How Do We Like this?

[The Ballzz]

Hi Folks,
Just curious as to everyone else's take on Modulus' "0" volt common grounding arrangement? Note that the only two connections to the chassis are the AC Mains ground lug and the termination lug near the input jacks. I know what I think, but want to hear what others think.
Thanks Folks,
Gene

 

[2L man]

That is simple method and work great! There are two shielded wire shields soldered to pot cases but that is fine when only one end is connect and no current can flow! Also shielded wire from input jack to tube socket and 68k resistor on socket end improve signal/noise ratio.

 

[Lynxtrap]

That is how all amps should be grounded IMO, including placing and grounding the filter caps locally with the tubes they supply.

 

[dan40]

I recently finished up a small 5f1 type circuit and decided to try this grounding method for the first time. To my surprise, the amp is dead quiet. I was in a hurry on this one so I didn't bother twisting the filament wires either but the overall noise floor on this little single ended amp turned out extremely low. My normal method is the split ground buss like so many of us here use but I think I'm going to try this method in more of my builds.

 

[The Ballzz]

Yes! I like the way the pot bodies actually get connected to the chassis ground, and the shields get grounded thru those pot bodies and therefore the chassis, in that manner, but none of the actual circuitry depends on the chassis for it's ground. In this case (pun sorta intended) the chassis, pot casings and wire shiald are just that: SHIELDS!
Thanks For The Comments Folks,
Gene

 

[Lowerleftcoast]

I know what I think, but want to hear what others think.

I am an advocate of a single reference to ground. There is no need to have ground current running through the chassis.
I also promote grouping each node together, much like Blencowe describes in his grounding article. This Modulus layout does not keep the nodes together. Below I have changed the layout to keep the nodes together betterer (or more like Blencowe describes them).

Besides the grounding...
This circuit has a cathode follower, so I would create a voltage divider at/near the reservoir cap and elevate the heater CT. The addition of another cap to filter the elevated supply may be desired. A pot can also be added as a humdinger if one so chooses.
On this layout, the depiction of the heater wires is not optimal, imo. I would route them differently. OMMV.

Some of the shock brothers describe all of these *extra efforts* as angels dancing on the head of a pin. Good layout of parts and wiring can result in quiet builds regardless of whether current flows in the chassis.

 

[mcentee2]

I am an advocate of a single reference to ground. There is no need to have ground current running through the chassis.
I also promote grouping each node together, much like Blencowe describes in his grounding article. This Modulus layout does not keep the nodes together. Below I have changed the layout to keep the nodes together betterer (or more like Blencowe describes them).

Besides the grounding...
This circuit has a cathode follower, so I would create a voltage divider at/near the reservoir cap and elevate the heater CT. The addition of another cap to filter the elevated supply may be desired. A pot can also be added as a humdinger if one so chooses.
On this layout, the depiction of the heater wires is not optimal, imo. I would route them differently. OMMV.

Some of the shock brothers describe all of these *extra efforts* as angels dancing on the head of a pin. Good layout of parts and wiring can result in quiet builds.

View attachment 936795

Could you point me to where I can find more info on the heater elevation voltage divider set up please - this is new to me and looks interesting, as I am starting a JTM50 build!

Edit: DOH! Of course, here it is:

Amp Mods

Generic Guitar Tube Amplifier Modifications

robrobinette.com

 

[The Ballzz]

I am an advocate of a single reference to ground. There is no need to have ground current running through the chassis.
I also promote grouping each node together, much like Blencowe describes in his grounding article. This Modulus layout does not keep the nodes together. Below I have changed the layout to keep the nodes together betterer (or more like Blencowe describes them).

Besides the grounding...
This circuit has a cathode follower, so I would create a voltage divider at/near the reservoir cap and elevate the heater CT. The addition of another cap to filter the elevated supply may be desired. A pot can also be added as a humdinger if one so chooses.
On this layout, the depiction of the heater wires is not optimal, imo. I would route them differently. OMMV.

Some of the shock brothers describe all of these *extra efforts* as angels dancing on the head of a pin. Good layout of parts and wiring can result in quiet builds.

View attachment 936795

While I wholeheartedly agree, in principle with your changes, the drawing requires five leads connected at that point "A" filter cap lug, which could become a bit cumbersome. Of course, elevating that filament center tap would remove one of those five wires from that "A" lug. And everything I build (at least so far) is cathode bias, so the bias feed ground is less consequential. Although, maybe the cathode cap/resistor ground could be better served by going separately to at least the "0" volt connection between the first two filter caps

While I have no problem with the speacker jack ground connected as Modulus shows, they should have specified the wire from "A" to that juction to be #18 gauge, instead of the #20 or #22 that typically comes in a kit for hookup wire.

