LED voltage (about 2V for red) all the way up to 45V. As it is a diode it also protects the LED from reverse current. It would require a little experimenting to get the resistor value just right. I don't know how much current the regulator requires so I can't calculate this value.
This sure beats spending over $40 at some over engineered device that doesn't really do what we need. Measuring voltage doesn't replace the charging light. It's a great addition but not a replacement. But then, we already have a gauge for that! The charging light indicates when the alternator is not charging. A voltage can still drop with large enough load even when the alternator is working as it should.
What do you think?
What would you say about using a circuit like this:
19147I believe this may have been mentioned earlier in the thread, but this is the first time I've seen a schematic for it. It looks a lot simpler than it sounds! I've got a breadboard I can test this out on. I'll report my findings back. I may be able to cram all that into the battery light space. With any luck I can make it a drop-in replacement with zero permanent modifications.
Frankly I'm just glad this isn't a "taboo subject" anymore, and once LED's become efficient and bright enough for headlights, we can finally have true 100% LED DeLoreans roaming around :)
Tomcio
05-19-2013, 01:55 PM
I believe this may have been mentioned earlier in the thread, but this is the first time I've seen a schematic for it. It looks a lot simpler than it sounds! I've got a breadboard I can test this out on. I'll report my findings back. I may be able to cram all that into the battery light space. With any luck I can make it a drop-in replacement with zero permanent modifications.
I don't think it will be very hard to do as our battery light is pretty large. If you use a wide angle SuperFlux LED something in the range of 120 - 140 degrees you should get great results. A LED like this will require 20mA of current. For this purpose look for this part: NSI45020AT1G. This is a 20mA CCR (Constant Current Regulator) and it's quite small. Package dimensions are: 1.6 x 2.7 x 1.2 millimeters. A SuperFlux LED is about 7.6 x 7.6 x 4 mm. You should be able to easily fit this entire circuit inside the battery light with room to spare and get a nice even light distribution.
The resistor might be the largest component here. You will need to create a drop of about 4V in order to light the LED to full brightness. Depending on the current required by the alternator it might be a large resistor. You need to experiment. I would start by replacing the original bulb with a resistor to find out the voltage drop. Get something like 100 ohms and measure across the resistor when you turn the key on without starting the engine. This will allow you to calculate the required value for a 4V drop.
Let us know what you find.
Frankly I'm just glad this isn't a "taboo subject" anymore, and once LED's become efficient and bright enough for headlights, we can finally have true 100% LED DeLoreans roaming around :)
That is my plan... no light bulbs at all!
There are LEDs out there that are brighter than HID.
Have you heard about Luminus? They make high power LEDs. Here's what they currently offer (just a reminder: HID D2S makes 3200 lumens, H4: 1200 lumens low, 1600 lumens high beam):
SST-90 - 30W - 2500 lumens (my choice)
CBT-140 - 75W - 3800 lumens
And many others...
They are already available!!
Sammy
05-19-2013, 05:11 PM
I honestly have not heard of the socket being a problem before :hmm:
A light bulb can cause corrosion if left in the same socket for 32 years. I recently changed bulbs for someone in their instrument cluster. We counted how many lights weren't working and drove to walmart to purchase that many #194 bulbs. We also purchased a #161 bulb for the battery.
Upon returning to the car we found that several indicator lights didn't work even after replacing the bulb. Upon inspecting the sockets I found corrosion. I cleaned the corrosion in the sockets and the lights started working again. Its a shame too because the owner of the car ended up buying a few more light bulbs than he needed.
Back on topic though, just put the correct bulb in the battery light socket and be done with it. You really have two options. Spend $2 on the correct bulb and be done with it, or spend tons of money trying to make a LED work and wonder if it is the source of your problem each time your car has an electrical glitch.
Tomcio
05-19-2013, 08:38 PM
Back on topic though, just put the correct bulb in the battery light socket and be done with it. You really have two options. Spend $2 on the correct bulb and be done with it, or spend tons of money trying to make a LED work and wonder if it is the source of your problem each time your car has an electrical glitch.
