VX Bulbs

[djkymar]

Etl. if you scroll down that page its almost in the middle and it cost only 6 bucks

[spoondeep]

Happy New Year everyone.

I thought I'd reboot this thread, and pass on the fruits of recent research and testing. Dec 2018~Jan 2019.

Side note to the mention of "Sylvania" above. I would strongly suggest avoiding Sylvania, I'm actually amazed they are still in business. I worked with their lamps for over 25 years, and usually having to repair and rebuild lighting modules due to the end user opting for cheaper Sylvania lamps over the recommended GE or Osram lamps. So many of those customers saved $10~20 per lamp, then faced repair bills of $700~$9000, due to Sylvania melt downs, explosions, or burning patterns damaging the other optical components. Granted my company made a LOT of money due to those situations, but it never sat well with me to profit under such circumstances, where really the end users just fell for the Sylvania marketing and quality promises.

Personally having seen that Sylvania crap over 25 years, and never seen that happen with GE or Osram, (assuming the user cleaned the envelope, and gloved up), I would never consider Sylvania as a vendor for anything you care about, or care to rely on.

OK, the purpose of this email wasn't to bash Sylvania, but rather to pass on some testing data and recommendation for the VehiCROSS 194 lamp, an LED option suited for high brightness for the parking lamps, with what is available in 2019.

I'd also set straight a common misconception on colour temperature. Generally the recent fashion seems to be people fitting lamps/leds, with "white" light, say 6000K or higher, even 10000K, sigh. (above 5000 is increasing through blue)
However if you're expecting to SEE things in the dark while driving, a lower temperature is far more effective. 4000K~5000K
This has NOTHING to do with lumens, lux or absolute level of emission, but rather how the human eye works.

Remember all those submarine movies, where they switch to red light 30~60 minutes before surfacing at night? Red is <1000K
Well, this is for very good reason. In low light conditions, the human eye peripheral is far better at "seeing" than when one looks directly at something. Using red light allows the eye to relax, and adjust to the dark, and the red end of the spectrum is ideal for this, rather than a broad spectrum light. To be clear, "broad spectrum" IS white light. With a relaxed eye, on low light conditions, (red), the iris will fully dilate, and this allows more light to hit the retina, and periphery of.
The problem with "white", (means to the human eye >4500K), is it triggers the eye reflex to close the iris down, limiting the reception of reflected light reaching the retina.

So, basically while white lamps might be sold as brighter, and measure as such on a lux-meter for example, it does not follow that you will see as well with them, especially at night.
In fact, with the colour temp down at say 4300K and say 5000 lux, you will see better than say a 7000K lamp with 8000 lux, assuming there are no other light sources, such as streetlights or other cars, with can and probably will be filling in the warmer spectrum.

So, I decided to do a little experiment with the driving lamps on the VehiCROSS, (lamp 194)
Initially I had problems with setting up the measurement rig, ambient light changes, and also battery drain on the Vehicross, (testing took several days), so I shifted testing to my lab, using 2x left side parking lamp modules (8972071920) and compared the original type 194 filament lamp and some 15 different leds. I used a benchtop PSU, set to 13VDC and current limited to 5A as a safety precaution. Note even the original 194 lamp only pulled about 0.3A, and all the leds a lot less.

Initially I started measuring at 1 meter which is a typically used distance, but found that to be problematic due to the reflector and scatter pattern of the Vehicross module, and measurements varied too greatly with just a slight shift or angle change on the lux meter.
I was not so concerned with absolute readings, but rather a stable, repeatable comparative reading.
So, more time was spent optimizing the test rig, and switching lamps between the two modules to check of comparative readings were possible.
In the end, I measured at 10cm directly in front of the module, modules were fixed in place, and a focal plain board fixed at 11cm from the top of the module lens. My lux meter has a base depth of 1cm, so absolute distance was 10cm to the top of the meter "lens".

I also shifted the lux meter around on the focal plain, and tweaked the angle slightly, <10deg to find the peak output from each led.

The key point to remember here, and in fact with ALL halogen lamp alternatives is, the light modules on your car have reflectors and lenses are designed for a halogen lamp. Those legacy lamps have a tungsten filament, a coil with a specific diameter and pitch, width, and protrusion depth. Also the glass envelope is in the case of a 194 is a spherical tipped cylinder, so the radiance pattern tends to be more focused out the top, compared to the side radiance.
Those lamps also tend to have a dead spot either end of the filament, which is why the lamp holder is keyed, as this points the dead spots into sections of the reflector that (should) try to compensate and back scatter to reinforce the dead spots.

Testing results of LEDs varied greatly, with the original halogen lamp kicking out about 900 lux, but the winner at this stage is not so surprising, as it incorporates an lens for forward illumination, (similar to the halogen lamp), but also an inverse parabolic reflector, that does a fine job of evenly radiating 90deg and that allows light to illuminate the light module reflectors, and those ensure the photons are emitted out the front of the module lens, rather than just bouncing around inside the light module.

Original filament (halogen?) lamps were 600~900 lux, I have no idea what make, brand those old lamps were, but even a new Bosch lamp was ca 850.

