Seperate Power Supply Chassis and Grounding

[lordglum]

Hello,
So I just finished rebuilding an amp that I built about 5 years ago. Its a Time bandit 2A3 more or less based on the 'simple' design. I wanted to use it as a headphone amp but the AC hum was pretty bad.

So I decided to move the power supply, rectifier tube, chokes into a separate chassis. I run 2 x 2.5 filament supply, 6.3V supply, 2x 330V and ground through the umbilical cord.

After some troubleshooting I finally got it working again tonight (right channel 4 pin tub socket pins were not contacting the 2a3).

My only issue is after adjusting the hum pots I still have hum in my grados.

The method I used to ground everything was follows:

I used bare copper wire in the power supply chassis and ground my IEC ground and every other ground to this buss bar. I then took a lead from the buss bar and put that into the "amp" section. From there I grounded everything to another bar in the amp section and then tied this to the "amp" chassis near the rca jacks.

My question is there a flaw in the logic I'm not seeing? Anybody have past experience on how to knock down a slight AC hum? On my Hawthornes its darn near silent, but in my ears not yet perfect. 🙂 I'm gonna play with the grounding a bit this weekend I was looking for some ideas to try.

Maybe shielded cables for the filament supplies?
Maybe I have some induced noise on the ground from running the AC with it from chassis to chassis?

Thanks for the help.
Jeremy

I have pics but I need to snag them off my phone. The schematic is attached.

PS sorry for the wall of text.

 

[Michael Bean]

Instead of using a ground "buss bar", use a star ground to establish a single "0 volt" common connecting point. A buss bar, no matter how large, will still develop voltage drops along its length when current flows.

Mike

 

[Richard Ellis]

My 12B4 under construction will be using the aluminum chassis plate at 5mm for a ground plane. To wit, the incoming green ground will be centrally located & grounds for signal will be on it.

_____________________________________________________Rick.........

 

[Conrad Hoffman]

Do you have anything with a transformer within about a foot of the cartridge? It may be unusually good at picking up magnetic fields, so shielding and fooling with grounds won't help. My preamp has a nice toroid, but if it's in the rack under the turntable, I get hum. Distance is the easy cure.

 

[rhone]

If I understand correctly there is flaw. If you regard your two chassis as being one amp then you only want the circuit to be connected to chassis ground at ONE point. So the circuit ground doesn't get connected to the chassis in the amp section. Pass the circuit ground through the umbilical to the buss in the PS section, where you have your chassis ground.
You MUST also have a separate "chassis ground/earth" wire in your umbilical that connects ROBUSTLY to each chassis (to create one chassis) and then the PS chassis is connected to the IEC earth. Now both chassis have a safety earth and the circuit has one ground point.

 

[Vinylsavor]

Hi,

try to isolate the hum source check this:

VinylSavor: Trouble Shooting Hum

Looking at the schematic, I would see two possible sources:

not enough B+ smoothing
AC filaments might be an issue with head phones

Thomas

 

[lordglum]

Thanks for the reply's 🙂 A bit more than I expected.

I have read many articles in favor of a star grounds as those that are opposed. Right now the amp is not set up for star ground.

Thanks for the hum troubleshooting link started to look through it this morning.

I have a couple general questions...

Shorting the inputs to see if hum goes away...I assume short them to ground?

It sounds like I need to ground the PS chassis on the top plate as well correct?

The 2.5V filaments center tap are not grounded. I am not sure why not?

Thanks
Jeremy

 

[mark02131]

The 2.5V filaments center tap are not grounded. I am not sure why not?

I believe the center tap from the filaments of a DHT has signal in it. If you were not using hum pots you could use that connection to your cathode resister/cap. Whats the hum look like on a scope, 60hz?

I was looking at your power supply, Check to make sure the HT is completly smooth, also your hum may be coming from the AC on the heaters at ether or both stages. If you have a bench supply you could try some DC on the 6SL7.

 

[Vinylsavor]

Hi!

Shorting the inputs to see if hum goes away...I assume short them to ground?

Yes

The 2.5V filaments center tap are not grounded. I am not sure why not?

