Any thoughts about "True Phantom"?

[abbey road d enfer]

Hmmm.... sorry, but I am unable to spot a constant voltage (source) [seen at the microphone] as there are - referring to the schematic posted above - these two 6k8 resistors which transform every voltage into a current.... even when the supply voltage is variable it stays a (variable) current source - at least this is what I have learned in school....

It's a long time I haven't checked this patent, so from my memory, the principle is that the voltage at pins 2 and 3 of the XLR are measured, which in turn controls the voltage upstream of the 6.8k resistors. This is a NFB process; since it tends to maintain the voltage constant irrelevant of current draw, it's a low resistance source now.
E.G. let's assume the mic draws 2mA -1mA per leg- the voltage drop across a 6.8k resistor is precisely 6.8V, so the system will set the voltage before the resistors at 54.8V.

Now, regarding these assertions
"..... the A-52 provides the following outstanding features:
• Safely and effectively accommodates dc-supply inputs ranging from +7 .5 volts minimum to +60 volts maximum.
• Automatically adjusts itself to power a 12-volt microphone only with any dc input from +7 .5 to just below +50 volts .
• Automatically adjusts itself to power a 12-volt or 48-volt microphone with any dc input from +50 to +60 volts.
• Acts as a regulator with any dc input above +13 volts; the dc source itself need not be inherently well regulated."
I think they must be taken with a pinch of whatever, since it's a current source. As I said, injecting 3.2mA in a mic that draws 1mA in normal operating conditions is not an ideal situation. Something has to give up; in that case the voltage on the XLR pins is higher than it would be with 6.8k resistors. Is it a problem? It depends on the rest of the circuit.
Just the same, if the microphone is designed to draw more than 3.2mA, the resulting voltage at pins 2 and 3 will be lower than expected, which may or may not have consequences.
P48 standard specifically requests 6.8k resistors: a constant-current system does not confirm to standard, neither a constant-voltage.
I'm not advocating the "True Phantom" system; I'm just saying it's a brilliant idea that solves a problem for which nobody asked a solution.

 

[Gus]

The way I read the patent is  lets use fig 4B in the patent
The voltages are sensed at the "top" of the 6.8K resistors not at pins 2 and 3
The sensed voltages are used to control the active  current sources

http://www.sumobrain.com/patents/wipo/Phantom-power-supply-microphone/WO2016171551A1.pdf

 

[abbey road d enfer]

The way I read the patent is  lets use fig 4B in the patent
The voltages are sensed at the "top" of the 6.8K resistors not at pins 2 and 3
The sensed voltages are used to control the active  current sources

http://www.sumobrain.com/patents/wipo/Phantom-power-supply-microphone/WO2016171551A1.pdf

There is an issue there; fig 4a & 4b contradict fig 3a & 3b. the integrator reverses the regulation process.
Anyway, tag 20 of the patent is clear; the voltage at pins 2 & 3 is regulated.
Although the voltages are measured elsewhere, it's still the voltage at pins 2 and 3 that is regulated. The voltage drop in feeding resistors  is compensated by the voltage drop in the current source resistors The fact that the voltages are controlled by current sources is irrelevant. In the end it's the voltage that is regulated.

 

[analogguru]

..... the voltage at pins 2 & 3 is regulated.
...... In the end it's the voltage that is regulated.

No..... thats wrong.  I would recommend to read the AES-paper 145 which is posted above in this thread.  It is written in a more understandable non-patent language:

[quote author=Blue Phantom ®]
.....
3 New Design Using Current Source

..... In Figure 1, the 48 volt power supply is now replaced by two 48-volt regulated current sources. Figure 2 shows a schematic of this design which be referred to as True Phantom ® throughout the rest of this paper. To ensure that the current sources will function properly, the voltage needs to be 14 volts above the 48 volts needed for phantom powering. This is why a total voltage of 63 volts is used in this design. Finally, a DC servo is used to regulate the output of the current source to be 48 volts.
[/quote]
And these (constant) 48V are connected to the 6k8 resistors which for itself are a current source.....

[quote author=Blue Phantom ®]
.....
6 Conclusions
1. The addition of a current source [rem: e.g. one like the AKG A52] to a phantom power supply, gives a substantial reduction of the distortion of the microphone at all signal levels.
2. The gain is 0.06 dB higher. [rem: wow, how incredible and audible]

......
[/quote]
Oh and not to forget:
[quote author=Blue Phantom ®]
.....
In the near future the True Phantom ® will be available as a chip.
The German chip manufacturer PREMA Semiconductor will make the chip.
.....
[/quote]
... a chip that noone really needs.

