What an in- rush current surge charging a 10.000uF caps can be ?
I just blew 3Amp rated bridge connecting 400VA trans 2x25V to a PSU with 2x 10k uF supplying positive and negative rails . Since the caps were few years old I used variac to take it easy at first try and did a few , charge discharge cycles, then left the circuit ON overnight. Next day I was occupied with other things turned it off until the evening. In the evening the first try resulted in shorted rectifier and burned PCB traces. Meh .
By the wattage of trans it suggest that I can expect up to 8 Amps in each winding right?
I just blew 3Amp rated bridge connecting 400VA trans 2x25V to a PSU with 2x 10k uF supplying positive and negative rails . Since the caps were few years old I used variac to take it easy at first try and did a few , charge discharge cycles, then left the circuit ON overnight. Next day I was occupied with other things turned it off until the evening. In the evening the first try resulted in shorted rectifier and burned PCB traces. Meh .
By the wattage of trans it suggest that I can expect up to 8 Amps in each winding right?
I know there are multiple online calculators and design assist apps that can be helpful here- However 3A is a really small rectifier for this application. Most designers just overkill the rectifier rather than calculate exact conditions- pick a nice big one, a 20A unit can be had for just a few dollars assuming you can make it fit in your PCB. Heat sinking the rectifier is typically overkill, but I have seen it done for insurance.
Most would use a 35A rectifier block for that.
Nelson Pass uses a 35A block for his amps, which all have far more capacitance than 10,000uF.
The transformer does not limit the current to VA/voltage.
Typical package:
https://www.digikey.com/en/products/detail/onsemi/GBPC3504W/1057462
Nelson Pass uses a 35A block for his amps, which all have far more capacitance than 10,000uF.
The transformer does not limit the current to VA/voltage.
Typical package:
https://www.digikey.com/en/products/detail/onsemi/GBPC3504W/1057462
The thing is the PSU will be supplying some 50-80 mA max so I just need to size rectifier for inrush current charging 10-15k sized caps
Inrush will be multiple amps. That is the cost of mongo capacitors. Either a big rectifier or drop the size of your filter capacitors significantly.
What exactly is your power supply powering? Is the 50-80mA just quiescent current? If the current demand is relatively low, you can use quite small input capacitors and regulate the output- If low ripple is desired, a regulator will do much better than huge capacitors. I think we assumed with such large capacitors that you would be powering a power amplifier, something that pulls too much current (amps) to easily regulate.
Since your design will be drawing such a low power, if you insist on using huge capacitors you should add a voltage bleed resistor from each capacitor to ground to safely discharge the caps in a reasonable amount of time (seconds). These alone will pull more current than the demand you indicated.
By the way- It is not uncommon for a power supply rectifier (or fuse) to blow after working for several previous attempts. With AC wall voltage, the voltage is continuously changing, and the precise amount of inrush current will depend upon exactly where the wall voltage is at when you close the switch. I built a power supply and used a low current fuse for initial testing, and the supply turned on 3 times correctly, and blew the fuse on the 4th try.
Since your design will be drawing such a low power, if you insist on using huge capacitors you should add a voltage bleed resistor from each capacitor to ground to safely discharge the caps in a reasonable amount of time (seconds). These alone will pull more current than the demand you indicated.
By the way- It is not uncommon for a power supply rectifier (or fuse) to blow after working for several previous attempts. With AC wall voltage, the voltage is continuously changing, and the precise amount of inrush current will depend upon exactly where the wall voltage is at when you close the switch. I built a power supply and used a low current fuse for initial testing, and the supply turned on 3 times correctly, and blew the fuse on the 4th try.
It's a Naim style power supply project for an output stage of a CD player working to supply a few opamps there..Very little normal logic behind it but according to many this approach works.
I bought a cheap LM317/ 337 Chi psu kit which already contained all the parts along with 3A rectifier and 2200 uF caps . I just substituted the caps with 10k uF thinking ii shouldn't be a problem once the caps are well formed. Now , the space on the board is limited and the rectifier is a flat type with 4 leads in a row so I'm not in a liberty to put 40 amps part there and need a part to work and fit that space.
