LED Bulbs Already Failing

[Rockers]

Update. First screen we exposed this morning, failed right away. What I`ve noticed now is that the small blue light on the exposure unit that indicates exposure comes on and the countdown kicks in too but the LED lights start up 10 seconds later. And my guess is because the room temperature is too low. This should really not happen on an expensive piece of equipment like this.

[jvanick]

If the timer isn't consistent this could explain your other issues too...

Do you know if there's a vacuum draw down sensor or some kind of sensor on the fans? 

Seems really wierd that a digital timer would be affected by colder temps.

[Rockers]

I don`t think the timer is affected. It`s rather a strange set up on the unit. Once the vacuum counter has reached zero the exposure timer kicks in regardless of the LEDs being on or not. There is just nothing that tells the timer if the LEDs are actually working. 

[ScreenFoo]

What strikes me as extremely odd is that the vast majority of electronics not only work more efficiently, but tend to last longer at temperatures near but above freezing. 

Almost sounds like you have an iffy power supply or driver...

[alan802]

What strikes me as extremely odd is that the vast majority of electronics not only work more efficiently, but tend to last longer at temperatures near but above freezing. 

Almost sounds like you have an iffy power supply or driver...

I thought that we might have had some sort of power issue because the north end of our screen printing building was underpowered in the summer months when we'd start running the personal AC units.  We had the electricians in and we essentially let them do everything they suggested and that included having a line ran dedicated specifically to the expo unit and only it.  So if there was a problem before, there isn't one now.  That day with the electricians ended up costing about $5K with all of the things they did.

Oh, and nothing changed with the expo unit unfortunately. 

[244]

What strikes me as extremely odd is that the vast majority of electronics not only work more efficiently, but tend to last longer at temperatures near but above freezing. 

Almost sounds like you have an iffy power supply or driver...

I thought that we might have had some sort of power issue because the north end of our screen printing building was underpowered in the summer months when we'd start running the personal AC units.  We had the electricians in and we essentially let them do everything they suggested and that included having a line ran dedicated specifically to the expo unit and only it.  So if there was a problem before, there isn't one now.  That day with the electricians ended up costing about $5K with all of the things they did.

Oh, and nothing changed with the expo unit unfortunately.

Power should not be an issue. The LED's run off of a power supply that should keep the power constant. Just a FYI

[ScreenFoo]

Power should not be an issue. The LED's run off of a power supply that should keep the power constant. Just a FYI

I'd agree most every power supply should keep the power constant, most especially constant current drivers used for LEDs.

I guess, it's not like the timer section couldn't screw things up when it's cold, would just seem like a much less likely possibility, IMHO.

FWIW, I wish I had some 'oh, probably a bad this or that' theory on Alan's board, but that just sounds like poor assembly or driver QC to me...

edit:  Just realized there are TWO threads here, and I'm in the wrong one. 
Sorry Alan, didn't mean to confuse you.   

[alan802]

We were grasping at straws honestly.  I can say I've tried everything in my power to make sure we weren't the problem. 

No problem, I was wondering what was going on but don't worry, I hadn't gone full-confused yet.

[blue moon]

We were grasping at straws honestly.  I can say I've tried everything in my power to make sure we weren't the problem. 

No problem, I was wondering what was going on but don't worry, I hadn't gone full-confused yet.

what about getting replacement strips from Vastex to see what that does?

pierre

[jvanick]

Pierre -- with all your testing goodies, do you know of a (low cost) device that will give a spectral output of a light from say 300nm up to 450nm

heck, what's even the name of a device like that? 

[Binkspot]

Depending on the transformer or power supply the output voltage can vary. Just a basic transformer that's wound for let's say 120vac in/ 4vac out may only be putting out 3.5vac when connected to 110vac. There may be another set of taps to connect the primary for different voltage ranges. Or use a "smart" power supply that can produce a constant output voltage by monitoring it self.
As far as Rockers, may have a two fold issue. The higher voltage and lower frequency of the power being supplied in Japan. This may not have been figured in the design.
Small flucations in voltage may cause the UV wave length to out of the ideal range of your emulsion.

[ebscreen]

Pierre -- with all your testing goodies, do you know of a (low cost) device that will give a spectral output of a light from say 300nm up to 450nm

heck, what's even the name of a device like that?

https://learn.sparkfun.com/tutorials/ml8511-uv-sensor-hookup-guide



Been meaning to get around to it.

[jvanick]

Pierre -- with all your testing goodies, do you know of a (low cost) device that will give a spectral output of a light from say 300nm up to 450nm

heck, what's even the name of a device like that?

https://learn.sparkfun.com/tutorials/ml8511-uv-sensor-hookup-guide

Been meaning to get around to it.

that's pretty cool.  while it only measures intensity, you could likely use this to build a light integrator.

I'm more looking for something that maps the output/intensity in NM

kinda like:

380nm = 0%
381nm = 2%
..
.
.
390nm = 100%
.
.
.
400nm = 0%

etc...

that way you could actually figure out if the spectral output is changing over time as the LEDs or light source ages.

would also be helpful for reverse-engineering why some of these units are providing such different (poor) results.

[ebscreen]

I'd imagine something that maps the spectrum ain't coming cheaply.

You could likely use physical filters to narrow your readings to the desired spectrum, and then measure intensity.

At the very least it would be useful for hot/cold spots.

[ScreenFoo]

I was thinking filters would be the way too--but I couldn't find any.  Then I started looking at this:
https://en.wikipedia.org/wiki/Fluorescence_spectroscopy

I keep thinking if you could find the proper responsive dyes/pigments to 'map' the different wavelengths, all you'd need is some geometry in a jig, a nice camera, and some PS chops.