Either way we slice it, the benefit I see here is the isolation of all circuitry "0" volt commons from the chassis, except at that one point near the input jacks. It does strike me as logical to basically "float" the whole cicuit with only one connection (through the chassis) to mains earth/ground for safety and shielding purposes, and not depend on the chassis for any current carrying, except in a fault condition! Given that scenario, it seems that the main "0" volt common to that buss at the top of the board, as well as the buss itself should be at least close to the same gauge wire as the mains feed. The last thing wanted in a catastrophic "dead short" condition is the safety ground wire becoming a fuse! Then again, proper fusing of the mains and HT provides an extra layer of protection.

Thanks for your thoughts and opinions on this.
Gene

 

[Lowerleftcoast]

the heater elevation voltage divider set up

As pointed out on Rob's site, the divider will act as a filter cap *bleeder* as well as a voltage to elevate the heaters. The total resistance of the two resistors will determine how fast the caps bleed off their stored voltage. Ime, 220k to 330k will bleed the caps in a reasonably short time. 470k and higher will work but the time to bleed the caps will be longer. Less than 220k creates more heat inside the chassis which is unnecessary.

FYI... a voltage divider/bleeder does use some power, it is usually not a problem. If you happen to be using an under-powered power transformer, the added current draw will tax the PT. Just something to be mindful of when making choices.

 

[Lowerleftcoast]

While I wholeheartedly agree, in principle with your changes, the drawing requires five leads connected at that point "A" filter cap lug, which could become a bit cumbersome.

Yes, but there are many workarounds for the *5 on one terminal* conundrum. The speaker ground could attach on the power tube pin1. The bias ground and the presence pot ground could share a terminal before point *A*. An individual B+3 cap could be mounted on the board rather than have two caps in one can. etc. My vision would have a terminal strip nearby where the elevated CT voltage divider resided along with a couple of the 4 remaining *A* terminal wires (just to keep things neat), but I do like the idea of another cap on the board. hah.

they should have specified the wire from "A" to that juction to be #18 gauge, instead of the #20 or #22 that typically comes is a kit for hookup wire.

In a catastrophic event wouldn't the OT fail with any of those sizes? Under normal operation there is very little current running in that wire.

 

[King Fan]

While I have no problem with the speacker jack ground connected as Modulus shows, they should have specified the wire from "A" to that juction to be #18 gauge, instead of the #20 or #22 that typically comes is a kit for hookup wire.

In a catastrophic event wouldn't the OT fail with any of those sizes? Under normal operation there is very little current running in that wire.

I'm not sure which wire we're talking about here, but FWIW here's Merlin in his Grounding chapter (my emphasis): "The secondary side of the output transformer (if one is used) should always be wired directly to the speaker jack using heavy- gauge wire. This is true no matter what ground scheme is used. A separate wire (which does not need to be heavy gauge) should then run from the negative connection of the speaker jack back to an appropriate star."

 

[Lowerleftcoast]

here's Merlin in his Grounding chapter (my emphasis): "The secondary side of the output transformer (if one is used) should always be wired directly to the speaker jack using heavy- gauge wire. This is true no matter what ground scheme is used.

Essentially this wire, Merlin is writing about, is the speaker wire. Since the speaker signal is AC, much of the current does not use the whole cross section of the wire. It travels more towards the outside of the wire. This is known as *skin effect*. The higher the frequency, the more skin effect. Larger gauge wire is suggested for this reason.

A separate wire (which does not need to be heavy gauge) should then run from the negative connection of the speaker jack back to an appropriate star.

This wire is the wire we are discussing in posts 8 & 10 above. In normal operation, it handles a small DC current. In unusual events, (like a catastrophic event), it may carry more current.

 

[King Fan]

Right. I was pointing out Merlin agreed with you. Or is it the other way around?

 

[The Ballzz]

@Lowerleftcoast , @King Fan & All,
Upon a closer look at that Modulus layout, I must embarrassingly admit to being at least partly "full of POO" and that some of my whining and ranting has been a moot point! I now see that the "0" volt common of the output transformer secondary (apparently orange in this case) is indeed shown as connecting directly to the speaker output jacks and then connected to that "0" volt buss, negating the need for the heavier wire that I previously alluded to. That does more closely resemble Merlin's suggestions. I totally missed that "orange" lead in the diagram!

And as for the catastrophic failure I spoke of, I was simply referring to a nearly apocalyptic catastrophe, which is what many safety regulation attempt to guard against

I Need To Pay Better Attention!
Gene

 

[King Fan]

@Lowerleftcoast , @King Fan &All,
Upon a closer look at that Modulus layout, I must embarrassingly admit to be at least partly "full of POO" and that some of my whining and ranting has been a moot point! I now see that the "0" volt common of the output transformer secondary (apparently orange in this case) is indeed shown as connecting directly to the speaker output jacks and then connected to that "0" volt buss, negating the need for the heavier wire that I previously alluded to. That does more closely resemble Merlin's suggestions. I totally missed that "orange" lead in the diagram

I Need To Pay Better Attention!
Gene

No, no, you were way ahead of me, at any rate.

