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: Self Bias vs. Fixed Bias (in the context of microphone design)  (Read 40763 times)

R-AP.SCI

  • Newbie
  • *
  • Offline Offline
  • Posts: 13
Self Bias vs. Fixed Bias (in the context of microphone design)
« on: September 19, 2010, 01:16:47 AM »

     Good day to all. I would like to hear some opinions regarding biasing.

    In an amplifying scenario such as a tube microphone circuit there is more than one school of thought in terms of design. One of the elements of said designs is the form of biasing. While I happen to enjoy the second order harmonics, earlier breakup and "warmth" many times associated with self-biasing in my microphones, many venerable microphones of fantastic design use fixed bias (U47, U67) and some have changed over time (M49 revisions). Obviously a "good" or "great" microphone is more than just the sum of its parts (sometimes and depending upon who you ask) and the overall sonic results are dictated by the complex interactions of the entire circuit, not just the influence of one design parameter. However my inquiry will be twofold:

1) Historic: Was the choice for self or fixed (in regards to "high end" condenser tube mics...i.e. the M49's revision c switching from fixed to self) made for a particular sonic reason or was it a budgetary concern? Or both?
In my opinion, all things being equal, it must have been a total design consideration, i.e., whether or not small or large amounts of NFB was being used (if any at all), tube considerations, headroom etc. This only seems reasonable assuming the designers realized and understood it as being a whole interactive system, not separate design parameters independent of one another that theoretically "should" work in harmony. It seems to me that Neumann and A.K.G. would see it as a total design function. I am not questioning the wisdom, merely the motivation, seeing as how it does have audible results.
Or was it simply what they figured was appropriate, given the resources and knowledge of the time?

2) Design: To the designers, technicians and modifiers participating in this forum I ask this:
Why do YOU choose one biasing scheme over another? Given the wide availability of parts and manufacture and significant access to great designs of the past and hopeful self-curation of microphone advancement, it seems to me that the biasing structure would be a decision of primarily sonic considerations.

Mr.Archut: Your CS-4\
Mr.Josephson: While I cannot think of any tube microphone you currently manufacture, your knowledge and opinion is still greatly valued and desired.
Mr.Bock: Your 507
Mr.Heyne: Any particular 67 or 49 modification stand out to you in this context? Or perhaps your KHE?

While this post does have a very wide arch, I hope the answers achieve some specificity, if not then just some insight and opinions of well informed individuals, hopefully all contributing to the understanding and love of tubed microphones!

This has been an area of interest of mine since I began my wonderful trip into the wonderful world of microphones. I hope, given this forum's wealth of collective knowledge, that not only can we see into the design wisdom of the past but also into the current designing spirit.

Thank you all!
Logged

Oliver Archut

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1125
Re: Self Bias vs. Fixed Bias (in the context of microphone design)
« Reply #1 on: September 19, 2010, 01:16:24 PM »

There is no real tube-microphone handbook, so we are lacking the proper names for the different ways of biasing. Using the standard terms is not proper because in tube mics there is the problem of the startup current, that is shorted out via the grid leak in normal tube design.

There are only a handful of tubes that were developed for microphone use and the rest are just normal tubes operated out of their recommended range. From a technical point of view, biasing is just a way of making the grid more negative than the cathode.

This can be done via:

* External source to cathode, like U47 current bias from the filament (CS-4 is done the same way).

* External negative voltage to grid, like U67 and C12.

* Mix of both- external source to cathode in conjunction with voltage divider, like  
  the M49/M49B.

* Self biasing via cathode resistor, like M49c and the like.

* Biasing via a capacitor or battery, used as a coupling cap that is charged via the  
  grid current; not done in commercial tube mics.

At the same time the start up current of the heated cathode will produce an negative voltage that drops via the grid leak resistor, that either works with the bias configuration (self-bias), or is forced to the given bias voltage (external source).

But from a practical point of view, the biasing is responsible how a mic behaves.

Best regards,

Logged
Oliver Archut
www.tab-funkenwerk.com

We are so advanced, that we can develop technology that can determine how much damage the earth has taken from the development of that technology.

