48v to 1.5v?

[sr1200]

I have an old stereo "dat" mic that was powered by a AA battery (1.5v). The output is not balanced and phantom will basically blow this thing up. Im wanted to try some experiments with the mic and have it use phantom power instead of the battery. Removing the battery compartment would also leave some room for a couple of small mic transformers to a 5 pin out, or i could do a transformerless output and design a really slim and long board to handle.
National makes an adjustable VR that can go down to 1.3v, just not sure how hot that will get esp inside a not vented case.

There shouldnt be a lot of current getting drawn, but i kind of suck at the math portion of electronics. (correct me if this is wrong) assuming we pulled 10ma from the phantom supply, we would have P=(48-1.5)x.01 = .465W? If thats correct, (which im going to assume its not since i dont think ive ever calculated anything correctly) what would that look like from a dissipation stand point? Would the VR be able to just handle that or would some kind of heat sink be needed (doubtful that would fit in the mic body... the VR MIGHT fit?

If anyone has any thoughts or ideas of what i can do with this, besides send it to a land fill, would be great! Thanks!

 

[ccaudle]

That mic has an electret capsule with built in JFET. It doesn't really have to run from 1.5V, the mic probably did that by default so it could run from a single alkaline cell. The typical configuration has a resistor from the power source to the drain of the JFET, and you scale the resistor based on the voltage of your power source.
Depending on how complex of a circuit you want there are several variations of circuit you can use with one of those capsules for running from phantom power. You can go all out and build a Schoeps style circuit with a few resistors, caps, and two transistors.
You can make an "Alice" style circuit with an op-amp (plus a few resistors and capacitors).
There is also a circuit with a name something like simplest phantom powered electret circuit that just uses resistors and caps, no other active devices. That circuit cannot drive very long cables, it isn't really recommended as a robust solution but might be fun to experiment with.

You should be able to find info on all those on this forum with some searches of the microphone forum, it just depends on how much work you want to put into playing with the mic. There are circuit boards available for most of those circuits, but they are usually sized for something like a large diaphragm microphone body, not many (if any) off the shelf that are sized to fit in the amount of space you probably have available.

 

[NOON]

Remember that the 48v phantom is fed via 6k8 resistors. If you short it to ground you will have 0v and about 7mA. The areas in between are simple voltage dividers with the 6k8, meaning as you increase the current draw the voltage will drop.
There are quite a few schematics around for converting battery electrets to phantom power, most have transformers. E.g. the Nakamichi CM300

 

[sr1200]

@NOON thanks the Nakamichi thing was an interesting read. Some cool stuff out there and very similar to my situation (except... 9v and 1.5v)
@ccaudle Thanks, Realized i didnt put the model of the mic im working on. Its an Audio Technica AT822 stereo dat mic.

 

[k brown]

The AT822 has very good capsules (same as the highly-regarded AT825 stereo mic); basically a slightly different version of those in the AT4041. Not landfill-worthy!

Each capsule has a small circular circuit board containing a 2SK660 FET.
https://groupdiy.com/threads/whats-a-good-raw-electret-capsule.82406/#post-1065030Just don't do what I mentioned in that thread, and add a center hole to the capsules - they sound better left stock.

 

[ccaudle]

Audio Technica AT822

That is definitely a step up from the Panasonic capsules I was expecting, but the same things should apply.

This post is an example of a Schoeps style circuit:
Small electret phantom Schoeps circuit

In that same discussion there are some posts talking about the Alice style op-amp buffer.
A good example of that is here from GDIY member rogs:
op-amp impedance converter info (Alice circuit)

The super simple circuit I mentioned is here:
Richard Lee "simplest electret circuit" sourced from instructables (PDF)

 

[k brown]

Roughly Schoeps-style circuits designed for the 2SK660 are found in the Samson C02 and Monoprice SC100, though you wouldn't be able to fit two of them into the 822 body.
https://groupdiy.com/threads/samson-c02-sdc-mods.35310/
If you're going to make something from scratch, in both the Samson and Monoprice the drain and source resistors are 5k, and there's 7-7.5v to the drain resistor; so you can adapt the Alice circuit to suit.

