I was reading through the spec sheet of the LM1085 linear regulator and thought about what they said regarding improving ripple rejection. No surprise, the value of the cap on the bypass pin would have to change depending on the input ripple frequency. I started wondering if using a number of paralleled caps - say 47uF, .22uF, and then .10uF - to "cover" a broader frequency range with the improved ripple rejection would actually work. It may be a dumb thought, but let me know if it is.
Hmm, in looking at the graph it shows 200uF for frequencies below 60Hz, and 25uF for those above 60Hz. Does that mean that if I paralleled these two together I'd get good rejection through the audio band?
Hmm, in looking at the graph it shows 200uF for frequencies below 60Hz, and 25uF for those above 60Hz. Does that mean that if I paralleled these two together I'd get good rejection through the audio band?
It's not dumb.
In theory, a capacitor that will work at one frequency, will work at all higher frequencies and what you suggest would not be necessary. But, in practice larger capacitors do not function well at high frequencies, and bypassing with smaller values can help.
IMO, only use quality capacitors (both) when paralleling values, and check your reg for oscillation after adding caps.
In theory, a capacitor that will work at one frequency, will work at all higher frequencies and what you suggest would not be necessary. But, in practice larger capacitors do not function well at high frequencies, and bypassing with smaller values can help.
IMO, only use quality capacitors (both) when paralleling values, and check your reg for oscillation after adding caps.
Hi,
the cap on the bypass pin is usually of small value, you quoted 25uF to 200uF.
These sort of values can be LESR and use the highest quality to suit the value of your circuit.
There should be no glitches appearing at the bypass pins, I think this is a voltage reference input terminal, therefore there is no need for a ceramic to minimise the glitch on the input voltage.
The input and output pins are a completely different ballgame. Glitches on the rails every time the output current steps from one value to another.
the cap on the bypass pin is usually of small value, you quoted 25uF to 200uF.
These sort of values can be LESR and use the highest quality to suit the value of your circuit.
There should be no glitches appearing at the bypass pins, I think this is a voltage reference input terminal, therefore there is no need for a ceramic to minimise the glitch on the input voltage.
The input and output pins are a completely different ballgame. Glitches on the rails every time the output current steps from one value to another.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.