Print

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

[soapfoot]

Wima FKP3 are available in much lower voltages and therefore smaller sizes.
Depending on the purpose these may be a good option, even paralleled for higher capacitance values.
They have the lowest distortion of all available (I built my rig now and can measure down to -154 dB or 0.000002 %).
Selected specimen come down to this figure, even unselected they're not too much worse.
The downside: they are really hard to find if you don't want to buy really large amounts.

Of course, if you want a cap to deliver a special "sound" others might be the choice.

Many values seem to be available from Mouser with no minimum order:

https://www.mouser.com/WIMA/Film-Capacitors/FKP3-Series/_/N-1z0zldhZ9x371Z1z0ye2a

[Kai]

The lower voltage/higher capacity versions are hard to find, unfortunately.
I had to oder from Indonesia, via Ebay lately.
They are worth the hassle, from a technical point of view if you're heading for minimum signal deformation.
I can't comment on the sound, I never A/B 'd them to something else.

But:
The use of technically superior parts does not necessarily lead into better sound or even better technical results.
Sometimes compensatoric effects are built into something, and constructions must be seen as a whole.
Advanced mic's like the ones from Neumann were not just put together with what was lying around, they're well thought out and thoroughly tested, and every of the few parts carefully selected I guess.

[klaus]

Wima FKP3 are available (...)
Depending on the purpose these may be a good option, even paralleled for higher capacitance values.

My experience has not been good with paralleled coupling caps. I first was THINKING (rather than listening!) that this made perfect sense, but abandoned this method years before WonderCap was launched (which was later bought out by Rel-Cap). I tested Rel-Cap's WonderCaps as well, and found them to be an overpriced naked emperor. Resonances MORE smear and no clarity.

When I noticed (late 1990s) that other modifiers copied my paralleling (1mfd film and foil with 10.000pfd styrene) I went back and listened and have since ripped out that combo from all of my earlier mods.

Here is a good explanation why paralleling caps is not working so well:
http://conradhoffman.com/hoffman_effect.htm

[Kai]

The mentioned theory is based on combining very different caps with high and low losses.
I could observe effects that can be explained by this when building my ultra low distortions measurement rig (besides finding out that a perfercly clean audio path does not exist).

If you parallel same sized very low loss caps, e.g. 2 x 0.22 uF or 4 x 0.15 uF to replace a 0.5 uF cap, this does not apply, and I could not observe any negative effects in that configuration.
When you choose 0.15 or 0.22 || 0.33 uF in that situation (to hit the nominal value of 0.5 uF more exact) it needs some practical "Golden Ear" evaluation if the different resonant frequencies of the 2 caps (far outside the audio range) does come into the way of perfect sound.

BTW: one has to be careful if a theory is based on models only, not practical evaluation.
Even the author states that his model does not work with caps of same parameters (division by 0 problem).

[AusTex64]

... not according to the page you cited.

I commissioned Bass Lim, in charge at Relcap, to roll me a few 1.0 mfd styrenes, so I speak from direct experience:
Not only do styrene caps sound lousy as coupling caps in values above 10.000mfd, they also would be so large at 1.0mdf (the standard in condenser mics) that they would not fit inside a U67.

Tin film & foil is the way to go, and even there, the specialty under-sized cap Relcap makes for me would not fit a U67 unless I reposition the tube and its socket (PITA).

Perhaps I misread this: Capacitance Values:   .01 to 3.0 mfd

I should have been more clear - I was referring to mic preamps, not buffers/impedance converters inside mics.

Also agree about bypassing caps - I think it’s BS.

Agreed that they would be too large for mics, and take your word on the sound. You’ve certainly done more critical listening of coupling caps in mics than I. Appreciate your observations!

[Jim Williams]

As much as I like my Audio Precision and the other test gear here, it does not tell the full story. In fact, it's very incomplete.

Sure, we can test down to the nano volt but it does not measure sonic colors nor qualify them. Similar measuring stuff doesn't sound similar in many cases. Noise gain tricks and FFT's will not show those colors. THD+noise and other distortion tests like CCIF IMD will not show the colors either. It will show related errors if measurable.

Dynamic response is also very hard to test. Violent musical waveforms would be the best test stymuli but that technology isn't developed. Higher output current devices also deliver a greater depth of low end response, something that also doesn't show up on audio tests until you get into power dissapation under load tests.

I like WIMA's scoopage design, it does overcome some capacitor errors but I find the brand to have sonic issues as well. The same values in Rel Cap RT caps have better resolution here. Exotic silver foil caps have even better resolution. I have also had Bass Lim make me their copper foil polyprop caps, I also found those wanting. They are available from Rel Cap if you want to try them out.

