Dual Pultec Build Ground Noise

[RoadrunnerOZ]

The terminal block can be fastened to the chassis with either a bolt or double sided mirror adhesive tape.

 

[Janalex]

Awesome thanks! So Ill need 6 jumpers to get 12 V out to 3 boards..?

And your circuit board explanation is the exact same wiring as the schematic just with different cap values..? Just want to be 100@ clear

 

[RoadrunnerOZ]

Awesome thanks! So Ill need 6 jumpers to get 12 V out to 3 boards..?

And your circuit board explanation is the exact same wiring as the schematic just with different cap values..? Just want to be 100@ clear

Yes - the 7806 needs no other components.
Those terminal blocks easily take 2 wires in each post (you can buy different gauges - 2mm sq, 2.5mm sq, 4mm sq and so on) so a four way is all you really need. They are sold at electrical or electronics stores. You can buy a long one and cut it to the length you want. They come with mounting holes as well that take a reasonable sized bolt.

 

[RoadrunnerOZ]

Also as there is little current drawn in the off-board LED and other ancillary components you won’t need a heat sink on the 7806 - those regulators can get pretty hot before you start needing. Check your voltage and polarity before connecting.
How do you identify which of the red twisted pair is + and which is ground for those DC runs to the LED’s etc?

 

[ccaudle]

I connected the jumper between 6.3V pin on the tube and the unbalanced input hot.

You put 6V DC on the input to the audio circuit? I don't follow what you were attempting to do.

Do you have a schematic for this bypass board you have mentioned? Relay coils are isolated, so it seems like there should be no reason for a bypass to introduce a power connection anywhere else (although one end of the coil could certainly be tied in to a common circuit node on the board, and that could be a potential problem source).

do I need more capacitance after the rectifier to get actual DC first?

If you are referring to that 330nF capacitor in front of the regulator, then yes, you need more capacitance. I don't know what kind of current you need, probably 100uF would be a good starting point for low current, but you should calculate based on current draw.
Oh, I see as I was typing this RoadRunner recommended 1000uF. That works, should be around the same size and cost and will handle more current if you need it.

The last thing I will be testing is if any noise appears after I connect in the red twisted pair seen in my initial pics that goes to the unbalanced input on the srpp boards.

That is another thing which does not make sense in the original design. The one big advantage which transformers have is ultra high common mode impedance, which means that transformers can reject common mode noise even when there is a fairly large imbalance in source impedance between each leg, such as from an unbalanced output. Having a separate unbalanced input which bypasses the input transformer is taking the single thing which can make the biggest improvement in noise rejection on an unbalanced input and throwing it away. It is exactly the opposite of what would make sense.
If you want to connect an unbalanced output to this EQ just make an adapter cable and use the input transformer for what it does best.

 

[RoadrunnerOZ]

You put 6V DC on the input to the audio circuit? I don't follow what you were attempting to do.

Do you have a schematic for this bypass board you have mentioned? Relay coils are isolated, so it seems like there should be no reason for a bypass to introduce a power connection anywhere else (although one end of the coil could certainly be tied in to a common circuit node on the board, and that could be a potential problem source).



If you are referring to that 330nF capacitor in front of the regulator, then yes, you need more capacitance. I don't know what kind of current you need, probably 100uF would be a good starting point for low current, but you should calculate based on current draw.
Oh, I see as I was typing this RoadRunner recommended 1000uF. That works, should be around the same size and cost and will handle more current if you need it.



That is another thing which does not make sense in the original design. The one big advantage which transformers have is ultra high common mode impedance, which means that transformers can reject common mode noise even when there is a fairly large imbalance in source impedance between each leg, such as from an unbalanced output. Having a separate unbalanced input which bypasses the input transformer is taking the single thing which can make the biggest improvement in noise rejection on an unbalanced input and throwing it away. It is exactly the opposite of what would make sense.
If you want to connect an unbalanced output to this EQ just make an adapter cable and use the input transformer for what it does best.

