UA100 Mystery Hum/Human Antenna

[StarTrucker]

I have a UA100D Preamp I built from scratch PtP in a Hammond aluminum chassis. When I connect it to an interface without a power cable (UA xlr out, Interface XLR in) it is quiet, but when I plug an IEC power cable into the UA I get a buzz at around 120hz with harmonics, like buzzy static.

Just to be clear, this hum is while the UA is still powered off, but just introducing an IEC cable introduced a buzz. Both the UA and the interface are plugged into the same Furman. With the UA off and an IEC plugged in, I touch the chassis with one hand, the buzz gets louder, and when I power up the unit the buzz becomes 10x louder with a new 60hZ component, but this time touching it makes no difference.

The UA sounds fine with a microphone in the input but there is always this hum/buzz.

Is this a shielding problem or a ground loop? I have gone over it for ground loops but the problem of powered off with the IEC plugged in is stumping me so, TO THE FORUM!

 

[radardoug]

You have no ground. Investigate why.

 

[StarTrucker]

Found that after unplugging the UA from power the hum was still there. With it opened up I disconnected all the wires from the output jack. Attaching any of the three wires causes the unit to hum - OT lead 1, OT lead 2, or Shield (tied to circuit ground).

Also any time I approach the output jack wires with a soldering iron there is an increase in hum volume. The rest of the circuit is earthed through a 3 prong IEC and I idiot checked it over and over again.

What can create a signal with no power flowing around the circuit? The Output transformer is a UTC Ouncer O-6.

 

[radardoug]

All the symptoms tell us that your unit is not grounded. So something is still wrong. Do you have a multimeter?

 

[StarTrucker]

Yes, I checked continuity between my grounds. What should I look for?

 

[radardoug]

Is the chassis connected to the mains earth? Is the B- of the circuit connected to the mains earth? ave you got continuity on the XLR out pins 2 and 3?

 

[StarTrucker]

Yes I have the following grounds:
1. IEC earth direct to chassis
2. Heater center tap to chassis (different point than above)
3. XLR Input pin 1 to chassis
4. XLR output pin 1 to chassis
5. #3 and #4 and tied together and go to the reservoir cap
6. V1 cathodes are grounded to the filter cap for that stage
7. v2 cathodes are grounded to the filter cap for their stage
8. Filter caps are tied together and go to the reservoir cap
9. Volume control goes to the XLR pin 1s (directly to reservoir)

The output of xlr pin 2 and 3 are floating balanced from of the transformer secondary. I tapped off the inverted output and used it for NFB, so no ground on the OT. Primary is single ended.

Thank you again for the brain power you're spending helping me with this!

 

[radardoug]

No ground on the OT? Case and sheild should be grounded, same on input tranny.

 

[ccaudle]

Sounds like a classic case of mixing up shield and circuit reference connections.

IEC earth direct to chassis

That is proper (especially if you do not have a double insulated power transformer).

3. XLR Input pin 1 to chassis
4. XLR output pin 1 to chassis

Correct, but should only connect to chassis. You indicate later that you then take the shield connections and connect them into your audio circuit, which is going to inject shield noise currents directly into your audio. To be avoided.

Heater center tap to chassis

I don't typically deal with tube circuits, but this seems questionable to me. If you have a separate power transformer winding for heater and you want to have the heater voltage centered around 0V, then connecting back to the power supply 0V reference seems like a better place to connect.

#3 and #4 and tied together and go to the reservoir cap

Don't inject shield noise currents directly into your power supply. That is asking for hum. Assuming you have a conductive chassis just tie the shields directly to the chassis, and connect the power supply to the chassis at one point.

Filter caps are tied together and go to the reservoir cap

That is just another way of saying that the local ground/reference nodes at those tubes gets tied back to the power supply 0V point, right? If so that seems correct.

Volume control goes to the XLR pin 1s

Pin 1 is shield. Shields have noise currents flowing on them due to differences in power supply potentials and due to antenna action in the presense of RFI. Nothing in the audio circuit should reference a cable shield directly.

The schematic I have seen for UAD100 is fixed gain, so it is not clear where you have located the volume control. As a general principle you would usually locate a volume pot close to the next buffer stage, and the lower leg would connect to the reference point for that stage.

output of xlr pin 2 and 3 are floating balanced from of the transformer secondary. I tapped off the inverted output and used it for NFB

That isn't going to work properly. A transformer isolates primary and secondary side, so for the secondary side signal to be usable for NFB it has to reference the primary side circuit reference, e.g. one side of the secondary has to be "grounded." That is why the original UAD100 schematic shows an output transformer with a separate winding for NFB, so that the NFB winding can be connected to circuit ground while leaving the output winding floating.

 

[abbey road d enfer]

9. Volume control goes to the XLR pin 1s (directly to reservoir)

The output of xlr pin 2 and 3 are floating balanced from of the transformer secondary. I tapped off the inverted output and used it for NFB, so no ground on the OT. Primary is single ended.

As ccaudle mentioned, tehre are two significant flaws.
The bottom leg of the potentiometer should be connected to "ground", i.e. to a reference point located electrically between the stage that delivers signal and the the stage that receives it.
This could introduce hum/buzz only when the unit is powered.

The output transformer's secondary should be left floating. Any connection to the rest of the circuit makes it unbalanced.

 

[StarTrucker]

Ok, I fixed everything listed by ccaudle. Volume pot replaces the grid leak resistor before the second stage and gets grounded to the first stage filter cap.

Now my suspicion is in the power supply, which is a triad transformer and a full wave rectifier. Its grounds are:

1. IEC Earth/AC ground : Tied directly to chassis on its own screw next to the IEC Jack.

2. Pwr Xfmr Shield : Tied to chassis

3. B+ winding CT, Heater CT, Reservoir Cap and Circuit Negative are all connected at single point, and this is tied to the chassis at point #2 with the power xmfr shield.

Is this right?

 

[abbey road d enfer]

Ok, I fixed everything listed by ccaudle. Volume pot replaces the grid leak resistor before the second stage and gets grounded to the first stage filter cap.

Now my suspicion is in the power supply, which is a triad transformer and a full wave rectifier. Its grounds are:

1. IEC Earth/AC ground : Tied directly to chassis on its own screw next to the IEC Jack.

2. Pwr Xfmr Shield : Tied to chassis

3. B+ winding CT, Heater CT, Reservoir Cap and Circuit Negative are all connected at single point, and this is tied to the chassis at point #2 with the power xmfr shield.

Is this right?

This is all good, but not enough to guarantee quietness. Ground circulation for interstage connections is also very important.
Now it could also be magnetically induced hum. Does roating teh power xfmr changes the level of hum?