I bought a cheap LM317/ 337 Chi psu kit which already contained all the parts along with 3A rectifier and 2200 uF caps . I just substituted the caps with 10k uF thinking ii shouldn't be a problem once the caps are well formed. Now , the space on the board is limited and the rectifier is a flat type with 4 leads in a row so I'm not in a liberty to put 40 amps part there and need a part to work and fit that space.
Another option is to put an inrush limiter on the AC line. Would also eliminate the light dimming when you turn it on.
It is a brute force approach🙂 I just don't know how brute I supposed to be. Transformer was generating a strange oscillation like mechanical buzz, pulses as well from time to time when the circuit was working . Maybe the layout and grounding concept of this double sided PCB simple Chinese regulator is not optimal for such project. It blew before I pulled the scope to check what's going on . It's a textbook LM317 circuit with protection diodes and 10uF decoupling caps
If you are using regulators, you do not need big caps. Big caps are used to limit the amount of ripple coming out of the rectifier. The regulator does the same thing, but much better- it's OK to have ripple going into the regulator, as the regulator does a good job and there will be virtually no ripple on the voltage output. The original 2200uF caps were plenty, and the 10,000uF caps are doing nothing to improve performance while stressing the transformer and rectifier.
On power up the capacitor is basically a short with current only limited by the transformer.
I use a 35 amp bridge rectifier on most of my gear.
Have a play with ltspice and a capacitor on a power supply and you will be shocked at current draw at turn on.
I use a 35 amp bridge rectifier on most of my gear.
Have a play with ltspice and a capacitor on a power supply and you will be shocked at current draw at turn on.
It may be a problem or it may be not . Lm317 is cheap as chips but it's not the reason I used it. OEM is using it. I will use biggest rectifier I can fit. OEm is using 800-1200W transformer to supply 5W of power with banks of capacitors followed by LM317. I had a Plitron 400W trans on hand so I thought I will check the concept although the trans is " undersized "
If the current draw really is that low, the supply is regulated, and for whatever reason you are using oversized caps, just use RESISTORS ahead of those big caps. Limit the inrush current to say 10 amps then any run of the mill flatpack bridge rectifier will be FINE. The resistors will even help reduce ripple.
Not strictly true. The "big caps" are to hold energy between the mains waveform peaks, perhaps they supply power for 60 to 80% of the time in the cycle so have to sustain things for 8ms or so. For example 100µF cap with 60mA load will discharge by 6.4V in that 8ms, 10mF cap with 5A load will discharge by 4V in 8ms.
The size of the caps depends on the load whether or not regulators are used (although having regulators means some drop-out voltage, so typically a larger cap value would be used with a regulator than without)
It happens to be the case that most power amp designs don't use regulators, so they have large cap values like 10mF or whatever, and most pre-amp designs use +/-15V and a few tens of mA using regulators, so they need caps of perhaps 100µF--220uF or so (this would be true without regulators of course).
If you had regulators for a power amp then large capacitors are still needed, before the regulators, and probably significantly larger ones to allow for drop-out voltage - this is one reason not to bother with regulators for power amps (the other being that good power amp circuits have good PSRR making regulation unnecessary).
Does it really need this big transformer, capable of 8 A per secondary winding, to supply just 50 to 80 mA?
Best regards!
Best regards!
And if it's also a toroidal transformer that big, it will create huge inrush all by itself regardless of caps.
I instantly burned up my two nicest Pomona jumper leads hooking up a ~400VA toroidal to a big 500 Ohm resistor...
The inrush caused by a decaying DC component in the core doesn’t flow through the rectifier. The secondary doesn’t even need to be connected to anything. It’s confined to the primary, and can burn up switch contacts, blow fuses or breakers. Since it causes momentary core saturation, it’s worse than the surge caused by hungry caps.