Testing for phantom power, blocking it...

Recording Techniques, People Skills, Gear, Recording Spaces, Computers, and DIY

Moderators: drumsound, tomb

User avatar
chconnor
takin' a dinner break
Posts: 170
Joined: Thu Jan 04, 2007 4:23 pm
Location: Bellingham, WA
Contact:

Testing for phantom power, blocking it...

Post by chconnor » Fri Feb 02, 2007 5:17 pm

Hi -

I want to build or buy a simple tester for phantom power. I've forgotten what little I used to know about designing circuits, and was wondering if anyone had some guidance:

My first inclination was to just buy a male XLR jack and wire 3 different LEDs across the 3 pins (1-2, 1-3, 2-3), but 48V and up to 10mA (according to what I've heard of the phantom power spec) seems like it would burn them out pretty quick? There must be a simple way to throw in a resistor or something to knock it down to something the LEDs can handle, right? At least for a few seconds at a time? My theoretical result is a tiny device that I can plug into a jack which will light up if phantom power is present across any two of the terminals.

Alternately, I'd be satisfied with something that just blocked phantom power (if present) on an XLR line, but something about what I've read of phantom power makes me think this may not be quite so trivial.

All the XLR cable testers I saw didn't seem to report on whether the power was present, just if the wiring was wrong, and besides, they were larger than I'd like and why spend $30 when I can do it for $5, etc. But, if you know of a small, cheap, durable product, I'm all ears.

Ideas? Thanks!
-Casey

User avatar
Scodiddly
genitals didn't survive the freeze
Posts: 3957
Joined: Wed Dec 10, 2003 6:38 am
Location: Mundelein, IL, USA
Contact:

Post by Scodiddly » Fri Feb 02, 2007 5:34 pm

The Swizz Army cable tester (and as stolen by Behringer :roll: ) will show phantom power status. But aside from that, it's fairly easy to put a resistor in series with an LED to match just about any voltage.

Ohm's Law is your friend here:
A typical LED wants around 20mA current, and at least a couple volts. Solving for current...
I (current) = V (voltage) / R (resistance), or R = E/I.
So, R = 48/0.02 = 2400 = 2.4kOhm

Blocking phantom power is easy enough - either use a transformer, or put a couple of high-value capacitors in series with pins 2 & 3.

User avatar
chconnor
takin' a dinner break
Posts: 170
Joined: Thu Jan 04, 2007 4:23 pm
Location: Bellingham, WA
Contact:

Ok...

Post by chconnor » Fri Feb 02, 2007 6:00 pm

Thanks for the help. I did some research; so given your math:

> A typical LED wants around 20mA current, and at least a couple volts. Solving
> for current...
> I (current) = V (voltage) / R (resistance), or R = E/I.
> So, R = 48/0.02 = 2400 = 2.4kOhm

...don't I want to calculate the "voltage drop" over current? So if the LED wants 20mA and a few volts, say 2.5V, I would do (48-2.5)/.02, right? Maybe you just left it out because 48V is big enough that 2 or 3V doesn't matter much in that particular calculation... ?

Then my other question is the wiring: shall I wire a resistor and a LED in series between pins 1 and 2, and repeat the same between 2 and 3, and also between 1 and 3? I'm not clear on the phantom power wiring, so maybe this is silly, but the hope was that my three LEDs would then show if anything was present.

I'm not particularly concerned that they accurately reflect anything. E.g. if all three always light up at once, etc. I just want to plug in my little device, see no lights, and trust that a given XLR jack is totally cold across all pins.

Thanks!
-c

The Scum
moves faders with mind
Posts: 2745
Joined: Thu Jul 03, 2003 11:26 pm
Location: Denver, CO
Contact:

Post by The Scum » Fri Feb 02, 2007 7:45 pm

...don't I want to calculate the "voltage drop" over current? So if the LED wants 20mA and a few volts, say 2.5V, I would do (48-2.5)/.02, right? Maybe you just left it out because 48V is big enough that 2 or 3V doesn't matter much in that particular calculation... ?
You're right, the 2.5V forward drop is about 5% of the 48V, so the overall error isn't too great. And for a long time, 5% was a reasonable tolerance for analog electronics.

You're probably not going to smoke an LED if it gets 25ma...and it might still light with maybe 10 ma, just not as brightly.
Then my other question is the wiring: shall I wire a resistor and a LED in series between pins 1 and 2, and repeat the same between 2 and 3, and also between 1 and 3? I'm not clear on the phantom power wiring, so maybe this is silly, but the hope was that my three LEDs would then show if anything was present.
Phantom is tied to both 2 and 3, WRT 1.

