Re: electrolytic mystery
From: Ross Herbert (rherber1SPAMEX_at_bigpond.net.au)
Date: 08/23/04
- Next message: Watson A.Name - \: "Re: A Golden Opportunity!"
- Previous message: Sal Brisindi: "Re: Wanted 7448 0r 74LS48 BCD to 7 segment drivers ic's"
- In reply to: Walter Harley: "electrolytic mystery"
- Next in thread: Walter Harley: "Re: electrolytic mystery"
- Reply: Walter Harley: "Re: electrolytic mystery"
- Messages sorted by: [ date ] [ thread ]
Date: Mon, 23 Aug 2004 12:19:01 GMT
On 22 Aug 2004 22:50:02 GMT, "Walter Harley"
<walterh@cafewalterNOSPAM.com> wrote:
|Perhaps y'all can help me with a mystery. All electrolytic capacitors have
|a "working voltage" rating. The question is, what are the consequences of
|exceeding that voltage? Now, before you jump to the obvious answers, read
|on:
|
|Audio devices commonly have opamp-based output stages, followed by a DC
|blocking cap. The rating of the cap is more often than not either 16V or
|25V, probably because of the typical +/-15V supply rails; the polarity
|orientation of the cap varies, or rarely it may be bipolar. A typical
|example would be the circuit at http://www.rane.com/pdf/old/pe15sch.pdf
|(chosen for no other reason than that I used one last night).
|
|Now, balanced outputs from audio devices often get hooked up to mixer inputs
|that are expecting to see microphones. And those inputs often are supplied
|with phantom power, which is equivalent to a 48V DC source in series with
|6.8k. (An example, again from the excellent and helpful Rane web site, is
|http://www.rane.com/pdf/ms1bsch.pdf.)
|
|So, if one plugs a typical balanced output into a typical phantom-supplied
|mic input, as is very frequently done, we get this circuit, where I have
|been optimistic about the polarity of the capacitor:
|
| 6.8k
| ___
| .--|___|--.
| | | +
| --- ### 15uF
| - 48v --- 25V
| | |
| '---------'
|
|
|QUESTION: WHY DOESN'T THIS BLOW UP??
|
|Just for grins, I've got this precise circuit sitting there on my test bench
|right now. I measured the cap after about 2.5 hours and its ESR and
|capacitance were unchanged. I'll leave it running for a few days and
|measure again. But I already know that a gazillion musicians and audio
|engineers get away with this on a daily basis.
|
|So, why does this work, and for how long should it be expected to work,and
|how should the eventual failure manifest? What does the working voltage
|rating really mean?
|
Here's my guess;
It is probably easier to analyse if you look at asimpler mic amp
circuit such as
http://www.siliconchip.com.au/cms/A_102159/article.html
In this design the blocking caps are 10uF 16V NP.
Imagine that the balanced output is plugged into a mixer input
sourcing 48V phantom power.
The 48V phantom supply in the mixer desk will have a 6k8 resistor to
each of the XLR input pins (pins 2 & 3) thus its output impedance is
effectively 3k4 ohms. When fed down a cable to the mic amp output this
voltage is recombined via the two 1K resistors then via a blocking
diode and the 12V zener regulator to ground. We now effectively have a
series connection of 3k4 ohms + 500 ohms + 12V zener with 48V behind
it. Thus total current from the 48V supply into the mic amp output
will be (48 - 12)/3900 = approx 9mA. Thus 4.5mA will flow in each of
the 1K resistors meaning a volt drop of 4.5V across each of them. If
we add Zd = 12V + 4.5V we get 16.5V wrt grnd appearing at the
capacitor terminals connected to pins 2 and 3 of the output connector.
Since the op-amp output is referenced to Zd/2 = 6V the maximum DC
voltage which can appear across the output caps is around 11V which is
safely within their 16V rating.
- Next message: Watson A.Name - \: "Re: A Golden Opportunity!"
- Previous message: Sal Brisindi: "Re: Wanted 7448 0r 74LS48 BCD to 7 segment drivers ic's"
- In reply to: Walter Harley: "electrolytic mystery"
- Next in thread: Walter Harley: "Re: electrolytic mystery"
- Reply: Walter Harley: "Re: electrolytic mystery"
- Messages sorted by: [ date ] [ thread ]
Relevant Pages
|
