R/E/P Community

Please login or register.

Login with username, password and session length
Advanced search  

Pages: 1 [2] 3  All   Go Down

Author Topic: Km-56 vs Km-54 noise  (Read 3231 times)

klaus

  • Moderator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1414
Re: Km-56 vs Km-54 noise
« Reply #15 on: April 23, 2016, 11:28:49 am »

No Tantalums on my forum!

Only half kidding. They fit, for sure, but they suck, ranking at the very bottom of the pile by the 'Golden Ears' . Even Neumann stopped using them in one of their last discrete component mics, the U87Ai.
Logged
Klaus Heyne
German Masterworks
www.GermanMasterworks.com

Kai

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 250
Re: Km-56 vs Km-54 noise
« Reply #16 on: April 23, 2016, 07:00:17 pm »

...any effect on response or 'phase coherence' is well outside any humanly possible audible perception! ... good tantalum types are often used...
There are good reasons to use DC-coupled stages in amplifiers, but this is not possible in the "c" circuit.
If you have more than one of these microphones, it would be a good idea to mod all of them, because in stereo application phase response should match as close as possible.

But keep in mind, an unmodded microphone has a higher value to collectors, even though it might perform weaker in a technical sense.
Anyway you will not get real low noise microphones out of those.

So an alternative would be not to mod them but just repair the one that is broken, as close as possible to original state.

Tantalum - why not, they perform much better then AL-electrolytics in some areas.
Plus-they are historical correct, but maybe not at this model.
I have to admit that low-ESR-types come close in performance, but judging them on the audio performance is out of my range.
If you ask me I prefer foil types to electrolytics of any make.
This would again mean stay with the old circuit, which does not need the cap at all.
Logged

klaus

  • Moderator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1414
Re: Km-56 vs Km-54 noise
« Reply #17 on: April 23, 2016, 09:49:07 pm »

Modding from fixed to cathode bias, as the opening post envisions, has several effects, some objectively measurable, some very subjective.

The objective one, and the reason why Neumann switched all of its triode mics to cathode biasing: improved noise floor, and better rejection of hum/stray interference.

The subjective one: a cathode-biased mic, all other things staying the same, sounds noticeably different. It's cleaner, clearer sounding, but it loses some woolliness and mid range texture in the process. In KM5x I see no issue with converting, whereas with M49, M269 it is a much more discretionary, subjective decision of taste, where, despite the noise floor advantages, some may prefer the sound of fixed bias.

An analogous example is the C12 (fixed) and C24 (cathode). These are clearly different sounding mics despite all major components and voltages being identical.
I even experimented with U47 and U67 and installed cathode biasing there, and in both cases the mics lost too much of their personality, sounding too glassy, "clean" and lacking in personality.
Logged
Klaus Heyne
German Masterworks
www.GermanMasterworks.com

soapfoot

  • Full Member
  • ***
  • Offline Offline
  • Posts: 209
  • brad allen williams
Re: Km-56 vs Km-54 noise
« Reply #18 on: April 27, 2016, 10:45:13 am »

C12 (fixed) and C24 (cathode)

I did not realize this difference between these two mics. Thanks, Klaus.
Logged

Jim Williams

  • Hero Member
  • *****
  • Online Online
  • Posts: 542
Re: Km-56 vs Km-54 noise
« Reply #19 on: April 27, 2016, 11:52:38 am »

I checked both C-12 and C-24 schematics and they look very close. Both use a 200 meg grid bleed resistor to ground, not to the cathode.

Am I missing something here?
Logged

klaus

  • Moderator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1414
Re: Km-56 vs Km-54 noise
« Reply #20 on: April 27, 2016, 12:40:46 pm »

You are missing that the C24's 6072 tube uses classic cathode biasing (bias resistor and and capacitor between cathodes and ground). The C12's 6072 is fixed biased, via pin 4 from the power supply.
Logged
Klaus Heyne
German Masterworks
www.GermanMasterworks.com

AusTex64

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 67
  • Real Full Name: Robert Mokry
Re: Km-56 vs Km-54 noise
« Reply #21 on: September 24, 2017, 11:20:19 pm »

Sorry to post to such an old thread. Will someone please explain the implications of C4 and R1 in KM54? It looks like feedback to me. Is this to help tame the HF rise on the KK54 capsule? Thanks.
Logged

uwe ret

  • Newbie
  • *
  • Offline Offline
  • Posts: 34
  • Real Full Name: Uwe Sattler
Re: Km-56 vs Km-54 noise
« Reply #22 on: September 25, 2017, 12:08:28 am »

In the KM 54 the R1/C4 combination (250kΩ/160pF) provides negative feedback and suppresses tendencies of high frequency oscillation in the circuit caused by the complex transformer load. The phase correction becomes effective above frequencies above 4kHz. (f = 1/2π x R x C)
Logged

AusTex64

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 67
  • Real Full Name: Robert Mokry
Re: Km-56 vs Km-54 noise
« Reply #23 on: September 26, 2017, 01:08:20 pm »

Uwe, thanks for your reply.

