Looking for 24V to 48V conversion for phantom power

[Potato Cakes]

Hello, Everyone,

I am looking for a simple, elegant way to integrate DC-DC conversion circuit to taking 24V and doubling it for phantom power to handle a single condenser mic or active DI. Something based around an IC like the ICL660S would be ideal, however this chip and many like it only work with input voltages that range from 5-10VDC and output up to 32V. I looked into do a buck-boost option but the test circuit I built for this drops voltage to zero when presented with a load. I found a company that makes very small DC-DC transformers (Ultra Miniature DC-DC Converter Transformers - Input Voltages of 5V, 12V, 24V, and 48V | Pico) that convert 24V to 48V (1.5 and 3W options), but I am waiting to see if what the minimum order will cost. If it's expensive I'll want to stick with an IC option that I can fit in a small foot print. I also don't know if a regulated circuit is necessary since the PSU for the entire unit is regulated.

Thoughts?

Thanks!

Paul

 

[moamps]

About 20 years ago I used a small SMPS 12-48V to power a stereo microphone from SLA batteries for field recording. It was based on the LM2577adj design which is still available cheaply. I put a few more RC filters at the power supply output for extra smoothing and it worked without a problem. (The value of the inductor should be recalculated according to the input and output voltage)





https://www.ti.com/lit/ds/symlink/lm2577.pdf

 

[Potato Cakes]

Thanks, moamps! I'll check this out.

Thanks!

Paul

 

[JohnRoberts]

consider a search, this has been discussed before.

One downside to DC to DC is magnetic fields. There is another technology called charge pumps (oops I just did a search and most charge pumps are too low voltage. )

around 15 years ago I made a novel phantom supply based on the old cap doubler/tripler topology, but instead of powering it from 60Hz unregulated transformer AC, I grabbed a HF square wave from a Dc to DC switcher already in the product. Because I was running at very HF I was able to reduce the size of the caps to tiny 0.1uF SMD. It was small on the PCB and cheap, but I already had the HF square wave. One could chop the 24V DC to make the HF square wave... or not.

DC to DC Switcher chip solutions are getting cheaper and probably easier than my oddball scratch design.

JR

 

[mjrippe]

Those Pico modules are very expensive, $80-90 each. A quick search on Mouser found two options for half the price. Larger size of course, so depending on your application they may not be ideal.

https://tinyurl.com/DCconverters

 

[ccaudle]

charge pumps

Analog Devices has a line of high voltage charge pumps:
AD high voltage charge pumps

At least one of those is just a controller and needs external transistors.
Just noticed at least one of them is a little misleading in the table, it says max input voltage 32V, but that must be close to the breakdown voltage of the internal transistors, because the max input voltage is 32V if you use it as an inverter, but if you use it as a doubler max input is only 16V.

LTC7821 can handle up to 36V in for a doubler, but that is the one which needs external devices (four MOSFETs).

This one looks nice, up to 55V input supply, up to 50mA output current. At first glance should work pretty well to double 24V supply and run a couple of mics:
LTC3290 datasheet

 

[JohnRoberts]

Good find, I have no idea if there are snakes in this woodpile but it looks promising. While there will not be a megnatic field to deal with, there will be current spikes related to the switching si careful layout matters.

Good Luck

JR

 

[moamps]

... I'll check this out...

Of course, you can take another similar route, buy a cheap module like this,

https://www.ebay.com/itm/2552832902...UWEUP2S7VchfAyfHXkcv0C14utE9|tkp:BFBM7Ii9spFg
change the switching diode and one resistor, and modify it to generate 48 V. The operating frequency is 400kHz. You add a few more passive parts to the output and you have a small cheap compact converter.

https://datasheetspdf.com/pdf-file/775384/XLSEMI/XL6009/1

 

[Potato Cakes]

consider a search, this has been discussed before.

One downside to DC to DC is magnetic fields. There is another technology called charge pumps (oops I just did a search and most charge pumps are too low voltage. )

around 15 years ago I made a novel phantom supply based on the old cap doubler/tripler topology, but instead of powering it from 60Hz unregulated transformer AC, I grabbed a HF square wave from a Dc to DC switcher already in the product. Because I was running at very HF I was able to reduce the size of the caps to tiny 0.1uF SMD. It was small on the PCB and cheap, but I already had the HF square wave. One could chop the 24V DC to make the HF square wave... or not.