That is not the right thing to say here!
You might as well say that we should all drive Toyota's rather than spending tons on money on fixing some old classics and wondering what is going to go wrong next. We like it! We like tinkering with those cars and we like to make improvements. I do not look at this as a problem of spending less by using the bulb. I like the challenge, I like the fun of making something better, I like creating new things while improving the car.
As a side note I might add that it will be a very light "ton" of money. An LED costs a few cents, a resistor is even cheaper and they sell them here in packs of 10 for 3 cents. The CCR will be the most expensive component at about 50 cents. So, altogether this circuit should cost less than $1! That's right, less than ONE dollar. I can afford that while having fun by doing what I like to do.
BTW, you're ruining all the fun here :wiggle:
I don't think it will be very hard to do as our battery light is pretty large. If you use a wide angle SuperFlux LED something in the range of 120 - 140 degrees you should get great results. A LED like this will require 20mA of current. For this purpose look for this part: NSI45020AT1G. This is a 20mA CCR (Constant Current Regulator) and it's quite small. Package dimensions are: 1.6 x 2.7 x 1.2 millimeters. A SuperFlux LED is about 7.6 x 7.6 x 4 mm. You should be able to easily fit this entire circuit inside the battery light with room to spare and get a nice even light distribution.
The resistor might be the largest component here. You will need to create a drop of about 4V in order to light the LED to full brightness. Depending on the current required by the alternator it might be a large resistor. You need to experiment. I would start by replacing the original bulb with a resistor to find out the voltage drop. Get something like 100 ohms and measure across the resistor when you turn the key on without starting the engine. This will allow you to calculate the required value for a 4V drop.
Let us know what you find.Thanks for the info! I'll get crackin' at this :)
When you say measure across the resistor [with a multimeter], do you mean measure voltage before the resistor and again after the resistor, or measure voltage of the circuit both with and without the resistor? Just want to make sure I'm measuring the right way.
and wonder if it is the source of your problem each time your car has an electrical glitch.I won't have to wonder. I'll know my soldering job needs to be redone and that'll be that. You're forgetting I had the entire rolling chassis completely 100% redone at DPI. Currently there are no electrical glitches.
Although I do fully intend on road-testing by driving in a 2-3 mile loop several times until I drain the tank down 1/4 of the way or more, which is only a couple miles away from my house, all on minimally busy side roads. If it doesn't work for whatever reason, I can take the binnacle off, swap it with the bulb I took out, and have a family member come and give me a jump to get home. If it works then, all things considered, it should work for many many years. If it doesn't, it'll only take me a couple minutes to swap the bulb as I will make sure to only rest the binnacle in place without securing it. I will be designing it as a plug-in replacement using the original bulb contacts, so to revert it to stock, I quite literally unplug the LED and plug in the incandescent.
(Cost by the way will be reduced even further as I will be producing a batch of these and offering them for sale, and most small electronics like this have price breaks in larger quantities. I'll go so far as to say that it may actually be cheaper than the incandescent bulb. Lower cost, higher reliability, what's not to like?)
Tomcio
05-20-2013, 05:31 AM
Thanks for the info! I'll get crackin' at this :)
When you say measure across the resistor [with a multimeter], do you mean measure voltage before the resistor and again after the resistor, or measure voltage of the circuit both with and without the resistor? Just want to make sure I'm measuring the right way.
Shep, when I say measure across I mean exactly that. Take your meter, set it to volts DC, connect a resistor in place of the bulb and attach the meter leads to both ends of the resistor. Turn the key on without starting the engine to energize the circuit. The meter will show you a voltage drop across the resistor. For the LED to light up to full brightness you will need to see at least 4V. I don't know if that will be enough to get the regulator going though. You might need to see more! That is why you need to experiment with the value of the resistor. To see if it is enough you might start the engine and see if the voltage across the resistor drops to 0 or almost 0. If it doesn't change it will mean the voltage drop was too low to start the alternator and you will need to LOWER the resistance. Lower resistance = higher current = larger voltage drop. Keep in mind that at some point the current might be high enough to exceed the power handling capacity of the resistor. Calculate the power to make sure the resistor doesn't burn up. The most common type resistors with color coded values are 1/4 watt.