Geoff Emerick de Fake

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 348
Re: Self Bias vs. Fixed Bias (in the context of microphone design)
« Reply #2 on: September 20, 2010, 04:32:45 PM »

In the very early days of "radio" (the word "electronics" was not used yet), almost everything was fixed-bias, because the cost of a resistor and a cap exceeded the cost of a bias battery. For many designers, that was the way they had been told, so they kept on.
On a purely technical standpoint, there is no significant difference of performance between fixed and self-bias on a low-level preamp stage. Since there is a cost penalty to fixed-bias,
self-bias is the logical choice.
Mic head amps have the peculiarity of operating with extremely high value grid resistors, which makes them extremely sensitive to grid current. The relative voltages between heater and cathode have a subtle but noticeable influence on grid current. Some designers have favoured the fixed-bias scheme on the assumption that the voltage relationship was a more predictable one than self-bias. In fact, both designs can be optimised equally well.
Personally, I can't hear a difference between a fixed-bias and a self-bias low level gain stage.
Logged

David Bock

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 333
Re: Self Bias vs. Fixed Bias (in the context of microphone design)
« Reply #3 on: September 20, 2010, 05:35:01 PM »

Testing a mic on high impact sources like drums can reveal the difference between self & fixed bias mic amps.
Fixed bias amps are generally much more sensitive to the sound of the specific individual tube being used in the mic. So flaws show up sooner.
Self bias (in the most commonly implemented cathode bypass cap method) introduces overshoot in the cathode, see Valley & Wallman chapetr 3.
The type (or brand) of cap in the cathode can affect the sound.
Since self bias is the easiest, cheapest, most forgiving to implement, & no one complained at the time, I'm guessing it was an obvious "nothing but benefits" decision to make.

R-AP.SCI

  • Newbie
  • *
  • Offline Offline
  • Posts: 13
Re: Self Bias vs. Fixed Bias (in the context of microphone design)
« Reply #4 on: September 20, 2010, 07:00:43 PM »

Thankyou Mr.Archut and very glad to see others of whom I regard highly participating as well....

    For me, while the fixed bias scheme of external negative voltage may slightly increase headroom somewhat, it still may raise odd order harmonics, that coupled with the incorporation of additional parts (me being a minimalist I naturally feel that it complicates unnecessarily) it might not be worth it. And I enjoy the softening of self bias. Unless the circuit NEEDS this type of biasing due to other circuit considerations. A great example of this type of biasing is the U67, after one considers other aspects of its design (the feedback/forward & e.q.) to get the most out of the tube its plate voltage is run higher than the "average" 120v. I am not sure of any other mic that uses an EF86 at around 200v. Being that plate dissipation at idle is set by the bias level, is it reasonable to assume its instantaneous dissipation and average dissipation is also influenced by these parameters in conjunction (plate voltage + plate dissipation being influenced by biasing)....therefor the difference revealed by highly dynamic sources as stated by Mr.Bock.

    What is your consideration of this in the context of its design? However the U47 biasing scheme from the filament sounds interesting because the tube is underheated...do you find this makes any difference in terms of headroom and harmonic distortion (pleasing or otherwise)? Is the Telefunken tube in your CS4 similarly underheated? Mr.Bock What of your 507?

    I have heard of battery/capacitor biasing but have not experimented any with it...If I can, I will do so eventually in a mic capacity. Have any of you?

    I am aware that in comparison to other considerations and implementations in a mic the audibility of this design element is small or diminished, however I do think it is important nonetheless. I also am aware that I might be "overthinking" this...however I am a bit persnickety about my mics.

    It is good that you pointed out the fact that most tubes were not developed for microphone usage. I think mic design has many other considerations and variables in the usage of its tubes opposed to power amplification and other large gain applications. I have done basic preliminary research (not as extensive as yours) into developing and having manufactured tubes specific for microphone applications...it has proved to be most daunting!

Thankyou again Mr.Archut, Mr.Bock and Mr.Geoff Emerick de Fake for your participation in this thread! I look forward to hearing some more of your and others estimation.
Logged

Klaus Heyne

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 3154
Re: Self Bias vs. Fixed Bias (in the context of microphone design)
« Reply #5 on: September 20, 2010, 09:02:43 PM »

There is another consideration when using self-biasing:
Self-biasing is decidedly not a linear process, but a sound and frequency-shaping one, because of the low frequency phase shift introduced by the bypass capacitor's cut-off frequency.
Logged
Klaus Heyne
German Masterworks
www.GermanMasterworks.com

David Bock

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 333
Re: Self Bias vs. Fixed Bias (in the context of microphone design)
« Reply #6 on: September 20, 2010, 10:55:59 PM »

10Hz?

R-AP.SCI

  • Newbie
  • *
  • Offline Offline
  • Posts: 13
Re: Self Bias vs. Fixed Bias (in the context of microphone design)
« Reply #7 on: September 21, 2010, 03:14:46 AM »

     A true assessment Mr.Heyne, one that is many times easily overlooked. (thankyou for participating as well!) and Mr.Bock both. I am glad this was mentioned because now we are entering the "marriage of compromise" and "diminishing returns" issue that all designers and technicians face!

