Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[soapfoot]

one of my favorite quotes from the venerable H.H. Scott:

"If it sounds good and measures bad, it's good. If it sounds bad and measures good, you've measured the wrong thing."

[Uwe]

... if you can not measure it, you may have to use a different instrument or test procedure ...

[Jim Williams]

... if you can not measure it, you may have to use a different instrument or test procedure ...

What is going to work better than an Audio Precison test rig? I have that and it will not measure small audible differences in many cases. A perfect example is cables. Many hear differences with those but the AP will not detect their differences. Use an Agilent or HP network anayzer and you will start to see those differences at a much higher than an audio bandwidth. Ray Kimber has that $50,000 piece in his lab used to design his spectacular cables.

If there were tests or proceedures (a collection of tests with pre-programmed perameters) that would measure every sonic difference heard by the end user, I would already own it. Part of the problem is that the test stimulus used by audio analyzers are fixed waveforms, sine or square. Music is full of violent, radical waveforms not at all like the oscillator outputs of the best analyzers.

Until actual music waveforms can be used as the test stimulus, we will only find the best way to measure simple waveforms, not musical waveforms. There is going to be some stuff missed.

[klaus]

Until actual music waveforms can be used as the test stimulus, we will only find the best way to measure simple waveforms, not musical waveforms. There is going to be some stuff missed.

AMEN. (Starting to feel like the Church of Klaus here…)

[panman]

AMEN. (Starting to feel like the Church of Klaus here…)

Klaustianity?

[dark fader]

klausology 

[brightmillion]

so do the electrolytic caps (such as the one {output cap in C5??} in question in the photo that started this "cap sound difference" debate) wear out over time? sorry if this is a dumb question.. I know replacing the caps in my fender amps really get things sounding good again, but I'm sure thats a completely different reasoning compared to mics.
Maybe this question belongs in the other recent thread about what makes an "all original" mic,, but, can you guys clarify for me if there are components in vintage mics that are "regular maintenance" type parts, like the car example of changing spark plugs.? Are there components that over time do lose sound quality and should (must) be replaced, or is it always a matter of when things start to sound off then something has failed, but otherwise don't touch anything until it's broke?
for example if someone found a brand new, never used old stock u47 in a closet, would it sound as good today as it did when it was manufactured? or would even an unused specimen benefit from some new (but authentic/correct) parts replaced?

I realize there is plenty of debate about changing things and if it helps or hurt tone compared to the original design, but i'm speaking in purely making a vintage mic sound the same as it did when manufactured (whether that is the best or not opinion wise) - and i guess i'm asking as a newbie to having a couple vintage mics, what should I expect for maintenance with my new friends (mics)? - do they need regular tune ups?

[Jim Williams]

Wet electrolytic caps do dry out over time, they loose mfd's as a result. Usually you detect a high pass function when they dry out.

As far as audio quality is concerned, el caps are probably the worst passive component you can shove quality audio through. They absorb and filter electrons, the source of sound.

This is why many use a quality film cap alone or as a bypass to allow those electrons to pass through that would be absorbed by the electrolytic cap.

An audio system without the use of blocking caps sets the bar much higher, you can get spoiled by that sound. Still, some don't like that, "too much music". Many are conditioned to "recorded sound" and want their work to sound similar. No problem, most commercial gear will get you that sound. The open sonic effects of a direct coupled audio path are much harder to find.

[soapfoot]

When bypassing an electrolytic with a film cap of a smaller value, isn't it true that the (typically) higher ESR of the electrolytic will let higher frequencies (cutoff determined by film cap's value) take the "path of least resistance" around the (presumably) lower-ESR film cap?

[klaus]

If that theory of electrons taking the "path of least resistance" in a dual-cap paralleled arrangement can ever be proven to be true, it then would need to be examined how perfect the electrons' behavior is. For example, if some electrons at a certain frequency cross over get confused which path to take, what happens?

From my personal experience, I have reversed my opinion stemming from my original impression that bypassing coupling caps in microphones does any good (for a decade I religiously bypassed my f+f caps with styrenes of 10000pf or less, depending on circuit), and now believe that it does more harm than good: It smears the midrange (always more important to our ears than any high end benefits). Instead, I selected even better and more expensive f+f caps than were available 20 years ago.

