Power Supply Capacitors?

[jevs]

Is there anything wrong with overdoing the filter caps in your power supply? Depending on how I run the numbers I come up with different current ratings for my motors. For example if I use the rated power and the max voltage I am supplying I get one number. If I derive the current using torque specs etc I get a different number. I was wondering if I just size them for the largest value if that would be a problem if they actually draw less current. I would think it would just filter even better and that it won't hurt anything, but maybe there is some other factor I am not familiar with?

Thanks

[H500]

When it comes to filter caps, the rule of the thumb is the bigger the better. The calc's just give you the minimum you can get away with.

[jevs]

That's pretty much the way I figured it, just wanted to make sure.

[jevs]

Bought these big suckers:
http://cgi.ebay.com/ws/eBayISAPI.dll...tem=7504192697

120,000uF 200 Volts. These should take that ripple down nicely

[TigerPilot]

Originally Posted by jevs Bought these big suckers: http://cgi.ebay.com/ws/eBayISAPI.dll...tem=7504192697 120,000uF 200 Volts. These should take that ripple down nicely

Wow, they are big. You could connect them to a timer and turn off the power periodically and run your equipment for a few hours off them. Just kidding of course but it is not far from the truth. Ask me how I know.

Yoram

[Al_The_Man]

I have a copy of a physics paper I downloaded sometime ago (I have lost the link) that explains the downside of oversizing smoothing capacitors in linear power supplies, especially when being run at maximum load current.
I will try to find it and post it in pdf.
It gives the example of the capacitor recharge cycle in a typical full wave bridge, the bridge rectifier only conducts at the point where the ripple has dropped or decayed and meets the rising 1/2cycle ripple until the peak occurs, this period that it conducts in is shorter for a larger capacitor than one of a lower value, although the period is shorter the energy in the pulse is much larger and can cause greater heating effects.
In the example, the lower capacitor shows a pulse peak of 5 amps for a period of 1/5th of the 1/2 cycle, for the larger cap. the pulse peak is 20amps for a period of 1/20th the half cycle. The Kva requirement for the transformer is double for the larger cap. In the above example the load current is constant and the same in both instances.
Al

[jevs]

I doubt I will ever work the 3 KVA transformer very hard anyway. It is already bigger than the original transformer. What happened to the "bigger is better" theory on the filter caps? Also, how much oversized are we talking? I was needing at the bare minimum 33,000 uF for one supply. I will be watching everything closely at first anyway. If I see a heat problem with the transformer then I will try different caps.

[H500]

Bigger is better as long as the other components are sized to match. It's like the engine in a car. 400hp is better than 100hp, but 10000hp will probably tear the drivetrain apart.

Your transformer is VERY large, so it is unlikely that you will overtax it. Just make sure your rectifiers are extra big also. There's a calculation for it, but I don't remember. If you pick diodes with an amp rating to match your transformer, then you should be OK.

Take extra care when working with these steroid-enhanced parts. They can unleash a huge amount of energy when something goes wrong.

[sendkeys]

couldnt you just add in a ac capacitor befor the rectifier to take care of any of the ac peaks?

[jevs]

I will be using two 50 amp rectifiers

[Al_The_Man]

Actually what you are partially describing is a constant voltage regulator as is made by Sola et.al, these run a special AC rated capacitor in a tank circuit which keeps the transformer saturated, and any load you put on the secondary detracts from the saturation so the power is constant.
Because the transformer runs saturated, it producess ALOT of heat constantly.
The main benefit to this is that any spikes that may come down the line are not passed on the the secondary due to this saturation.
For the average linear power supply it is not practical, I would not recommend it.
Al

[sendkeys]

ahhh didnt know that thanks Al the Man