LED Lighting

owner

Oz '03 316CDI LWB ex-Ambo Patient Transport
Gah this is so annoying. All I want is my tail lights to be led for when I leave the park lights on. But because they are combined brake/tail bulbs its just one thing after another...

The latest is that once I established that I got a faulty pair the first time, I decided to buy some decent BAY15Ds.

I also hooked up the unused bottom rear fog light incandescent bulbs to come on with the brake lights. This means the ECU will still see the correct current when the top brake lights change to led. It should also be safer since its two more unique brake lights that come on, rather than just the combined tail/brake light.

But.

Now the fancy new bay15d led bulbs I bought don't work either. The problem seems to be that when in park/tail light mode, the brake light circuit also has around 1V on it, and somehow this is causing the led bulb to freak out. The led won't light, and it gets hot with an odd burning smell. If I insulate the pad for the brake circuit on then led bulb, it comes on in tail light mode just fine, but obviously no brake light mode working.

I just want it to work, but it is conspiring against me.
 

owner

Oz '03 316CDI LWB ex-Ambo Patient Transport
Upon initial investigation, it appears that when the taillights are on, the brake circuit is pulled to ground, presumably for some sort of monitoring purpose on the dual input bulbs.

But on these new led bulbs it basically shorts out the 30 ohm resistor that is on the tail light input, shorting it directly to ground via the direct brake light input.

More time is needed to get to the bottom of this for sure.
 
In looking thru some of the LED replacement bulb links I noticed a mix of 12V only lamps with other that have a stated range such as 9V-14V.

I would caution against using LED lamps that state 12V only. This is because most of these type lamps are not voltage regulated and require an external regulated 12V supply. Operating on 12.7-14V such as found in our Sprinters will shorten their life substantially. In most cases you can tell if they are regulated or not by simply observing the lamp with the engine running (14V) and with it not running and some load on the system such as headlights on to drag the battery below 12.7V. Unregulated LED lamps may also produce a very slight glow when they should be off when used in dim-able or electronically controlled applications because of a small residual current that may flow when they should be "off".

Lamps with internal voltage regulators will normally state a range over which they are designed to operate and they will produce the same light output across that range. Assuming the rating is at least 14V on the high end they will also last longer. They will not show any residual glow when when they should be off if used in dim-able applications. If there is any downside it is that unlike an incandescent lamp, when used in a dim-able circuit they will be either off or on and the transition will occur right around the lower voltage of their rated range. Many NCV3 do a soft turn-off of interior lights. With the regulated LED lamp the lighting goes from full on to full off very quickly, rather than slowly dimming as the dimmer control drops the average supply voltage below the lower voltage rating of the lamp.
 

Eric Experience

Well-known member
Walter.
Led lights are not voltage regulated, they contain a transistor which is connected as a constant current source. So if they have 3 LED at 3volts each plus a transistor and emitter resistor 1 volt, they will come on at 10 volt and go up to the power capacity of the transistor. Eric
 
Walter.
Led lights are not voltage regulated, they contain a transistor which is connected as a constant current source. So if they have 3 LED at 3volts each plus a transistor and emitter resistor 1 volt, they will come on at 10 volt and go up to the power capacity of the transistor. Eric
Current regulation (constant current) versus voltage regulation makes sense for this application. However the effect is the same to the user. Unregulated LED lamps will vary in intensity with input voltage and running them above the stated voltage on the device will shorten their life.
 

220629

Well-known member
With claims of LED lamp life being 25,000 to 50,000 hours, is reduced lamp life really a concern in a vehicle? Add to that the fact that the LED lamp output reduces after those hours. They don't "burn out" in the traditional sense.

:2cents: vic
 

avanti

2022 Ford Transit 3500
With claims of LED lamp life being 25,000 to 50,000 hours, is reduced lamp life really a concern in a vehicle? Add to that the fact that the LED lamp output reduces after those hours. They don't "burn out" in the traditional sense.
The thing is, though, that those are average lifespan figures. The variance appears to be pretty large with consumer-grade parts. This means that if you have a lot of LEDs you will see random individual device failures much sooner. Doesn't affect total light output all that much, but it starts to look tacky.
 

220629

Well-known member
The thing is, though, that those are average lifespan figures. The variance appears to be pretty large with consumer-grade parts. This means that if you have a lot of LEDs you will see random individual device failures much sooner. Doesn't affect total light output all that much, but it starts to look tacky.
Even with average...

