Power supply ground clarification.

[gar]

071015-0845 EST USA

Ken:

I believe you have motors motors that are rate at 36 V and 3 A continuous based on looking back.

36 V would then probably be the maximum RPM you want to run them at. However, for high acceleration you might want to exceed the max current rating of the motor by a large amount. This does not last long so may be OK. The average current needs are the primary factor in determining both the motor and power supply power rating. There will be some economic trade off between power transformer size (impedance) and filter capacitance for your peak loading.

If you want that maximum speed capability implyed by the 36 V, then at maximum peak current during acceleration of all axis at once you want a voltage above 36 V by some margin. Assume 4 V. Lets assume a 20% voltage drop from the power supply at minimum input line voltage and your design peak current, then we have 40/.8 = 50 V. Increase this another 10% for a low input line voltage and we are at 55 V DC at no load. Add another 2 V for diode drop and we are at 57 V DC. 57 * 0.707 = 41 V AC RMS secondary for a nominal 120 V AC input. The transformer VA rating should be about 60 *12 = 720 VA.

This a ball park guess on what you may need. Others that have worked with your particular servo drives may have some suggestions for variations on my assumptions.

.

[Ken_Shea]

gar,
Thank you for taking the time to help. What you have mentioned seems what would be ideal in a power supply for these motors.

Ken

[gar]

071015-1116 EST USA

Ken:

On regenerative braking the filter capacitors receive the deceleration energy and thus raising their voltage. I would suggest 150 V capacitors, and some means at about 90 V to switch in a deceleration resistor to protect your motor drives. This 90 V value would depend upon what the motor drivers will tolerate.

Your bridge rectifier should be a 400 V PIV rating and a more than adequate current rating. Current rating in bridge rectifiers in this range is cheap.

For example a
GBPC3504-E4/51 Vishay/General Semiconductor is about $6.

.

[Ken_Shea]

gar,
Would the bridge rectifier mentioned be the one I should get ?

On the 150V capicitors, they are rated in uf ???
What uf would you suggest?
Does it do any harm to have higher V in these?

[gar]

071015-1428 EST USA

Ken:

That bridge rectifier would work. See www.mouser.com . You might want a slightly different number. There were two versions -- one with wire leads and the other with Faston tabs. The tab version would be better to eliminate soldering.

Talk to Al about suggested transformers and capacitors. Both are expensive.

I do not buy these normally so my judgement on sources and prices won't help a lot.

You could buy one of the regulated supplies that you referred to above and just use the output from the transformer, rectifier, and filter capacitor to play with. Maybe a permanent solution with three and there would be no need to parallel. Use one for each motor. With one to use for experimenting you could learn more about what would happen to the capacitor voltage under your typical decel. I would not buy more than one at the start. At normal line voltage I would expect you will be near 50 VDC no load. The capacitor and rectifier may not be over-rated much and therefore there could be problems from the decel counter EMF.

.

[Al_The_Man]

If you are thinking of building your own, then I usually buy the a toroid transformer from ebay Antek (johnango) 270175297591.
I prefer a linear supply, even unregulated over a switching supply, regulated is not really necessary and switching is finicky to fix if problems occur.
Toroid is not a must if you can find a square laminated type, the reason I go with toroid is the ease of tailoring the winding if slightly off by a few volts etc.
Also, although I will usually pick up a 5v or 12v supply from the PC, I generally use 24v auxilliary drivers/relays etc for I/O, so I spec in a little more VA and wind a 24v winding on top (~40t).
Plugging in your specs to my trusty Cap calculator comes to around 12,000ufd. At that it shows 5% ripple at 12 amps.
Al.

[neilw20]

Originally Posted by gar 071015-0852 EST USA Ken: The power supply you referenced lists the input voltage as 100 - 240. My guess is that is an incorrect statement. More likely it should read 120 xor 240 +4%/-17%. I am guessing it should tolerate 125 V as a max. Most commerical linear regulated supplies run very hot at maximum input voltage with maximum output load current. Basically marginal design. For good reliability you need to derate the output current rating. If enclosed it is worse. .

There are switch mode supplies that are 'universal'.
Paralleling not a good idea. Choose one that is has enough capacity.
A good swithmode suuply will be very efficient. Good ones can exceed 90%.

[gar]

071015-1543 EST USA

Al:

When I looked up the bridge rectifier I checked a Mallory capacitor of 11,000 mfd at 150 V for about $50 to 60 at Mouser. Do you have any other suggestions?

.

[Al_The_Man]

I know 'Computer Grade' capacitors are expensive, I usually keep an eye out for deals like ebay 290169156659.
I probably have enough on the shelf now for 10 future projects.
Where was eBay when I started out.
Al.

[Ken_Shea]

oh boy........

I hate electronics,

You guys have given enough info here so I guess it is my turn to make up my mind and just do something even if it's wrong.

Al, on the capacitors you referrenced on eBay, from what I have read capacitors can be hooked together, as needed, to get the VDC and uf needed ?


Thanks again for the input.


Ken

[gar]

071016-0945 EST USA

Ken:

Not a very good idea to put capacitors in series. This will increase the voltage rating in some cases, but can be unpredictable. Two causes -- difference in leakage resistance, and thus current, and difference in capacitance. I will describe why if necessary, otherwise accept it as fact..

Connecting capacitors in parallel is fine. They do not even have to have the same voltage or capacitance rating. Your maximum voltage would be limited by the lowest rating of the group. The resultant capacitance will be the sum of all the individual capacitors in the group. Note the capacitance in normal capacitors is relatively independent of applied voltage up to breakdown.

.