What type of load draws 40A for 100ms?
How much voltage variation will that load tolerate during that 100ms?
Finally, a real question!
But not enough data...
What is:
the load resistance
the rail voltage you wish to maintain
Keep in mind that, for this duration of time (100msec), the size of the caps is only affecting the ripple on the caps in between charging pulses. The transformer is therefore a very important part of the power supply.
Pick one, both statements are incompatible.
The second one, besides, is lacking parameters for a meaningful answer.
...........................................................
Why "incompatible"?
Just a hint: 78V into 4 ohms=19,5 A peak/max/you name it , no matter how short the period.
You have nothing there which can pull 40A, not even for a microsecond .
I think the OP is thinking of +/-78Vdc rails. This gives around 160V between rails, so that into 4R the peak current could reach 40A (if you ignore losses and such). But a real supply will have real losses, and you can't ignore these.
In any case, the OP should realize that a capacitor alone discharges in an exponential fashion and can not deliver a stead 40A into a fixed load over any period of time. As I mentioned, the power transformer is really supplying most of the current over the span of 100msec and the caps are just helping to keep up the voltage in between charging pulses. A transformer that is rated several kVA will be needed so that the secondary voltage does not sag under the high current demand. But the caps do not to be as high as 1F. 100,000uF would result in only a few volts of ripple. Anything more is getting into the "diminishing returns" category.
Edit: it looks like you need something on the order of a 3kVA transformer, but only about 68,000uF per rail of capacitance to get around 1V ripple. Instead of a linear power supply, this would be a great application for a switching supply.
Last edited:
We still have no #####ng idea what OP is trying to do so it's all speculation fromhere.
all you need to know about inrush currents is here:
http://www.emeko.de/uploads/media/01-Press-TSR-Speech-2009-en.pdf
charging up the caps could be done ramping a PWM , or a resistor dc side with relay driven by a sensing the voltage across the R.
http://www.emeko.de/uploads/media/01-Press-TSR-Speech-2009-en.pdf
charging up the caps could be done ramping a PWM , or a resistor dc side with relay driven by a sensing the voltage across the R.
You can do the systems engineering at the beginning...
Thanks for that but, in the absence of a decent problem definition, that's about as useful as me saying
All you need to know about inrush currents is here:
Thanks for that but, in the absence of a decent problem definition, that's about as useful as me saying
All you need to know about inrush currents is here:
It requires a large wirewound resistor in series with the primary and a relay to short it out which is held on by one of 3 methods:
1: A seperate dc supply. I use this because it's a cooling fan supply.
2: A suitable dc main supply.
3: The mains direct.
In each case it's a relay with a low voltage dc coil of the voltage rating of the power supply OR the coil protected by a parallel zener diode. A delay is required from a dc source consisting of a series resistor which charges a capacitor, the time constant being determined partly by experiment. (Do this before connecting to mains). If you use the ac mains, use a single series rectifier diode (at least 400Vac rating) and definitely a zener diode across the relay coil to protect it.
I used a 1k 7W resistor for a 300W power supply but can't recall the capacitor value. It will be a low voltage electrolytic in most cases. I think a 1 second delay should be adequate.
1: A seperate dc supply. I use this because it's a cooling fan supply.
2: A suitable dc main supply.
3: The mains direct.
In each case it's a relay with a low voltage dc coil of the voltage rating of the power supply OR the coil protected by a parallel zener diode. A delay is required from a dc source consisting of a series resistor which charges a capacitor, the time constant being determined partly by experiment. (Do this before connecting to mains). If you use the ac mains, use a single series rectifier diode (at least 400Vac rating) and definitely a zener diode across the relay coil to protect it.
I used a 1k 7W resistor for a 300W power supply but can't recall the capacitor value. It will be a low voltage electrolytic in most cases. I think a 1 second delay should be adequate.
dazed2 solution with an ordinary rush in unit to first delay the transformer and another unit to use after the transformer but before the diode bridge and the condenserbank, must be the "columbie egg" to start up a big transformer/condenserbank.
In post 80 dazed2 gave us the schema:
He has design this unit according to the DC power and condenserbank used in his power amplifier. I intend to copy this for use in my Slewmaster. The transformer is 2.2KVA, AC out 2x50V. DC power=2x67V.
I have tried to contact dazed2, but contact during the summer hollidays can be difficult.
Being in short of time, I hope that may be other members can give me a helping hand with the correct parts to fit my Slewmaster for this unit.
Eivind S
In post 80 dazed2 gave us the schema:
He has design this unit according to the DC power and condenserbank used in his power amplifier. I intend to copy this for use in my Slewmaster. The transformer is 2.2KVA, AC out 2x50V. DC power=2x67V.
I have tried to contact dazed2, but contact during the summer hollidays can be difficult.
Being in short of time, I hope that may be other members can give me a helping hand with the correct parts to fit my Slewmaster for this unit.
Eivind S
- Status
- Not open for further replies.