(This is a re-post of something I submitted a few weeks ago before the Projects forum disappeared.)

I'm working on a project that needs more amps than I've worked with before. Can you let me know if my assumptions are correct, and even recommend the type of power supply I'll need?

The project: I have a "mechanical" digital clock where all the segments of the digits are controlled by solenoids. There are 34 solenoids - each has an initial draw of 12v @ 1.4amps, and to maintain the pull it takes 12v@450mA. The solenoids are controlled via a Basic Stamp and some other electronics: they're wired in parallel via a set of transistors. Pretty straightforward.

The problem: how big a power supply - and what type - will I need? Currently, my 12v @ 3amp old laptop brick doesn't seem to have enough juice to power all the solenoids - even though they update in "ripple" fashion (a slight delay between each solenoid firing to prevent too much current draw).

I suppose I'll need a power supply around 12v with around 15amps (34 soleniods * 450mA)... but will that be OK (i.e. safe) if only a few of the solenoids are active at any given time? Do I need to be concerned with "switching" or "no load" power supplies? Anything else I should pay attention to?

Any advice gratefully recvd.

 

Having more ampacity than needed is not unsafe at all. It is the load, not the power supply, that determines current draw.

Is your stagger interval long enough to allow the inrush current to subside completely? If not, then fifteen amps won't be nearly enough. You'll need to find worse case scenario (ie does 1:59:59 > 2:00:00 draw an extra five amps? An extra ten? Are other time change-overs even worse?)

 

Originally posted by thingmaker3@Mar 13 2006, 03:00 AM
Having more ampacity than needed is not unsafe at all. It is the load, not the power supply, that determines current draw.

Is your stagger interval long enough to allow the inrush current to subside completely? If not, then fifteen amps won't be nearly enough. You'll need to find worse case scenario (ie does 1:59:59 > 2:00:00 draw an extra five amps? An extra ten? Are other time change-overs even worse?)

[post=14923]Quoted post[/post]​

The power for the solenoids alone may not need tight regulation, so buying a high current regulated supply may not be a good choice. To design an unregulated supply there are two ways to go. One large transformer, or several small transformers in parallel. Once you make this decision the fabrication is relatively straightforward. For this unregulated linear supply there would be no minimul load issues.

The main issue would be how much voltage drop will there be when all the solenoids switch on and what will the voltage be when they are all holding. You will have to test the relays but I'll bet that they work fine in a range of say 9 to 18 Volts.

 

Switching supplies are more efficient and would be smaller but are more complex. Also the regulation they provide is probably a bit over the top.

Another way you might be able to get around the high current problem would be to pulse the soleniods at a frequency high enough to maintain their activation. I am assuming a basic stamp can be programmed to do this sort of switching.

 

For the sake of obtaining some meaningfull data, try connecting this device to a car battery in order to study the current requirements. Note the current levels durring the cycles of hours-minutes-seconds etc. Then you'll have a better grip on the actual current levels you need to produce/design your power supply.

 

I am wondering what size mechanical segment need a 6W solenoid to drive it.

 

Originally posted by pebe@Mar 14 2006, 10:41 AM
I am wondering what size mechanical segment need a 6W solenoid to drive it.

[post=14986]Quoted post[/post]​

...yes, I bet these solenoids probably could be activated and hold the plunger for less than 12v @ 450mA... I'll play around with that. Either way, though, when you've got up to 37 of these things flipping on and off every second, it will suck up a fair wattage - even with a 100ms delay between each "segment" switching from "off" to "on" (to minimize the 1.4amp initial surge).

Thanks everyone for their comments so far.

 

Originally posted by dhenshaw@Mar 14 2006, 07:36 PM
...yes, I bet these solenoids probably could be activated and hold the plunger for less than 12v @ 450mA... I'll play around with that. Either way, though, when you've got up to 37 of these things flipping on and off every second, it will suck up a fair wattage - even with a 100ms delay between each "segment" switching from "off" to "on" (to minimize the 1.4amp initial surge).

Thanks everyone for their comments so far.

[post=14987]Quoted post[/post]​

Have you thought of using a segment that's black one side and white the other, and turning it back and forth 180degrees?

 

What about my favourite cheap source of power - PC power supplies? The last one I purchased provides 17A @ 12V and cost only £17 uk - £1 / amp for a pretty good quality supply.

Only problem is that they can be a bit noisy.

 

Originally posted by bigbigblue@Mar 16 2006, 03:14 PM
What about my favourite cheap source of power - PC power supplies? The last one I purchased provides 17A @ 12V and cost only £17 uk - £1 / amp for a pretty good quality supply.

Only problem is that they can be a bit noisy.

[post=15067]Quoted post[/post]​

So that's what I eventually did -- a PC power supply that not only powers the solenoids at 12v, but also the microprocessor circuit at 5v. It was a bit expensive ($90) but that's mostly because it was an impulse buy. Either way, it appears to work just fine.