Another pitfall of LED retrofits

[TheIntruder]

Since I learned of Philips' attempt to provide legal, properly-engineered LED retrofits for the common automotive bulb types, with the X-tremeVision LED and VisionLED, I couldn't resist ordering a pair to try out, while at the same time keeping in mind the dependency on the fixture design (to a greater extent than I should have considered) and short "approval" list.

In short, subjectively speaking, they appear to perform no worse than the original P21W bulbs as brake lights, and more than slightly brighter in the CHMSL. All three positions are served by P21Ws, properly indexed sockets, and oblong shapes that suit the transverse filament (and LED) projections.

However, where the story goes sour is related to the design of the fixture, and the nature of diodes.

Diodes, being diodes, will only allow current to flow in one direction, and if the engineers at Magnetti Marelli/Seima lighting for some reason reverse the polarity of the current flowing inside a tail light fixture, only to one specific function, and only for the right side fixture, can one be forgiven for thinking it's a bit of an odd situation? Are they taking cues from The Prince of Darkness?

Yet, it explains why an XTV LED retrofit bulb will function in the CHMSL and left side, but not the right !?

I have confirmed via wiring diagrams, and with a meter, that this situation is due to the design of the fixture; positive/negative coming from the harness is as expected; it is the fixture where the fault lies.

Granted, these factors didn't need to be considered when the design was done, with only incandescents in mind, but they do present an issue in modern times.

Perhaps the newer VisionLED design might not be subject to the same requirements, but I highly doubt it.

I've contacted Philips Automotive to get their take, but this is a warning that not only does optical design present potential pitfalls for the adoption of retrofits, electrical design does as well!

I could salvage this experiment by sticking with the CHMSL application (are they as tightly regulated?), but otherwise, these puppies will be making their way to Hebron KY.

 

[more_vampires]

Philips X-treme Vision PY21W LED Retrofit

Please be aware that this range of Philips X-treme Vision Festoon LED bulbs is designed for interior use. They can be used in an exterior application but it is not road legal to do and, in addition, due to technical differences between LED and halogen technology, a vehicle warning light may be triggered depending on the configuration of your vehicle. This is more likely to happen if your car was manufactured after 2005. This pack contains two Philips CANbus warning cancellers for LED lamps to solve this issue.

I made it to:

Luxeon Rebel High Power LED

...and then I stopped. I'm not hooking these to any of my vehicles. Every "festoon LED" bulb I've tried is a complete failure. Festoon works by having a filament centered down the tube. You can't do that with a point emitter. 5x the light output is a flat out lie. No way is this true. An 18 watt incan festoon has defeated every single led festoon I've tried in a legitimate vehicular bulb holder and reflector housing.

http://philipsxtremevisionled.com/

Philips Vision LED lights have been tested with luminaires for specific functions of certain car models, and for these tested luminaires, the beam pattern measured fulfills the requirements as stated in the SAE standards.

Something about this is setting off warning bells for me. Only tested in 13 vehicles. Duplicates in the list to make it longer. 2 vehicles it's only reverse lights. Bypassing system warnings. Hmm.

Sound like a steaming pile of "garbage" to anybody else?

 

[Alaric Darconville]

Something about [this disclaimer] is setting off warning bells for me. Only tested in 13 vehicles. Duplicates in the list to make it longer. 2 vehicles it's only reverse lights.

It bothers me, too. If I had my 'druthers, I'druther the bulbs they say will replace an 1157 will replace an 1157 in EVERY application.

Bypassing system warnings. Hmm.

There's nothing at all ominous about this: Those "warnings" being "bypassed" are simply bulb-out indicators. It's nothing to hide in the corner and shiver about.

 

[-Virgil-]

The Philips Vision and Xtreme Vision LED bulbs actually work very well (meaning: safely, effectively, legally) in a lot of lamps. Not all of them, it's true, and that's troubling that they're injecting them into a market that has been set up and conditioned over many years to expect universal application interchangeability (an 1157 works in every-any lamp that takes an 1157, a 3157 works in every-any lamp that takes a 3157, etc). Most consumers aren't going to bother checking to see if their vehicle is on the tested/approved list. And Philips isn't going to test each and every possible application; those tests take time and cost money.

 

[more_vampires]

It bothers me, too. If I had my 'druthers, I'druther the bulbs they say will replace an 1157 will replace an 1157 in EVERY application.

Very concise. Exactly my issue with this product.

 

[Lightdoctor]

At one time, I was pretty excited about these LED bulbs from Philips. Not so much anymore. I think my next vehicle will probably have all exterior LED lighting anyway. All new lighting is going to be solid state very soon anyway.

 

[TheIntruder]

As already mentioned, the irony is that these retrofits can. and probably do work very well in a variety of fixtures, but just haven't been verified to meet the objective requirements, which would be a difficult and expensive task.

Subjectively speaking, in the application I'm trying them in, they are almost indistinguishable from the original bulbs, day or night, and from most angles. They do make real use of the original optics, and are a far cry from the typical eBay-variety junk that usually results only in a dim glow inside a lamp housing.

