Current for phantom power?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
The constant current source has a defined current while the voltage across it is not defined and depends on the load. A real current source has a limit of that "floating" voltage and in this case it is 48V. This also means that the output voltage at the CC source will be 48V for all loads less than 350mA in this case. This is not a normal work area for CC source but an end point, for which this SMPS is not actually designed. And very often that voltage is not accurate and can fluctuate.

Ok. So I think I get it now. The CC supplies are not designed to work at 48V. They're designed to work at a voltage below that so that it is operating in the normal work area for CC. You select a number of series and parallel LEDs to make some voltage, like say 40V.

With CC, it is designed to draw the full rated current at which point the voltage starts to drop. With CV, it is designed to draw less than the full rated current.

So the reason SoundSkulptor used CC was ultimately because the package was half the size. The equivalent CV supply is APV-16 which only goes up to 24V and makes up to 670mA.

The noise is actually probably higher with CC because they don't put as much effort into regulating the voltage (none below the rated voltage). But for audio it doesn't matter because that noise is at 60kHz, already low and easily attenuated with even a simple filter.

So this means that APC-16-350 is actually a good choice for phantom because you're not even going to find a supply that small that makes 48V.
 
Last edited:
These work perfectly.....
ECE05US48
Note that this is not isolated. So you cannot "stack" these. But for 48V, it might be ok. They're so small I would be weary and study the noise performance intently before committing. But they are so small it is very attractive.
 
Where do you see that they are not isolated?
I have built a few stereo mic amps now with these. Just add a common mode choke and
capacitance multiplyer and the jobs a good'n!
 
CC because they don't put as much effort into regulating the voltage

The voltage is not regulated at all, the current is regulated. A constant current supply measures the output current and feeds that back to the error amp to regulate the current the same as a constant voltage supply measures the output voltage and feeds that back to the error amp to regulate the voltage.
When running a constant current supply with an impedance high enough that the voltage is pegged to the high limit, the feedback network is in the non-linear region where it is not working at all. It is like running an amp at a DC level that continuously clips at the power rail.

The reason that constant current supplies are so widely available on the market now is because LEDs have a brightness rating at a particular current, and as I'm sure you are familiar, the LED voltage vs. current curve is very nonlinear, and varies with temperature. If you regulated to a constant voltage, the current could vary pretty wildly across device variation and especially temperature (and high power LEDs change temperature a lot as they go from off to full brightness).

You select a number of series and parallel LEDs to make some voltage, like say 40V

I doubt you would use parallel strings, because you have no way to guarantee that the current divides evenly between the strings. More likely you just use a string of LEDs that are rated for the output current of the power supply, and based on how many LEDs you want (how big you want the light fixture) you pick a power supply that has output compliance in the range that works for that number of LEDs in series.

It seems to me that if a particular constant current supply works out OK when running at the limit voltage, it is a matter of luck, since the operation is not specified at all in that region. A different model may behave differently, or different revisions of the same model may behave differently since the manufacturer makes no claims at all about behavior at the limit. I would not want to press my luck that way when you can just by supplies that are designed to regulate the output voltage.
 
The voltage is not regulated at all, the current is regulated.
That's not what I have observed. The output voltage is in fact a steady 48V. So below the rated current, voltage is regulated. It's only when you reach the rated current that current becomes regulated.

I doubt you would use parallel strings, because you have no way to guarantee that the current divides evenly between the strings.
They must. There are no individual LEDs that use 350mA. Power LEDs now are multiple individual LEDs on one "chip" that are wired in series and parallel. Older models did not divide current perfectly and you could sometimes see variations in brightness of individual LEDs. Newer "chip" LEDs are actually pretty well matched even at low currents.

It seems to me that if a particular constant current supply works out OK when running at the limit voltage, it is a matter of luck, since the operation is not specified at all in that region. A different model may behave differently, or different revisions of the same model may behave differently since the manufacturer makes no claims at all about behavior at the limit. I would not want to press my luck that way when you can just by supplies that are designed to regulate the output voltage.
I have used 5 different CC MeanWell SMPS and 4 different models. I can't recall testing each one like I did the APC-16-350, but my recollection is that they all worked the same way.

So CC wins simply because of physical size. I would much rather have a unit that puts out the maxium voltage below the rated current as opposed to a unit that puts out half that voltage over a range of current.

Note that the high voltage CC supplies might also be useful for tube amps because you could use a shunt regulator to adjust the HT voltage.
 
Of course it is isolated, it is AC input.
Right at the top of the datasheet: "Isolation • 4000 VAC Input to Output"
True. Good point.

But you can tell from the drawing that there's only one magnetic component. That has to be the transformer so it's a very basic device. But it's so small, that it's definitely something worth testing.
 
