I recommend a soft start for starting the transformer.
If the capacitor bank is big enough then a slow charge may be an advantage to limit the current pulse into the capacitors after start up. I suspect this will increase the life of the electrolytics which are generally not designed for fast charge nor fast discharge. There are special fast discharge caps sold for welding and similar duty.
Yes, two timers are needed.
100ms to 300ms for the soft start.
2s to 20s for the slow charge.
If the capacitor bank is big enough then a slow charge may be an advantage to limit the current pulse into the capacitors after start up. I suspect this will increase the life of the electrolytics which are generally not designed for fast charge nor fast discharge. There are special fast discharge caps sold for welding and similar duty.
Yes, two timers are needed.
100ms to 300ms for the soft start.
2s to 20s for the slow charge.
My personal opinion is that the diodes and caps have no problems to handle the start up. The major issue is the magnetizing current of the transformer itself. So the answer to your question is no. My solution is to have 100 ohms in series with the primary winding and the time 300-600 ms, somewhere under 1 sec at least. I have pcb's...
Limiting inrush current
If the capacitors are completely discharged they represent a very low impedance load and the initial current surge can be very high, particularly if the switch is closed near the crest of the AC sinewave. The power supplies in commercial computer systems usually require this initial inrush current to be limited. It can be damaging to the supply itself if the components are not rated for such peaks, but maybe more important is the effect this surge has on other nearby equipment. The current surge causes the AC line to dip which can cause other equipment to malfunction.
The design challenge is to have the inrush limiting circuit "disappear" after the caps are fully charged. Thermistors are usually used to add the required impedance and they are are often bypassed with a relay to completely remove them from the circuit during normal operation.
If the capacitors are completely discharged they represent a very low impedance load and the initial current surge can be very high, particularly if the switch is closed near the crest of the AC sinewave. The power supplies in commercial computer systems usually require this initial inrush current to be limited. It can be damaging to the supply itself if the components are not rated for such peaks, but maybe more important is the effect this surge has on other nearby equipment. The current surge causes the AC line to dip which can cause other equipment to malfunction.
The design challenge is to have the inrush limiting circuit "disappear" after the caps are fully charged. Thermistors are usually used to add the required impedance and they are are often bypassed with a relay to completely remove them from the circuit during normal operation.
I should add that two stages of limiting is not needed. The transformer (to a first order approximation) would only be present in the design equations as as a term representing its turns ratio. As the current will be smaller on the primary (because of the turns ratio) it will be more practical to limit the inrush there. Also, if the secondary is center tapped for a +/- output, you would need twice as many limiting components on the secondary side.
Hi,
inserting a current limiting resistance in the secondary circuit requires quite different component values from inserting it in the primary circuit.
To get effective current limiting of say 10A peak from a 50Vdc supply would need 5r. This is well within reach of a high current Power Thermistor.
To get the same 10A peak in the secondary would require a quite different Power Thermistor, or combination of Power Thermistors, to be inserted in the primary circuit.
I can't advise, but I can assure you a CL60 used in the primary will have a very different effect compared to a CL60 in the secondary. I can also assure you that a single CL60 in a 220/240Vac primary circuit is just about enough to prevent a time delayed close rated mains fuse from blowing. But it will have very little effect on the rate of charge of the smoothing caps across the secondary.
But is a slow charge required?
Posts 2, 3, 4 & 5 offer the opinion that the slow charge circuit may not be required, except in exceptional circumstances
inserting a current limiting resistance in the secondary circuit requires quite different component values from inserting it in the primary circuit.
To get effective current limiting of say 10A peak from a 50Vdc supply would need 5r. This is well within reach of a high current Power Thermistor.
To get the same 10A peak in the secondary would require a quite different Power Thermistor, or combination of Power Thermistors, to be inserted in the primary circuit.
I can't advise, but I can assure you a CL60 used in the primary will have a very different effect compared to a CL60 in the secondary. I can also assure you that a single CL60 in a 220/240Vac primary circuit is just about enough to prevent a time delayed close rated mains fuse from blowing. But it will have very little effect on the rate of charge of the smoothing caps across the secondary.
But is a slow charge required?
Posts 2, 3, 4 & 5 offer the opinion that the slow charge circuit may not be required, except in exceptional circumstances
Last edited:
A simple small toroid will reduce the peak inrush current...
You put it in series right before the cap bank... It will limit the peak inrush and then once saturated by the DC current it becomes just a a wire...
I have used it in similair applications to limit 17A inrush peaks down to 8 to 9 amp peaks, to meet inrush requirements... The toriod can be designed for a specific time delay based on size, turns and type of permeability...ect...
Chris
You put it in series right before the cap bank... It will limit the peak inrush and then once saturated by the DC current it becomes just a a wire...
I have used it in similair applications to limit 17A inrush peaks down to 8 to 9 amp peaks, to meet inrush requirements... The toriod can be designed for a specific time delay based on size, turns and type of permeability...ect...
Chris
If you're going to get into sequencing things, why not just use a small transformer to bring up the caps before energizing the main transformer. You can leave the small xformer secondary in parallel with the main xformer secondary. After the caps are charged and the main transformer is already mainly synced to the line, hit the big primary with the juice and go... It'd be a little more expensive than a thermistor, but extremely reliable. You would want to be reasonably certain of matched turns ratio, but if impedance is high enough and the ratio close, it wont matter. If you could manage to hold off linear output bias until the main transformer was online, the current spike at the starting point would be really small because all the little tranformer would be left doing if you waited long enough would be magnetizing the big core pretty much in phase with the line.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.