On the post where Janalex said they used the unbalanced input he showed a pic where they cut the track from the + input - I’m assuming it’s to be able to use the unbalanced terminal post unconnected to audio as a distribution point for the 6.3V DC.

 

[Janalex]

Probably by unsoldering and continuity testing to see which is plus and minus

And yes correct just used those unbalanced as posts but still introduced noise

 

[RoadrunnerOZ]

Best get all but the audio and the tube supplies off those audio boards.

 

[Janalex]

Thanks I'm ordering the parts for the rectifier board now.

 

[RoadrunnerOZ]

All good

 

[Janalex]

Bypass board 1/2. Gonna still have to tighten up those shells

 

[RoadrunnerOZ]

This is where it’s possible the XLR ground is coupled to the relay coil ground - you can check on the bypass board with a meter - this may have been an additional contributor to ground loops.
The ground back to the audio boards needs to be lifted from the audio terminal block on the bypass boards anyway so that the XLR shells and pin 1 are only chassis grounded here.
Your new DC supply for the relays should be grounded to the chassis at your star ground.
It looks as though your shells are not latched to the XLR bodies still - there should be a little concealed screw that rotates the latch at the front of the XLR.

 

[Janalex]

These xlrs were also not used during testing as the boards were not powered. My plans is to remove the 4 green wires coming off the bypass boards going to adjacent chassis xlr ground tabs. To only hook up pins 2 and 3 to the terminal blocks and solder the shield to each grounding tab adjacent to xlr. Pin 1 is jumped to shell already.

The entire led/bypass circuit will get one wire coming out of the negative of the new rectifier board going to my star ground point where 0 vac also attaches.

Got all the parts from digikey including the terminal barrier priority mail.

 

[RoadrunnerOZ]

The only other thing to do is keep the HT and Heater AC cabling well away from the near SRPP board until they get terminated at the connecting posts on the far board.

 

[Janalex]

Right it’s hovering well over the top

 

[RoadrunnerOZ]

Right it’s hovering well over the top

Probably radiating straight into the tube on the first board

 

[Janalex]

Gonna try this again but I may need to use new wires cause it will be longer coming from the terminal barrier. Is there a specific wire choice brand that would make a difference in terms of carrying AC and hum? What do you recommend ?

 

[RoadrunnerOZ]

2 core armoured/shielded cable rated to 250 to 300V - like used in thermostat and control switch wiring or you could go Belden 19364 - 3 core shielded mains cable not using the 3rd core but just 2 cores and the screen. There’s not much current draw so the 1.5mm sq Belden may be a bit of overkill, 20 or 18 AWG 2 core shielded would be fine as long as it’s rated for 250V for your HT AC. These cables are not so easy to source but you could try your local electrical wholesaler.
Second using existing type cable twisted pair, you could just use a short length of copper tube, inside edges of the ends smoothed to prevent cutting through the wire, grounded at one end with the cable running inside - small gauge tubing is available from hobby stores or hardware stores - mounted on cable tie mounts. I have used this method in various keyboards like Clavinets for stage use where radiated dimmer noise from stage lights breaks into audio and DC wiring.
Third option is conductive adhesive copper foil tape (I use this for shielding guitars and synth top cases - the adhesive is conductive as opposed to the non-conductive type making for a better enclosed shield). You just solder a ground wire to it at one end and if you wish to isolate it from shorting out anything, just slide a length of heatshrink over the cable after taping then bend into shape for the cable path - work out your bends and use a little heat to help the bend - the heatshrink helps stop the tape tearing at the outside of the elbow. Or forget the heatshrink, prebend the elbows and tape afterwards. Then mount it to cable tie mounts stuck to your chassis away from the boards.

 

[Janalex]

So the current AC twisted pairs are not shielded. Just 2 solid cores wires. Sounds like the shielding could be very helpful. Where do I solder the shield to on these?

 

[RoadrunnerOZ]

Chassis ground where the mains earth is connected.