    The consideration of these issues is exacerbated by individuals and companies who care about all aspects and facets of their products and services...(Like those currently participating in this thread...). Since we are speaking of the biasing portion of the circuit and not say... the output coupling cap to transformer, I can understand how this poses issue on many levels in terms of component selection, audibility and thusly consequence.

    As such  I offer this: Dependent upon the C (and quality of the cap) I can see how it is possible the shift can begin high enough in the low end or be drastic enough in deviation to affect the audible areas of the low end, (I believe this MIGHT be an extreme case or it might be more common than many of us realize, conversely, in a well designed circuit it might not occur audibly). My solution to this (while somewhat simplistic) is to connect a high quality film cap of a much lower value in parallel, thereby mitigating (but not completely eliminating) certain performance maladies. In my opinion this is a good practice in general as the value of the original bypass cap is normally high-ish (I've seen 50uf and more) and therefor usually an electrolytic. Size considerations warrant the usage of said electrolytic and being that it is not directly in the audio path many designers can be satisfied with it as such. However, I humbly posit that bypassing THAT cap with aforementioned film cap will address many issues...IF there are any....which would also be a total circuit consideration. What do you gentlemen think?

    The various lines of standards and considerations are at many times subjective and dependent upon many interactive variables.....however the decisions made based upon them have in many cases a practical affect, and it is important that forums such as this exist and entertain said considerations. In the hopes of curating and progressing the art of microphones.

    I am excited to hear more opinions and assessments from well informed participants! Thankyou all!
Logged

MagnetoSound

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2589
Re: Self Bias vs. Fixed Bias (in the context of microphone design)
« Reply #8 on: September 21, 2010, 08:05:15 AM »

R-AP.SCI wrote on Tue, 21 September 2010 00:00

A great example of this type of biasing is the U67, after one considers other aspects of its design (the feedback/forward & e.q.) to get the most out of the tube its plate voltage is run higher than the "average" 120v. I am not sure of any other mic that uses an EF86 at around 200v.




Not sure what you're getting at here. In fact the EF86 in a U67 has approximately 75v at the plate.

The incoming HT before the plate resistor is indeed higher than usual - one of the reasons the U67 has such great headroom - but the tube is running comfortably within normal parameters.


Logged

Music can make me get right up out of my chair and start dancing or it can get me so pumped up I have to walk around the block.
It can also knock me back and make me sit there and cry like a little baby. This shit is as powerful as any drug!!!
- Larry DeVivo

R-AP.SCI

  • Newbie
  • *
  • Offline Offline
  • Posts: 13
Re: Self Bias vs. Fixed Bias (in the context of microphone design)
« Reply #9 on: September 21, 2010, 12:04:23 PM »

...oops, sorry, I meant B+!! that's what I get for not re-reading my post. (I guess my father was right about the whole "measure twice cut once" thing. Thankyou sir.
Logged

Geoff Emerick de Fake

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 348
Re: Self Bias vs. Fixed Bias (in the context of microphone design)
« Reply #10 on: September 21, 2010, 12:36:32 PM »

In order to make things clearer in perspective, here are the time-delay responses of both configurations, i.e. one with the influence of the cathode cap (25uF over 2.2k, as per U49c) compred to the same without the phase-shift.
In green, with the original 25uF, in blue, with the res infinitely bypassed.
As expected, the amount of phase-shift introduced by the cathode cap is much more than the one introduced by the xfmr/coupling cap combo.
At 20 Hz, the time-smear introduced by the cathode cap is 2milliseconds, much less than the 5ms of the xfmr.

Bypassing the electrolytic with a film cap is a very common, although controversial scheme. But it has no effect whatsoever on low-frequency phase-shift. Cathode cap LF phase-shift is solely and utterly determined by the Rk.Ck product; increasing this product by a fraction of a percent is not significant.

Regarding the "unusual bias of the EF86 in U67, this is in fact very common practice. You have to remember that for negative going outputs, the tube has a relatively unlimited current capability, but for positive-going, it is determined by the plate resistor; the more positive the output goes, the less current it can draw. In order to balance the positive and negative, the quiescent current is increased, thus reducing the plate voltage. An additional benefit is that the noise performance is (very marginally) better. Both fixed and self-bias offer this possibility.

