The FrankenBebi Project

[stevensc]

Capt. Couillon, I though you might be interested to see the boat of a guy I know in Houma, Louisiana. The boat's name is the Maudit Couillon
Steve W

[Paul Elliott]

Originally Posted by capt-couillon

Evening Paul, thanks for your comments and interest.

While ballast resistors have been brought up several times, here are some of the factors that were considered prior to the decision not to use them. [...]

I understand. Can you disclose the Cree part # for the LED, and the binning specs? I would be interested in seeing the I/V curves for these LEDs. As I mentioned, if the slope isn't too steep perhaps current-sharing won't be an issue.

[capt-couillon]

Capt. Couillon, I though you might be interested to see the boat of a guy I know in Houma, Louisiana. The boat's name is the Maudit Couillon

Yeah.... nice ride, but de Darned Fool got dem fender on the wrong side de bateau.... Course he is from Houma.

For them what wasn't fortunate enough to grow up round 'Ti Mamou, here is a further definition of "Couillon" and my credentials

Can you disclose the Cree part # for the LED, and the binning specs

Sure Paul... We are using C503Ds with a Vf bin of 29. You can grab the spec sheet here

[Paul Elliott]

Originally Posted by capt-couillon

We are using C503Ds with a Vf bin of 29. You can grab the spec sheet here

OK I agree that you probably don't need extra ballast resistors. If I set one 3-LED string at minimum voltage spec, and the other four strings at max voltage spec, the one string gets about 30mA and the other four each get 18mA (rough calculations).

With this worst-case scenario you are running one string at maximum rated current. At a high ambient temperature you will exceed the spec. But, the chances of this worst-case diode distribution are microscopic. With normal tolerance distribution things should be fine.

[capt-couillon]

But, the chances of this worst-case diode distribution are microscopic. With normal tolerance distribution things should be fine.

Thanks Paul... glad to have someone else check (and validate) our assumptions.

[capt-couillon]

New news at the FrankenBebi Project

Cheops' Law: Nothing ever gets built on schedule or within budget
We have had some changes in our release dates. For the full story, follow the above link.

Documentation: Now available online
Documentation pages are now online at the website. This area contains zip archives of each posted version of the FrankenBebi up to and including the latest untested v0.11.03. Each file includes the native Eagle Lite sch and brd files, standard image files of the schematics and board traces and layouts, and additional files concerning licensing and appropriate use.

Download and enjoy.

As always, if you have questions or comments, please contact us via this thread, a PM, or via the Email address listed on the website.

[Bestathook]

Have you decided on a good filler for the completed light yet?

Also, have you considered using a larger PVC end cap to give a bit more working room? If the larger cap was only partially filled, there would be room to mount it on a piece of pipe

Beeswax or the wax from a toilet bowl sealing ring make a good semipermanent seal on PVC slip fittings, that can be released with heat.

[capt-couillon]

Have you decided on a good filler for the completed light yet?

We will be using a commercial epoxy based potting compound due to cost considerations. For the kit (and or your own DIY design) we expect to be recommending plain old filled marine epoxy (not polyester) resin with a slow or tropical hardener. You can fill the epoxy with cabosil, silica, or whatever filler you have on hand as long as it is stll thin enough to flow around the board and into the base of the cap. Do Not Use fast hardener, or pour in over 70 degree temperatures. The exothermic reaction properties of curing epoxies can damage the electronic components and or stress the solder joints due to over expansion. For the same reason, we will be recommending that when using standard filled epoxies, the cap be filled by pouring at least 3 equal layers with 5-10 hours cure time between layers. 2" of fast cure epoxy in a paper cup on a hot day can generate enough heat to set the cup on fire and expand to 4 times its initial volume. Thin layers, slow cure, cool environment = no problems.

Also, have you considered using a larger PVC end cap to give a bit more working room? If the larger cap was only partially filled, there would be room to mount it on a piece of pipe

You've been reading our mail !..... From early on in the project, the cap size was increased from the original OWL size of 1" to a size of 1.5". The actual inside dimension of a 1.5" solvent well pvc pipe cap is 1.9". This leaves plenty of room for our 1.4" PCB and allows you to insert the board with the LEDs already soldered in place and then gently press the LEDs into the holes. The trick is in the little spring-bend we put in the LED legs prior to soldering. How to make this bend is included in the Kit instructions, but we gotta keep some secrets As you suggested the potting compound will stop ~0.5 inches below the lip of the cap so a standard PVC pipe thread adapter could be glued in place.

However, this type of mounting will preclude the use of the integrated day-night switch as the photo-transistor actually is mounted in the open end of the cap. You could install the mounting adapter but it would "blind" the phototransistor resulting in the LEDs being on all the time. One option would be to hard mount a basic non photo-switch version and use an external photo-switch. We are hoping to be able to release a stand alone version of our photo-switch sometime in the future as it seems to be a nice reliable low cost piece of gear that could have other uses... The mosfet we are using as the actual "switch" is rated to 10A so it could be used for existing incandescent anchor lights as well.

If we take too long in getting around to developing this particular product, it's open source so you can use our prior work to package your own product

As always, thanks for the interest and suggestions.

[hellosailor]

Cap-
I thought epoxy was asking for a UV problem? And that something like acrylic or urethane casting compounds were recommended to withstand UV exposure? No?

[capt-couillon]

You are correct, discoloration, chalking and loss of "gloss" due to UV exposure can be issues with epoxies when used in applications where aesthetics are an issue. However, these effects are not really relevant when epoxies are used as a potting compound.

Dark pigmented epoxy resins are more resistant to these issues than light or non-colored resins. The commercial compounds we are using are already colored black as purchased. If you are using a slow cure marine epoxy resin and are concerned about discoloration, West Systems research dept has recommended the addition of "One-Shot™ Lettering Enamel" in a loading ratio of 10% by weight as an effective method of coloring their clear resins with no significant changes in physical properties. DO NOT USE West 423 Graphite as a coloring compound in this application. The conductive properties of graphite will wreak havoc with the embedded electronics.

[hellosailor]

Do not use graphite? You mean, surrounding the LEDs with black potting compound might not be a good idea? <G>

[colemj]

The problem is in surrounding them with an electrically conductive potting compound, not a black one.

Mark

[Bestathook]

"If we take too long in getting around to developing this particular product, it's open source so you can use our prior work to package your own product "

If you take much longer than I do, you will come to a complete stop.

I like the development of a device, and the physical building of it.

But when it is as perfected as I can get it, and I have made myself as many as I need, I'm off to a new project.

But no one has ever accused me of excessive speed in the process.

[hellosailor]

Yes, mark. I was ignoring the electrically conductive aspect, because if the potting compound is black, well, whether there's any conduction going on inside is a moot point, isn't it?

[colemj]

I guess I don't understand your point about the color of the potting compound. We have electronics on board potted in all sorts of colors, including black.

Mark