You can email me with your findings if you need help figuring out the math for this.
(Cost by the way will be reduced even further as I will be producing a batch of these and offering them for sale, and most small electronics like this have price breaks in larger quantities. I'll go so far as to say that it may actually be cheaper than the incandescent bulb. Lower cost, higher reliability, what's not to like?)
Mouser Electronics has those CCRs for 42 cents when you buy 1. If you purchase an entire reel of 3000 the cost drops to less than 15 cents.
http://www2.mouser.com/ProductDetail/ON-Semiconductor/NSI45020AT1G/?qs=%2fha2pyFadug2wDjEaGdXEcxGtIAJ%2fVstzZnvq3GdoR LAxMJcRWBHDQ%3d%3d
Don't go crazy with trying to get the price as low as possible. Use high quality LED and high quality resistors. I would rather pay $10 for the highest quality set than $5 for the cheapest parts available. High quality LEDs give even, bright light and are worth the few more cents in price. The price will be low enough so anyone interested will be able to afford it.
Bitsyncmaster
05-20-2013, 06:28 AM
When you measure the voltage. Note if the polarity changes when the generator is working. I was going to do this test way back when I was thinking of a circuit to work with an LED.
Tomcio
05-20-2013, 07:33 AM
I seriously doubt the polarity will change. The light comes on when the alternator is not charging. This means that current will flow in one direction to activate the regulator. Then, as it starts charging the current stops flowing and the light goes out. If the engine stalls of your belt brakes the current starts flowing again but it must be in the same direction... There is no reason for it to flow the other way or else the light would come on. It comes on only in one situation and it's always the same situation. That should not be a problem.
Shep, when I say measure across I mean exactly that. Take your meter, set it to volts DC, connect a resistor in place of the bulb and attach the meter leads to both ends of the resistor.Oh oh oh, I misunderstood. When you said voltage drop, I thought that meant the voltage had to be different in two different measurements, not that the voltage had to be 4v for one measurement. Makes perfect sense now.
Mouser Electronics has those CCRs for 42 cents when you buy 1. If you purchase an entire reel of 3000 the cost drops to less than 15 cents.
http://www2.mouser.com/ProductDetail/ON-Semiconductor/NSI45020AT1G/?qs=%2fha2pyFadug2wDjEaGdXEcxGtIAJ%2fVstzZnvq3GdoR LAxMJcRWBHDQ%3d%3dJust want to make sure: is that the right link? The price for that is 39 cents each individually and 13.8 cents each in reels of 3000, but that's the same model number you posted before.
I tend to order from Digikey, but I did find that part here (http://www.digikey.com/scripts/dksearch/dksus.dll?vendor=0&keywords=NSI45020AT1G) (the difference being purely packaging). I also looked for some LED's from there, and my broad search thus far is this (http://www.digikey.com/scripts/dksearch/dksus.dll?FV=fff40008%2Cfff801b9%2C940001%2C87c001 d%2C89c0029%2C89c002a%2C89c002b%2C89c002c%2C89c002 d%2C89c002f%2C89c0030%2C89c0031%2C89c0032%2C89c003 3%2C89c0034%2C89c0037%2C89c00b2%2C89c00b3%2C89c00b 5%2C89c00b7%2C89c00b9%2C89c00bc%2C89c00be%2C89c00c b%2C89c00cd%2C89c00d1%2C89c00d3%2C89c00e2%2C89c00f 1%2C89c0103%2C89c010a%2C89c01be%2C89c0201%2C89c021 4&mnonly=0&newproducts=0&ColumnSort=-206&page=1&stock=0&pbfree=0&rohs=0&quantity=1&ptm=0&fid=0&pageSize=25) (red LED, 20ma, viewing angle from 120 degrees to 200 degrees, sorted by candela descending). That top one looks pretty good and quite bright, which for any warning light is most definitely a bonus. Anything in particular I should use to narrow it down further?