I'm one of those guys who believe in single ground anchors -- except in the common small simple well-tested amps we tend to build here, where their theoretical advantages don't seem audible. By the time an amp gets this complex, better grounding is, um, better...

 

[mcentee2]

I am an advocate of a single reference to ground. There is no need to have ground current running through the chassis.
I also promote grouping each node together, much like Blencowe describes in his grounding article. This Modulus layout does not keep the nodes together. Below I have changed the layout to keep the nodes together betterer (or more like Blencowe describes them).

Besides the grounding...
This circuit has a cathode follower, so I would create a voltage divider at/near the reservoir cap and elevate the heater CT. The addition of another cap to filter the elevated supply may be desired. A pot can also be added as a humdinger if one so chooses.
On this layout, the depiction of the heater wires is not optimal, imo. I would route them differently. OMMV.

Some of the shock brothers describe all of these *extra efforts* as angels dancing on the head of a pin. Good layout of parts and wiring can result in quiet builds regardless of whether current flows in the chassis.

View attachment 936795

I notice you moved the speaker 0v/ground from to directly connected to cap can point A, and away from its board turret which is connected A anyway.

What's the reason for this move, as there no other ground destined connections made on that wire between A and the turret board where the speaker wire originally connects ?

 

[schmee]

"Note that the only two connections to the chassis are the AC Mains ground lug and the termination lug near the input jacks. "
That's the method that generally works. So I like it fine.

 

[Lowerleftcoast]

I notice you moved the speaker 0v/ground from to directly connected to cap can point A, and away from its board turret which is connected A anyway. What's the reason for this move, as there no other ground destined connections made on that wire between A and the turret board where the speaker wire originally connects ?

I agree the proposed termination of the *speaker wire ground* is electrically the same place as on the Modulus layout.

The reason for the move is to group the *speaker wire ground* and the PI ground together at their node. It also makes for shorter wire runs. The Valve Wizard grounding article describes it as,
"If global feedback is used then the speaker ground should be returned to the local star of whichever stage the feedback happens to be applied to, which is usually the phase inverter (e.g., fig. 15.14)".

The wire running from *A* to the next filter cap, on the board, can be viewed as a ground bus. I have noticed, just like this Modulus layout, many layout designs will attach ground wires indiscriminately to a ground bus. If a true star ground scheme is to be implemented, the ground wires will be grouped into nodes (stars).

The ground wire is at *0* volts but it is not quiescent. There is current running in the wire. All wire has *some* resistance, so the placement of the several grounds and nodes can make a difference.

If you notice, I also moved the ground wire from the master vol to the node it is associated with.

(To those of you who view this theory as angels dancing... Sorry for the minutiae and being didactic.)

 

[mcentee2]

I agree the proposed termination of the *speaker wire ground* is electrically the same place as on the Modulus layout.

The reason for the move is to group the *speaker wire ground* and the PI ground together at their node. It also makes for shorter wire runs. The Valve Wizard grounding article describes it as,
"If global feedback is used then the speaker ground should be returned to the local star of whichever stage the feedback happens to be applied to, which is usually the phase inverter (e.g., fig. 15.14)".

The wire running from *A* to the next filter cap, on the board, can be viewed as a ground bus. I have noticed, just like this Modulus layout, many layout designs will attach ground wires indiscriminately to a ground bus. If a true star ground scheme is to be implemented, the ground wires will be grouped into nodes (stars).

The ground wire is at *0* volts but it is not quiescent. There is current running in the wire. All wire has *some* resistance, so the placement of the several grounds and nodes can make a difference.

If you notice, I also moved the ground wire from the master vol to the node it is associated with.

(To those of you who view this theory as angels dancing... Sorry for the minutiae and being didactic.)

Ah, ok! I hadn't spotted the star for the speaker was at the PI node, dual cap cans always confuse me

I am building a Modulus jtm50 (with a single point ground scheme like the jtm45 layout below) where the speaker ground comes all the way across to join the presence pot ground wire at the pot lug then to the separate cap can node for the PI that sits off the side of the chassis then to the preamp cap node on the board.

Phew

 

[2L man]

I have used a shielded cable for NFB and used its shield for ground reference soldering its both ends, other to loudspeaker jack and other to ground bus to the stage where NFB hot goes.

Although there flow high current signal between OT secondary and loudspeaker the current in NFB is very small.