Phantom power test circuit - what makes sense ?

[clintrubber]

What kind of functionality would you suggest for a little gizmo
that plugs into a mic-pre input and does a few checks for not
too unlikely potential phantom power errors ?

This will be overkill for me, but to keep ribbon-mics alive and to check
if there's a healty voltage available that little phantom test circuit
from Jensen came to mind:

http://www.jensen-transformers.com/as/as038.pdf

So I was wondering if it would make sense to add some stuff
to it to detect eventual unbalances in the pin-2 & pin-3 voltage
while equal currents are drawn from each leg.

Thought behind this is to detect the unlikely situation that both legs
are not created equal (one of the two 6k8 resistors bad for instance)
and that a ribbon-mic could be hurt, even while it's not 'ground-referenced'.
I can imagine this could actually happen by transients
when the 'wrong' voltages are stored on AC-coupling caps.


So I'm all ears for suggestions, any hints & sanity checks welcomed.
Like do we really need a check for DC-unbalance ?

Eventual suggested functions will be easily addible to the circuit,
but as said I'm mainly interested in what one actually really needs
before going crazy with checks for all kinds of highly unlikely events.

If we come to the conclusion that nothing needs to be added actually
then it'll be the Jensen-circuit and I'll apologize for using Drawing Board
bandwidth :wink:


Thanks / regards,

Peter

 

[clintrubber]

[quote author="mediatechnology"]Peter: I'm assuming in the drawing that "-" is actually ground not signal minus or ring. I wrote about DC imbalance quite a bit in my DC-Coupled preamp project (that I'm still working on) but for different reasons than you cite. [/quote]
Thanks Wayne for responding.
I'll check your words about DC-imbalance.
Right, "-" must be ground. When seeing this circuit for the first time I more or less assumed right away that this would be best implemented in an XLR-plug, but the Jensen-file doesn't mention this.

The circuit as drawn could have two test leads (to be inserted in a female XLR-socket: #1&#2 and #1&#3) and the testbutton could be skipped.

Or indeed have a male XLR-plug and a on-off-on switch connects the top-side of the circuit to either XLR-#2 or XLR-#3. This will be fine for most stuff I assume, but it doesn't detect a difference between both hot inputs.
But I assume that when something is wrong it'll be really wrong, so not as subtle as showing 45V vs 47V under load. The check of each leg separately will be fine then.


Hmm, I now realize this circuit as is CAN'T sniff the phantom voltage 'after' the 6k8 resistors. Current draw is 7mA, so there'll be a full 48V already across those 6k8...
OK, let's keep the topology but adjust the dimensioning.

Seems like the most likely damager to a ribbon mic is a bad cable where either tip or ring is shorted to ground or when input coupling caps dump a huge charge into the mic.

So when connecting ribbons it'd be best to mute eventual speakers, power-up the mic-pre, plug in the cable at the mic-pre side, check the other end with the resulting gizmo of this thread and only then plug in the (cheap Chinese :wink: ) ribbon-mic.

Regards,

Peter

 

[sodderboy]

I have built little jammies with a MXLR and LED's. The next design is to be a multicolor LED, that will be violet when P48 is correct. If not, it will be either blue, red, or dark. It really only tests the presence of voltage, but it really catches any wiring mistakes or problems.
Mike

 

[clintrubber]

[quote author="sodderboy"]I have built little jammies with a MXLR and LED's. The next design is to be a multicolor LED, that will be violet when P48 is correct. If not, it will be either blue, red, or dark. It really only tests the presence of voltage, but it really catches any wiring mistakes or problems.
Mike[/quote]
Hi Mike,

Sounds interesting & like a fairly compact thing addressing the relevant stuff. Is it using some window-comparators? Did you made up the circuit yourself ? In any way related to the Jensen-circuit ? Can we have a look at the schematic ?