In 1st place: These babies kicked out 7950 lux, (warm white) which was more than double the next best.
https://www.ebay.co.uk/itm/2-x-W5W-T...zOhj:rk:1:pf:1

In 2nd place, at 3520 lux, (white) but the key point in the failure of all similar types is lack of reflectors and lenses, preventing the light from reaching the reflector surfaces inside your light module.
https://www.ebay.co.uk/itm/VW-SUPER-...72.m2749.l2649

3rd place was a glass enveloped COB led. 2150 lux (warm white)
https://www.ebay.co.uk/itm/2x-501-W5...72.m2749.l2649

I won't post a list of all the others tested, because really they were so crap in comparison, and all just leds on boards, with just the led count/type varying the light output generally.


I am still waiting on 3 more led orders to arrive, so will post a followup if any of those perform well in the Vehicross parking lamp module.

But equally, I will do some more testing on the other Vehicross lamp modules, the driving light, reversing, and number plate lights. As these have their own unique reflectors and lenses, so some of these runner up leds might measure well in those other fittings. It's all about the ability of the led to emulate the original filament lamps, but equally how well the light module reflector and lenses were designed.
One needs to consider in cars, a lot of the design is constrained by aesthetics and design "theme" and the intended style.

If lights were all perfectly optimised for "perfect" illumination, they would all be round, with a glass/crystal parabolic reflector, and high purity crystal lens. So really anything that isn't at least round for starters is already heavily compromised in terms of optical efficiency.

I have a spare rear light module, number plate module, and also a USA Vehicross side indicator, so I might also get some amber leds tested.
But the key thing in this post was more about driving lights, as those are illuminating the environment, and we're expecting to SEE WHAT they illuminate, (same for reversing lights), while indicators are intended to be directly viewed, i.e. we look at THEM, not what they light up.

If anyone has any other LED alternatives you feel perform well, please let me know, and I can probably test them if I have not already done so.
Cheers, happy night driving.



Which brings me back nicely to my point above, sure your 10000K headlights might look brighter when you look AT THEM, but that is not the purpose headlights are serving.

P.S. Updated this as I realised I'd pasted in the data offset by 1 place.

Also, some other contenders for comparison.

Very bright when viewed by eye, end on etc, but actually only measured 910 lux.
https://www.ebay.co.uk/itm/2x-T10-50...72.m2749.l2649


This did not appear very bright end on, but did quite well in the Vehicross parking light module, so I include it here, as this illustrates how the number and orientation of the LEDs, can benefit from the Vehicross reflector, but even so, as dim as it looked compared to the led above, it also only measured in at 910lux.
https://www.ebay.co.uk/itm/Vw-Super-...72.m2749.l2649


I'm waiting on these 2, potential contenders, and the first two from Philips I will have tested, although I suspect aside from the different packaging, they will be identical.
https://www.ebay.co.uk/itm/PHILIPS-G...72.m2749.l2649

https://www.ebay.co.uk/itm/Philips-X...72.m2749.l2649


And finally, this is the other yet to arrive, and I don't hold high hopes, but along with the warm white leds, and the other COBs tested, I expect to to benefit from gains in the Vehicross reflector, and punch above its weight.
https://www.ebay.co.uk/itm/10-X-T10-...72.m2749.l2649

Update to follow.

[ipd]

Good info. Thanks for sharing. Keep in mind that ebay links will decay and it may be better to just type the brand/model of the bulb used. Amazon links usually just show "no longer available", but ebay's tend to die entirely.

I've used Sylvania blue LED's in my dash previously, and was quite satisfied with the result. I can only imagine their white LED's are even brighter. They looked similar to the Phillips ones you posted the link to.

I'm currently using some Chinese-made LED's because I wanted to use purple fog/turn/horn lights--and purple LED's are still like hen's teeth. There's a significant sacrifice to brightness with purple--and the high-end companies don't seem to offer this color...at least yet.

I have these in my front turns at present, and they are quite a bit brighter than the 194 style purple bulbs, and are well suited for the task. You WILL have to replace the stock blinker module with an aftermarket one for LED use--unless you like strobe-fast blinkers.

https://www.amazon.com/gp/product/B0...?ie=UTF8&psc=1

Keep in mind that I have smoked, clear lenses on my turns. I would love to find some clear side marker lamp lenses so that I could replace the amber with purple there also.

Oddly, the bulb that I have in the climate control and shifter is SIGNIFICANTLY brighter than the ones behind the instrument cluster.

https://www.amazon.com/gp/product/B0...?ie=UTF8&psc=1

The bulbs I have in the dash are the 2-stack varieties of these:

https://www.amazon.com/gp/product/B0...?ie=UTF8&psc=1
https://www.amazon.com/gp/product/B0...?ie=UTF8&psc=1

I had a 1-stack variety, but they were lackluster. I still have the 1-stack kind in my license plate and the light quality is sub-par. The 3-stack kind are used in the fog lamps and i believe the horns as well.

Pics for reference (crappy phone):





p.s.
Cree are supposed to be REALLY good LED's. I have these currently installed--replacing my "brite-*** LED's" which failed in <2 years.

https://www.amazon.com/gp/product/B0...?ie=UTF8&psc=1