DC Voltage Doubler From Phantom Power

[sr1200]

I'm trying to figure out the best way of going about getting somewhere near 100v to polarize an old mic with the least amount of components.
Can I build something to double the phantom source, or should i build a dedicated power supply? (was thinking a DC-DC boost, but cant find one that will take 48v, but could knock down the 48 to something lower to work with the constraints of something like a UC3843/UC3845.
Other details: Its an old Grundig Mic that needed around 100v to polarize but didn't need a "pre" pre.

 

[ccaudle]

Don't forget that the 48v is sourced through 6.8k resistors, so it is not actually 48V available to use once you start drawing any current.
If you want something simple Analog Devices has some high voltage charge pumps which may work. Might depend on how "near" to 100V you actually need, but doubling voltage is pretty straight forward with a charge pump.

 

[ruffrecords]

Many condenser mics have a simple very high frequency dc/dc convertor based on a few CMOS gates.

Cheers

Ian

 

[sr1200]

Don't forget that the 48v is sourced through 6.8k resistors, so it is not actually 48V available to use once you start drawing any current.
If you want something simple Analog Devices has some high voltage charge pumps which may work. Might depend on how "near" to 100V you actually need, but doubling voltage is pretty straight forward with a charge pump.

I have no clue how much current this would draw. The high end of things pulls 10mA (earthworks). Since this thing is probably in the neighborhood of 70 years old, there isn't much documentation on it. The capsule supposdly has a 50-15K response. Im just really curious as to what it sounds like more than anything. Not expecting anything great at all.

 

[sr1200]

Many condenser mics have a simple very high frequency dc/dc convertor based on a few CMOS gates.

Cheers

Ian

Anything you know of that gets up to about 100V?

 

[ruffrecords]

Anything you know of that gets up to about 100V?

The Rode NT1A for example creates 73V from an internally derived 15V supply.

Cheers

Ian

 

[ccaudle]

Anything you know of that gets up to about 100V?

The design is scalable, it is a CMOS gate configured as an RC square wave oscillator, and the AC waveform drives a diode/capacitor ladder to form double, triple, quadruple, etc. voltage.
This is a copy of the article which inspired most of the designs:
https://gyraf.dk/schematics/Voltage_multipliers_with_CMOS_gates.pdf
You can make that style of design with one or two gates if you want to use single gate packages and really save space, but when the original article came out single gate packages were not available, so using hex inverters was an easy way to also get additional current capability (which isn't needed if you are just biasing a microphone condenser capsule).

Its an old Grundig Mic that needed around 100v to polarize but didn't need a "pre" pre.

By using the term polarize I assume you mean it is a condenser mic, but what do you mean by it doesn't need a pre pre? Any condenser mic needs a buffer stage because of the high impedance of a condenser capsule. Or are you referring to something else?

 

[sr1200]

The design is scalable, it is a CMOS gate configured as an RC square wave oscillator, and the AC waveform drives a diode/capacitor ladder to form double, triple, quadruple, etc. voltage.
This is a copy of the article which inspired most of the designs:
https://gyraf.dk/schematics/Voltage_multipliers_with_CMOS_gates.pdf
You can make that style of design with one or two gates if you want to use single gate packages and really save space, but when the original article came out single gate packages were not available, so using hex inverters was an easy way to also get additional current capability (which isn't needed if you are just biasing a microphone condenser capsule).



By using the term polarize I assume you mean it is a condenser mic, but what do you mean by it doesn't need a pre pre? Any condenser mic needs a buffer stage because of the high impedance of a condenser capsule. Or are you referring to something else?

Im going off of the limited amount of info i have found out there in crazy internet land about it.

The mic is rated at 2.5mV output (without needing the usual pre-amp boost) - way higher than the comparable moving coil mics Grundig produced at the time. This was apparently achieved by the HUGE 70mm diaphragm.

Pinout for a GCM mic with a 3 pin DIN plug:
the signal is on pin 1 and supply voltage on pin 3.