The Alice op amp circuit wouldn't apply, as it's only for capsules w/o FETs.

Here's the schem (w/o parts values) of the AT825 (same as the AT822, but with phantom/balanced outs). I haven't measured all the values, but I can tell you that when phantom powered it puts 8.5v on the FETs' drains, and I measured the source resistors (R1 / R2) as 5.6k.

Something puzzling about the schem: I got two 822s and an 825, and none of them had the caps on the FETs' pins that are shown on the schem.

I will warn you though, the 822/825 capsules are rather noisy by today's standards, and vary a bit in that regard.

 

[ccaudle]

The Alice op amp circuit wouldn't apply, as it's only for capsules w/o FETs.

The capsules were described as having a FET on a small circuit board, so would be relatively easy to modify to remove the FET.
You can also use the capsules with FET with just a small modification of the Alice style circuit. See this TI app note for one example:
phantom powered op-amp buffer TI app-note

That app note shows symmetrical output, but would be even easier to build as an impedance balanced asymmetrical output, like Rogs showed in one of his Alice style circuit variants.

 

[Admiral-dk]

I don't know this exact electret - but the majority can't withstand more than 10Vdc. - something to bear in mind ...!
I see that the AT825 circuit has a Zener Diode to protect against high voltage - but no values are given there ....

Per

 

[k brown]

I don't know this exact electret - but the majority can't withstand more than 10Vdc. - something to bear in mind ...!
I see that the AT825 circuit has a Zener Diode to protect against high voltage - but no values are given there ....

Per

As stated above, the FET on those capsules is 2SK660 (the capsules aren't the cheapy type with FET built into a stamped aluminum housing; the FET is on a board mounted onto the back of the capsule), and the voltage delivered to it's drain by that circuit is 8.5v when phantom power is used.

 

[k brown]

The capsules were described as having a FET on a small circuit board, so would be relatively easy to modify to remove the FET.
You can also use the capsules with FET with just a small modification of the Alice style circuit. See this TI app note for one example:
phantom powered op-amp buffer TI app-note

That app note shows symmetrical output, but would be even easier to build as an impedance balanced asymmetrical output, like Rogs showed in one of his Alice style circuit variants.

Possible to remove the FET, but not fun - the FET is mounted on the capsule side of the board, so to remove it you'd have to hacksaw off the back of the capsule where the board is attached.

 

[user 37518]

I have an old stereo "dat" mic that was powered by a AA battery (1.5v). The output is not balanced and phantom will basically blow this thing up. Im wanted to try some experiments with the mic and have it use phantom power instead of the battery. Removing the battery compartment would also leave some room for a couple of small mic transformers to a 5 pin out, or i could do a transformerless output and design a really slim and long board to handle.
National makes an adjustable VR that can go down to 1.3v, just not sure how hot that will get esp inside a not vented case.

There shouldnt be a lot of current getting drawn, but i kind of suck at the math portion of electronics. (correct me if this is wrong) assuming we pulled 10ma from the phantom supply, we would have P=(48-1.5)x.01 = .465W? If thats correct, (which im going to assume its not since i dont think ive ever calculated anything correctly) what would that look like from a dissipation stand point? Would the VR be able to just handle that or would some kind of heat sink be needed (doubtful that would fit in the mic body... the VR MIGHT fit?

If anyone has any thoughts or ideas of what i can do with this, besides send it to a land fill, would be great! Thanks!

Is the voltage regulator a switching regulator or a linear regulator? if the former, the dissipation will be much lower, if the latter, yes, the dissipation will be nearly 1/2 W for the current you mention. Perhaps you could consider a two-stage switching plus linear regulator, drop the majority of the voltage with the switching regulator, then use a linear regulator as a final stage.

 

[MicMaven]

Its an Audio Technica AT822 stereo dat mic.