 

[abbey road d enfer]

No..... thats wrong.  I would recommend to read the AES-paper 145 which is posted above in this thread.  It is written in a more understandable non-patent language:

You're correct. My memory was set on another paper where the actual voltage delivered to the mic was controlled.
So it turns out that this patent concentrates on reducing the loading of the 13.6k. That would allow mic preamps to offer a significantly higher differential impedance. It is said to be beneficial to some mics, particularly ribbons.
My experience tells me that there is nothing to gain by increasing the input Z above about 5k, but I haven't tried every mic in the world.
In that respect the AKG N52 would not be particularly brilliant, with 1.2k of additional load.

 

[analogguru]

....
In that respect the AKG NA52 would not be particularly brilliant, with 1.2k of additional load.

Sorry, but it is 1.36k....

And this additional load only occurs when you run the A52 from 12V (without a center tapped mic-transformer) [which True Phantom ® is unable to do].  If you run it from 48V or above you can always increase the feeding resistor from 680 Ohm to 6k8 and the additional load will be: 13.6kOhm.

BTW, I have seen phantom power supplies which have a switch 12V/48V but don't change the feeding resistors..... 12V phantom power with 2x 6k8 resistors.... how ingenious....

 

[abbey road d enfer]

Sorry, but it is 1.36k....

Dunno, the picture is hardly legible. I read one 620r resistor and another 600, which I assumed was actually another 620. You tell me they are 680, which it should indeed be for P12.

And this load only occurs when you run the A52 from 12V (which True Phantom ® is unable to do).  If you run it from 48V or above you can always increase the feeding resistor from 680 Ohm to 6k8 and the additional load will be: 13.6kOhm.

Information I have about NA52 is too parcellar, however the claims do not mention the fact that the circuit must be tuned to the specifics of mics for optimum performance.
Believe me I have the utmost respect for AKG's engineers - I have collaborated with Bernhard Weingartner for a time - but even them could not get around the basivs of Ohm's law. 

BTW, I have seen phantom power supplies which have a switch 12V/48V but don't change the feeding resistors..... 12V phantom power with 2x 6k8 resistors.... how ingenious....

In many cases, particularly with modern mics, the user won't notice the possible difference. OTOH, a U87 won't work with 12V phantom, whatever the feed resistors...

 

[analogguru]

....You tell me they are 680, which it should indeed be for P12.

It's not me, it's the schematic .....

 

[analogguru]

.... and the reality.....

 

[analogguru]

..... Information I have about NA52 is too parcellar, .....

There is a remedy for that, this....

 

[analogguru]

...and this....

 

[analogguru]

...and this:

 

[Gus]

Joost Kist


Would you be upset if I post a screen shot of a sim I made of fig 4B from the patent?

 

[abbey road d enfer]

Would you be upset if I post a screen shot of a sim I made of fig 4B from the patent?

That would be interesting. I tried to sim the circuit of fig 3d and I get silly results.

 

[analogguru]

Yes, that would be really interesting.  Maybe the secret can be lifted which positive effect it has, when the voltage of a 48V-zener-diode (must be a selected one) is integrated with a 2M2 resistor and a 1µF capacitor....

 

[abbey road d enfer]

Yes, that would be really interesting.  Maybe the secret can be lifted which positive effect it has, when the voltage of a 48V-zener-diode (must be a selected one) is integrated with a 2M2 resistor and a 1µF capacitor....

I have a feeling this is to provide identical terminations to both the inverting and non-inverting inputs of the opamp.

 

[analogguru]

I have a feeling this is to provide identical terminations to both the inverting and non-inverting inputs of the opamp.

Yes, I thought of this too, but with a corner frequency of ~ 0,07 Hz (or integrator time constant of ~ 14 sec) ?  I am really curious.....

 

[MediaWorksOKC]

https://www.prema.com/index.php/en/assps-en/audio-ic/item/download/264_9b8e12bb14b80959d13a0e610d92eeb7

 

[kingkorg]

https://www.prema.com/index.php/en/assps-en/audio-ic/item/download/264_9b8e12bb14b80959d13a0e610d92eeb7

Now, having in mind 30$ Schoeps based mics like mxl603 have stock 0.1% thd at some serious SPL, and can be biased further to 0.02% thd. SO WHAT ARE THEY TALKING ABOUT HERE!?!?

Capsule's physical thd is larger than that.

 

[analogguru]

.... SO WHAT ARE THEY TALKING ABOUT HERE!?!?

Maybe only sales-BS for a product that noone really needs ?
I bet that the chip will cost more than € 10,--/pcs (in quantities).

But it could be funny to buy one chip only to prove that the claimed 6db better noise performance in reality doesn't exist - compared to a decent (not patented) phantom power supply....