hi all,
i am building my own outdoor landscape lighting supply/controller. i have a problem that needs a solution. the whole thing only gets powered on from dusk to dawn using a photocell as a sensor for dusk to dawn. i am to buy 4 Cotek DN style power supplies (120vac in, 12vdc out). each DN has a inrush of 30A and then will run at 1A or less.

i dont know the actual inrush, but from spec that would be 120A of inrush if all powered on at the same time.

anyone have idea on how to cascade turn on, maybe 1sec between each? the DN's have a output relay that closes when the output turns on (output itself has small delay turn on). one idea i had was to daisy chain the DN's using opto SSR, so #1 DN relay will turn on a SSR and that SSR feeds 120vac to the next DIN.

another idea i had was to use a cap-resistor with a LM339 (multi threshold config) where each threshold controls SSR.
kiss model applies.

your ideas??

 

or what about simple 4 component using series resistor-cap for the delay, and center tap that with a zener that biases a small fet or bjt? the fet or bjt control the SSR opto side current? this is simple.

something like this?

 

I thought "Why not just let each transformer turn the next on via a control relay. The two arrows go to the lights while the relay doesn't switch the next transformer on until the previous has already turned on. By the time it's outputting it's voltage the transformer has finished turning on. And each relay is going to add its own delay (not much).

Are ALL your transformers on a single light sensor? If they're on individual sensors (one for each transformer)? If individual then chances are they won't all trip at the same time. And if two DO happen to trip you can always put a very thin dab of black RTV (gasket sealer) over one of them and it will turn on earlier and stay lit longer.

 

the light sensor is used to turn on the 1st power supply, from there i have 12v to use for turning on the rest. i would not use mechanical relays because this stuff is outdoor stuff, shielded from weather but temp swing is ~25F to ~115F, and, it will cycle every day, etc.

keep in mind, the power is a industrial 12vdc DIN mount power supplies. i believe the Cotek DN series has a delay between inrush and 12vdc output, so i can probably take use 12vdc from the 1st to turn on the 2nd via opto SSR, the issue is, SSR's that are 40A inrush rated are fairly large and pricey. i only need 2-4A(nominal) 120vac SSR.

 

the light sensor is used to turn on the 1st power supply, from there i have 12v to use for turning on the rest. i would not use mechanical relays because this stuff is outdoor stuff, shielded from weather but temp swing is ~25F to ~115F, and, it will cycle every day, etc.

keep in mind, the power is a industrial 12vdc DIN mount power supplies. i believe the Cotek DN series has a delay between inrush and 12vdc output, so i can probably take use 12vdc from the 1st to turn on the 2nd via opto SSR, the issue is, SSR's that are 40A inrush rated are fairly large and pricey. i only need 2-4A(nominal) 120vac SSR.

TRIACs are rugged, pretty cheap, and can take a pretty high MT2 to MT1 current hit for a little while, with no damage, so if you can find out from Kepco how long the supply inrush transient lasts, you might be able to do something, like this, with a four quadrant TRIAC:



From here

 

or what about simple 4 component using series resistor-cap for the delay, and center tap that with a zener that biases a small fet or bjt? the fet or bjt control the SSR opto side current? this is simple. something like this?

Looks good to me, but I recommend the LM339 approach.

If you change to darlington transistors you will get more snappy on/off operation, and require less base current. That means the timing capacitor can be smaller because its source resistor can be larger. For outdoor/every day operation, I recommend solid state relays over mechanical ones. Since they take only 3-10 mA to operate, you can drive them directly with an LM339 quad comparator for more precise operation. Note that the R-C charging ramp is non-linear. Equal-value setpoint resistors will cause a non-constant time delay between each stage, but that probably doesn't matter in your application as long as all 4 stages are on within, what, 1 second?

Or for super-KISS, no transistors or LM339: Each supply's 12 V output drives an R-C-SSR delay/switch, and the SSR output drives the next supply. 3 R's, 3 C's, 3 SSR's, no perfboard.

ak

Note: No 555's, PICs, or Arduinos were consumed in the creation of this response.