It's an interesting situation for Philips, or any other legitimate company to balance the cost and effort of the engineering side with the business and marketing side. There has to be a business case to sustain these kinds of developments, and to hamper the effort in what is already a niche segment by explicitly discouraging broader application would kill it, and future efforts, for sure. (To replace a $2 bulb, who spends $15 on something that looks the same, and has few other tangible benefits, even if it lasts 12 years?)

For most consumers seeking to replace a burned-out bulb (if they happen to notice at all), the usual standard is "it fits, and lights up." Even so, the number of blister packs torn open and hastily taped and replaced on parts store shelves suggests that even that may involve a little too much trial and error.

The marketing may be lamentable, and all too familiar, but I will give credit to Philips for a solution that works very well, even if not in every application. Contrast that to the equivalent Osram products, the aforementioned eBay junk, or the dazzlingly bright and dazzlingly costly products that some retrofit dealers push.

 

[Alaric Darconville]

For most consumers seeking to replace a burned-out bulb (if they happen to notice at all), the usual standard is "it fits, and lights up." Even so, the number of blister packs torn open and hastily taped and replaced on parts store shelves suggests that even that may involve a little too much trial and error.

Bingo. They don't even look at the manual long enough to determine they need an 1157, and go ahead and buy an 1156, and then discover the 1156 doesn't work in the 1157 socket. So does anyone REALLY think they'll look to see if their vehicle is on Philips' approved list?

But they'll buy on price, anyway, and instead of getting the Philips, they'll get the PUTCO ("Where the putz goes!") bulb instead.

 

[TheIntruder]

Has anyone ever interacted with Philips Automotive Lighting NA?

As I mentioned earlier, I contacted them to get confirmation about the polarity situation I encountered, and confirm that the newer variant is similar.

I've provided the name/model line/type/part # of the items in question, and now they've written back requesting pictures of the parts I referred to?

I don't know if it's because they have a 3rd party company handling their customer service, but either way, shouldn't they, or Philips have basic knowledge of their own products, or at least the wherewithal to go to Philips' own site, or as a last result, use Google to identify the parts being referenced?

Was that really too much to expect?

 

[more_vampires]

I've provided the name/model line/type/part # of the items in question, and now they've written back requesting pictures of the parts I referred to?

Having done some tech support over the phone before, asking for pictures of YOUR parts in YOUR hand can go a long way to resolving confusing questions.

It can help identify counterfeits and other things, too. I've had customers send me pictures of things that I don't sell before, as well as pictures of counterfeit stuff they didn't get from me.

If they want a picture of the packaging and device, I'd suggest indulging them... especially if they're trying to talk to you.

 

[-Virgil-]

The polarity question is an interesting one. I recently received some interesting LED bulbs to examine, made by this Taiwanese company. The bigger bulbs (to replace 1157, 3157, etc.) don't produce enough light, but might eventually get there. The one that made me perk up my ears was the D009 designed to replace 194, 168, W5W, etc. Not only is its light distribution surprisingly even, but it's also not polarized; it lights up equally no matter which way around the test leads are connected. I tested its function in the front side marker light of a Chevrolet with its front side markers connected between turn signal and parking lamp (as described here). With polarized LEDs, this doesn't work; you can install the bulb so the side marker function will work or flip it over in the socket so it flashes with the turn signals, but not both because this hookup depends on the bulb lighting up no matter which way current flows through it. This unpolarized LED bulb worked just like the incandescent original, no matter which way it was installed in the socket. That's a pretty neat trick, and I imagine more of tomorrow's LED retrofit bulbs will probably be designed that way.

As for the Philips items, mine are now several layers deep in the box pit below my test bench, but I will dig them out and do some polarity experimentation to see what I see.

 

[Alaric Darconville]

The one that made me perk up my ears was the D009 designed to replace 194, 168, W5W, etc. Not only is its light distribution surprisingly even, but it's also not polarized; it lights up equally no matter which way around the test leads are connected.
---
This unpolarized LED bulb worked just like the incandescent original, no matter which way it was installed in the socket. That's a pretty neat trick...

A neat, but rather necessary trick, I'd say. Saves headscratching and hard-to-notice, harder-to-track failures.

 

[-Virgil-]

I agree. "Neat trick" might not have been the best choice of phrase; I think universal (un)polarity would be a basic requirement at least for wedge-base bulbs, and if they're going to design the drive circuit that way for those, they might as well also use it in the bayonet-base bulbs.

 

[Alaric Darconville]

As already mentioned, the irony is that these retrofits can. and probably do work very well in a variety of fixtures, but just haven't been verified to meet the objective requirements, which would be a difficult and expensive task.

Verifying that they meet the objective requirements in a variety of fixtures should be a comparatively *easy* task, by plotting the bare LED bulb's output against the original bare bulb. Once you demonstrate the LED 1157 replacement bulb has the same characteristics as the filament 1157, you can rest assured that it'll perform properly in any lamp assembly designed for the 1157.

 

[TheIntruder]

Philips CS (or more accurately, Archway) requested to pictures of the bulbs I was "referring" to, as a "favor."

That doesn't sound like they suspect I'm using knockoffs, selected the wrong application, or have other concerns; they just don't seem to know what I'm talking about.

