For a power amplifier I am contemplating a multi-Farad buffer to the power rails. That would complement some aplifiers great That now also have 56.000 uF big-can capacitors like from Mallory, big, blue, screw Mount. Some designs just love such buffers!
While I can easily cycle the power supply to an on-state = going from 0 volt to 14 volt, I am worried that the capacitors might not ‘like’ that and are in fact made fo allways-on use.
Such as being used parallel to a (car) battery. Then it virtually never goes ‘off’ to unloaded.
On the other hand, in an amplifier I would see it stop gracefully in about 5 minutes, once the power is switched off.
So in actual use it would cycle.
If that is is not wished, then I would make a relais (the DC-type contacts as used in cars) to engage them. Then they would stay charged When the amp is not in use.
What is your advice for this?
While I can easily cycle the power supply to an on-state = going from 0 volt to 14 volt, I am worried that the capacitors might not ‘like’ that and are in fact made fo allways-on use.
Such as being used parallel to a (car) battery. Then it virtually never goes ‘off’ to unloaded.
On the other hand, in an amplifier I would see it stop gracefully in about 5 minutes, once the power is switched off.
So in actual use it would cycle.
If that is is not wished, then I would make a relais (the DC-type contacts as used in cars) to engage them. Then they would stay charged When the amp is not in use.
What is your advice for this?
I read that the smaller tyoe cells themselves have no memory effect and can be charged and recharged a million times.
But how do these big ones handle it, having a very low Rdc? Like 100 mohm?
But how do these big ones handle it, having a very low Rdc? Like 100 mohm?
"Buffer" is not the correct term. Decoupling may be better?
Caps aren't batteries and don't mind turning on and off, but the circuitry feeding "multi-Farads" will mind. Why don't you just sneak up on it with another set of 56000uF?
Caps aren't batteries and don't mind turning on and off, but the circuitry feeding "multi-Farads" will mind. Why don't you just sneak up on it with another set of 56000uF?
Yes that is right: you can't just switch this, e.g. to a battery. The wire will explode (50 amps?). I have seen several such 20-50 Farad buffers where the wire has a small switch, that you have to engage, standard is 0,5 ohm for instance, then the switch shorts that and the cap stays connected to the car battery all the time.
What I have in mind is a SMPS that starts not with a hiccup (then it will hiccup forever) but one where the overload protection is a constant current. So turning that on will not harm the power source. Example Meanwell HRP-150 series with its
OVERLOAD Protection of Constant current limiting, that recovers automatically after fault condition is removed and starts once about 105 ~ 135% of the rated output power is reached, in 20V mode that is 7 A.
The question is: will the multi-Farad banks destroy them selves if they are turned on and off a couple of times a day with that charge of say 8-10 Amps?
OK I should just buy the thing and test, I know.
What I have in mind is a SMPS that starts not with a hiccup (then it will hiccup forever) but one where the overload protection is a constant current. So turning that on will not harm the power source. Example Meanwell HRP-150 series with its
OVERLOAD Protection of Constant current limiting, that recovers automatically after fault condition is removed and starts once about 105 ~ 135% of the rated output power is reached, in 20V mode that is 7 A.
The question is: will the multi-Farad banks destroy them selves if they are turned on and off a couple of times a day with that charge of say 8-10 Amps?
OK I should just buy the thing and test, I know.
That is generally true; however, as always there are some exceptions, when you push the envelope to the max. For example, photo-flash Ecaps have a special construction to prevent the formation of parasitic dielectric layers growing each time it is subjected to a very high discharge current. A normal Ecap of identical voltage and capacitance ratings would only last a few tens of times.
Capacitors used in large discharge current applications (not Ecaps) can suffer from electromigration in some sections.
Shallow metallization caps (like X-rated caps) have their metallization near the electrodes eroded each time they are charged or discharged through a low impedance, like a true short.
The mains impedance doesn't count as a short because the wiring has a characteristic impedance in the tens of ohms.
If you repeatedly discharge a X cap through a dead short, you will decrease its capacitance.
Etc, etc : you can certainly find many other examples
Capacitors used in large discharge current applications (not Ecaps) can suffer from electromigration in some sections.
Shallow metallization caps (like X-rated caps) have their metallization near the electrodes eroded each time they are charged or discharged through a low impedance, like a true short.
The mains impedance doesn't count as a short because the wiring has a characteristic impedance in the tens of ohms.
If you repeatedly discharge a X cap through a dead short, you will decrease its capacitance.
Etc, etc : you can certainly find many other examples
Well - they might 😉 : 1 million times in one day (24h) would be 11.5 times every second.
That would give you some 18A constantly applied either charging or discharging (with 14V delta and 56000uF).
The heat generated by this for a prolonged time may overheat the capacitor.
But besides this you are right - in general capacitors don't mind being turned on and off and do not suffer from the wearout that a battery would.
And then there are the exceptions as noted in #6.
Then I'll go ahead. It is for a new implementation of a amplifier Le Monstre. What the designer said in another circuit: the 1F supercapa makes the sound seem dull but it is darker and has more depth.
Thanks for doing the math! I was going to say "thousands" but that didn't seem extreme enough to make my point that it's not really about the caps but about how to control the inrush...
Attachments
Charging every minute would already heat them up beyond repair . . I have several capacitor banks from (mil) airplanes that had a 22 ohm resistor and 10x 56.000 uF. All in a case that slid right into a recess somewhere.
That array (6x 2.7V 500F caps in series) has an energy storage capacity of 10kJ or so, and a theoretical AC impedance at 20Hz of about 0.0001 ohms (ignoring ESR and wiring). Unless your amp is driving loads of 10 milliohms or less at currents of 1000A or thereabouts you don't need that much capacitance for a mains derived supply. 2 to 3 orders of magnitude too much probably. Bit like using a 100 tonne anchor on a 40 foot yacht...