I'm guessing the forward voltage of the LED is irrelevant since it's wired parallel to the resistor rather than after it.
By the way, I'm thinking the ultimate test to verify the bulb lights when it should would be to start the engine, ensure it's not lit, then have a helper and/or video camera focusing on the dashboard while I manually grab the alternator belt and force it to slip, wearing thick leather gloves and grabbing just downstream to avoid injury. If it's done right, it should logically light up correctly. Or would just unhooking the alternator wire trigger the light too? That's the first test I'll do, and if all goes well, then I'll drive it around the block several times as the second test, with the headlights, A/C, and other battery-draining components running to quicken the results. I can also do that when testing the initial voltage to figure out the right resistor value and whether or not the polarity reverses at any point, although considering it has a shared ground, I don't see why (or how) the polarity will reverse.
Tomcio
05-22-2013, 06:50 AM
Oh oh oh, I misunderstood. When you said voltage drop, I thought that meant the voltage had to be different in two different measurements, not that the voltage had to be 4v for one measurement. Makes perfect sense now.
You can measure the voltage in two separate points and subtract but why bother if you can just attach the leads to the resistor and get the result in just one measurement. Make it easy for yourself :)
Just want to make sure: is that the right link? The price for that is 39 cents each individually and 13.8 cents each in reels of 3000, but that's the same model number you posted before.
Yes, the link is correct but the prices are different. Mouser, it seems, is giving me higher prices because I'm a damn pollack ;)
They direct me to a European page with prices in Euro. When I try to switch to the US site it goes to an international site with those higher prices. Just be glad that you will pay less!
I tend to order from Digikey, but I did find that part here (http://www.digikey.com/scripts/dksearch/dksus.dll?vendor=0&keywords=NSI45020AT1G) (the difference being purely packaging). I also looked for some LED's from there, and my broad search thus far is this (http://www.digikey.com/scripts/dksearch/dksus.dll?FV=fff40008%2Cfff801b9%2C940001%2C87c001 d%2C89c0029%2C89c002a%2C89c002b%2C89c002c%2C89c002 d%2C89c002f%2C89c0030%2C89c0031%2C89c0032%2C89c003 3%2C89c0034%2C89c0037%2C89c00b2%2C89c00b3%2C89c00b 5%2C89c00b7%2C89c00b9%2C89c00bc%2C89c00be%2C89c00c b%2C89c00cd%2C89c00d1%2C89c00d3%2C89c00e2%2C89c00f 1%2C89c0103%2C89c010a%2C89c01be%2C89c0201%2C89c021 4&mnonly=0&newproducts=0&ColumnSort=-206&page=1&stock=0&pbfree=0&rohs=0&quantity=1&ptm=0&fid=0&pageSize=25) (red LED, 20ma, viewing angle from 120 degrees to 200 degrees, sorted by candela descending). That top one looks pretty good and quite bright, which for any warning light is most definitely a bonus. Anything in particular I should use to narrow it down further?
From what I have found the original #161 bulb emits about 10 lumens @ 14V, 190mA and 2.66W. That is quite bright! I would not go with any round LEDs. Either use an array of SMDs or one SuperFlux. They will be bright enough to give you nice results. Any round 3 or 5mm LED will give you a bright hot spot in the middle and that would not look good. That is exactly what I was trying to avoid when designing my door lights.
If you want to have exactly the same load as the bulb you will need a resistor value: R=U/I; R=14/0.19; R=73,68ohm. Remember that this resistor will have to handle about 2.7 watts! You will need to use one that is at least 4W or else the resistor will heat up quickly.
The nearest higher value resistor that will be available is 75 ohms.
I'm guessing the forward voltage of the LED is irrelevant since it's wired parallel to the resistor rather than after it.
