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: Underheating tubes in microphones  (Read 15606 times)

AusTex64

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 75
  • Real Full Name: Robert Mokry
Underheating tubes in microphones
« on: May 07, 2015, 09:53:19 PM »

I would like to hear other's experiences with underheating tubes in mics. I understand there are only benefits - reduced self noise, longer heater/tube life, and higher impedance presented to the capsule in most cases. I've underheated the Telefunken PCC85 in my CS-1 build from 9V to 7V and it sounds great with very low self noise. I run the EF802 in my U47 build at 5.5V instead of 6V, it sounds great and is the most quiet tube mic I've listened to so far. Quieter in fact than my mid '70's U87. Learned about underheating from the late, great Oliver Archut.

Does anyone know how much more life one can get from the tube by underheating? I know with tungsten light bulbs, underheating the filament by 5% will increase life by 40% (saw this in a GE paper 20 years ago, I'm in the lighting business). Since tubes have tungsten heaters, I don't see why the same metric would not apply.

Do any of the manufacturers underheat tubes in mics like C12, M49, M147, Manley, etc?

Robert Mokry
Logged

soapfoot

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 285
  • brad allen williams
Re: Underheating tubes in microphones
« Reply #1 on: May 08, 2015, 08:53:06 AM »

Underheating the tube will change its transfer characteristics somewhat. Klaus or someone else will correct me if I'm wrong, but I'm not sure it's advisable to underheat a tube unless the mic was designed around the tube in its underheated configuration.

Also, there's a point beyond which underheating is detrimental to tube life, not beneficial. Up to 10% is usually safe for the tube.

Also worth noting-- preserving the life of the filament itself is not the primary concern. The filament is rarely the part to fail when a tube fails.
Logged

Jim Williams

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 599
Re: Underheating tubes in microphones
« Reply #2 on: May 08, 2015, 11:29:26 AM »

There are a series of articles in the old Audio Express magazine on this subject. Most of the subjects involved 6.3 volt heaters. After several tests the authors determined that a slightly reduced 6.1 volts did add significant life to the tube without any audio problems.
Logged

klaus

  • Moderator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2210
Re: Underheating tubes in microphones
« Reply #3 on: May 14, 2015, 12:24:04 PM »

(...)After several tests the authors determined that a slightly reduced 6.1 volts did add significant life to the tube without any audio problems.
Confirmed, from 30+ years of observation.
Logged
Klaus Heyne
German Masterworks®
www.GermanMasterworks.com

Piedpiper

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 76
  • Real Full Name: Tim Britton
Re: Underheating tubes in microphones
« Reply #4 on: May 15, 2015, 05:03:54 PM »

Is that to say that 6.1 would be optimal but any less is not advisable? As noted earlier, I would imagine this to be circuit dependent beyond a certain point, yes?
Logged
row row row your boat...

Pied Piper Productions

klaus

  • Moderator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2210
Re: Underheating tubes in microphones
« Reply #5 on: May 15, 2015, 07:28:47 PM »

I find a heater voltage measured at the tube of slightly above 6 VDC, but below 6.3VDC ideal in the use as an impedance converter (not amplifier).

But all of this stuff is "seat of the pants": who has the time and smarts to figure out an objective, absolutely identical test setup where you heat not the same, but a similar tube, at many different levels and, enough of them to have a statistically relevant sample size? Not me. So I just go by empirical data.

With that in mind, I have found through the years a lot more busted tubes where the heating voltage was on the high side, and a few busted/noisy ones on the extreme low side, and the least amount of tube failures or noisy ones in that 6.1VDC range.
Logged
Klaus Heyne
German Masterworks®
www.GermanMasterworks.com

AusTex64

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 75
  • Real Full Name: Robert Mokry
Re: Underheating tubes in microphones
« Reply #6 on: May 17, 2015, 10:24:15 PM »

Thanks to all for the thoughtful comments. I Googled "vacuum tube under heating" and found this: http://www.vacuumtubes.net/How_Vacuum_Tubes_Work.htm

I thought this part especially relevant to this thread:

Tube life is sharply dependent on temperature, which means that it is dependent on filament or heater operating voltage. Operate the heater/filament too hot, and the tube will give a shortened life. Operate it too cool and life may be shortened (especially in thoriated filaments, which depend on replenishment of thorium by diffusion from within the filament wire).