[Jim Williams]

Ask any RF or power supply designer about bypass caps, they will explain it better than I. There is a reason all your audio gear uses a larger el cap combined with a small cap as a bypass in it's power supply, there are limits to what an electrolytic cap can pass or filter, they have internal resistance (ESR) that rises with increasing frequency, an effect the better film caps avoid. Their dialectric absorbtion rates are also very high compared to film caps.

Switch mode power supplies must use small caps to aborb and filter switching transients and artifacts. Bob Dobkin, head of Linear Technology has done papers on this subject. They specify certain types and even brands of film capacitors to assure their switching mode power supply chips perform up to spec.

As I mentioned before, these audio types of 'tests' and evaluations must be done using an el cap, a film cap and then no cap in a circuit to fully understand the sonic effects of capacitors on your audio. Most have never done those three comparisons so they may not have the background to make any assertions one way or another.

[klaus]

Jim,
Your examples of using bypass caps are from different applications than small signal, close to zero current, coupling caps in condenser mics. I am not questioning their use in power supplies, especially with polarized electrolytics. 

I agree with you that only comparative tests can decide whether bypass caps in the coupling position of microphones are appropriate or not. I found their sonic contribution to be negative when combined with very good film and foil caps. So I now remove the bypass caps from mics previously modified by me when I did not know any better, and I upgrade the f+f caps while I am at it.
The before-and-after effect is clear to my ears, every time I undertake this change.

[soapfoot]

But what about in signal applications where the use of an electrolytic is unavoidable (or all-but-unavoidable?)

I'm putting together some EMI/REDD 47 preamps right now that specify a 100µF cap for cathode bypass of both tubes.  Since 100µF film/foil caps are scarcely available (if at all), I chose to increase the values to 220µF and 470µF with a nice Nichicon KZ cap, bypassed with 2.2µF and 4.7µF polyprop film/foil, respectively. I have high hopes that this will perform better than the stock 100µF electrolytic alone, but experimental listening will tell the tale, when I get that far (a few weeks away perhaps). 

This is sort of drifting off-topic now, so my apologies... but in the general context it is relevant.

[Kai]

I now remove the bypass caps from mics previously modified by me when I did not know any better, and I upgrade the f+f caps while I am at it.
The before-and-after effect is clear to my ears, every time I undertake this change.

A bypass cap cannot improve (bass- and midrange-) distortions caused by a (much bigger) bad electrolytic cap.
These distortions are measurable under certain conditions, not just imagination!
Look here for an overview:
http://www.co-bw.com/Audio_Capacitor_Distrotion_Mechanisms.htm
See here for details:
http://waltjung.org/PDFs/Picking_Capacitors_1.pdf
http://waltjung.org/PDFs/Picking_Capacitors_2.pdf
So going the route of replacing bad caps by better ones sounds 100% senseful to me.
A small bypass cap only improves the HF impedance of a bigger one, that's all.
It doesn't cure any other problems.

E.g. without those (ceramic caps for HF blocking), no modern computer ever since would have worked. But that's in a frequency range of 10 MHz to 4 GHz, far beyond audio.
Ceramics, BTW, have very bad distortion figures, don't use them for audio.

For audio, bypassing even bigger foil caps with smaller ones of the same make, can improve low impedance circuits. JBL demonstrates this for ages with their passive speaker x-overs.
 
In a tube mic no cap is used in a real low impedace audio configuration, where the situation is most critical. The (usually 0.5 - 2 uF) coupling cap to the output transformer is what comes closest to this.
Here the 12 dB/Okt high pass function in conjunction with the transformer can bring a bad cap into a situation where it causes distortion, although it's DC bias helps a lot.
The standard U47 0.5uF paper-metal foil cap e.g. is not suspicious at all.
But it's big, this size cap does not fit into every kind of mic.

Regards
Kai

[klaus]

All I can tell you is that my story is not about bypassing bad caps, but very good caps, and the result, in hindsight of a decade or more, was a decrease in resolution and an increase in mid range smear. I let other people's talk get in the way of making a decision based on my own experience and what I hear. Removing the bypasses relaxed the sound of virtually every mic that had one installed considerably, without any discernible loss in audio quality.

This may be a different story in "bad cap" situations. However, I cannot foresee a microphone application where I would keep a low quality cap in the circuit in the first place, to mitigate its lossy performance. I would simply replace it with the best quality cap my ears can discern and my pocket book can afford.