It is great that audio engineers have always used not only their ears to listen but also their mind and their heart.
Any world class microphone such as the Sennheiser MD21 can sound bass heavy to our mind due to many psychoacoustic effects of our brain perception.
The Fletcher Munson curves and the voice effort curves are two reasons and there are more. And of course directional microphones have the proximity effect too which is not the case with omnidirectionals such as the MD21.
While testing in the past various Pleiades filters (inductors in parallel with the microphone) an inductance of 85mH was found to produce very natural results on male voice for the MD21 HL. The voice is very clear, not bass heavy and psychoacousticaly it sounds deep presumably as the mind is now not disturbed by exaggerated lows.
Assuming a 200Ω microphone output impedance this inductance corresponds to aproximately a 380Hz cutoff. The filter created by the R and L is first order and gives a gentle slope of 6dB per octave in reducing bass.
While testing the Pleiades V6 preamplifier with a flat input trasformer (Altec 4722) and a Pleiades output trasformer that gave a similar cutoff when driven directly by the EF183 electron tube the same pleasing and natural effect was observed on voice.
The Pleiades V6 is possibly the simplest microphone amplifier using an input trasformer, an electron tube and an output trasformer directly connected to the anode.
A coupling capacitor connects the input trasformer to the control grid so as not to disturb the grid potential.
A high megohm resistor from anode to grid, 8.2MΩ, neutralizes the grid from its negative self assumed potential due to electron cloud emitted violently from the cathode.
This Rag resistor allows operation of the tube at a typical plate voltage of 5V.
Operation is very quiet as it is battery powered with same battery (6V in these experiments) for both heaters and plate circuit. The Vb is connected directly to the other side of the primary of the output trasformer.
The cathode is connected directly to ground. This maximises gain due to no negative feedback cathode resistor and further increases simplicity.
2 setups were used for sound comparison on male speech and singing voice.
1st setup:
MD21 HL at Low Z - Pleiades 82mH - Sony TCD5 PRO - Sennheiser HD580
2nd setup:
MD21 HN at Low Z - Pleiades V6 preamplifier - Sony TCD5 PRO - HD580
Anode current on the triode connected vari μ electron tube was from 100μA to 200 μA. This was varied by varying the rheostat in series with heaters to battery.
Although the heater voltage is specified at 6.3 volts only for example 3 volts is a typical value used and varrying around this.
It is known that underheating an electron tube cathode gives a very nice sound quality to a microphone amplifier and a legendary example is the underheated VF14 tube on the Neumann U47 microphone.
The sound appears much more focused and the low noise produced is a joy to observe.
The apparent reduction in noise may be explained by less cathode to grid induced current (see earlier euroelectron posts). The cathode electron cloud may resemble a boiling water steam cloud. This seems relevant at such small input signals to be amplified. On the Neumann U47 operating at higher plate voltage there must be other phenomena too, reduction of secondary emission? Etc. it was also observed in previous experiments that at some sweet value of cathode temperature perceived voice output maximizes.
This maximum occurring at slightly less than Ia=100μA was objectively verified as a 3dB increase on the VU meter of a Realistic disco mixer on previously described setups. The best way to observe gain increase while cathode temperature lowers is as follows. The heater rheostat bypassed. The cathode heats at full 6 Volts. Then the heaters are disconnected. (Actually whenever heaters are disconnected an jump is anode current is observed). With heaters disconnected the anode current is normally decreasing from about 400μA. When it reaches aproximately 100μA the gain increases by about 3 dB. This experiment was conducted on previous setups while a 440Hz square wave constant acoustic source was at constant distance to a Grampian ribbon microphone.
Continuing on today's experiment, while varying the heater voltage, i.e. the cathode temperature, it was found that the low frequency cutoff point was changing by listening on the headphones.
It was then confirmed by connecting an on attenuated portable Sony CD player with as test disc, connected to the Altec transformer. Response was observed on the Sony TC-D5 VU meters. Throughout all tests the TC-D5 was switched to mono operation so that the voice could be heard on both channels.
The change in 3dB cutoff point varied from 300Hz to less than 500Hz.
Values greater than 300Hz produced with less heater current gave a more focused sound on voice and a natural quality that is hard to describe with words.
The change in frequency responce is explained by the change in the output resistance of the electron tube as gm (transconductance) reduces with less anode current.
So the heater voltage can be used as a methods to adjust for compensation in perceived bass increase.
With a heater voltage that gave the same subjective bass naturalness on both setups their sound could now be compared more effectively.
While the 1st direct to Sony setup gives a world class sound quality there seem to be 2 advantages on the 2nd setup.
It sounds much more noise quiet and when the microphone is connecte one is surprised by low loud and sensitive the microphone sounds.
On the 1st setup the Sony recording level potentiometer has to be at 10 (maximum value).
On the 2nd setup,it has to be at just 2 for the same VU meter responce. So any noise generated by the Sony preamplifier is almost completely minimized.
The Pleides V6 with the above mentioned techniques is very quiet.
In fact the sensitivity of the MD21 is now comparable to the AKG C24 which has to be operated too at comparably very low potentiometer setting on Sony TCD5.
Although the bass responce on both setups is now similar the 2nd setup seems to have a more focused quality with the voice ready to jump at the listener.
This may be explained by the big signal, that drives the Sony input trasformer and input electronics.
Further tests need to be done.
And it must not be overlooked that setup 1 is much more simple and consumes much less power.
A very portable setup is using the Pleiades K117 JFET preamplifier inside the mini Tuchel connector of the MD21 HL. The output of the MD21 HL (Hi Z connection) internal step up transformer is connected directly to gate. The resistor and power supply comes from the device it is connected to, eg iPad. The iPad when using GarageBand introduces a similar around 400Hz cutoff. This makes a world class and simplest setup that is very suitable to the MD21 HL.
The world class MD21 HL was used at aproximately 1 inch and at 45 degrees from mouth.
No comments:
Post a Comment