A few researchers have observed that the lifetime of an oxide-cathode tube can be greatly increased by operating its heater at 20% below the rated voltage. This USUALLY has very little effect on the cathode's electron emission, and might be worth experimenting with if the user wishes to increase the lifetime of a small-signal tube. (Low heater voltage is NOT recommended for power tubes, as the tube may not give the rated power output.)

Operating the heater at a very low voltage has been observed to linearize some tube types-- we have not been able to verify this, so it may be another worthy experiment for an OEM or sophisticated experimenter.
Logged

klaus

  • Moderator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2210
Re: Underheating tubes in microphones
« Reply #7 on: May 17, 2015, 11:16:33 PM »

That advice contradicts my experience and measurements for all microphone tubes but the Telefunken VF14 and UF14 steel tubes.
Underheating the filament of a 6.3VDC EF86 by 20% would mean a heater voltage of barely over 5VDC, and for the AC701 3.2VDC.

There's no chance of any acceptable noise floor for tube mics with these tubes powered that way.

But microphone tubes, like the VF14, whose filament is underheated by 40%, work superbly.
Logged
Klaus Heyne
German Masterworks®
www.GermanMasterworks.com

AusTex64

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 75
  • Real Full Name: Robert Mokry
Re: Underheating tubes in microphones
« Reply #8 on: May 18, 2015, 12:58:15 AM »

Interesting, especially the comments about increased noise with under heating. I run the EF802 in my AMI U47 build at 5.05V per Oilver's instructions, and it has lower self noise that my '70's U87. See this schematic, where Oilver suggests this voltage: http://www.tab-funkenwerk.com/id68.html
Logged

Jim Williams

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 599
Re: Underheating tubes in microphones
« Reply #9 on: May 18, 2015, 11:11:47 AM »

Noise has a stew of contributors, low frequency hum and its harmonics, midrange grunge, shot/spot noise plus white noise from resistors, tubes, transistors and random electron movements.

Any comprehensive noise tests/comparisons are best done with a graphical analyzer like Audio Precision that can show a 20 hz~20k hz spectral view of the noise. Wideband S/N measurements will sum all those noise sources together but will not tell which is the main contributor.

If you can see the dominent noise source then you can formulate a plan to reduce it. White noise/hiss is attacked with lower source noise tubes or semiconductors plus resistors. Low frequency hum and harmonics are usually from power supply and or grounding causes.

A few hours with that test gear and a few test subjects would remove all doubts about heater voltages, noises and their relationship. So far I've not seen that done so guessin' is the next option.
Logged

klaus

  • Moderator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2210
Re: Underheating tubes in microphones
« Reply #10 on: May 18, 2015, 01:16:29 PM »

I have not found it much easier using test gear in determining microphone-generated noise than using ears. And I use relative dB comparisons mostly when communicating with clients (they seem to feel better when disaster is clothed in decibels.

For one thing, extracting external noise from the equation, i.e. separating unwanted sound that enters through the capsule, from internal, microphone contributed, noise is harder than just reading numbers off a display. It takes extreme isolation (professionally designed anechoic chambers) to eliminate low-level environmental noise bleed. Disconnecting the capsule, and substituting with a fixed capacitor, as is often recommended to those who don't own a six-figure anechoic chamber, also has its drawbacks: the mechanical-to-electrical gain structure of the two sections is not accurately reproduced with a fixed capacitor: different capsule types and brands have different gain before the impedance converter.

Between the two, the anechoic chamber test is preferable, but also has other flaws: comparing noise between mics and weighing its relative annoyance is hard. A 3dB rise in one frequency range might not be as noticeable or as unpleasant to the ear compared to a 1.5 dB rise in another.

I plead again, as I have done so often, for acoustic, subjective testing and the confidence for the listener that derives from that.

An example: I test a lot of clients' VF14, to determine whether they will be usable "in the field". There are specimens of this tube which have a relatively minor (in dB) rise in tube hiss, but the spectrum of that noise is centered in an annoying range, which would be disturbing to the ear in quiet recording passages. Other VF14 whose noise, again measured in dB, is centered in an octave below the previous example pass the subjective test.

The true test for microphone tube noise is whether the noise interferes with quiet music passages, i.e. low spls. The same subjective approach can be used with determining tube microphonics. I have not found an objective test one could refer to at what level and/or frequency microphonics are unacceptable.
Logged
Klaus Heyne
German Masterworks®
www.GermanMasterworks.com

Jim Williams

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 599
Re: Underheating tubes in microphones
« Reply #11 on: May 19, 2015, 11:20:53 AM »

Ears are a rather non-linear way to determine small noise differences. Plus they sum all the noises together so seperating low and high frequency noise and their relationship is difficult. If ears were completely reliable, test equipment wouldn't be necessary. That is akin to global signal to noise measurements in the test gear where all the noises are summed to give a broad band reading. Break that down into a spectral image and then you see everything.

