Is there be a good reason to use a voltage 10 times more than needed ?
I saw two huge 33µF 450V caps in mic preamps for phantom power isolation.
I wonder why such large, pricy caps where 48V only is needed.
I am puzzled because this is from a designer who does know about high quality mic preamps.
I saw two huge 33µF 450V caps in mic preamps for phantom power isolation.
I wonder why such large, pricy caps where 48V only is needed.
I am puzzled because this is from a designer who does know about high quality mic preamps.
May be. There was two design flaws and the peak led could not work at all.
However the PCB was very good, the preamp a best at low noise.
Well, I'll use 63V phantom caps and forget about this.
I wonder too, about a common mode choke at the inputs. I have doubts about its actual effectiveness and only find those made for mains power.
This was about four mic preamps boards and steppers from a small company in Australia specialized in pro audio gears.
The design is a clone of "Cohen double balanced mic preamp", same as MonteGeneroso from Samuel Groner. A well proven very good design that is easy to find and well documented on the web.
The peak lead part was not populated, I understood why, later, it was a wacky design that could not ever work.
First flaw: The trimming to adjust balancing was way too sensitive. I had to change resistor values.
Second flaw: In the second stage that simply does A - B and B - A, from the first stage, with two op-amps and height equal resitors, two of those were of wrong value at some wacky trimmers to achieve perfect balancing. A trimmer joke that could not ever do its purpose. However, no big deal, it could work well without such a super duper balancing. I left as is.
I like so much, this design, that I will remake it, with:
_No fancy AC and DC balancing at the second stage.
_Servo to replace the trimmer that balances the first stage and to accept no so well matched transistor pairs.
_Vu meter driver.
_Correct peak lead.
The design is a clone of "Cohen double balanced mic preamp", same as MonteGeneroso from Samuel Groner. A well proven very good design that is easy to find and well documented on the web.
The peak lead part was not populated, I understood why, later, it was a wacky design that could not ever work.
First flaw: The trimming to adjust balancing was way too sensitive. I had to change resistor values.
Second flaw: In the second stage that simply does A - B and B - A, from the first stage, with two op-amps and height equal resitors, two of those were of wrong value at some wacky trimmers to achieve perfect balancing. A trimmer joke that could not ever do its purpose. However, no big deal, it could work well without such a super duper balancing. I left as is.
I like so much, this design, that I will remake it, with:
_No fancy AC and DC balancing at the second stage.
_Servo to replace the trimmer that balances the first stage and to accept no so well matched transistor pairs.
_Vu meter driver.
_Correct peak lead.
The wacky peak led.
The intention was: Red blinking when too hot. Steady green when mic open.
This with a tricky use of a JFET. Could not work because of loose charateristics of JFET. Well another of these designs that has worked once, on the corner of a kitchen table..
The four instances of this contraption were acting in four different ways....
The intention was: Red blinking when too hot. Steady green when mic open.
This with a tricky use of a JFET. Could not work because of loose charateristics of JFET. Well another of these designs that has worked once, on the corner of a kitchen table..
The four instances of this contraption were acting in four different ways....
Hello,
Well, according to testing done by Cecil Bateman (believe linear audio web site has his articles posted), the higher the voltage rating of a electrolytic, the less distortion induced.
By making a blocking cap much bigger than needed, with a higher voltage rating (and ideally bi-polar) you can knock the distortion down to to an amount that is quite had to even measure.
Well, according to testing done by Cecil Bateman (believe linear audio web site has his articles posted), the higher the voltage rating of a electrolytic, the less distortion induced.
By making a blocking cap much bigger than needed, with a higher voltage rating (and ideally bi-polar) you can knock the distortion down to to an amount that is quite had to even measure.
While that is true, it matters when the ac signal voltage across the capacitor is large, for a phantom power circuit the capacitor ac voltage is tiny (much less than the microphone output signal), only the dc component is large, so I think its just cargo-cult circuit design.
Where signal voltages are large across a capacitor its a known issue (line level filter circuits for one).
Where signal voltages are large across a capacitor its a known issue (line level filter circuits for one).
I do not think the "tiny ac voltage" is a valid point, because a tiny distortion can result in a substantail distortion ratio.
Nevertheless, I plan to use polypropylen film caps, despite high cost and large size when over 10µF.
May be, electrolythic caps are acceptable for phantom isolation based on actual mic distortion and ac voltage across the cap. I am interested to know more about this.
I am interested to know about distortion comparing microphone preamps using electrolythic caps versus film caps for phantom power isolation.
I know the consensus about caps. Assertions like lythics are evil, film is good are of no interest to me.
Caps requirements mostly depend of what the cap actually does in the circuit.
I am interested in facts, objective facts preferably, about phantom power isolation caps.
I know the consensus about caps. Assertions like lythics are evil, film is good are of no interest to me.
Caps requirements mostly depend of what the cap actually does in the circuit.
I am interested in facts, objective facts preferably, about phantom power isolation caps.
Indeed 33µF 450V caps in mic preamps for phantom power isolation is likely non sense. 63V suffices and is way smaller.
I plan to use 33µF 63V 5% film caps. They are quite pricey.
Because electrolytics are smaller and less expensive, is it acceptable to use lythics, there, distortion wise ?
What about this answer ? " the amount of distortion will be vanishing even if the cap had a square-law, if the DC offset is 48V and the ac signal 5mV... Do the math, run an emulation..."
I do not see what maths or simulation ( about DC offset 48V and ac signal 5mV ) would be relevant to evaluate distortion by the phantom isolation caps.
I plan to use 33µF 63V 5% film caps. They are quite pricey.
Because electrolytics are smaller and less expensive, is it acceptable to use lythics, there, distortion wise ?
What about this answer ? " the amount of distortion will be vanishing even if the cap had a square-law, if the DC offset is 48V and the ac signal 5mV... Do the math, run an emulation..."
I do not see what maths or simulation ( about DC offset 48V and ac signal 5mV ) would be relevant to evaluate distortion by the phantom isolation caps.
5mV signal riding on a 48V offset is 1 part in 10000. That's 80dB below any conceivable intercept point. Just how non-linear are you expecting this capacitor to be?
OK, 79.64542466 dB, so what ?
This result is totally irrelevant to calculate distortion, whatever the capacitor non-linearity be.
Forget it, this goes nowhere.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.