Capacitor for smoothing of AC rectification

[ccherry.1225]

Hea,

We're building a custom power supply to power some servo controllers (Copley Control Junus). Unfortunately we don't have any 7.5 kw power supplies to power these things, and don't want to spend the 5-10k necessary to get them.

We're taking mains, isolating it via transformer, rectifying it with a diode bridge, and then smoothing it with a capacitor. My question is how big of a capacitor should we be using. It's my understanding that larger capacitors provide better smoothing, but what are the drawbacks of using a huge one, and what would be a good size to use for smoothing 7.5kw of power at 110v

Thanks,
Chris

 

[rbj]

7500 watts at 110 volts? are you building an electric chair or something?

well your DC current will be around 7500/110 ≈ 70 amps! you're going to have some mondo transformers and rectifiers and capacitors to smooth that out. so the peak voltage after rectification will be about 110 × $\sqrt{2}$ ≈ 150 volts. the "ripple" (the unsmoothness) will be this instantaneous voltage dropping from 150 to whatever it is from your capacitors discharging and you have to decide how much ripple voltage you can tolerate.

since the bridge rectifier is a full-wave rectifier, you will have 120 sinusoidal pulses per second charging the capacitors to nearly 150 volts. the fundamental volt-amp characteristic of a capacitor is

$$i(t) = C \frac{d v(t)}{dt}$$

which we will approximate with

$$i = C \ \frac{\Delta v}{\Delta t}$$

so

$$C = i \ \frac{\Delta t}{\Delta v}$$

or

$$C = 70 \ \frac{1/120}{\Delta v}$$so you pick how small the ripple, $\Delta v$, has to be and solve for C. your capacitor will be on the order of Farads and, when charged up to 150 volts, there is enough energy stored in them to kill an elephant!.

be careful. big-@ss power supplies like this are friggin' dangerous.

i don't know where you're going to get the parts (the transformers, rectifiers, and capacitors) to do this, but be very, very careful. one spurious metal strap that falls across those capacitor terminals might bring the world to an end.

 

[sophiecentaur]

I don't think you can do this sort of thing on the cheap. Either you need to bite the bullet and buy something off the shelf or pay for some professional advice for your specific application. Nothing you can get off this forum (with respect to all the excellent knowledge that is available here) can be relied on to give you a bomb-proof answer to your query because you will have no guarantee of the quality of the particular opinion that you happen to choose. No one worth their salt is likely to commit themselves to risking giving you advice on this except in very general terms - which is not what you need. You will almost certainly have to pay someone who will be indemnified against any risk involved.
Like rjb says - you could kill an elephant and the elephant could be you.

 

[rbj]

i think that, except for regulation circuitry, that a power supply can be simple enough that the average pedestrian electrical engineer can make it and make it work. and if you know what you're doing and put it into the right cage, it might even be safe.

but delivering 7500 watts to a load is no small deal.

 

[Studiot]

that the average pedestrian electrical engineer can make it and make it work. and

I think he'd be lucky to make it work first time.

Have you calculated the inrush current to these capacitors?

 

[sophiecentaur]

I would definitely station myself outside the door when the gear was first switched on!
You would need some 'soft start' circuitry I think.

 

[psparky]

Like they said...it's ok to get information from this board for informational awareness perhaps...

But it sounds like you will need stamped drawings from a professional engineer that meets the electric code of your area.

Also...do things the right the first time (spend the 10 K)...they tend to cost less in the long run.