Pweh, four question masks already, that'll do for now :wink:

Thanks,

Peter

 

[sodderboy]

Hi Peter,
There is no schematic. Just two series circuits, each a LED and resistor between pins 3 and 1 and 2 and 1 loaded into a MXLR. Use ohms law to calculate required current for LEDs, and grind-down the locking tab in the XLR for easy in/out. Here is a picture of it on my bench:

I blow down studio or console panels with it and if I do not get both LEDs, it is out with the Fluke.
Another way to go is with a bi-color I saw in Rat Shack. The LED biased one way is green, and reverse biased is red. That way you can include a wire reversal check. Like the version you linked.
The three LED version you linked is cool, and truth table should create a simple circuit to make using "dead bug" or 3D SOIC construction to slam it into an XLR. I might even be only LEDs and diodes.
The new version will use a single multi-color like I described, which will stick out of the boot of an epoxied Neutrik. I think I will incorporate the Jensen circuit as well. I like that.
Mike

 

[clintrubber]

[quote author="sodderboy"]Hi Peter,
There is no schematic. Just two series circuits, each a LED and resistor between pins 3 and 1 and 2 and 1 loaded into a MXLR. Use ohms law to calculate required current for LEDs, and grind-down the locking tab in the XLR for easy in/out.
Mike[/quote]
Hi Mike,

Thanks for posting! Right, simple resistors & straightforward use of Ohms law, I guess it'll do the job just fine for most situations.

About the wire-reversal check, I assume you mean that either 2 or 3 are reversed w.r.t. 1, correct ?
Or simple having the two normal LEDs and as a third LED the bicolor between 2 & 3 to see if there's a significant unbalance ?

The Jensen circuit is a nice gadget indeed, but after having looked at the various circuitvalues I realized it can't be intended to 'snif' after the 6k8 resistors, since it'll never reach its threshold. Which also explains they don't mention any XLR-connection; it must be intened for checking 'raw'-pre-resistor 48V points.

Bye,

Peter

 

[SSLtech]

Actually, you can probably do away with series resistors... you've already got them in there with the 6.8Ks by default...

If you want to add another couple of kilohms there's no harm in that, but themax current is 14mA as it is, and that won't hurt an LED. It's probably better to add about 4.7kΩ though, to make things safer... -The different forward voltage drop of green and red LEDs means that statically, you'd have a big DC imbalance across your transformer primary with no additional limiting resistors

Assume for a moment that green =3V forward drop, red=2V forward drop... with no additional series resistance, that puts 1V across the transformer primary... which will be pulled down a lot by the transformers primary DC resistance betwen the two points, but it'll still be bad for business in terms of DC current through a transformer primary, for as long as you leave the tester plugged in.

Keith

 

[clintrubber]

[quote author="SSLtech"]Actually, you can probably do away with series resistors... you've already got them in there with the 6.8Ks by default...

If you want to add another couple of kilohms there's no harm in that, but themax current is 14mA as it is, and that won't hurt an LED. It's probably better to add about 4.7kΩ though, to make things safer... -The different forward voltage drop of green and red LEDs means that statically, you'd have a big DC imbalance across your transformer primary with no additional limiting resistors

Assume for a moment that green =3V forward drop, red=2V forward drop... with no additional series resistance, that puts 1V across the transformer primary... which will be pulled down a lot by the transformers primary DC resistance betwen the two points, but it'll still be bad for business in terms of DC current through a transformer primary, for as long as you leave the tester plugged in.

Keith[/quote]
As I understood it the LEDs that each are connected between #2 / #3 and #1 are of the same colour. If these would have been bi-colored then I assume they'd each be oriented identically.
So while the DC-current through a TX-primary looked indeed like an overseen consequence of this tester at first reading, I don't see it any longer now.

Maybe we should start adding little drawings because despite the circuits being very simple, we might have different interpretations.