Here is what I'm trying to figure out. I understand KK54 and Gefell M94 to be very similar, but not identical. The freq plots for M94 show a +5dB boost around 8kHz. I built 4 mics, using MV691 bodies, M94 capsules and a custom tube impedance converter. Topology is cathode biased, plate follower design with output transformer. I call them "my poor man's KM54's".

The issue is my mics seem pretty bright, probably because of the +5dB HF rise in the M94 capsule. However, I see no such rise in freq plots for KM54. Does KK54 not have the HF rise present in M94? If it does, how does KM54 eliminate it in order to achieve pretty flat freq response, at least until falloff starts around 12kHz? I'm assuming KK54 does have the rise due to the similarity to M94, and the feedback through the R1/C1 combination is what was making the correction. But perhaps my assumption is incorrect. Thus my question.

Also, the schematic for KM54a shows the KK54 capsule's capacitance as 45pf, yet the KM54c schematic shows the KK54 capsule impedance as 37pf. Yet NF loop values are the same. ?

I'm sure I'm missing a piece of the puzzle here. Thanks for your patience.
Logged

klaus

  • Moderator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1414
Re: Km-56 vs Km-54 noise
« Reply #24 on: September 27, 2017, 02:20:33 am »

(...) the schematic for KM54a shows the KK54 capsule's capacitance as 45pf, yet the KM54c schematic shows the KK54 capsule impedance as 37pf.
That's an easy explanation.
The schematic date for the KM54 "c" is 1965, at which point Neumann had moved on from the nickel K54 to Mylar (the same capsule would later become famous as the heart of the KM84 and KM86 with nominal 34pf, not 37pf capacitance.

To further complicate the transition from KM54 to KM64 and beyond: the earliest KM54 with an embossed "c" behind the serial number already had the "c" cathode biasing but still used the original nickel capsules from all prior KM54 generations. Later KM54c with Mylar skins used a decoupling ring to facilitate the grounded backplate connection of all KK64, KK74, and KK84 capsules.

In this regard, the Neumann schematic Uwe posted is incorrectly deleting the components and circuit of the decoupler, as there are no 34pf Mylar heads connected that way.

Boy, what a long-winded story.
Logged
Klaus Heyne
German Masterworks
www.GermanMasterworks.com

David Satz

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 93
Re: Km-56 vs Km-54 noise
« Reply #25 on: September 27, 2017, 08:46:47 am »

Klaus, are you saying that some KM 54c were manufactured (i.e. as direct products of the factory) using a capsule with Mylar membrane, rather than being retrofitted for repair? Martin Schneider has said on the Neumann Pinboard, "No KM54/254 was originally produced with anything but nickel membranes. There is/was a conversion available from Neumann which uses current production K84/184 mylar gold capsules in conjuction with an electronic subassembly under the capsule head." Later he added for clarification, "When KM5x capsule production ran out, the corresponding KM6x capsules with a mechanical adaptor were used to repair defective KM5x's."

--Separately he made another comment which is relevant to this thread: "(O)nly the earliest KM54s had the mentioned 5 dB rise, the later KM54a/b/c had a basically flat frequency response (modified circuitry), with -3dB at 15kHz."

--best regards
Logged

klaus

  • Moderator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1414
Re: Km-56 vs Km-54 noise
« Reply #26 on: September 27, 2017, 01:47:12 pm »

Neumann's KM54 capsule history is complicated.

1. I don't believe the schematic showing a KM54c with 34pf capsule (reply #5) is correct. Neumann's nickel capsules were always 45pf, probably because this metal membrane is so incredibly thin (under 1) that arcing would ensue too easily were it placed any closer than the 45pf indicates.

The schematic showing a 34pf capsule on a KM54 was issued in September 1966, more than a year after K54 with 45pf were replaced by the KM64 with 34pf Mylar capsule. There is also no schematic I have in my collection showing a Neumann pencil mic Mylar capsule with anything other than a 34pf capacitance- that capsule was manufactured unchanged from 1965 to the late 1980s.
I guess this was a schematic specifically issued for servicing the transition KM54 with K64 heads retrofitted with an adaptor.

