i'm open to all suggestions.
but may have a problem on the switches.
i'm limited to round rockers so they fit my housing.
srj24a3bbbnn is what i'm using now.
i prefer to keep most things on the circuit board instead of added labor of extra wires and associated confusion during assembly.
maybe it'll be safe with 60v fets and 15v zeners.
let me try this circuit first when the boards get here.
thanks

 

i'm open to all suggestions.
but may have a problem on the switches.
i'm limited to round rockers so they fit my housing.
srj24a3bbbnn is what i'm using now.
i prefer to keep most things on the circuit board instead of added labor of extra wires and associated confusion during assembly.
maybe it'll be safe with 60v fets and 15v zeners.
let me try this circuit first when the boards get here.
thanks

Cool, I'm pretty optimistic that the circuit you've got now will be just fine. I only mentioned the alternate switches as a tangent from DECELL's comments and your concern after them. If you're ok with the size and parts count of the current project, I certainly don't have any problem with it.

There's definitely something to be said for using easy, generic, interchangeable parts too. If the alternatives require special MOSFETs or rare switch configurations, then you're far more likely to have supply chain issues at some point in the future.

As it currently sits, your entire design (or at least everything but the LED driver) is made of very standardized parts that you could get from several manufacturers and a gazillion vendors!

 

still waiting on boards but i thought about trying this.
will it work? what do you think?
all the mosfets and zeners add too much cost and complexity.
and i'm still scared of blown zeners and fets.
lower forward voltage allows me to use ss14 instead of Nfet for polarity protection.
the driver chip has open and short protection. would this arrangement stress the chip?
i only lose 1 function of being able to turn off half of string 1 and i can live with that.
i know current in string 1 driver chip would need to be doubled.
i thought of this myself...is this old hillbilly dreaming?
thanks

 

still waiting on boards but i thought about trying this.
will it work? what do you think?
all the mosfets and zeners add too much cost and complexity.
and i'm still scared of blown zeners and fets.
lower forward voltage allows me to use ss14 instead of Nfet for polarity protection.
the driver chip has open and short protection. would this arrangement stress the chip?
i only lose 1 function of being able to turn off half of string 1 and i can live with that.
i know current in string 1 driver chip would need to be doubled.
i thought of this myself...is this old hillbilly dreaming?
thanks
View attachment 152937

I must confess that I'm having a hard time keeping up with ever changing user requirements here! Nevertheless, it's totally up to you what functions you do and don't need, so I won't question that.

One thing that jumps out at me immediately is that you've got two strings of LEDs in parallel with each other, but you've got nothing to determine how they share the current. LEDs are known for not sharing current well. If one string has slightly lower Vf, it will draw more current, which heats it up, lowering its Vf further, making it draw even more current, etc. This can lead to a very lopsided current distribution, with one string getting much hotter than expected.

There are some people on this forum who disagree, but it seems that the majority agree you should have a resistor in series with each string to insure that they divide the current between them fairly evenly (or have dedicated LED driver circuits for each string if you don't want the increased heat and inefficiency of series resistors.)

I don't know enough to comment either way on changed protection schemes. Hopefully crutschow will have some insights.

 

original goal was to have a choice of turning on or off 2 LEDs out of 6.
with former arrangement overhead was insufficient.
that's why i asked about replacing diodes with fets.
i've run parallel strings of LEDs with no problem with a current control driver chip.
this design eliminates overhead problem if there isn't some other problem i don't know about.
thanks