Is this a Diode?

[t1d]

I am reverse engineering a boat navigation light, in order to repair it for a friend. It is a typical LED circuit. All LEDs are 5mm/3.2~3.4v. All Resistors are 1k. The VR is a 78L12. The thing which resembles a diode is the mystery...

I have used a diode icon as a place-holder on the schematic. However, I am not certain what I think the part may truly be... It is the common shape of a black, cylindrical diode, but it has absolutely no markings on it, not even a polarity band. It does measure as polarized and has 2M of resistance.

A diode for upstream electrical protection seems a bit unlikely, as a diode drops a volt. A fuse doesn't fit the criteria, as it would not be polarized.

The PCB has multiple holes for this component, to allow various parts to be sourced, based on pricing. The back of the PCB has SMD pads for it. That is marked M1.

Any ideas what it might be? Thanks for the help!

[Gyro]

Are you sure this is operating on a 12V system? I suspect it's running on 24V. The 78L12 won't start regulating until the input voltage gets to nearly 15V. It's fairly common to use a series diode for cheap reverse polarity protection and if running on 24V its voltage drop would be of no consequence. I suspect that the PCB allows direct connection for 12V systems and adds the 78L12 for 24V input.

[jeroen79]

It does measure as polarized and has 2M of resistance.

Did you measure resistance in both directions?
Did you measure the voltage drop with the diode test as well?

[doktor pyta]

HI,

this should be a Schottky diode with big junction.

P.S. The schematic is drawn in so unintuitive way that even I cannot stand it.

[Circlotron]

It is probably drawn the same way the pcb is laid out.
Or it is meant to look similar to a typical car wiring diagram. 

[Powermax]

It was hard to figure out that schematic, but it appears that it is just 4 strings of 2 LEDs and a resistor powered by a 12V voltage regulator. I would suspect it is a diode, and also that if you plan on operating just on a fixed voltage, that regulator is not necessary. LEDs are supposed to be powered with constant current LED driver anyway, not a low impedance constant voltage output of a 7812.

If you are repairing it anyway, I recommend hacking in brighter, better LEDs. I am not a fan of the crap 5mm LEDs, other than for use as indicators. Consider a few 1W LEDs, or the better 5mm LEDs you can find (by that I mean not the ones on eBay, which are garbage.).

[t1d]

Thanks for all the replies!
@ Gyro - It is a straight 12vdc system.
@ jeroen79 - I measured the resistance in both directions. That's how I discovered the polarity. It measured open and 2M. I did not measure the voltage drop, but I could, if you needed me to; I was just speaking of the typical voltage drop across a diode.
@ doktor pyta - Well, I have learned that some manufacturers use the "M1" designation for their Schottky diodes. Please tell me more about using a Schottky diode in this type of circuit. What do you mean by a "big junction?" My apologies for the layout of the schematic, but I was not trying to draw a proper schematic, I was trying to understand the circuit.
@ Circlotron - Yes, see above.
@ Powermax - I agree with your analysis... Unless doktor pyta teaches me some significant benefit to having a Schottky diode in the circuit, I will just configure the circuit, in a typical manner.
Thanks to all for your help!

[batteksystem]

Thanks for all the replies!
@ Gyro - It is a straight 12vdc system.
@ jeroen79 - I measured the resistance in both directions. That's how I discovered the polarity. It measured open and 2M. I did not measure the voltage drop, but I could, if you needed me to; I was just speaking of the typical voltage drop across a diode.
@ doktor pyta - Well, I have learned that some manufacturers use the "M1" designation for their Schottky diodes. Please tell me more about using a Schottky diode in this type of circuit. What do you mean by a "big junction?" My apologies for the layout of the schematic, but I was not trying to draw a proper schematic, I was trying to understand the circuit.
@ Circlotron - Yes, see above.
@ Powermax - I agree with your analysis... Unless doktor pyta teaches me some significant benefit to having a Schottky diode in the circuit, I will just configure the circuit, in a typical manner.
Thanks to all for your help!

Junction = p-n Junction, However, Schottky is partly metal, partly semi-conductor abomination. I think the big just mean it is high power. Depends on the current of LED, I would say either a SS14 or SS16 would fit right in.

[t1d]

Depends on the current of LED, I would say either a SS14 or SS16 would fit right in.

Thanks, batteksystem. I will look for those Schottkeys...

Progress:
I have laid out the board in a more sensible manner, increased the trace widths, added top and bottom ground planes, for heat dispersal, tweaked the board dimensions and ordered the new boards from OSH Park. I have also ordered the LM78L12 voltage regulator from Mouser.

[Seekonk]

If 24V I wouldn't be surprised if it was a zener to reduce the voltage drop on the regulator and reduce heating.  It would also protect regulator from higher voltages the system might see.

[t1d]

If 24V I wouldn't be surprised if it was a zener to reduce the voltage drop on the regulator and reduce heating.  It would also protect regulator from higher voltages the system might see.

This is the confusing part of the existing design. The light is built for recreational boats in the USA and they operate on 12vdc. If this is a diode, and I really think that it is, any voltage drop seems wrong, for this many LEDs...

[Gyro]

That's what's confusing me - regardless of any diodes, an LM78L12 regulator will not regulate with 12V input, it needs around 2.5V min input - output differential. At best it's sitting there dropping voltage, but it isn't regulating.

[LaserSteve]

"crap" discrete 5 mm LEDS are often used in Nav lights for redundancy and a uniform 90, 135 or 180 degree pattern that Meets Coast Guard and 72 ColRegs  Spex.  Sometimes it's very difficult to meet the "Acient" pattern specifications with modern leds.

S.

[StillTrying]

I don't get it, what's the point in building something that's not going to work properly.

[ConKbot]

It's probably just a diode to keep it from getting trashed if it's hooked up wrong when someone is installing it. "Try it, if it doesn't work, try the other way" and a 3 cent diode is a lot cheaper for the manufacturer then dealing with returns and whatnot. Plus in an older boat would the polarity even be marked on the wires running to a incandescent bulb?

[t1d]

Thanks for all the great insights...

That's what's confusing me - regardless of any diodes, an LM78L12 regulator will not regulate with 12V input, it needs around 2.5V min input - output differential. At best it's sitting there dropping voltage, but it isn't regulating.

Well, the regulator makes some small amount of sense, as the circuit will see about 14.6vdc, with the motor and alternator running... Higher spikes are possible, and likely...

[Gyro]

Well, the regulator makes some small amount of sense, as the circuit will see about 14.6vdc, with the motor and alternator running... Higher spikes are possible, and likely...

Hmm, good point, although a 78L12 has lousy input overload voltage capability and there is no input resistor or capacitor to help mop up the spikes.

I think I would treat this as an opportunity to design it 'properly' - remove the regulator, size the resistors correctly to keep the LED currents at a safe level over a 12-14V input range. Then Include an input resistor (fusible or PTC?) and Zener clamp to take care of spikes (and reverse polarity at the same time).

[t1d]

I think I would treat this as an opportunity to design it 'properly' - remove the regulator, size the resistors correctly to keep the LED currents at a safe level over a 12-14V input range. Then Include an input resistor (fusible or PTC?) and Zener clamp to take care of spikes (and reverse polarity at the same time).

So very true...