Grounding them would short out the cathode resistor which sets the op point of the output tube. Don't ground them

Thomas

 

[DF96]

The method I used to ground everything was follows:

I used bare copper wire in the power supply chassis and ground my IEC ground and every other ground to this buss bar. I then took a lead from the buss bar and put that into the "amp" section. From there I grounded everything to another bar in the amp section and then tied this to the "amp" chassis near the rca jacks.

My question is there a flaw in the logic I'm not seeing?

A bus ground can work fine if everything connects in the right order. You haven't told us what order you used so we can't comment on it. All you have said is that you have two busses (one in PSU, one in amp) and they are connected together.

You may have mixed up safety ground and signal ground. The IEC ground must go the chassis, not the bus. There must be a chassis ground connection between the two boxes in addition to the signal ground. Then somewhere you have one and only one connection between chassis and signal ground.

 

[lordglum]

Ok. So I'm learning on the fly so humor me. I appreciate the help. I have the ps chassis apart. On the bar I grounded the iec, then 300v center tap, then 6.3 tap then the umbilical cord that goes to the amp chassis.

So i need to add a chassis ground and get the iec off signal ground to PS chassis ground?

I'm a bit confused when both chassis need to be grounded but then I have 2 paths to ground? I can't have only 1 ground from chassis to signal if both chassis need grounded.

I will get the amp section grounding order and pics later.

Thanks for the help!

 

[H E Pennypacker]

Also as Mark mentioned, are you certain it is mains hum?

Does it still hum while using a battery powered MP3 player to feed the preamp?

 

[mark02131]

So with ground, they are right you need a chassis ground connection on both boxes connected together. For what I call Audio ground, In the power supply box I would take a connection off of the chassis ground through a resister/cap filter to create audio ground. This is the connection that would not be connected to chassis ground anywhere else but the one connection off the chassis. This will isolate noise off the chassis ground to the rest of your circuit.

I do wonder if it's a ground or just a flow coming off the heaters or some other inductor induced hum. From what I've seen with low powered DHT's and AC heaters there will always be a little bit of hum. If you don't want that DC will help you get there.

 

[lordglum]

So 1 ground that ties the chassis together and 1 ground only for audio. Tie the iec ground to the chassis grounds.

Possibly look into filter circuit for audio ground.

Got it....I think lol

Thanks for the help. Time to tweak.

 

[DF96]

On the bar I grounded the iec, then 300v center tap, then 6.3 tap then the umbilical cord that goes to the amp chassis.

Where are the cap negatives grounded? It is important that the connection from 300V CT to cap negative does not go through the bus (or to a star, if you had one, which you don't but others do). It should be CT to cap -ve, then from there to the bus or star. This keeps charging pulses away from signal ground.

 

[mark02131]

Where are the cap negatives grounded? It is important that the connection from 300V CT to cap negative does not go through the bus (or to a star, if you had one, which you don't but others do). It should be CT to cap -ve, then from there to the bus or star. This keeps charging pulses away from signal ground.

Good Stuff, So center tap to first cap negative defiantly, what about C-LC-LC-C where should the 2-3-4 cap ground go ? back to the first cap or to the star

 

[DF96]

Best to treat the cap negatives as their own bus. CT to one end. Signal ground to other.

 

[lordglum]

Quick update:

I removed the bar on the ps chassis. 1st filter cap and both ct's now go to the same signal ground in the ps chassis.

I connected the chassis together on their own circuit and brought them to the iec ground. Then there is 1 line that goes to the ps chassis gnd.

There is 1 wire that joins the ps chassis gnd to the signal ground that goes through the umbilical cord.

Hum is less but still present.

 

[lordglum]

1 more thing...shorting the inputs to ground did not do anything. Next step appears to be shoring input tube to gnd.

 

[lordglum]

I think my next changes are to move the ps filter caps off the buss bar, they are near the end of the line and I think they are playing a role. Hum is in much better shape and I am a bit smarter. Wish I had came here before I put it together.

Dumb question if I have a buss ground & 4 ps filter caps on it how important is the order they are attached to the bus bar?

Thanks for the help!

Ps pic attached