Any and all red LEDs will have a voltage drop in the range of 1.8 to 2V. The voltage drop across the CCR will be another 1.8V. If you add the two you will see that the voltage drop across the resistor will have to be at least 4V. It's ok if it's more but it cannot be less as this will be the voltage going to the LED+CCR.
By the way, I'm thinking the ultimate test to verify the bulb lights when it should would be to start the engine, ensure it's not lit, then have a helper and/or video camera focusing on the dashboard while I manually grab the alternator belt and force it to slip, wearing thick leather gloves and grabbing just downstream to avoid injury. If it's done right, it should logically light up correctly.
Or would just unhooking the alternator wire trigger the light too?
DON'T EVEN THINK ABOUT GRABBING THE BELT!!!
It doesn't matter how thick the glove is, you will loose your hand! This is a VERY BAD idea.
You get exactly the same result by just turning the key on and verifying the battery light comes on. Then start the engine and make sure the light goes out. If you really need to see if the light comes back on just stall the engine. If it's a manual transmission it's simple to do. Put it in gear, step on the brake and release the clutch. The battery light will come on. If it's an automatic you can simply cover the air intake and block the air going in. Same thing, it will stall and the light will come on. Do not try to stop the alternator with the engine running.
It's also a bad idea to disconnect the alt cables. The battery works as a load for the alternator. The regulator might not be able to keep the voltage within limit without the battery. The voltage might rise to over 60V and that will surely burn out anything electrical still attached to the alternator or the regulator itself. DON'T DO IT.
That's the first test I'll do, and if all goes well, then I'll drive it around the block several times as the second test, with the headlights, A/C, and other battery-draining components running to quicken the results. I can also do that when testing the initial voltage to figure out the right resistor value and whether or not the polarity reverses at any point, although considering it has a shared ground, I don't see why (or how) the polarity will reverse.
I don't see how and why the polarity might reverse. It can't happen in this circuit.
You can do the testing without actually driving the car. Just attach a volt meter to the battery and see if the voltage rises with the engine running.
If you attach the meter directly to the alternator the effect will be easier to see as the voltage change will be quicker and larger.
Good luck!
Just don't get near running belts. PLEASE!!
................while I manually grab the alternator belt and force it to slip, wearing thick leather gloves and grabbing just downstream to avoid injury. If it's done right,.............
I can't believe what I'm reading!
No, it can not be done right, no way!
Don't even think on doing it, go get yourself some sleep while you can.
LEVY
Michael
05-22-2013, 07:37 AM
All this so you can have an LED battery light?:rolleyes1:
With some alternators the voltage WILL run "backwards" if the alternator is putting out but the battery has a problem (has a lower voltage), by design. Others will turn the light on (for the same reason) but handle it with circuitry within the regulator -- You guys are missing this function.
What you have so far will only tell you if the alternator is putting out more voltage than the battery has (and that doesn't mean a lot if both are bad....). And it will probably glow or flash when the battery gets fully charged (to detailed to explain...)
If you just gotta have it, you could adapt this , but it's not worth the effort to build it much less adapt it IMHO.
19202
Hope this helps.
(Look/google around someone has probably done it ...)
Tomcio
05-22-2013, 05:21 PM
With some alternators the voltage WILL run "backwards" if the alternator is putting out but the battery has a problem (has a lower voltage), by design. Others will turn the light on (for the same reason) but handle it with circuitry within the regulator -- You guys are missing this function.
Ron, there's something I'm not getting here. Please explain. By design the battery will always have a lower voltage unless the alternator is bad and is not charging. Am I wrong?
I agree that the current might flow "backwards" but not on the battery light lead. It's actually impossible. According to the DeLorean electrical schematic the battery light is connected to positive 12V (with ignition on) and the other side goes to the alternator. When the alternator is not charging this terminal gets grounded and the light comes on. With the alternator charging this terminal gets to + 12V and the bulb actually has positive on both sides - light goes out. The current cannot flow "backwards" here. Unless I'm missing something...
