3 phase rectifiers.

[Al_The_Man]

The only problem in sizing the Cap/amp is that the percentage ripple will vary depending on the final DC voltage.
For an example, considering 200VDC at a current of 54amps (duty cycle figured in) for 28vp/p = 5% ripple, would require approx 2,600µfd on a three phase full wave supply.
IOW a 1000µfd cap on a 200vdc 1amp supply will be far less % ripple for the same capacitor on a 12v 1amp supply.
Al.

[Karl_T]

Al,
How is it that you know so much? At least for the subjects I'm interested in, you allways seem to be the one with the answer.

Green with envy

Karl

[NC Cams]

Using the posted recommendations:
"...Rutex recommends 1000 uf per amp..."

and the given
"...30 amps per drive...",

the following relationship seems intuitively obvious:

1000 x 30 x N = cap needed

where N = numbers of drives

This would give a reasonable MINIMUM cap value - one could increase capacitance as one can afford. "Afford" is the keyword here as at 200wvdc, 30K of capacitance per drive is NOT going to be cheap.

I suspect the Rutex cap factor is based upon a typical single phase rectification assumption. Thus, the "30" factor can probably be 'derated' as chances are that one would never draw 30 amps simultaneously from all drives AND there will be less ripple due to 3 phase rectification. A reasonable SWAG derating would be 66%.

BTW: My Fanuc's run 3 phase SCR drives and the ONLY filtration is an inductor - there is NO capacitance WHATSOEVER on the DC side of the drive. Hence, volatage ripple is not insubstantial and they run just fine.

It should be noted however, that EACH drive is operating off of a SEPARATE transformer rather than a common AC buss feeding all the drives.

Keep in mind that DC motors running at 200v are going to be a bit less sensitive to ripple than a sensitive electronics sensing circuit operating at 5-12vdc.

[Al_The_Man]

Originally Posted by NC Cams Using the posted recommendations: "...Rutex recommends 1000 uf per amp..." BTW: My Fanuc's run 3 phase SCR drives and the ONLY filtration is an inductor - there is NO capacitance WHATSOEVER on the DC side of the drive. Hence, volatage ripple is not insubstantial and they run just fine. .

The feature of SCR drives is such that they do not need a Power supply, that is if you discount the 3 phase provided by the Power company.
You do not fit a capacitor between drive and motor, any more than you would fit a Cap. between a PWM output and the motor.
The incoming three phase (power supply) has to be kept pure AC in order for the SCR action to work in this mode.
So what does the Choke or inductance do to minimize ripple?
The choke is in series with the motor armature, if the DC resistance is measured it will probably be found to be less than 1ohm, but an inductance possesses something called Inductive Reactance (measure in Ohms) one of the values it is dependant on is frequency, in this case 3600hz ripple. As opposite to Capacitive reactance, the inductive reactance value increases with frequency.
The Inductive Reactance will appear to be in series with the DC resistance, (say in this case say 1000ohms).
So if you imagine 2 resistors in series with the motor and one is 0.5 ohms and the other 1000 ohms.
The DC portion of the SCR supplied power only sees the 0.5 ohms, but the ripple frequency 'sees' the total, 0.5+1000 ohms. So you can visualize the effect on each.
On the question of Capacitor sizing, if a guide of 1000µfd/amp is used, that is OK for one particular voltage if the object is to attain a certain % ripple.
For e.g. , if a 200v supply at 1 amp uses a 1000µfd cap. This will produce x% ripple.
If the same standard is used on a 24v 1amp supply and again 1000µfd is used the same p/p ripple will result but the percentage ripple will be magnified proportionately.
I have the full formula plugged into a spread sheet, I will dig it up and post it for reference.

[najnielkp]

thank you al and nc.