Author Topic: PSU 48v for microphone  (Read 4278 times)

0 Members and 1 Guest are viewing this topic.

Offline harpsTopic starter

  • Regular Contributor
  • *
  • Posts: 73
PSU 48v for microphone
« on: February 01, 2018, 09:00:04 pm »
Hello there  i hope to power microphones that require 48v. ( 1 or 2 mics to begin with, then maybe 8 later on)

I have a PCB that i have 63v AC going in.

using a regulator I now have the desired 48v DC going out ( 47.75 to be a bit more precise)

I include a picture. attached below.

I thought to ask if this is good and safe enough, just in case i need a current limiting resistors or other things I am not very sure of.

i used a fixed resistor instead of a variable, just in case it ever got knocked.(64.7k)

thanks for help. : )



« Last Edit: February 03, 2018, 09:37:55 am by harps »
 

Offline floobydust

  • Super Contributor
  • ***
  • Posts: 6924
  • Country: ca
Re: PSU 48v for microphone
« Reply #1 on: February 01, 2018, 09:39:17 pm »
Problem with the TL783 is it has a minumum output current spec of 15mA.
You would need a dummy load that would make 0.72W of heat; could use 10mA worth to an LED and the rest to TL783 divider resistors; and the TL783 wastes additional 0.72W
It's a 1.5W of heat doing nothing.

Microchip HV linear regulator like LR8 might be an option.

For phantom power, consider a discrete regulator with TIP47-TIP50 and a zener is popular or something like Elliot sound Project 96. Current limiting is the only hassle there.

870uF is huge for your filter capacitors, and you have lots of (ripple) headroom with 96VDC coming in.
A current-limiting resistor on the output is needed or you can damage gear hot connecting things.
 
The following users thanked this post: harps

Offline harpsTopic starter

  • Regular Contributor
  • *
  • Posts: 73
Re: PSU 48v for microphone
« Reply #2 on: February 02, 2018, 12:28:50 am »
Thank you very much. so glad i asked.  : )  I include an updated picture.

I have these 870uf caps very much welded in, i fear removing them. ( i just presume it will take a while to discharge, no harm can be caused?)

I wonder, are you suggesting that if i add an led for the front panel and maybe an led inside the case I should be o.k to protect the microphone/s?

I do not mind wasting 1 - 2 watts.

Regards the current  limiting resistors , are they the resistors for the leds as pictured below, or do i need to add current limiting resistors in series on the output of the pcb like a 50ohm resistor 1w watt? or are the resistors on the leds o.k?

thank you very much. for your suggestions!!  : )

« Last Edit: February 02, 2018, 12:30:33 am by harps »
 

Offline floobydust

  • Super Contributor
  • ***
  • Posts: 6924
  • Country: ca
Re: PSU 48v for microphone
« Reply #3 on: February 02, 2018, 02:06:31 am »
Staying with the TL483, maybe check your resistor calcs.
For 48V output R1=172R if 6k4 used to GND. It dissipates 0.34W so use 1/2W-1W part for the 6k4, and total is about 7mA load. R1 can be 1/4W, it only sees 1.25Vref.

Here is sample Elektor circuit for phantom power like yours. They used 5k6 and 150R, giving 0.4W (use 1W resistor) and 8.3mA load so they added an LED. For the LED, 4k7 gives around 10mA and it need to be a 1W resistor. Your 1k value is too low for LED's. (48V-VLED 2.1V)/1k=45.9mA!

I'm saying the divider load and LED should be at least 15mA. TL783 will need a heatsink too.

Usually you have 6.8k 1/2W 1% resistors to the balanced mic (+), (-) which give current limiting if there is a short on the load or a bad mic cable
 
The following users thanked this post: harps

Online Zero999

  • Super Contributor
  • ***
  • Posts: 19345
  • Country: gb
  • 0999
Re: PSU 48v for microphone
« Reply #4 on: February 02, 2018, 11:02:08 am »
How about the LM317, plus a couple of transistors?