I'd do one between 1 and 2, and another between 1 and 3. If they both light, you know that both lines have power, so they're both connected. If ony one lights, you'll know you've got a bad connection somewhere on that side of the run. If neither lights, there's no phantom present.

You could be sneaky and use a 3-terminal bi-color LED: orange is good, red or green is miswired, and dark is no phantom. And a single T1-3/4 size LED would fit in an XLR boot.

Putting one between 2 and 3 would be mesauring any imbalance in the 6k8 resistors in the phantom supply, and in a proprtly wired system, shouldn't have much potential difference, thus very little current between them.

User avatar
chconnor
takin' a dinner break
Posts: 170
Joined: Thu Jan 04, 2007 4:23 pm
Location: Bellingham, WA
Contact:

Ok...

Post by chconnor » Fri Feb 02, 2007 11:19 pm

Beautiful; so:

> I'd do one between 1 and 2, and another between 1 and 3. If they both light,
> you know that both lines have power, so they're both connected.

The reason for this device is to make sure that a device I plug into a snake without a DI at a strange club will be safe... so I wanted to check 2-3 as well, in case there was some crazy wiring mistake, so I don't fry my box. Is checking 2-3 just excessive paranoia and I shouldn't worry about it?

If I'm not mistaken, blithely adding a third LED across 2-3 would cause all three to light up if even only one of the pins had voltage WRT another pin, right? Though perhaps not as brightly because two would be in series, etc etc...

I guess I could just keep it simple. :-)

I probably do the 3-lead bicolor idea, thanks again for the help,
-c

The Scum
moves faders with mind
Posts: 2745
Joined: Thu Jul 03, 2003 11:26 pm
Location: Denver, CO
Contact:

Post by The Scum » Sat Feb 03, 2007 12:17 pm

You'll need to go back and figure out what your requirements really are: are you simply verifying that phantom is turned off, or are you trying to diagnose all of the potential wiring problems that can exist?

The proposed LED gizmo will work for detecting phantom...assuming the systems you are working with were built with off the shelf cables and the like (or custom built by someone with a clue), it should suffice.

For all of the possible bad wiring combinations that could possibly exist, it's hard to come up with a simple device that will diagnose them...a cheap multimeter might be more effective than a little tester, if that's what you really want to diagnose.

User avatar
Scodiddly
genitals didn't survive the freeze
Posts: 3957
Joined: Wed Dec 10, 2003 6:38 am
Location: Mundelein, IL, USA
Contact:

Post by Scodiddly » Sat Feb 03, 2007 12:39 pm

I think it's a great idea. Lots of times I've seen people hook stuff up without worrying about whether the phantom is on or not - and in situations where phantom would definitely cause problems.

User avatar
chconnor
takin' a dinner break
Posts: 170
Joined: Thu Jan 04, 2007 4:23 pm
Location: Bellingham, WA
Contact:

...

Post by chconnor » Sat Feb 03, 2007 1:45 pm

The question I was hoping to answer was "does this jack possess any ability to hurt my box", which I was hoping to answer by asking "is there voltage across any two of the three pins", which I was hoping to answer by two or three LEDs and resistors strapped between the pins.

For this tester, I don't really care if the jack is wired wrong, as long as it doesn't have power coming down the wires.

So if I put a third LED/resistor combination across pins 2-3, I was thinking it would disable the ability to tell which leg of phantom power was missing, because if just one pin has voltage to pin 1, it would make a circuit through all three LEDs, but that would be OK with me. All I really want to do is plug this thing into a jack, see some light(s), and say "oops, better not plug my box in until that's figured out." So I thought three LEDs would provide maximum protection, even if they give a little less information. Just not sure if the resultant circuit would work (e.g. "light something up") in the various possible combinations (e.g. normal phantom, one leg missing, weird wiring, etc).

Much thanks,
-c

The Scum
moves faders with mind
Posts: 2745
Joined: Thu Jul 03, 2003 11:26 pm
Location: Denver, CO
Contact:

Post by The Scum » Sat Feb 03, 2007 10:17 pm

is there voltage across any two of the three pins
Hmmm...I've been puzzling over this all day, and I think I've hit on the solution: 6 LED/resistor pairs.

Draw a triangle to represent the 3 pins. Use 3 LEDs so the "arrows" they make go around the triangle clockwise. Use the other 3 to go around the triangle counterclockwse.

If there's any significant potential difference between any pins, something will light...significant meaning enough surpass the forward voltage, and deliver enough current to make the LEDs light. By puzzling over which ones light up, you can decipher what's hot.

But this is still only between the pins on the mic input. It wouldn't detect differeing AC mains potentials, AC ground miswiring or loops, etc. (the old "I get shocked every time I sing" phenomenon)

User avatar
chconnor
takin' a dinner break
Posts: 170
Joined: Thu Jan 04, 2007 4:23 pm
Location: Bellingham, WA
Contact:

Ahh soo...

