RCA BA-2C Buzzing Issue

[mjrippe]

Use your meter to test from pin 1 of the XT-3 to the metal case. It should show continuity. If not there may be a broken wire inside. It seems odd to me that they didn't put the shield on it's own pin but rather tied it to one side of the secondary. Maybe look up some H-K schematics that used these and see if they did anything "different"?

 

[Cranehazard]

Those look like awesome sheilded transformers and should not buzz. Make sure you connect the shielding pin to ground

 

[Cranehazard]

Wow! Saw the schematic, sheilding connect to secondary. Still, try grounding the unused pins one by one. One might be the sheild.

 

[mattrad]

Just tried to ground the remaining pins one by one, no change in the buzzing. Also, hooked up the input transformer on alligator clips and moved it around to see if it had any affect on the buzz- Even with the transformer over a foot away from the chassis the buzz was unchanged. In fact, when i moved it closer to the power supply the buzz wasn't noticeably worse until it was an inch or two away from the PT. So i'm assuming its well shielded.

The case of the XT-3 is plastic so cant do the pin 1 continuity test (assuming the shielding is underneath) but I do have 2 of them and i've tried both with the same result. Trying to look up HK schematics that may use the XT3 or similar Mic transformers, seems like the "Commander" line of mixer amps has the correct 9 pin sockets for Mic Ins but no dice in finding a schematic thus far. Thanks for everyones input!

 

[mjrippe]

The case of the XT-3 is plastic so cant do the pin 1 continuity test (assuming the shielding is underneath) but I do have 2 of them and i've tried both with the same result.

Huh? I have two XT-3 on my bench right now and I can assure you they are not plastic. Dig your meter probe into the grey paint and you will measure zero ohms between the case and Pin 1.

Anyhow, searching for the problem, are any of the pins 5, 6, or 8 attached to ground? They should not be. Also, where is pin 1 attached to ground - at the same point as pin 1 of the XLR or somewhere else?

 

[mattrad]

Ha! Right, metal. Yeah theres is continuity from pin 1 to case. None of 5, 6, or 8 are attached to ground. At this point, yes both XLR pin 1 and XT3 pin one are going to the same ground, i've tried grounding XLR 1 to chassis via the XLR ground lug with the same result. As of now all grounds, except for the green wire of the AC, are grounded to chassis at the same point, the negative lug of of the first filter cap. I figure this would rule out ground loops. All grounds show continuity to the AC earth prong.

 

[radardoug]

Are you sure that transformer is good? Is it really Harmon Kardon? Remember, HK wrecked lots of good companies, maybe they cant make a good transformer!

 

[mjrippe]

Are you sure that transformer is good? Is it really Harmon Kardon? Remember, HK wrecked lots of good companies, maybe they cant make a good transformer!

It is a vintage transformer from the days when tubes ruled. They have tried two of them with the same result.

Harmon has ruined pro audio, but maybe Kardon is still ok?

 

[mjrippe]

Ha! Right, metal. Yeah theres is continuity from pin 1 to case. None of 5, 6, or 8 are attached to ground. At this point, yes both XLR pin 1 and XT3 pin one are going to the same ground, i've tried grounding XLR 1 to chassis via the XLR ground lug with the same result. As of now all grounds, except for the green wire of the AC, are grounded to chassis at the same point, the negative lug of of the first filter cap. I figure this would rule out ground loops. All grounds show continuity to the AC earth prong.

I've been trying to look at the photos to see if anything catches my eye, but my eyes are not so sharp today! I see the shielded wire from transformer pin 9 to the 6SN7, but is the shield connected at the tube end? I can't quite tell. If it is, try connecting it to pin 1 of the transformer instead.

 

[mattrad]

Yes, the shield was connected to a socket lug on the tube end. Just switched it to ground at pin 1 of the XT3, still no change.

Sort of interesting, when the shield is disconnected, when i turn on my soldering iron the buzz gets 3 times louder. I turn it off, back to the normal, constant buzz. With shielded cable, the soldering iron being on or off doesn't add any noise at all but the normal buzz is still constant. If i'm not mistaken this would seem to display that this isn't a shielding issue. If the noise from the soldering iron noise is effectively blocked going into the stage one i would assume other environmental noise generators would be as well. But how else would PSU noise get into the signal?