8 hours per day for 365 days is about 3000 hours. Most vehicle lamps will see a much lower average than that. So even when using the lower claimed number, given 50% = 12,500 hours will give many years of service. Under typical vehicle service the lamps should outlast any rusting Sprinter. Just musing a bit.

vic
 
The large lifespans quoted are largely meaningless. I have yet to see commercial LEDs spec sheets define the parameters of those "lifespans".

1) Is it actually life or MTBF? They are very different.
2) What is the distribution of failures and how many are expected to have already reached end of life by that point?
3) What are the electrical and environmental conditions required for that lifespan?

I have some experience in this field though not specifically with LEDs.

I found that unchecked many a manufacturer will quote things such as life = the time when 100% have failed, where the average may be way way less than half that and current/temperature operating conditions to reach even that stated "life" are well below what is featured in the product literature for typical operation. So dont assume 30,000 hours applies to what you have. Under the right (meaning wrong) conditions it could be a tenth of that, and if you are using an unregulated cluster of them such as in a single "lamp" the odds of a short life are even higher.

LEDs and semiconductors have a non-linear failure rate with respect to temperature and current. Some can exhibit an inverse square profile of reliability with increased current or temperature. I have personally seen high current LEDs fail within a few hours when current is just 10% beyond the max continuous rating and others that dimmed dramatically within days when operated at the rated max current.

Of course they CAN be conservatively specified and designs can avoid pushing either the current or temperatures to maximize life, but these are likely not going to be cheap lamps. As I said, looking for a lamp that is regulated is a good step in the direction of getting one that is well engineered and will last, but even that is no guarantee these days since the parts in a good design can still be crap.
 
Last edited:

avanti

2022 Ford Transit 3500
The large lifespans quoted are largely meaningless. I have yet to see commercial LEDs spec sheets define the parameters of those "lifespans".
I don't think this is right at all.

1) Is it actually life or MTBF? They are very different.
Huh? Its not that they are "very different", they are apples and oranges. MTBF is a *measure* of typical life. It is a descriptive statistic of a population of devices. Only an actual, individual part has an "actual life" and we can't know it until it fails. OF COURSE, the ratings are MTBF, and that is the appropriate statistic.

2) What is the distribution of failures and how many are expected to have already reached end of life by that point?
The normal distribution is almost always an adequate approximation of post-infant-mortality device failure curves. Any QA engineer would make this assumption absent extenuating circumstances. What we often don't know is the standard deviation of the curve. If we had that, we would be able to predict the expected number of failures at any point in time. But even without it, the mean lets us estimate the lifetime of a typical part.

3) What are the electrical and environmental conditions required for that lifespan?
Obviously, the stated specs of the device, which are almost always thoroughly specified by the manufacturer.

I have some experience in this field though not specifically with LEDs.

I found that unchecked many a manufacturer will quote things such as life = the time when 100% have failed, where the average may be way way less than half that and current/temperature operating conditions to reach even that stated "life" are well below what is featured in the product literature for typical operation.
In all respect, I think that this is nonsense. First of all, if what you say were true, the specified lifetime would always be "infinity", since as a practical matter, you will *never* get 100% of a large sample of LEDs to fail, ever. More generally, what you are describing would violate very long-standing industry standards to the point of constituting false advertising. Any manufacturer engaging in such practices would be sued out of business by their large customers. Now, what a random retailer on eBay might claim is another matter, but anyone who relies on those claims deserves what he or she gets.

So dont assume 30,000 hours applies to what you have. Under the right (meaning wrong) conditions it could be a tenth of that, and if you are using an unregulated cluster of them such as in a single "lamp" the odds of a short life are even higher.
If you use a device outside of its published electrical or environmental parameters then, yes, any published spec will be meaningless. But that is a critique of you, not of the specs.

LEDs and semiconductors have a non-linear failure rate with respect to temperature and current. Some can exhibit an inverse square profile of reliability with increased current or temperature. I have personally seen high current LEDs fail within a few hours when current is just 10% beyond the max continuous rating and others that dimmed dramatically within days when operated at the rated max current.
Yes. So? See above.

Of course they CAN be conservatively specified and designs can avoid pushing either the current or temperatures to maximize life, but these are likely not going to be cheap lamps. As I said, looking for a lamp that is regulated is a good step in the direction of getting one that is well engineered and will last, but even that is no guarantee these days since the parts is a good design can still be crap.
Neither the specs nor your design have to be "conservative", merely correct. If so, you can expect the published lifetime specs to be a reasonable predictor of time to failure. Such specs are far from "meaningless".
 