2. You and Schneider are at least partially right. As far as I have seen, only very few KM54 were issued ex factory with the adaptor to accommodate the then new K64 Mylar heads, but plenty of KM54 were retrofitted with adaptor and K64 head. Many of these went to the German Broadcast system in the mid 1960s.

3. 1965 was a transition year. KM64 with the Mylar capsule were already shipped in spring of that year (the capsule was developed in 1964). U64 followed in 1966, and by February 1967 the first KM84 were shipped. Yet, KM54c were not immediately discontinued, but shipped as late as July 1966 to major German Broadcasters. I own one of the KM54c shipped to WDR in Cologne, and it sports the Mylar K64 head with adaptor.

4. 1965 was not the end for the K54 nickel capsule. The KM88, made from 1969 to 1986, used two of them (left-overs from the SM2/23 period).

5. Addressing the question of the KM54's frequency response:
I have never seen an "early" KM54 mic or schematic with a stronger response curve in the higher frequencies. I would love to see at least a schematic pre-dating my earliest one (1957) that would explain such rise (Uwe?). A possible explanation for an "early" KM54 with stronger high end may be the lack of  feedback network Uwe mentioned. High-cut and output networks were introduced gradually in the mid to late 1950s in all Neumann condenser mics with AC701 tubes.
Logged
Klaus Heyne
German Masterworks
www.GermanMasterworks.com

uwe ret

  • Newbie
  • *
  • Offline Offline
  • Posts: 34
  • Real Full Name: Uwe Sattler
Re: Km-56 vs Km-54 noise
« Reply #27 on: September 27, 2017, 10:09:46 pm »

The earliest on in my files dates from Oct. 15 of 1957 and shows the capsule capacitance to be 45 pF (nickel diaphragm. The values of the feedback components corresponding to R1 and C4 of the later schematic versions (up to -c version) are identical at 160 pF and 250 kΩ. Except for the bias scheme and later capsule capacitance all schematic diagrams show identical values, including the transformer, indicating that the frequency response of nickel- and mylar diaphragms match closely.
The frequency response plots are published in the respective user manuals for KM53a/KM54a and KM53c/KM54c and seem to confirm this. However, I a slightly different frequency response for the KM54a can be found in the manual for the initial release. Also, some general KM54 information is available as well and can be found at:
http://www.neumann.com/?lang=en&id=hist_microphones&cid=km54_publications
Logged

uwe ret

  • Newbie
  • *
  • Offline Offline
  • Posts: 34
  • Real Full Name: Uwe Sattler
Re: Km-56 vs Km-54 noise
« Reply #28 on: September 27, 2017, 11:41:17 pm »

Older KM54a and KM54c frequency response plots:
Logged

AusTex64

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 67
  • Real Full Name: Robert Mokry
Re: Km-56 vs Km-54 noise
« Reply #29 on: September 28, 2017, 01:22:36 pm »

Wow guys, this thread is an awesome learning experience. Thanks to all for the info. The documents on the Neumann site are quite useful.

It looks like my assumptions may be correct regarding the HF rise being similar in the KK54 and M94 capsules, based on Martin's comments that David Satz posted.

Regarding M94, Gefell says "The M294 is corrected for mid to far field recording, and is characterized with an emphasized presence peak (+5dB) in the 8kHz region." Wonder why Neumann decided to flatten it out? My guess is they were using it for broadcast dialog and perhaps the users requested a flatter response. I can see that peak maybe being problematic with siblance for dialog.

In any case, I'm going to try the negative feedback approach used in KM54 in my mics. The Neumann docs call for a "Z 54 testing capsule", which I imagine is a 45pf (or higher) capacitor to replace the nickel capsule during test. Neumann also lists "M 73k test apparatus for testing microphone preamplifier and power supply unit". Any hints on what that comprises?

Seems to me one would attach the capacitor in place of the capsule, insert test signal through the calibration input (C3), monitor the mic output with a frequency analyzer presenting the correct load to the mic output transformer  (1500 ohms or greater), adjust the value of R2 via 30K trimpot until -5dB at 8K frequency response is achieved, then solder in a corresponding fixed value resistor(s) very closely matching the value determined by the trimpot. I don't speak German, but Google says the word "abgleichen" next to R2 in the schematic means "to compare with/to level off/to match/ to tune". And since I don't have an anechoic chamber to use white noise into the capsule, it seems the best I can do is try to adjust out the capsule +5dB HF rise with a -5dB HF droop.

Thoughts?
Logged
Pages: 1 [2] 3  All   Go Up