 

[MisterCMRR]

The noise being a buzz (rather than hum) and repetition frequency of 120 Hz is consistent with "pulsed RFI" created by the reverse recovery (rapid snapping off of conduction after voltage has reversed) of the silicon rectifiers. Most garden-variety rectifiers have no maximum listed for t sub-rr and some are micro-seconds long. It's why you'll often see 10 nF capacitors wired across each rectifier on lots of gear. Because the "snap" creates lots of ultra-sonic ringing in the power transformer winding, it can couple via unusual routes - including radiation. Just be sure the added caps have a voltage rating that's substantially higher than the peak AC voltage involved.

 

[Murdock]

Have you tried changing the tube?
I once had a humming unit and changing the tube got rid of it. I don't know if it was just the reseating because something was loose or if it really was a faulty tube. But worth a try maybe.

 

[inkster]

Follow Ian's grounding scheme carefully and that may reveal a wiring flaw...the age old rule of console and outboard wiring is that balanced shields connect to chassis...unbalanced to 0V, I would check that with the transformer case isolated from all grounds and then clip lead the case to different ground potentials to see which is quieter.
https://drive.google.com/drive/fold...iSDM1N0k?resourcekey=0-9o5zVSCaSCX1VYW-A1v-wQ

 

[mattrad]

The noise being a buzz (rather than hum) and repetition frequency of 120 Hz is consistent with "pulsed RFI" created by the reverse recovery (rapid snapping off of conduction after voltage has reversed) of the silicon rectifiers. Most garden-variety rectifiers have no maximum listed for t sub-rr and some are micro-seconds long. It's why you'll often see 10 nF capacitors wired across each rectifier on lots of gear. Because the "snap" creates lots of ultra-sonic ringing in the power transformer winding, it can couple via unusual routes - including radiation. Just be sure the added caps have a voltage rating that's substantially higher than the peak AC voltage involved.

This put me on the right track! Alligatored up some 10nf capacitors across the rectifier and the buzz is less aggressive. Interestingly, if I move the capacitors around the chassis I can find places where the buzz is gone completely- usually when one of the capacitors is near the middle of the preamp section. Not sure what thats about. I'll solder them in to the PSU and see where i'm at.

 

[MisterCMRR]

Be sure to connect the caps, with short leads, directly across the diodes - the transition time of the voltage the diodes generate essentially makes it RF, so lead length is important. The caps basically slow the turn-off times of the diodes, suppressing the "snap."

 

[Winston OBoogie]

There is a scheme whereby an RC series snubber is used across diodes instead of a lone C. With an RC the Q is damped further.
I don't remember the exact method used and have only done it empirically myself but...

 

[MisterCMRR]

There is a scheme whereby an RC series snubber is used across diodes instead of a lone C. With an RC the Q is damped further.
I don't remember the exact method used and have only done it empirically myself but...

In my experience RC snubbers in power supplies are generally placed across devices that actively switch to an "on" state (like a switching transistor or a TRIAC for example). It's a way to use the capacitor to slow the rise of voltage across the switch but use the resistor to avoid the very high peak current the switch would see when it turns on (and the capacitor is charged to the "off" state voltage). In most cases, I believe lowering the Q of the resonant circuit is incidental. It's mostly about dV/dt at turn-off and di/dt at turn-on. With the rectifier it's a case of limiting the turn-off dV/dt and R isn't a concern because turn-on of a rectifier is always slow - due to the very slow dV/dt of 50 or 60 Hz supply voltage.

 

[Winston OBoogie]

That makes sense Bill.
I'm confusing two separate issues there I think. Diode switching noise vs ringing due to the resonant circuit of the secondary winding and rectifier.

 

[mattrad]

Big thanks to everyone for helping out.

Got the capacitors soldered in- experimented with a few values and ended up with 100nf as it was the quietest. Still a slight PSU hum left over that I need to figure out, interesting that I could cancel the noise completely by moving the caps around on gator clips. But noise is much more manageable now!

Reading that some different diodes may further help me out... does anyone have experience with UF4007s as opposed to 1N4007s in rectifiers?

 

[Cranehazard]

Sound like power supply noise is the issue. Try putting a choke in the power supply in addition to the anode choke, if you haven't already. test the current of the circuit and find a choke as close as possible with at least 5 henries that and a dc filament supply ( voltage doubler plus lm 317 with lots of heatsink) and it should be studio quiet.