Post by chconnor » Sat Feb 03, 2007 11:28 pm

Draw a triangle to represent the 3 pins. Use 3 LEDs so the "arrows" they make go around the triangle clockwise. Use the other 3 to go around the triangle counterclockwse.
Doh! This whole time I was forgetting: LEDs are polarized! I was thinking they worked in either direction like an incandescent... Sorry for the confusing questions that resulted.

Anyway, so with the 6-LED plan we get what I was hoping would happen with three imaginary non-polarized LEDs... sounds good. I'll probably implement it with 3 bi-color LEDs to save space (and add to the confusing fun when it comes to decoding the output. :-))
It wouldn't detect differing AC mains potentials, AC ground miswiring or loops, etc.
Ah, good point. Maybe if I used 3 small incandescent lights instead I'd have a shot at picking some of that up as well?

And, before I buy the resistors: I'm going to calculate the resistance as described before, and basically buy the closest available resistor not less than that value, yeah? If I recall, there's nothing too weird to watch out for in resistors in terms of type, etc...

Almost ready to solder! :-)
-c

The Scum
moves faders with mind
Posts: 2745
Joined: Thu Jul 03, 2003 11:26 pm
Location: Denver, CO
Contact:

Post by The Scum » Sun Feb 04, 2007 12:32 am

Ah, good point. Maybe if I used 3 small incandescent lights instead I'd have a shot at picking some of that up as well?
No. They'd still only be comparing the 3 pins on the mic connector to one another. They wouldn't be checking them against the mains ground/neutral on the stage.

That's one of those conditions that takes an electrician to solve, too, not something you can fix when taking the stage.

To protect against it, you can use transformers with the ground lifted.

What is the mystery device we're trying to protect here? Does it even connecto to mains?
And, before I buy the resistors: I'm going to calculate the resistance as described before, and basically buy the closest available resistor not less than that value, yeah? If I recall, there's nothing too weird to watch out for in resistors in terms of type, etc...
I could point you at a catalog with probably 40 pages of resistors...

...but for this application, they're aren't particularly critical. The garden variety 5% carbon comp resistors will be fine.

But there are a few gotchas here, too...
  • -phantom power is usually current limited using 6k8 resistors at the source...so they're in series with the 2k4. A phantom powered mic needs to be able to run off 15 mA (since phantom is on 2 & 3, they can each deliver ~7mA).
    -the 2k4 was calculated for 48V phantom...but there are cheapo systems out there with 12V and 24V phantom.
    -if my late night math is right, drawing 20 mA from a 48V source gives you nearly a Watt of power dissipation, so you'll need 1W resistors...but limited by the 6k8s, that only gives you about 1/3 Watt.
    -if there's something really, really wrong with the system, like 120VAC on a mic connector, it could kill the LEDs as soon as you plug the thing in. So they'd stay dark, and everything would look safe. Not that that's a common situation.
To just check for phantom, it's not too hard. To be maximally paranoid, this device is going to be complex.

User avatar
chconnor
takin' a dinner break
Posts: 170
Joined: Thu Jan 04, 2007 4:23 pm
Location: Bellingham, WA
Contact:

...

Post by chconnor » Sun Feb 04, 2007 1:57 am

You've been so helpful.

I found this:

http://www.allelectronics.com/cgi-bin/i ... EGS_.html

...which even shows a sort of yellow with AC, though you said that's not a concern as far as this project is concerned (it could be if something really crazy had happened, yeah?). I'm thinking three of those...
What is the mystery device we're trying to protect here? Does it even connect to mains?
It's just a Presonus Firebox, and it is connected to mains, but it's also connected to my laptop (running on batteries)... I'm hoping to avoid frying either of them, but generally it'd also be nice to more safely plug other things... (keyboard, etc.) And to confirm that phantom is on if the mic signal isn't working, etc... The big idea is to run its low-Z balanced 1/4" outputs into an XLR snake and into the board's 1/4" line-ins, but first to make sure that the sound tech has switched the correct ends of the snake to the line-ins (I provide adapters) on the board (i.e. that they aren't still connected to phantom-powering mix pres)

This is all moot with a couple DIs, of course, but this way is cheaper and avoids another box in the signal path (though it involves using adapters).