What you have so far will only tell you if the alternator is putting out more voltage than the battery has (and that doesn't mean a lot if both are bad....). And it will probably glow or flash when the battery gets fully charged (to detailed to explain...)
If you just gotta have it, you could adapt this , but it's not worth the effort to build it much less adapt it IMHO.
I agree that this is too much effort but I believe you are missing the point here. We are not trying to improve on the charging circuit. We are just trying to replace the light bulb with a LED. Whenever the bulb would come on the LED will come on. That's it.
From what I see the circuit you are proposing is a differential volt indicator. That is a nice idea but it's not what we are trying to accomplish. Yes, the best way to monitor the alternator/battery condition would be to indicate the direction of the current flow but it is not what the battery light is suppose to indicate. Is it?
If the voltage drop across the proposed resistor is high enough the light will not glow not flash. Remember that by design the alternator voltage is always higher than the battery voltage. I don't see a problem here.
If you want to have exactly the same load as the bulb you will need a resistor value: R=U/I; R=14/0.19; R=73,68ohm. Remember that this resistor will have to handle about 2.7 watts! You will need to use one that is at least 4W or else the resistor will heat up quickly.
The nearest higher value resistor that will be available is 75 ohms.Maybe I'm missing something, but where does the 14 volt measurement come from? It's a 12 volt system, and the regulator and LED need a minimum amount of voltage, so wouldn't overestimating the resistance go below that minimum voltage? 12 volts would equate to about 63.16 ohms and 2.28 watts, and 11 volts would equate to about 57.89 ohms and 2.09 watts, so I'd think somewhere in the ballpark of 60 ohms would be the correct value. Can we rely on the alternator putting out a minimum of 14 volts on a 12 volt system? :hmm:
Ron, there's something I'm not getting here. Please explain. By design the battery will always have a lower voltage unless the alternator is bad and is not charging. Am I wrong?
You are right (as long as the engine is running ;-)
I agree that the current might flow "backwards" but not on the battery light lead. It's actually impossible. According to the DeLorean electrical schematic the battery light is connected to positive 12V (with ignition on) and the other side goes to the alternator. When the alternator is not charging this terminal gets grounded and the light comes on. With the alternator charging this terminal gets to + 12V and the bulb actually has positive on both sides - light goes out. The current cannot flow "backwards" here. Unless I'm missing something...No, there are several ways it can. Easiest example I can think of and explain easily is a common problem where the light is on when the key is off...say one of the 3 positive diodes in the bridge rectifier shorted and blew the diode trio feeding the regulator with one of its diodes being shorted also. On a D, current would flow from battery (+) => rectifier heat sink (+) => shorted rectifier diode => shorted trio diode => bulb => anything connected to the bulb's feed circuit (EG oil light, gauges, etc, etc.)
19204
Also, if one of the battery cells goes open, the light will come on with most cars...the battery is surely not powering the bulb here...It's coming from the alternator, fed "backwards" through the bulb (until it eventually blows the regulator).
I agree that this is too much effort but I believe you are missing the point here. We are not trying to improve on the charging circuit. We are just trying to replace the light bulb with a LED. Whenever the bulb would come on the LED will come on. That's it.
From what I see the circuit you are proposing is a differential volt indicator. That is a nice idea but it's not what we are trying to accomplish. Yes, the best way to monitor the alternator/battery condition would be to indicate the direction of the current flow but it is not what the battery light is suppose to indicate. Is it?
If the voltage drop across the proposed resistor is high enough the light will not glow not flash. Remember that by design the alternator voltage is always higher than the battery voltage. I don't see a problem here.
No, I am not missing the point -- You made that perfectly clear in a previous post. I like playing with circuits and thought the idea was cool and offered help. It was as said- something you might adapt, not a proposal (solution). It can be done easily even with a relay, if you feed the bulb independently off of the switch (as you may have guess from the above), but that is ridiculous IMHO, and it too would cost loosing functionality ("light on when charging" don't cut it to me. I want to know if the battery is taking it. The battery can develop a short, where it would not not exceed 12V. With most integral alternators, the regulator can see that the voltage is lower than its cut-in voltage and will turn the light on (by grounding it).