I do these tests here against my current ultra high quality capacitor base standard, a 1/2" piece of Kimber Black Pearls solid silver wire.

Yes, I am a bit anti-capacitor here as I spend much of my bench time on recording consoles and as such removing them by the thousands. The superior sonics of a well designed direct coupled audio system cannot be overstated. Most of my mics here are down to one capacitor in the signal path.

I also add selected bypass caps to larger el caps and some film caps. Not only myself but my customers hear an improved transient reponse and they always tell me the general sonics improve more with them than without. In this case, the customer's are right.

[Kai]

As much as I like my Audio Precision .... Noise gain tricks and FFT's will not show those colors..

Audio Precision isn't capable measuring the low distortions from high quality caps without the mentioned trick.
My system can, just straight analysis of an extremly clean generator signal.

This does not mean it can replace listening tests, but it's a good starting point.

...I do these tests here against my current ultra high quality capacitor base standard, a 1/2" piece of Kimber Black Pearls solid silver wire.

Can you explain that further, how can this extremly small cap (few pF) resulting from a 13 mm piece of cable serve as a reference for bigger caps?

Yes, I am a bit anti-capacitor here as I spend much of my bench time on recording consoles and as such removing them by the thousands.

In the aera of large scale analog consoles DC blocking caps were used all over in an inefficient attempt to avoid clicks of the mechanical switches and DC noise of controls.
Very often polarized electrolytics were used out of specs (for AC), making the situation even worse. Most of those are obsolete.

[Jim Williams]

The Kimber Black Pearls solid core silver wire is 26 awg, .0009% pure and coated with a 10,000V rated teflon dialectric. It's around $50 per foot.

It is my quality standard for audio electron transfer and easily beats any capacitor ever made for transparency. It is also used inside my mics to connect capsules to jfet gates or tube grids. I also use their 19 awg silver stranded wire here. That is used to connect mic pcb's to XLR connectors. I don't wear jewelry, I listen through it.

Millions of electrolytic capacitors were used in analog recording equipment, usually tossed in without checking the DC offset polarity. That was a cheap way out from designers that used them to stop the DC offsets from 072 and 5534 opamps. One will pay a sonic price for using them.

Modern opamps don't have those DC offsest in many circuits. Then those caps can be removed. Alternative techniques are DC servos and offset trimpots, very rare in audio gear.

[Brian Campbell]

In the U67 schematics,  C17 is shown as variable 80-160pf. Is this to allow for variables in capsule, transformer or both?
Also what would be the range of frequency change with these values?

[klaus]

As the test input only measures amplifier performance, without taking the variables of a hand-made capsule into consideration, it was probably the sum of all amplifier components-passive, tube, transformer- affecting the high frequencies. Hence a simple way was found to keep the highs within specs during final testing and adjusting.

You asked the same question on another forum, where at least one person speculated that the U67's transformer was so complicated that an adjustment for its variations in HF transmission alone would be necessary. I have my doubts as all C17 installed in U67 of a given period were always of the same value: either 80pf or 100pf (nothing above 100pf was ever installed in U67 I have seen).

Both values are so close, that I rather believe C17 was another way to show the German broadcast brown-book folks that the mic can be quickly adapted to their strict HF cut requirements. As the feature was retained in the broadcast version of the U67, it might have helped in the end to pass the mic. (All of this is just more speculation, I am afraid)

[Brian Campbell]

thank you

[Brian Campbell]

For example the initial U67u 00-00-00S. (Considering that 17 is the highest number for a capacitor in this model, the designation  as C17 may be an indication that is may have been an afterthought...):

Interesting to note in that schematic that the heater supply is +6.3V on pin 4 not -6.3V.


(Pardon me for bringing this thread up again but I find this mic fascinating)

[klaus]

A clerical mistake, not a design change: the second value on the schematic, a bit closer to the tube filament, shows the correct negative voltage.

[gtoledo3]

Absolutely fascinating! Your schematic is dated one day after the first series I have (21.6.1960) which sports the C17.

Your schematic is curiously labeled "U67u", so let me speculate: In addition to the NWDR-rejected model U67 (later re-submitted as M269), Neumann initially contemplated also offering an export/U.S.- version of the U67, without the HF broadcast attenuator ("u" is a Neumann suffix often used for export power supplies, specifically for the North-American market; also note that the U67u schematic is using English language and U.S. component measurements/symbols).

Old topic, but I noticed it because it was linked within the top thread in the forum.

Perhaps the “U” in U67u is for “Umbau” (modification), the same as the rare M49u and M249u.