Guru
Joined: 31/12/2012
Location: New ZealandPosts: 2552
| Posted: 05:05am 09 May 2019 |
the LM2596 is a buck (step-down) regulator:
http://www.ti.com/lit/gpn/lm2596
the LM2577 is a boost (step-up) regulator:
http://www.ti.com/lit/gpn/lm2577
with a bit of trickery one can be persuaded to act as the other, or to even act as a buck-boost regulator, but the result is usually suboptimal.
it looks like you are driving a water valve, with two separate coils - one to turn on the valve, the other to turn it off. just had a look at the datasheet for a IR-S-985-3W, which i presume is very similar.
i'd be using the full 50v (or as close as possible), as the energy stored in a capacitor is (C x V^2) / 2 where C is the capacity in farads. so going from 18v to 36v gives you four times the stored energy, going from 18v to 50v gives you nearly eight times the stored energy.
the absolute maximum switching current of the LM2577 is 6 amps. assumung 12v input supply, this translates to 1.5 amps output at 48v (actually a bit less current due to the efficiency). so we can get away with a 47 ohm limiting resistor feeding the capacitor. as an aside, i'd not want to be drawing this current for any length of time without any heatsink on the regulator, so make sure you don't leave either relay closed for more than a second.
charge time to achieve about 65% of the supply voltage is R x C. lets say a 50v supply, a 47 ohm resistor, and a 4700uF capacitor. this gives:
50 x 4700 x 10^-6 = 0.24 seconds.
so to charge the 4700uF capacitor to close to the full 50v is going to be a few seconds at most. i presume this fits in with your planned operation? the datasheet for the valve i've looked at doesn't give any information on cycle times or maximum lifetime cycles for the device. but at a guess, i'd say it is likely to be around the 20,000 actuations mark.
cheers,
rob :-)