When measuring the tube and system noise, bypassing the capsule is needed to remove the acoustic contributions to the measurements. Then the readings are illuminating as nothing gets by an analyzer with a 140 db measurment range. The human ear is limited to about 110 db of differentiation, far less than that test equipment. Anyone that has spent some time with that level of test gear will realize it offers a microscopic view of non-linearites. The human ear will have a very difficult time matching and repeating of the same data without errors or human subjective input.
Logged

klaus

  • Moderator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2210
Re: Underheating tubes in microphones
« Reply #12 on: May 19, 2015, 02:06:21 PM »

If ears are so unreliable as bullshit detectors of noise, offensive frequencies, etc. as you claim, why use them at all? Why not just sit in front of your TEF analyzer and watch Beethoven's Ninth that way?

The world of sound and other realms of sensory perception does not respond so well to its analysis by means of test equipment. It is therefore not analogous to the world of the hard sciences- physics, chemistry, etc.  Take medicine, where the detection and elimination of pathogens, measurable only with sophisticated equipment, can make the difference between life and death. The ultimate destination of all things audio is our hearing, in all its seeming limitation and, from a measuring standpoint, "unlinear" beauty. It is not compliance with an objective standard.

And, if the human ear is limited to a dynamic range of ca. 110dB, as you say, what's the point of measuring another 30dB beyond that? If you can show that that extra lowered noise floor which your machine indicated will serve the our hearing of music, I'd like to understand how.

Yes, ears are indeed very non-linear in sound perception. And, thankfully so. If linearity of audio transmission were the goal, B&K mics would rule every recording venue. But their (truly helpful) role is pretty much limited to the precise detection of sound levels (noise measurements of aircraft, to define local noise ordinances, for example).

Bypassing the capsule's contribution to noise is not logical for the reasons I had stated earlier: capsule gain is a significant part of the overall s/n of a mic, and its contribution to overall mic noise, then, is neglected or distorted by substituting it with a fixed capacitor bypass.


In sum, if the human ear is such a poor judge correctly identifying what sounds 'right' from what sounds 'wrong', or what is an objectionable noise level from an acceptable one, what then is the point of "objectifying" sounds via measuring equipment, if the bottleneck (hearing) cannot be eliminated?

I continue to insist in my work and my approach to microphone repair and refinement that the ultimate end control a microphone must pass must remain the human ear.

But I also agree that if, for example, a certain type of distortion is measured by using analyzing equipment it sometimes can speed up troubleshooting. However, if the tester is not capable, or does not trust his/her hearing, as final quality control, all is lost in the end.
Logged
Klaus Heyne
German Masterworks®
www.GermanMasterworks.com

Jim Williams

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 599
Re: Underheating tubes in microphones
« Reply #13 on: May 20, 2015, 11:27:00 AM »

Test equipment is not intended to replace hearing for subjective analysis. It is intended to find faults and errors, some below the human hearing abilities. In that realm, it is not only useful, but required for manufacturing a consistant product.

If test equipment like TEF, Audio Precision, Rhodes and Schwartz and Prism were not useful and required, companies like Sennheiser wouldn't use them.

Anyone attempting to refine a design or minimize noise contributions without that level of test equipment will be missing a lot of data. I'll leave that to the readers to determine whether that is the wisest decision. I realize these test rigs are not cheap, many I know would love to have them but the price of admission is very high.
Logged

radardoug

  • Full Member
  • ***
  • Offline Offline
  • Posts: 100
  • Real Full Name: Doug Jane
Re: Underheating tubes in microphones
« Reply #14 on: May 20, 2015, 04:52:42 PM »



When measuring the tube and system noise, bypassing the capsule is needed to remove the acoustic contributions to the measurements. Then the readings are illuminating as nothing gets by an analyzer with a 140 db measurment range. The human ear is limited to about 110 db of differentiation, far less than that test equipment. Anyone that has spent some time with that level of test gear will realize it offers a microscopic view of non-linearites. The human ear will have a very difficult time matching and repeating of the same data without errors or human subjective input.
Where do you get that 110 dB from?
Logged
Pages: [1] 2 3  All   Go Up
 



Site Hosted By Ashdown Technologies, Inc.

Page created in 0.074 seconds with 21 queries.