DC to DC Switcher chip solutions are getting cheaper and probably easier than my oddball scratch design.

JR

JR,

I have been searching for a solution to this for sometime for a particular project that I've been working on for the last couple of years. I was actually one of those guys that have initiated this same discussion, but then I get sidetracked with work and other life challenges then time went by and I still haven't solved this issue. The bigger problem is that I absolutely hate dealing with power supplies but have been forced to try to figure out a solution that is compact and still get various components the power they need. I've sorted out the various voltage requirements except for phantom power. I know that for actual designers my challenges amount to basic entry level circuit design, but I seem to have run out of mental bandwidth to tackle this last issue.

That said, thanks to ccaudle for pointing me to the LTC3290. This is the most promising lead I have seen to date to solve my issue.

Thanks!

Paul

 

[JohnRoberts]

Charge pumps are routinely used to generate higher bias voltaegs from 5v logic to avoid extra power supplies. I have been out of the design trenches so am glad to see the technology has been advancing.

FWIW a well designed DC to DC switcher should be possible, but look for manufacturer app notes and do what works for you.

JR

 

[ccaudle]

pointing me to the LTC3290. This is the most promising lead I have seen to date

Keep in mind I haven't actually used that, I just found it based on some vendors I knew had products that might work.
I feel your pain on the power supply problem, there are a few switching devices that might work, but some would require winding a custom transformer, which is a big learning curve for a one-off project. Others might be able to use off-the-shelf inductors but still are a lot of layout work to get quiet (electrically) for audio use.

One thing I noticed looking over the datasheet for that charge pump is that it seems to change switching frequency based on load, which can sometimes cause a problem with audio. If you have additional filtering on the output, when the switching frequency decreases the ripple frequency might move below your filter cutoff, so a supply that was quiet enough with multiple loads might get noisier with lighter loads.
Look through the datasheet and see if there is a fixed frequency mode you can force, or if that is not possible make sure you look at the lowest and highest switching frequencies it will use and size your output filter appropriately.

 

[Potato Cakes]

I'll drop a note to Analog Devices and see if they respond with any helpful pointers, especially in regards to filtering and characteristics under load.

Thanks!

Paul

 

[Potato Cakes]

Here is another option that I am considering:

https://www.mouser.com/datasheet/2/308/1/CS5171_D-2310541.pdf
In figure 46 it shows a 4VDC in to 100VDC out at 10mA. The question I would have is do I just use less 1N4148's to get the output to be closer to 48V?

I will have to email them as well as this chip is more cost effective than the LT3290.

Thanks!

 

[JohnRoberts]

5V output?

JR

 

[Potato Cakes]

5V output?

JR

It depends on the design around the chip. There are several figures below that show examples which yield +12V, +/-12V, 100V, etc.

I've emailed Onsemi to see if they have any further advice.

Thanks!

Paul

 

[AudioIngenia]

I am looking for the same
I have found some solutions:

 

[thor.zmt]

I am looking for a simple, elegant way to integrate DC-DC conversion circuit to taking 24V and doubling it for phantom power to handle a single condenser mic or active DI. Something based around an IC like the ICL660S would be ideal, however this chip and many like it only work with input voltages that range from 5-10VDC and output up to 32V.

How about MC34063 with a cascode Mosfet to allow for > 40V?

The same principle can be applied with any switcher IC, just select the right Mosfet (or BJT).

Thor

 

[Potato Cakes]

I had a board made with that schematic but I haven't tested it yet.

 

[ruffrecords]

Back in the 70s, Neve had a phantom module that converted the 24V supply to 48V phantom power using what we would probably now call a SMPS. The schematic is on the web,

Cheers

Ian

 

[Graham Nystrom]

Its small and it's its own voltage, so you won't have issues with your 24v.

https://www.jlmaudio.com/shop/48v-15w-315ma-smps.html?display_tax_prices=1
which I believe to be this unit:

https://www.jameco.com/webapp/wcs/s...oreId=10001&catalogId=10001&productId=2106693
I bought 4 from Jameco and have implemented them in Langevin AM16 builds so I can confirm that they do indeed work.