Behringer MX3282A Phantom Power Voltage issue

[Gord]

Hi once again,

I replaced the LM317 voltage regulator because there was over 70 volts at the test point on the power supply board. This fixed the over voltage issue. I let the board run for almost 20 hours with a load on the phantom power, and there was no issues, everything worked great.

Took the board out to use it for a small Teen concert, plugged it in, turned it on an poof the board started to squeal. Took the power supply apart again and saw that the input pin on the Phantom LM317 was discolored and that the case was burnt.

The Question I have is: Should I be using a LM350 ipo the LM317 or did I miss something up stream of the regulator on the input side?

Thanks.

Gord.

[Jazz P Bass]

What regulator was installed from the factory?
The LM317 is good for 1.5amps.
The LM350, 3 amps.
That is how much current they can provide at the output.
These are adjustable voltage regulators.
The maximum output voltage stated in the datasheet is 33 volts.
Strange that they would use this item for a 48 volt phantom supply.
What is the voltage at the input pin?
What is (was) the voltage at the output pin?
Seeing that there are not any schematics available, can you draw the circuit & post it.
I have attached a diode protected circuit from the datasheet.
Is it even close to what you have?
If you circuit has a capacitor on the output pin, replace it.
A short on the input or the output will destroy the ic, unless it has the diodes as in the circuit posted.

[Gord]

Thanks for the reply Jazz P Bass,

I have the schematics for the board and the power supply and I originally wondered the same thing - Why install a regulator that is rated lower than the expected voltage, but I am not an electronics technician, just someone who likes to fix things.

The regulator that was in the circuit was an LM317 on the circuit board itself it calls for a LM317 as does the schematics. After I changed out the regulator the first time I got just under 48 volts at the test node. The three other regulators tested within the design limits call for on the drawings.

I did not check the voltage at the input but will tonight

I did check the caps for shorts and opens but found none, but did not check the diodes.

Thanks.

Gord.

[doctor]

Thanks for the reply Jazz P Bass,

I have the schematics for the board and the power supply and I originally wondered the same thing - Why install a regulator that is rated lower than the expected voltage, but I am not an electronics technician, just someone who likes to fix things.

The regulator that was in the circuit was an LM317 on the circuit board itself it calls for a LM317 as does the schematics. After I changed out the regulator the first time I got just under 48 volts at the test node. The three other regulators tested within the design limits call for on the drawings.

I did not check the voltage at the input but will tonight

I did check the caps for shorts and opens but found none, but did not check the diodes.

Thanks.

Gord.

In accordance with schematic

1. Potential at LM317 input with reference to GND is

Vin = 48Vx 1.41 = 67.7V

2. Potential at LM317’s ADJ pin with reference to GND is

Vadj = 48Vx9100 / (9100+240) = 46.8V

Therefore, input voltage applied to LM317 is equal

Vin – Vadj = 67.7V – 46.8V = 20.9V.

[km6xz]

Given the repeated symptoms I would check whether any of the input pc board's 6.8k phantom pull up resistors is shorted. They are tiny surface mount parts so if any input was encountered that had a higher potential across it, such as a XLR input connected instrument that had been unbalanced, and the chassis of the ungrounded offending unit was hot could have caused the original problem. That happens all the time in little bars and casual venues were ground lifts are used or where extension cords are strewn across the stage. Instruments need isolating DI's before safely attached to any mixer that you care about.

[Jazz P Bass]

I read the datasheet.
The V out formula is:
V out= 1.25 (1+ R2/R1)
R2= 9K
R1= 240
V out = 48 Volts.

[km6xz]

The regulator is probably working fine, but something is pulling high current if the regulator is getting temperature discoloration. There is nothing in the circuit that would only be able to pull current in operation, yet not on the bench...(that eliminates the bypasses and ribbon cables). The pull up resistors are phantom power standard of 6.8k which can't allow enough current to notice. Add one shorted pull up chip resistor and all that changes, the regulator out put current is limited only by the external circuit that is plugged in. If the XLR used is connected as unbalanced, there it the short directly on the 317 output. If a dynamic mic, the mic itself might have been damaged long before regulator.
What other scenarios would allow high current only while in operation in a venue?

I notice a lot of attempts to re-engineer designs that have worked for years, and change the circuit, such as putting in a LM350 or changing grounding when, clearly, there is a defect. Why? The circuit conditions were not looked at any further after a damaged part was found. That should have been the beginning of the search, not the end.

[Enzo]

This is phantom power, not light stalks, it doesn't need more than a 1.5A regulator.

They needed to drop 20v, well within 33v or whatever spec., the voltage to ground is irrelevant, no part of the regulator is grounded, it never sees 70v. You don;t need to alter the circuit or re-engineer anything.

If you fix the power supply and get 48v on the phantom lines, then I'd call the power supply fixed. I'd wager on a problem outside the mixer.

You fixed it and ran the board for 20 hours, then as soon as you hit the gig, iot blew the circuit. SO what was different at the gig? What was plugged into the board?

[mozwell]

I agree, the issue seems to be outside your mixing board. You could fit a 100 ohm 1/2W resistor in series with D8.
This way if there is an external short, the resistor will burn out, and leave your 48V supply intact. Otherwise, a 0.5A fuse instead of the resistor may help.
It looks like each mic channel has phantom permanently connected. Each channel has a 1k resistor from 48V, which then feeds through two 6k resistors to the XLR mic input.
Check each XLR mic input (with the desk powered up & no inputs connected) is there 48V between pins 2 & 1 & between 3 & 1 of each XLR

[km6xz]

48v would be expected if the phantom decoupling was working or not. A better test would be to make that check with a 6.8k resistor between 1 and 2, and 1 and 2. If the decoupling is correct they potential across the temporary load will be 24 volts or in the range, and it an external problem was blowing the 1.5 amp regulator the drop across the load would be 48volts. In this latter case, it explains the regulator problem and also says that there is a low resistance path around the 6.8k pull up resistors in the phantom decoupling circuit. That would mean a problem in the jack board and possibly, but not necessarily, a active load on the XLR jack presented by an external circuit. If there is a low resistance path around those 6.8k resistors any unbalanced jack plugged in would cause the regulator to over heat. These are chip film resistors on the input board, look for solder bridges as the most likely cause. The 6.8k test load will spot that for you very quickly.
The problem has to account for the facts at hand....signs of over current operation of the regulator, 6.8k determining peak current flow, and something in the remote location that triggers the problem. An unbalanced plug inserted will not draw any more than 7ma as the industry standard for phantom power specifies. A transformer coupled dynamic mic presents that to the board all the time without damage to either. So obviously the phantom supply being able to source high current is a problem in the board, possibly that is not seen until an unbalanced connection in made in the venue.

[curfewboy]

i would like to thank the op and everyone who replied as i just sold my old board to get one of these from a church and it has the same exact issue. it tested at about 76 volts. i have orderd the lm317t so hopefully it will be resolved otherwise i crack open the case on the board itself. channel 11 is also showing signal but not sending to aux or sub/main busses.