..I have built a few stereo mic amps now with these. Just add a common mode choke and
capacitance multiplyer and the jobs a good'n!
Hi Kevin,

I need to build a 48 volt supply for a three channel mic pre. Would this device provide enough current for that? Would you have an example of the circuit using this with a capacitance multiplyer and common mode choke, or know where to point me to one?

- Richard
 
That's not what I have observed. The output voltage is in fact a steady 48V. So below the rated current, voltage is regulated. It's only when you reach the rated current that current becomes regulated.

Just seeing this as a new post put on this thread. But thought worth saying that the behaviour described above wrt Constant Current supplies isn't voltage regulation in the sense that it widely understood. Rather the supply is operating at the limit of its voltage compliance due to the load impedance.
 
Of course it is isolated, it is AC input.
Right at the top of the datasheet: "Isolation • 4000 VAC Input to Output"

Be careful with those isolation specs in general though. The figure in kVs is usually rated for, say one minute. Usually related to transient / burst conditions such as tested for regulatory compliance eg EMC specs'. Similar applies to ratings on optoisolators.
It does, of course, mean that there is AC/DC isolation and at 'normal' voltages they should be fine.
 
Hi Kevin,

I need to build a 48 volt supply for a three channel mic pre. Would this device provide enough current for that? Would you have an example of the circuit using this with a capacitance multiplyer and common mode choke, or know where to point me to one?

- Richard
Hi Richard, yes it will work fine for a 3 channel mic pre as the absolute maximum current through available through two 6.8 kOhm phantom resistors is only 14mA, so if all three pres had shorted inputs the max will be 42mA.
The circuit I used for filtering is at work, I’ll try to remember to look it up Monday.
 
Hi Richard, yes it will work fine for a 3 channel mic pre as the absolute maximum current through available through two 6.8 kOhm phantom resistors is only 14mA, so if all three pres had shorted inputs the max will be 42mA.
The circuit I used for filtering is at work, I’ll try to remember to look it up Monday.
Hi Kevin,

That's very kind of you - much obliged,

Richard
 
Hi Kevin,

That's very kind of you - much obliged,

Richard
Here is the circuit I used. The choke was from RS Components order number 863-3807. It may not be the best one, but it does reduce the switchiong noiuse considerably.
The transistor I used, BSS51, is a Darlington type, so you will drop a couple of volts from nominal 48v. I used it because I had some in the draw!
There are probably better implimentations of a CM filter but it worked for me! :)
 

Attachments

  • Filter.jpg
    Filter.jpg
    460.7 KB
Hi Richard, yes it will work fine for a 3 channel mic pre as the absolute maximum current through available through two 6.8 kOhm phantom resistors is only 14mA, so if all three pres had shorted inputs the max will be 42mA.
The circuit I used for filtering is at work, I’ll try to remember to look it up Monday
The datasheet says: max. output current is between 21mA and 25mA. With your 42mA you are out of specs.
Best regards!
jokeramik
 
For a 3 channel mic pre these ranges aren´t so good, beecause the power isn´t enough.
??? The datasheet is clear, 0.1A at 48V. It should be capable of powering at least 10 mics (or supporting 7 shorted connections).
Generally speaking, I would just be a little concerned about how it behaves with a very light load, which would happen with only 1 mic connected. However, the datasheet says: "No minimum load required", so it should be a non-issue.
 
I would just be a little concerned about how it behaves with a very light load, which would happen with only 1 mic connected. However, the datasheet says: "No minimum load required", so it should be a non-issue.
I would very carefully test a new SMPS module under low load for sure. I would not be shocked if the noise floor changed and you got whistling in the audio range under light load. The datasheets are sparse on details like that but it does say the non-5V models have a max ripple of 1% which in this case would be 480mVpp which is quite high.

I like this module a lot because it's the smallest (2.54cm square) at 48V I've seen that looks viable for pro-audio. My guess at this point would be that, under this particular confluence of circumstances, this module deserves a PCB for a capacitance multiplier (which will attenuate the 120kHz switching noise so no choke necessary) with a copper area stitched with vias large enough to dissipate the 2-3W of a shunt regulator that asserts a minimum load of 20mA or so. A low-class version might just add an power resistor with a lot of space around it (which might also be used as the dropper resistor for an indicator LED).

You might also try to find the open-frame version and hack it to make 50V to make up for the voltage drop of a proper two transistor CM but that's probably being pedantic in this scenario.
 
Last edited:
As I said previously, they work perfectly. No whistling on no/low load that I can hear or measure, and scoping the output showed no evidence of hiccuping.
I did indeed use a power resistor with an LED as bit of an extra load just to be sure. :)
The next mic amp I build with one, I'll just use the CM without a choke and compare.
 
Back
Top