As to  "Testing a mic on high impact sources like drums can reveal the difference between self & fixed bias mic amps", one has to be extremely cautious to make sure that the "all other parameters being equal" rule is enforced.
As an example, comparing an M49 (w/ fixed-bias AC701) and an M49c (self-bias AC701) is not a proper comparison, since the 49c runs at 60% more quiescent current (0.73mA vs. 0.45), lending it a 4dB advantage.
Logged

Geoff Emerick de Fake

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 348
Re: Self Bias vs. Fixed Bias (in the context of microphone design)
« Reply #11 on: September 21, 2010, 12:46:40 PM »

As an exercise, here is the time-delay introduced by the cathode-cap (green) compared to infinitely bypassed (red) (the xfmr is no more in circuit), showing 1.5ms at 20Hz.
I have favoured the time-delay display over phase-shift because it is more instinctive, methink.
Logged

Klaus Heyne

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 3154
Re: Self Bias vs. Fixed Bias (in the context of microphone design)
« Reply #12 on: September 21, 2010, 01:59:11 PM »

I hear the phase aberration of bypass capacitors that were not chosen carefully through hearing experimentation, and I design remedies accordingly, regardless of what one thinks what should or should not theoretically be audible.

So, I am very much insistent that, if we discuss electronic circuits or esoteric phenomena in this forum, we relate them to what we experience as audible about them (or not.) If that is not  done, the discussion cannot possibly lead to better sounding mics.

Logged
Klaus Heyne
German Masterworks
www.GermanMasterworks.com

R-AP.SCI

  • Newbie
  • *
  • Offline Offline
  • Posts: 13
Re: Self Bias vs. Fixed Bias (in the context of microphone design)
« Reply #13 on: September 22, 2010, 06:12:55 AM »

     This is true Mr.Heyne...whilst theory should be the functioning framework, we still have to use our ears! And at many times make a subjective decision based upon a careful balance of cultivated preference and measurement, hearing and experimentation, because the goal (for some, myself included) is better SOUNDING mics! I am of the mind that while certain time and frequency based phenomena should not be audible (according to todays current standards and theories-whatever that means), I think we all have encountered an instance of audio perception upon which we hear (or think we hear) something which theoretically should not or cannot be heard.

    Either way, if I replace a component and I experience a change, then I include it as a variable of some sort that might produce sonically pleasing results...(of course this goes hand in hand with some repeatable testing methodology). I also believe that we lack a total comprehension and development of methods of testing as well....in the sense of if I am constantly perceiving something that according to my measurements has not occurred or should be outside the realm of my perception, then I question BOTH my myself (expectation bias, etc...) and my test! After that, hopefully a refined sonic aesthetic born of discernment through as you put "hearing experimentation" can be a suitable guide.

    Thankyou for those measurements good sir! They are of serious consideration and really help to put things into some perspective...it also indicates that one cannot consider aspects of these designs independently of one another. Just to add another log on the fire:...You make the well informed point of the relationship between the plate resistor, positive output, current and plate voltage...which would naturally (for some) bring us to the issue of transconductance. This is an important factor when considering tubes in general and given this relationship should also be a factor when considering a biasing scheme. In my opinion this bias-transconductance factor (for many, a preference for high transconductance) could be a very good indication as to why pentodes are so favorable in many amplification applications in general (I personally have an affinity for the 6SJ7 in mics), even when sometimes connected as triodes! Even though triodes can (for some) be looked at as more universal.

    As an aside, on the issue of audibility, I at many times attribute these phenomena to something of a "black hole effect"...i.e....while something (like a black hole or phase shift at 10-20hz) may or may not be directly perceivable, it exerts changes and influences upon its surrounding environment (in a diminishing capacity relative to its distance from the "point of origin") and will have consequences and affects reaching into areas that ARE directly perceivable. And with enough information (carefully scrutinized and analyzed) one can realize, with some degree of accuracy, what the "point of origin" of these aberrations is...and propose viable solutions. But thats just my opinion and more importantly a topic for another thread.

    This was the type of exchange I always looked for in various internet fora and this reason why I joined this one....I would like to hear if Mr.Heyne or Mr.Bock (or anyone else) have any preferences towards any particular type of biasing, what they are and what their opinions are of them sonically. Have any of you experimented with say modifying a 67 to have a self bias?
Logged

Klaus Heyne

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 3154
Re: Self Bias vs. Fixed Bias (in the context of microphone design)
« Reply #14 on: September 22, 2010, 12:43:21 PM »

As I mentioned in a previous thread dealing with this issue:
changing fixed bias in U67 and U47 to cathode bias is detrimental to the character of these mics. They "thin out"; and while they become quieter, they also become too sterile and listless and transparent. (Yes, too transparent can be a problem: you look right through the leaves of the tree!)

That's another reason why modifications of U47 mics to run with EF14 tubes are questionable: they often include self-biasing, due to feeding a separate heater voltage to the mic. The tube's sound is already a bit anaemic, even with fixed biasing.
Logged
Klaus Heyne
German Masterworks
www.GermanMasterworks.com
Pages: [1] 2 3  All   Go Up
 

Site Hosted By Ashdown Technologies, Inc.

Page created in 0.033 seconds with 20 queries.