Pin 2 is normally ground or screen, i don't know exactly what this should be connected to. But, as long as that whopping supply voltage is not touching it, experimentation would be par for the course. Caution advised, obviously!

It is not an electret condenser, and requires a supply voltage. Originally 100v. Apparently 48v should work.
Purists can apparently incorporate a "voltage doubler" on a 48v phantom supply (Caution again!), OR, it's said, there are designs online for a 100v supply, OR, as stated above, most of the oldest (1950's) Grundig valve tape recorders are compatible and put out that 100v supply, on pin 3.
The compatible Grundig models are:
700L, 700C, 500L MKI/II,
TK5, TK7, TK8, TK12, TK15, TK16, TK20, TK24, TK25, TK30, TK35, TK55, TK60 (Stereo), TK820, TK830-3D, TK919
TM20, TM60
TR3.

As it puts out 2.5mV it does not require the "pre" pre-amplifier that other condensers usually have built-in.
To use it with the usual low impedance inputs on pro and semi pro gear, it needs a high to low impedance conversion. An audio transformer of appropriate spec (try about 100:1?) is probably best for this job. Converting the unbalanced signal to balanced may also be advisable, either electronically, or with another transformer. (not sure how common transformers which do both jobs at once, are - but may be best solution).

Those Grundig models can be found dirt cheap, and if the amp parts are working, a (valve-amplified) monitor signal output and/or a low-wattage low-impedance speaker output (easily adaptable to line level) can be taken from the machine.

Disclaimer: the above post is meant for entertainment only!
Working with voltages can be dangerous, make sure you know what you're doing first, any actions are at your own risk!

Good luck!

My thought was to get a transformer and build this voltage doubler/(tripler if i pare down the 48v to something lower) into a small hammond enclosure with the DIN3 on one side and XLR on the other.

 

[jp8]

I'm trying to figure out the best way of going about getting somewhere near 100v to polarize an old mic with the least amount of components.
Can I build something to double the phantom source, or should i build a dedicated power supply?

A CMOS charge pump design using two single-gate inverter ICs can be found on my website as part of the KM84++ design here. Expand with a few more C + D stages to obtain 100V. Circuit draws ~2.5mA at 18V. I've designed an 80V version on a 10 x 20mm PCB, so it can be made very small.

Jan

 

[ccaudle]

That sounds like just a condenser capsule that doesn't have the buffer amp in the housing.
You want one of Rog's op-amp impedance converters with a voltage multiplier circuit.
Check here for his example of how to build that on stripboard:
https://www.opic.jp137.com/index-ldc.html

 

[sr1200]

That sounds like just a condenser capsule that doesn't have the buffer amp in the housing.
You want one of Rog's op-amp impedance converters with a voltage multiplier circuit.
Check here for his example of how to build that on stripboard:
https://www.opic.jp137.com/index-ldc.html

Thanks this looks like a perfect solution. I think ill try the impedance converer and a transformer and see which one works better. The voltage multiplier is great though!

 

[ccaudle]

ill try the impedance converer and a transformer

An impedance converter will provide a high impedance input, a low impedance output, and a gain of 1. So the output will be the same voltage as the capsule output at a much lower impedance, while providing a 1G Ohm impedance to the capsule.

The transformer will transform the impedance of the pre-amp to the capsule side based on the square of the turns ratio. So for the given 100:1 transformer example the impedance to the capsule will be 10x the impedance of the pre-amp, so probably on the order of 30k Ohms, the output impedance will be 1/10 the impedance of the capsule, so probably on the order of 40k Ohms (that is a big guess on my part), and the output voltage will be 1/100 of the capsule voltage, so reduced from 2.5mV to 25uV.
A 40k Ohm output impedance will additionally be attenuated to less than 1/10 when connected to a typical modern preamp with 3k input impedance, so now you are down to 2.5uV.
That would be a pretty poor spec for a condenser mic (not to mention that a 100:1 transformer is an order of magnitude more expensive than a good TI op-amp).

 

[sr1200]

Youre probably right. I just realized that the opamp is balancing as well. Probably just easier, less noisy too. Thanks again!