You might want to use a lower voltage zener diode, than 39V though, try 33V or 27V.
 
The following users thanked this post: harps

Offline Alex Nikitin

  • Super Contributor
  • ***
  • Posts: 1150
  • Country: gb
  • Femtoampnut and Tapehead.
    • A.N.T. Audio
Re: PSU 48v for microphone
« Reply #5 on: February 02, 2018, 03:06:23 pm »
The load current from this 48V supply should be less than 10mA, so a parallel low noise regulator will be a better and much safer option. Something like seven LM329 in series and a ~12mA current source (i.e. BSP129 DFET and 100 Ohm resistor), supplied by 60-90V unregulated DC, in case of 63V AC (or about 87V rectified) a quality 5W 1K5 W/W resistor in series would be a good idea.

Cheers

Alex

« Last Edit: February 02, 2018, 03:17:43 pm by Alex Nikitin »
 
The following users thanked this post: harps

Offline harpsTopic starter

  • Regular Contributor
  • *
  • Posts: 73
Re: PSU 48v for microphone
« Reply #6 on: February 02, 2018, 10:01:17 pm »
thank you everyone for your great ideas !!  :  P

i would like to stick with what i started as i have the pcb already populated.

i wonder if this design tweak attached below would be o.k?

eevblog member "floobydust" says i should use an led to help protect the microphones. i have added two leds in the design, as i fear one might blow and then a microphone may be harmed.  I may leave the second led inside the case.  ( Could i just use a load resistor?  is an led needed?)

Thank you very much for help as i fear blowing up my microphones : )

« Last Edit: February 02, 2018, 10:05:18 pm by harps »
 

Offline floobydust

  • Super Contributor
  • ***
  • Posts: 6924
  • Country: ca
Re: PSU 48v for microphone
« Reply #7 on: February 03, 2018, 12:41:01 am »
I little misunderstanding, the LED is just to keep the TL783 regulator happy as part of its minimum 15mA load requirement. The LED does not protect microphones; other than discharging the large filter caps when power is turned off. One LED is enough.

The 870uF output cap is huge and will make a large high-current spark as the 48V output gets connected, or if shorted in a bad cable.
You must have the two 6.8k resistors somewhere, or things like the mic or mic preamp can get damaged.
How are you connecting this phantom power to the mics and pre-amps?

For a bullet-proof PSU, there are three protection diodes needed. I just remembered you have ~90VDC raw and a shorted TL783 would damage your microphones. You have one diode D1 good, and need one across R1 (cathode to +48V) and another across the PSU output as in the prev. circuit. These are in the same position as vanilla LM317 regulator IC protection diodes.

 
The following users thanked this post: harps

Offline harpsTopic starter

  • Regular Contributor
  • *
  • Posts: 73
Re: PSU 48v for microphone
« Reply #8 on: February 03, 2018, 09:35:38 am »
I little misunderstanding, the LED is just to keep the TL783 regulator happy as part of its minimum 15mA load requirement. The LED does not protect microphones; other than discharging the large filter caps when power is turned off. One LED is enough.

The 870uF output cap is huge and will make a large high-current spark as the 48V output gets connected, or if shorted in a bad cable.
You must have the two 6.8k resistors somewhere, or things like the mic or mic preamp can get damaged.
How are you connecting this phantom power to the mics and pre-amps?

For a bullet-proof PSU, there are three protection diodes needed. I just remembered you have ~90VDC raw and a shorted TL783 would damage your microphones. You have one diode D1 good, and need one across R1 (cathode to +48V) and another across the PSU output as in the prev. circuit. These are in the same position as vanilla LM317 regulator IC protection diodes.


Hi

thank you once again!!

i have added 2 more diodes
lowered the output cap from 870uf to 47uf.
removed 1 led.

The other picture included shows the preamp already has 6.8k resistors for the phantom , perhaps this is what you were refering too?

hopefully this looks a bit better.