=====
IIRC- With the 194 bulb, you can put a diode in parallel, in reverse, of the LED to protect it, and a 257* ohm resistor also in parallel. It will work, but will make the bulb flash when the battery recovers after starting the engine (My guess is it happens during the cut in/out transitions??).
* The math they were using worked out to 257, but for most cars, they said 250 ohm keeps the cut-in/cut-out voltages closer to specs. The math didn't work out that way for me (or make sense with a quick look), so I marked it up to adjustments they made when actually installing it an getting it all to stay within specs, and moved on....
Again this is from memory -- Google it. There were several examples AND complaints of flashing when I was looking at this a while back, including bridge rectifier designs.
FWIW- If I understand the way you are doing it above, 50 ohms is what people were having luck with.
Keep plugging, there IS a good way...
Have fun!
EDIT: This is a job for Bitsy or Elvis...if you can talk one of them into it is another story, lol.
Tomcio
05-23-2013, 06:41 AM
Maybe I'm missing something, but where does the 14 volt measurement come from? It's a 12 volt system, and the regulator and LED need a minimum amount of voltage, so wouldn't overestimating the resistance go below that minimum voltage? 12 volts would equate to about 63.16 ohms and 2.28 watts, and 11 volts would equate to about 57.89 ohms and 2.09 watts, so I'd think somewhere in the ballpark of 60 ohms would be the correct value. Can we rely on the alternator putting out a minimum of 14 volts on a 12 volt system? :hmm:
Well yes, it is a 12 volt system. But, the 12V is just a very generic term. When your battery goes down to 12V it is below dead!
A fully charged battery has a voltage of 12.66V
Alternator voltage should be between 13.6 and 14.2V with an average charging voltage of 13.8V
When you do calculations you should always prepare for the upper limits as not to overpower your circuit.
Example:
A LED requires 20mA of current at about 2V. Let's say you'd like to use it in your car. If you assume that a 12V system has exactly 12V your current limiting resistor will be R=voltage drop/current = 10/0.02 = 500 ohms. Then you start your engine, the alternator brings the voltage up to 13.8V and the result is I=11.8/500 = 23.6mA. The LED heats up and dies soon. Then people blame the Chinese LEDs for being poor quality because they burn out after just a few hours instead of the advertised 100,000 hours.
Always base your calculations on the upper limits.
When doing your calculations, you have to take in consideration that you battery also acts as a voltage regulator if you battery is dead or bad, it would allow the alternator to provide it's full capacity,which is over 16 v, that is the reason you should not run your engine without a battery as we did on old vehicles, before the electronic ignition era.
LEVY
Dang it, I had to look ...
19205
Schematic showing the diode trio.
Note that in my example above a shorted trio diode is NOT actually needed since they will already pass current in the direction required to feed the bulb when a main (+) diode is shorted (but same results).
Also, note the difference in the they way the bulb is fed...hmmm
==========
For curiosity-
Another-
19206
Schematic showing regulator internals.
Note diode added in separate bulb feed....
==========
With all of the different alternators we use on our Ds, the only thing that makes sense to me would be to use a stock Motorola (to cover the most people), connect a diode across the LED to protect it and use a variable resistor to start taking measurements of the cut in/out voltages. (Probably work it out about the time someone locates a working schematic, or, one the previously mentioned says, "Hey! Just do this..." ;))
Tomcio
05-23-2013, 02:32 PM
You are right Ron. The current might flow the other direction if the light is getting grounded when the ignition is off. I didn't think of this kind of failure and was just thinking about regular operation.
To solve this problem and have the light come on in a situation where one of the positive diodes shorts out we can connect the LED through a bridge rectifier. In this way the LED will come on regardless of the direction of current flow.
A protection diode is not needed if a CCR is used. The CCR allow current to flow through the LED in one direction only.
