Phantom Power squeal being generated on mic outputs of EZ1290s

[corgan4321]

Hi all,

I have two pairs of EZ1290s "Neve" preamps in separate rack chassis being powered by a Vintech PSU. I used to have two separate different power supplies powering each of my EZ1290 pairs, but I switched recently to the Vintech.
I noticed today there's a very audible squeal when I have more than one of the phantom switches on. The frequency of the squeal tone gets lowered with each switch that gets turned on - take a look at the screen shots.

These are viewing the output of one microphone. Is this due to inadequate filtering of the phantom rails in the Vintech? I did modify the Vintech PSU to ensure I had one star ground with a 10ohm resistor going from the audio 0v reference to the chassis ground per the EZ1290 instructions. Maybe the specific regulators in the Vintech doesn't like this connection?

Any help would be much appreciated.

-Will

 

[Bo Deadly]

So does the other PS result in squeal?

I would first pull everything from the rack, physically move the whole rig to a different place, plug into a different mains outlet and measure again. Try just one, the other, then both. Just mains > PS > cable > pre > audio interface. Repeat with a different mic cable. The point being that if you change as much as you reasonably can about the environment you reduce the number of possible causes.

Sounds like it could be a ground issue. Is the 10R between phantom ground and supply?

With nothing but a PS > cable > pre, open the PS (let the caps drain before poking around) and measure resistance between pin 1 of the XLR in of the pre and the 0V at the down-stream filter cap in the supply PCB (incidentally is this your "star" point?). What does your meter measure?

Note that when measuring low resistances it's important to fiddle with making good contact. You might need to stab at the XLR pin a bit and fiddle until the reading settles on the lowest value.

Ultimately I don't think you want the 10R in there. Is there also a cap across it? Not sure why they would instruct you to do that. It could be that there's supposed to be another path between the chassis of the pre and 0V that does not have 10R and something is miswired.

More generally there are a number of potential failure contacts between pin 1 of the XLR and the 0V filter cap. You have all sorts of cable contacts and panel screws and such. Something could be a little corroded. It could be as simple as you need to just re-tighten some screws to improve the phantom return path.

 

[corgan4321]

Thanks for your reply Bo Deadly. I did some testing and found that if I turned the "LIMIT" mounted pot in the 48v circuit the squealing sound changed. The sound is intermittent so I've decided to just turn the current limiter all the way one way and see if it ever comes back. If so, maybe I'll try turning it the other way I wonder if the current limiting circuit here gets upset with such a small draw. The supply has .5A max in the 48V module and with 4x condensers I am only drawing 30-40mA max probably...

By the way, for anybody's future reference, these Vintech PSUs are using one each of these modules:

https://internationalpower.com/wp-content/uploads/2018/01/IHB24-1.2.pdf
https://internationalpower.com/wp-content/uploads/2018/01/IHB48-0.5.pdf

 

[corgan4321]

Ultimately I don't think you want the 10R in there. Is there also a cap across it? Not sure why they would instruct you to do that. It could be that there's supposed to be another path between the chassis of the pre and 0V that does not have 10R and something is miswired.

Yes, the 48V supply 0V reference is also grounded to the chassis through the 10R resistor. Maybe the 24V should be referencing ground through the 10R but not the phantom supply? I'm not sure I know why Neve did it that way, but that is how they did it.

 

[Bo Deadly]

Actually it sounds like pin 1 of the XLR is just not connected to the chassis at all and that pin 1 is connected directly to 0V. Usually pin 1 of the XLR is connected to the chassis at the input to shunt RF and then 48V returns through the chassis. But Neve is an older design and being pluggable modules it's kinda hard to use the chassis as part of the circuit. So the chassis is only connected to 0V through the 10R. Is that right? Or is pin 1 connected to the chassis as well?

If you could answer my question, it might be clear. Do you not have a DMM? What is the exact resistance reported by the DMM between pin 1 of the XLR and the 0V point on the output of the PS?