 : )
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 16545
  • Country: us
  • DavidH
Re: PSU 48v for microphone
« Reply #9 on: February 03, 2018, 12:30:33 pm »
The feedback divider can supply the minimum load current if necessary.

At a minimum, I would include foldback current limiting commensurate with the expected load current.

If a power LED is included, then I would include an LED to indicate current limited operation.  A power LED on the output should include a zener diode so that it shows that the output voltage is not only present but proper.
 
The following users thanked this post: harps

Offline DougSpindler

  • Super Contributor
  • ***
  • Posts: 2093
  • Country: us
Re: PSU 48v for microphone
« Reply #10 on: February 03, 2018, 02:09:39 pm »
I would look at a 48v POE injector for Ethernet.
 

Offline Cliff Matthews

  • Supporter
  • ****
  • Posts: 1910
  • Country: ca
    • General Repair and Support
Re: PSU 48v for microphone
« Reply #11 on: February 03, 2018, 02:36:01 pm »
I would look at a 48v POE injector for Ethernet.
Would it be good practice trusting mics costing hundreds of dollars to noisy PoE injector's? Maybe not.
 

Offline harpsTopic starter

  • Regular Contributor
  • *
  • Posts: 73
Re: PSU 48v for microphone
« Reply #12 on: February 03, 2018, 03:05:18 pm »
The feedback divider can supply the minimum load current if necessary.

At a minimum, I would include foldback current limiting commensurate with the expected load current.

If a power LED is included, then I would include an LED to indicate current limited operation.  A power LED on the output should include a zener diode so that it shows that the output voltage is not only present but proper.

Hi Thank you David

Do you mean make a test circuit to say whether the voltage is either too high or too low for the front panel? (48v O.K.  49v = fault )

 i made a circuit picture below, is this what you mean?  i would have to breadboard to find the resistor values as i have not used a zener before. i presume a 47v zener might work for both 48v and 49v and above detection.
 
thanks for your input.

 

Online Zero999

  • Super Contributor
  • ***
  • Posts: 19345
  • Country: gb
  • 0999
Re: PSU 48v for microphone
« Reply #13 on: February 03, 2018, 05:50:39 pm »
thank you everyone for your great ideas !!  :  P

i would like to stick with what i started as i have the pcb already populated.
How about swapping it for the LM317HV or LM117HV? Their minimum load current is lower than the TL783 and they're pin compatible so no modification to the PCB is necessary.
http://www.ti.com/lit/ds/symlink/lm317hv.pdf

The feedback divider can supply the minimum load current if necessary.

At a minimum, I would include foldback current limiting commensurate with the expected load current.

If a power LED is included, then I would include an LED to indicate current limited operation.  A power LED on the output should include a zener diode so that it shows that the output voltage is not only present but proper.

Hi Thank you David

Do you mean make a test circuit to say whether the voltage is either too high or too low for the front panel? (48v O.K.  49v = fault )

 i made a circuit picture below, is this what you mean?  i would have to breadboard to find the resistor values as i have not used a zener before. i presume a 47v zener might work for both 48v and 49v and above detection.
 
thanks for your input.
That circuit doesn't take the forward voltage of the LEDs into account. I think you'll struggle to get a zener circuit to work with that kind of accuracy. The LEDs will not have a sharp turn on threshold, but slowly get brighter, as the voltage is increased.
 
The following users thanked this post: harps

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 16545
  • Country: us
  • DavidH
Re: PSU 48v for microphone
« Reply #14 on: February 03, 2018, 06:02:00 pm »
Do you mean make a test circuit to say whether the voltage is either too high or too low for the front panel? (48v O.K.  49v = fault )

I made a circuit picture below, is this what you mean?  i would have to breadboard to find the resistor values as i have not used a zener before. i presume a 47v zener might work for both 48v and 49v and above detection.

That is the idea.  I figure as long as the power LED is running off of the output, it might as well indicate something more than *some* voltage is present.  I might put a real power indicator on the input as well.