To be honest, it's just interesting, since it's all kinda overkill to begin with. :-)
But there are a few gotchas here, too...
Ok, so hopefully I understand the implications of the gotchas:
A phantom powered mic needs to be able to run off 15 mA (since phantom is on 2 & 3, they can each deliver ~7mA).
...does this imply that there may not be enough current for the 20 mA LEDs? Probably not, huh... my lack of electronic know-how is showing, here... You were maybe just giving background here...
the 2k4 was calculated for 48V phantom...but there are cheapo systems out there with 12V and 24V phantom.
...in which cases the LEDs may fail to light, or be dim, etc. Ok.
if my late night math is right, drawing 20 mA from a 48V source gives you nearly a Watt of power dissipation, so you'll need 1W resistors...but limited by the 6k8s, that only gives you about 1/3 Watt.
Ok, so the wattage is a rating that reflects the current load the resistor can handle... So if your late night math and my abject guesswork are correct, something like this:
http://www.mouser.com/search/ProductDet ... J500-2.4K may do the trick, since they are 1/2 Watt rated.
if there's something really, really wrong with the system, like 120VAC on a mic connector, it could kill the LEDs as soon as you plug the thing in. So they'd stay dark, and everything would look safe. Not that that's a common situation.
Thankfully! My plan is to "test the tester" on a trusted phantom source, then check the snake jacks I'm hoping to use, and then, if no lights appear, I could always check it again in the trusted source to make sure they didn't just burn out.
To just check for phantom, it's not too hard. To be maximally paranoid, this device is going to be complex.
Yeah, maximal paranoia is not necessary. I'm hoping for maximal crude utility limited by requirements of simplicity and economy. :-) If can detect a couple cases of weirdness at the same time, so much the better. If not, I'll live without the extra features.

Since this tester isn't going to be plugged in for more than a few seconds at a time, perhaps the circuit could be designed to be sensitive enough to detect 24V phantom and it wouldn't be too hard on the LEDs for such a short time... I guess using a 1k2 resistor instead of the 2k4 would double the current draw to like 40 mA... maybe that's too high, but perhaps a 2k0 resistor or something in between.

Anyway, I'm just a dude plugging his stuff into your average clubs, nothing too crazy or exotic (although I've used some "ye olde mixing boarde" systems for sure)... if it's splitting hairs to worry about 24V (how common is that, anyhow? Or systems that have 48V but less current available?) then it's not a big deal.

Ok!
-c

User avatar
Scodiddly
genitals didn't survive the freeze
Posts: 3957
Joined: Wed Dec 10, 2003 6:38 am
Location: Mundelein, IL, USA
Contact:

Post by Scodiddly » Sun Feb 04, 2007 8:42 am

LEDs do vary in brightness depending on the current, which in this case will depend on the voltage through the dropping resistor. I'd guess that 120v would make the LEDs glow very brightly, and you'd want to disconnect fairly fast to avoid burning them out. But they'd likely survive.

Here's a thought: In your little test box, have one more LED between pin 1 and the chassis/casing. If the phantom test LEDs don't light up, touch the casing to a grounded metal bit on your source to see if that "grounding check" LED lights up.

Another project I haven't gotten around to yet is a phantom power voltage meter - some of my mics aren't very happy with less than the full 48v phantom power. I was thinking of something with a few LEDs in a bargraph...

User avatar
chconnor
takin' a dinner break
Posts: 170
Joined: Thu Jan 04, 2007 4:23 pm
Location: Bellingham, WA
Contact:

grounding...

Post by chconnor » Sun Feb 04, 2007 1:57 pm

Scodiddly wrote:Here's a thought: In your little test box, have one more LED between pin 1 and the chassis/casing. If the phantom test LEDs don't light up, touch the casing to a grounded metal bit on your source to see if that "grounding check" LED lights up.
Ok, just to make sure I'm clear. :-) So the idea here is to see if there's any voltage inappropriately wired into Pin 1 that could ground through my box or the singer or whatever. I'm planning to build this thing within the body of a male XLR jack only, and if the XLR body I use for the tester has a metal housing then pin 1 is already wired to it's casing, right? And whatever I plug it into will connect its ground to the body of the XLR connector as well, yeah? Does that imply that I would want one with a plastic body so that I can effectively do this grounding test?
Scodiddly wrote:Another project I haven't gotten around to yet is a phantom power voltage meter - some of my mics aren't very happy with less than the full 48v phantom power. I was thinking of something with a few LEDs in a bargraph...
In case it's useful, I ran across this: http://www.electronics-lab.com/projects ... index.html
...maybe it's overkill. But it is small.

Thanks!
-casey

User avatar
Scodiddly
genitals didn't survive the freeze
Posts: 3957
Joined: Wed Dec 10, 2003 6:38 am
Location: Mundelein, IL, USA
Contact:

Post by Scodiddly » Sun Feb 04, 2007 2:31 pm

Any decent XLR male will have a separate lug for the chassis.

Post Reply

Who is online

Users browsing this forum: No registered users and 43 guests