Um ... ahem ... er ... whilst I am NOT an expert in these matters, AudioTechnica IS and the user manual for that microphone clearly states:

The AT822 is designed for battery operation only; install the battery​

before attempting use. WARNING: Do not attempt to use when​

phantom power is present. Possible damage to the microphone​

may result. (Please note, however, that the presence of a bias​

voltage – from a DAT recorder, for example – is acceptable and will​

not harm the microphone.)​

Just saying for what it might be worth. James -K8JHR

 

[ccaudle]

Um ... ahem ... er ... whilst I am NOT an expert in these matters, AudioTechnica IS and the user manual for that microphone clearly states:

The AT822 is designed for battery operation only; install the battery​

before attempting use. WARNING: Do not attempt to use when​

phantom power is present. Possible damage to the microphone​

may result. (Please note, however, that the presence of a bias​

voltage – from a DAT recorder, for example – is acceptable and will​

not harm the microphone.)​

Just saying for what it might be worth. James -K8JHR

He's not talking about using phantom power with the mic as-is, the discussion is about ripping out the guts and putting in new circuitry which is designed to use phantom power.

 

[ccaudle]

the dissipation will be nearly 1/2 W for the current you mention.

The first post was not accounting for the voltage drop across the 6k8 resistors on the pre-amp side. 10mA will drop 34V across the feed resistors, so no more than 12.5V across the regulator, so 1/8 W. Assuming that the 10mA figure is correct, which seems high for a single JFET circuit.
In any event there are other ways to get there, Schoeps style circuit probably being the most reasonable choice.

 

[sr1200]

Is the voltage regulator a switching regulator or a linear regulator? if the former, the dissipation will be much lower, if the latter, yes, the dissipation will be nearly 1/2 W for the current you mention. Perhaps you could consider a two-stage switching plus linear regulator, drop the majority of the voltage with the switching regulator, then use a linear regulator as a final stage.

This is an interesting idea. How complex would that be? (thinking circuit size since this would need to fit into a very thin mic body)

 

[sr1200]

The first post was not accounting for the voltage drop across the 6k8 resistors on the pre-amp side. 10mA will drop 34V across the feed resistors, so no more than 12.5V across the regulator, so 1/8 W. Assuming that the 10mA figure is correct, which seems high for a single JFET circuit.
In any event there are other ways to get there, Schoeps style circuit probably being the most reasonable choice.

Yes, I figured on the HIGH side of things as 10mA is the suggested top end of draw for a phantom powered device. Better to over estimate than under i guess.

 

[sr1200]

Roughly Schoeps-style circuits designed for the 2SK660 are found in the Samson C02 and Monoprice SC100, though you wouldn't be able to fit two of them into the 822 body.
https://groupdiy.com/threads/samson-c02-sdc-mods.35310/
If you're going to make something from scratch, in both the Samson and Monoprice the drain and source resistors are 5k, and there's 7-7.5v to the drain resistor; so you can adapt the Alice circuit to suit.

The Alice op amp circuit wouldn't apply, as it's only for capsules w/o FETs.

Here's the schem (w/o parts values) of the AT825 (same as the AT822, but with phantom/balanced outs). I haven't measured all the values, but I can tell you that when phantom powered it puts 8.5v on the FETs' drains, and I measured the source resistors (R1 / R2) as 5.6k.

Something puzzling about the schem: I got two 822s and an 825, and none of them had the caps on the FETs' pins that are shown on the schem.

I will warn you though, the 822/825 capsules are rather noisy by today's standards, and vary a bit in that regard.

Thanks for the schematic on this.

 

[user 37518]

The first post was not accounting for the voltage drop across the 6k8 resistors on the pre-amp side. 10mA will drop 34V across the feed resistors, so no more than 12.5V across the regulator, so 1/8 W. Assuming that the 10mA figure is correct, which seems high for a single JFET circuit.
In any event there are other ways to get there, Schoeps style circuit probably being the most reasonable choice.

It doesn't matter he is dropping the voltage with the resistors, that just means the resistors will be the ones dissipating most of the power. You are exchanging dissipation in a VR with dissipation in resistors.

Edit: I am not sure what Schoeps circuit you refer to. But if it is the one I am thinking of, It might work. I am not sure about the application the OP is referring to.

 

[user 37518]

This is an interesting idea. How complex would that be? (thinking circuit size since this would need to fit into a very thin mic body)

The largest component will probably be the inductor. There are pre-made solution ICs by AD or TI, if you do a Google search, you will find many ICs.