Given there's an identical current drawn from both legs (same LED-color, same added resistance) a third LED (a bicolor one) between #2 & #3 could now detect big (BIG) unbalances. It'll detect an unbalance which was already there (and makes the LED lighting up). Dunno how much use such a LED would be since the thresholds will be Volts and indeed, as Keith said, be unequal.
So make that unbalance-detection of use I think we need so use a simple window-comparator or some other diff-pair-ish thing that's looking for smaller unbalances and fires up a LED which doesn't cause new & unwanted side-effects.

Regards,

Peter

 

[Gus]

I would use three pin jacks (the kind test probes fit in) and a DMM

one at pin 2
one at pin 3
one at pin 1

I would also used a hand matched set of resistors for what ever current draw you want.

The voltage measurement between 2 and 3 should be very close to 0 volts This is a more direct test of an imbalance than measuring 2 to 1 and then 3 to 1.

 

[clintrubber]

[quote author="Gus"]I would use three pin jacks (the kind test probes fit in) and a DMM

one at pin 2
one at pin 3
one at pin 1

I would also used a hand matched set of resistors for what ever current draw you want.

The voltage measurement between 2 and 3 should be very close to 0 volts This is a more direct test of an imbalance than measuring 2 to 1 and then 3 to 1.[/quote]
Thanks for chiming in. It's of course a compromise between accuracy & completeness on one side and speed & ease of use at the other side. The original approach was to have a simple plug-in gizmo that you add to your small carry-around box of mike-pads, earplugs, picks, bottle-opener & other small utils. The DMM will of course be more accurate but is a bot more hassle.

W.r.t. #2 vs #3, indeed, the comparison should be made as direct as possible.
Initially was thinking of two times the Jensen-TL431-circuit (which might be accurate enough already with the given values & tolerances or when not then tuned for drawing matched currents from #2 & #3). And added to that some diff-pair that makes a third LED light up when the voltage difference between #2 #3 exceeds a certain value. But I wasn't sure whether that would be worth the hassle - in other words: how (un)likely such an imbalance would be and if it was worth testing for.

Regards,

Peter

 

[Gus]

Some DMMs are cheap so it could be a small cheap DMM wired to a 3 way rotary switch and then to the connector. Taped together installed in a box etc.

2 to 3
2 to 1
3 to 1

 

[sodderboy]

If I am making a phantom feed in a mic pre, I pair 1% resistors matched to less than .1%. With my tester, I am checking for the existence of phantom at the end of an installation, after I have used a cable tester to test all the wiring.
If the concern is phantom imbalance, you would want to test P48 voltage through the mic lines under load with a multimeter and no quick fix tester.
For an in-house user, do this test and stop worrying about it. For a journeyman, if you bring mics, bring your own spec'ed phantom power and do not worry about what the house has.
In close to 20 years in the trenches, I have NEVER experienced any damage to any microphone due to bad, mis-wired or mis-used phantom power.
Engineer error however. . .

This screen was on a C-12VR when it took a 5 foot drop.
Mike

 

[clintrubber]

[quote author="sodderboy"]If the concern is phantom imbalance, you would want to test P48 voltage through the mic lines under load with a multimeter and no quick fix tester.
For an in-house user, do this test and stop worrying about it. For a journeyman, if you bring mics, bring your own spec'ed phantom power and do not worry about what the house has.
In close to 20 years in the trenches, I have NEVER experienced any damage to any microphone due to bad, mis-wired or mis-used phantom power. [/quote]
Well said! Can't do otherwise than agreeing with the above, I'm convinced.

Engineer error however. . .

A sad pic !

Thanks,

Peter

 

[CJ]

Get a DMM.
Put it on 200 ma scale.


dead short pins 1 to 2 and 3 with the ammeter and read current.

if no 6.81 K's you blow a fuse. no biggy.

 

[gyraf]

This screen was on a C-12VR when it took a 5 foot drop.
Mike

..I have an AKG "TheTube" screen that looks exactly like that.. :razz:

Jakob E.