Huh? Its not that they are "very different", they are apples and oranges. MTBF is a *measure* of typical life. It is a descriptive statistic of a population of devices. Only an actual, individual part has an "actual life" and we can't know it until it fails. OF COURSE, the ratings are MTBF, and that is the appropriate statistic.
MTBF, or its reciprocal, referred to as AFR (Annualized Failure Rate) is the steady state bottom of the failure bathtub curve between infancy and wear out or end of life. I would suggest this is NOT what is meant by a manufacturers data sheet for LEDS as this value should be in the millions of hours, not tens of thousands. Most HDDs have MTBF figures of 300,000-1.000.000 hours. LEDs steady state failure rate (MTBF) should be able to do better. I contend that what the manufacturers are stating as lifetime is supposed to be wear out or end of life.

Obviously, the stated specs of the device, which are almost always thoroughly specified by the manufacturer.
Its not obvious to me. Please show me an LED datasheet that lists its "lifetime" number as being based on the upper limits of its published current and temperature range. I would like to see one that mentions what current and temperature the lifetime number is based on, no matter where it is.
 

e8hffff

2007 Sprinter 315CDI
...Now the fancy new bay15d led bulbs I bought don't work either. The problem seems to be that when in park/tail light mode, the brake light circuit also has around 1V on it, and somehow this is causing the led bulb to freak out. The led won't light, and it gets hot with an odd burning smell. If I insulate the pad for the brake circuit on then led bulb, it comes on in tail light mode just fine, but obviously no brake light mode working.

I just want it to work, but it is conspiring against me.
Note LED lights have a polarity. Other words a '+' side and '-' side, so they must be put in the right direction. The reason is they are diodes, light emitting diodes. You probably already know this, but if you don't then strange things will happen if you put them in wrong direction. Use a 9 volt battery to test polarity and then multimeter the wiring in the van to get right direction.

If people want to buy LED's on cheap, click here to join DHGate China Sales.

I will add more images soon on the newly installed led festoons I got for the cargo area, door lights and glove-box.
.
 
Last edited:

e8hffff

2007 Sprinter 315CDI
Cargo bay uses 4 components of 41mm 12 SMD's festoons in this ex-ambulance van. Maybe too white. Looks like a space craft


Door light uses a 39mm 9 SMD's festoon. So much brighter than incandescent


Glove compartment uses a 39mm 9 SMD's festoon


Here are the festoons I used to archive the new lighting. 39mm and 41mm festoons, using 5750 SMD's.
 
I replaced the 4 cabin dome lights in '12 with these: http://www.amazon.com/gp/product/B0045XQ3LS/ref=oh_aui_detailpage_o08_s00?ie=UTF8&psc=1 they are "warm white", which I think means around 3500K. I used 4500K lamps in the dome light over the dash and in the ground illumination lamps of the doors. I like the more daylight feel of the 4500K in those locations, but as I mostly use the cabin lights at night when camping, the warmer 3500K is nice here.

Oh, the 37mm length is shorter than the stock lamps but I really wanted the warm white and it was not available in longer festoon packaging. It is a simple matter to slightly bend the dome light lamp contacts to accommodate this shorter lamp.
 

dickknapp

dickknapp
Anyone have good source of warm white led rope lights? I'm thinking the ones mounted in clear PVC tubing would be more durable for my purposes (run under cabinets for ambient lighting). I've yet to find what I'm looking for.
 

e8hffff

2007 Sprinter 315CDI
Anyone have good source of warm white led rope lights? I'm thinking the ones mounted in clear PVC tubing would be more durable for my purposes (run under cabinets for ambient lighting). I've yet to find what I'm looking for.
The color temperature for Warm White is (2700-3500K). If you search for 'warm' 'led' 'pvc' and, 'rope' or 'flexible' or 'cord', then you should find results.

example;
http://www.amazon.com/s/ref=nb_sb_noss_2?url=search-alias=aps&field-keywords=led+warm+rope+pvc

Consider how you would cut the led strips or rope, as that may include soldering or other means to supply power. Also think about the lumen value, as most of the rope leds are old technology meant for mood rather than visibility.
 
Last edited:

e8hffff

2007 Sprinter 315CDI
I finally received LED head light replacements. I will soon take photo's on-road. Yet to check how they will fit in my Sprinter 315, 2007

Boxed item.


Unboxed. Quality is quiet good.


Testing components.


Very bright and hurting eyes whilst testing.


.
 

Top Bottom