I will do some experimenting when I get my car running again. For now I can only guess.
Thank you for your input Ron.
Okay I got lost, like, five posts ago. Tomcio, you obviously have the motivation, the knowledge, and the skills to pull this off. I have the motivation, my skills are debatable, and I have a hell of a lot to learn before I have the knowledge. Best of luck to you in making this a reality. I'll definitely buy one when you finish it. You're making the first LED side marker compatible with the blinker mod, so I have no doubt you can do this. Godspeed! :thumbup:
Side note: my battery light does NOT come on at any point, even when I turn the key to just "on". The oil light works as it should, but not the battery light. I've driven about 150 to 200 miles without any electrical issues, and have witnessed my voltmeter going from 12v down to about 9v or 10v when my brake lights come on, but I can't trust that it'll remain that way for long. I didn't realize it was even supposed to come on when the key was turned to "on" until you said something, so thank you very much for pointing that out! :biggrin: Now I'll pull my binnacle this weekend and save myself from being stranded and needing a jump :wiggle:
Tomcio
05-24-2013, 06:15 AM
Okay I got lost, like, five posts ago. Tomcio, you obviously have the motivation, the knowledge, and the skills to pull this off. I have the motivation, my skills are debatable, and I have a hell of a lot to learn before I have the knowledge. Best of luck to you in making this a reality. I'll definitely buy one when you finish it. You're making the first LED side marker compatible with the blinker mod, so I have no doubt you can do this. Godspeed! :thumbup:
The only thing missing that I really need is the time! Maybe I should build a time machine... ;)
But really, I need to put my car back together. Once I can drive it again I can start working on the battery LED. I'm working on my car everyday now...
Side note: my battery light does NOT come on at any point, even when I turn the key to just "on". The oil light works as it should, but not the battery light. I've driven about 150 to 200 miles without any electrical issues, and have witnessed my voltmeter going from 12v down to about 9v or 10v when my brake lights come on, but I can't trust that it'll remain that way for long. I didn't realize it was even supposed to come on when the key was turned to "on" until you said something, so thank you very much for pointing that out! :biggrin: Now I'll pull my binnacle this weekend and save myself from being stranded and needing a jump :wiggle:
When you turn the key on you should see two red lights - oil and battery. Those are two very important lights and they come on as a test to show you that they are working. Once you start the engine both should go out. If your battery light doesn't come on you should not drive the car until you correct the problem. Otherwise you will kill the battery. And "kill" means that you might have to replace it soon. Charge it as soon as you can. Don't count on the alternator to recharge a dead battery. Alternators are meant to maintain a battery and not to recharge a dead one.
The dash volt gauge is not very accurate but still, if it goes below 12V you have a problem.
There are at least three things that might cause what you are describing:
1. Burnt out battery bulb
2. A break in the battery light circuit
3. Bad alternator/regulator
I would first check to make sure that all the wires are still securely attached to the alternator. Look closely at the thin wire as this is the one going to the light. They tend to corrode and brake right at the point where the wire enters the connector. Make sure there is no corrosion there. Then make sure that the little nut securing this wire is tight but do not over tighten it. You do not want to brake the stud.
If you want to save yourself the trouble of ripping the dash apart you can disconnect this thin wire from the alternator and just ground it with the key on - touch the alt casing. If the light comes on it is good and you'll probably need to replace the alternator.
But whatever you do, please DON'T START THE ENGINE WHILE WORKING ANYWHERE NEAR THE BELTS!!
Good luck!
Tom
Bulb's good, alternator's new, thank god for warranties :)
:icon_smile_duh:
Correction: bulb's fine, alternator's fine too, and the battery does charge, but no light. (It would help if I actually tested the alternator before posting...) Doesn't seem to pose a problem, but I'm working with DPI anyways for good measure :)
Anyways, don't mind me, carry on as if I wasn't here! :wrenchin:
Powered by vBulletin® Version 4.2.5 Copyright © 2024 vBulletin Solutions Inc. All rights reserved.