And it was a very simple question -- are the VisionLED units polarized like the X-tremeVision units?

I understand that in providing support, context is very important, to gather as complete picture of the circumstances as possible, and determine what the real issue is. But in this case, I think it's a very simple, pointed question that anyone with even the most basic electronics knowledge should be able to answer, free of any other context. Are these things sensitive to polarity, yes or no? I'm not trying to determine if I have the correct bulbs, or why they won't work in the vehicle I'm testing them in; all of that has already been verified.

Granted, there is a possibility that a light went off and they were wondering "hey, this is a unique situation we've never heard of before, tell us more about the vehicle and other details." Or, "We'd like to see a picture of the bulbs in question, so that we might be able to visually determine a product revision that isn't reflected in any other fashion," but that's not the level of sophistication of the correspondence we're having.

While I'm glad, and thankful, that they didn't just blow me off (like many companies would), I just don't believe that this arrangement they have with outsourced support is set up to provide this level of support. I'm sure they'd have no problem providing general information, store location, or looking up the proper application from their script. But, they're likely trying to act as intermediaries between myself and some actual contact at Philips, and that's often not a very efficient, nor effective way of communication or problem solving. But, that's the nature of modern customer support, so it must be accepted.

 

[more_vampires]

Oh, nevermind then. Level 1 tech. Got it. That's no insult to anyone who has posted. It just sounds like what happened.

 

[TheIntruder]

Verifying that they meet the objective requirements in a variety of fixtures should be a comparatively *easy* task, by plotting the bare LED bulb's output against the original bare bulb. Once you demonstrate the LED 1157 replacement bulb has the same characteristics as the filament 1157, you can rest assured that it'll perform properly in any lamp assembly designed for the 1157.

So, then the real obstacles are allocating the resources to perform such testing, obtaining the variety of test subjects (or at least a pool larger than the rather pathetic "approval" list they have now, especially given they're had these product lines for at least a couple years), or the fact that these retrofits don't really match the characteristics closely enough? Or all of the above?

It's already been noted that the plastic wedge versions of the Vision remain sensitive to polarity, and I don't really expect Philips to be able to provide a different answer, if they ever do, but I still like to have all the cards on the table. And ponder the reasons they've gone from a single Rebel (which they're retained for the "white" P21 intended for reverse lights, and have added to their 12-year guarantee coverage) to the multi-LED array.

Cost is certainly one factor, but I'm curious about the others, and how performance compares as well.

 

[-Virgil-]

Verifying that they meet the objective requirements in a variety of fixtures should be a comparatively *easy* task, by plotting the bare LED bulb's output against the original bare bulb. Once you demonstrate the LED 1157 replacement bulb has the same characteristics as the filament 1157, you can rest assured that it'll perform properly in any lamp assembly designed for the 1157.

So, then the real obstacles are allocating the resources to perform such testing

Not yet. That's in the future when there's an LED bulb designed specifically to produce the same light distribution from a light-emitting surface the same size, shape, and position as the original filament(s). Osram's working on it, but we don't have it (from anybody) yet.

the fact that these retrofits don't really match the characteristics closely enough?

That's right -- it's why they only work in some lamps, not others.

 

[Alaric Darconville]

So, then the real obstacles are allocating the resources to perform such testing, obtaining the variety of test subjects

What I'm saying is that for any bare 1157 (not mounted in a lamp assembly designed to redirect the light anywhere just to measure it), there's a rather predictable output that could be plotted, and then the LED would have to be designed to try to replicate that.

Essentially, any filament bulb has a hotspot putting light out more or less evenly in all directions-- not quite 4∏ steradians, because the bulb has a base, and the filament itself has part of its output shaded by the contact wires and such... So the problem is to develop an LED device that emits the light in exactly the same pattern.

Replicate the "bare bulb" pattern and you guarantee compatibility in every lamp using that same bulb you've replicated.

 

[TheIntruder]

Not yet. That's in the future when there's an LED bulb designed specifically to produce the same light distribution from a light-emitting surface the same size, shape, and position as the original filament(s). Osram's working on it, but we don't have it (from anybody) yet.

While I'm glad that Philips (and Osram, apparently) are working toward the possibility of universal retrofits, I do wonder whether there's a business case for them.

In other applications where incandescents are being displaced, there are some tangible benefits, aside from the regulatorly push.

For automobiles, in theory, LEDs might result in a marginal improvement in fuel efficiency, but any decrease in power requirements is probably more than offset by the multitude of both functional and comfort equipment in a modern car.

And the retrofit market is limited by cost, and appeal.

Essentially, any filament bulb has a hotspot putting light out more or less evenly in all directions-- not quite 4∏ steradians, because the bulb has a base, and the filament itself has part of its output shaded by the contact wires and such... So the problem is to develop an LED device that emits the light in exactly the same pattern.

Replicate the "bare bulb" pattern and you guarantee compatibility in every lamp using that same bulb you've replicated.

What do you make of the move from a single Rebel (which more closely approximates a point source) to the array in the newer product?

I suppose that, in a signaling lamp, focus isn't as vitally important as in a headlamp.