If the LED forward voltage drop is 2 volts, and there is 3 volts across the current limiting resistor, then the zener diode should be 48-2-3=43 volts which is the standard value below 47 volts and for 5 milliamps, the current limiting resistor is 2V/5mA=600 ohms.  560 or 680 ohms would work fine.

Actually when taking into account the standard zener diode tolerance of 5%, a 39 volt zener diode should be used with a roughly 1.4k series resistor.

I would not bother indicating high output voltage.  What I might do if I wanted to protect a sensitive or expensive load is add an SCR crowbar to the output which triggers under conditions of high voltage and has its own fault indicating LED.  This would work well in combination with foldback current limiting.
 
The following users thanked this post: harps

Offline harpsTopic starter

  • Regular Contributor
  • *
  • Posts: 73
Re: PSU 48v for microphone
« Reply #15 on: February 03, 2018, 06:15:06 pm »
thank you everyone for your great ideas !!  :  P

i would like to stick with what i started as i have the pcb already populated.
How about swapping it for the LM317HV or LM117HV? Their minimum load current is lower than the TL783 and they're pin compatible so no modification to the PCB is necessary.
http://www.ti.com/lit/ds/symlink/lm317hv.pdf


great thank you hero999,  i shall do,  if it makes it all more safer to use.  seems like a good idea.  : )   

 

Offline harpsTopic starter

  • Regular Contributor
  • *
  • Posts: 73
Re: PSU 48v for microphone
« Reply #16 on: February 03, 2018, 06:20:44 pm »
Do you mean make a test circuit to say whether the voltage is either too high or too low for the front panel? (48v O.K.  49v = fault )

I made a circuit picture below, is this what you mean?  i would have to breadboard to find the resistor values as i have not used a zener before. i presume a 47v zener might work for both 48v and 49v and above detection.

That is the idea.  I figure as long as the power LED is running off of the output, it might as well indicate something more than *some* voltage is present.  I might put a real power indicator on the input as well.

If the LED forward voltage drop is 2 volts, and there is 3 volts across the current limiting resistor, then the zener diode should be 48-2-3=43 volts which is the standard value below 47 volts and for 5 milliamps, the current limiting resistor is 2V/5mA=600 ohms.  560 or 680 ohms would work fine.

Actually when taking into account the standard zener diode tolerance of 5%, a 39 volt zener diode should be used with a roughly 1.4k series resistor.

I would not bother indicating high output voltage.  What I might do if I wanted to protect a sensitive or expensive load is add an SCR crowbar to the output which triggers under conditions of high voltage and has its own fault indicating LED.  This would work well in combination with foldback current limiting.


Thank you again David.

i think i understand , select a 39 v zener instead of a 47v, as the tolerance variations in production are not that great, and then use a higher resistor as needed for the desired result.   i need to bread board, im more hands on.  i cant talk technical terms too well,  but its all very interesting, thank you, I  shall give it a go : )
 

Offline daedalux

  • Contributor
  • Posts: 34
  • Country: es
Re: PSU 48v for microphone
« Reply #17 on: February 04, 2018, 09:29:21 pm »
In a phantom power supply the main problem I've found is not a a good regulation (mics are always internally regulated) or a significant amount of current capability is needed. But more likely a huge amount of filtration with decoupling resistors and caps is needed so it won't make and audible sound through the line.
 

Offline harpsTopic starter

  • Regular Contributor
  • *
  • Posts: 73
Re: PSU 48v for microphone
« Reply #18 on: February 05, 2018, 02:37:19 am »
In a phantom power supply the main problem I've found is not a a good regulation (mics are always internally regulated) or a significant amount of current capability is needed. But more likely a huge amount of filtration with decoupling resistors and caps is needed so it won't make and audible sound through the line.

hi Daedalux

do you think i can add anything  in particular?....on post " 38 "  above ^  i have two circuit pictures, should i add some more capacitors to it? 

thanks for help. : )
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf