Hello friends,
I'm in the layout planing for a tubed amp. There will be input diff amps with a bus bar between the two paired sockets as a local signal ground. It would be nice if the sockets' central spigot could be used as mechanical support for the bus bar.
The general grounding scheme is as follows:
XLR input/1:2 input trans ground > input stage bus bar > input regulator ground > diode bridge/mains PS cap ground > mains earth/chassis bond
The output stage "star" also joins the input star point at the diode bridge/mains PS cap ground. Finally the negative speaker terminal joins the mains earth/chassis bond.
In his book "Building Valve Amplifiers", 2nd ed. (2014), Morgan Jones writes :
If a valve socket has a central spigot it should be bonded to chassis as this reduces capacitance between valve pins.
As you can see, my grounding scheme does bond the spigots to chassis eventually, but indirectly through low impedance wiring. Would that provide the capacitance reduction as well? Does my grounding scheme makes any sense?
Thanks in advance for any insights, and a nice week-end to all.
- Joris
I'm in the layout planing for a tubed amp. There will be input diff amps with a bus bar between the two paired sockets as a local signal ground. It would be nice if the sockets' central spigot could be used as mechanical support for the bus bar.
The general grounding scheme is as follows:
XLR input/1:2 input trans ground > input stage bus bar > input regulator ground > diode bridge/mains PS cap ground > mains earth/chassis bond
The output stage "star" also joins the input star point at the diode bridge/mains PS cap ground. Finally the negative speaker terminal joins the mains earth/chassis bond.
In his book "Building Valve Amplifiers", 2nd ed. (2014), Morgan Jones writes :
If a valve socket has a central spigot it should be bonded to chassis as this reduces capacitance between valve pins.
As you can see, my grounding scheme does bond the spigots to chassis eventually, but indirectly through low impedance wiring. Would that provide the capacitance reduction as well? Does my grounding scheme makes any sense?
Thanks in advance for any insights, and a nice week-end to all.
- Joris
Last edited:
...Depends on circuit configuration and working impedances. Traditionally the McMurdo/phenol bases were worse in every parameter as M Jones rightly mentions.
However, with tube bases, on new amp builds I avoid the older McMurdo/phenol bases and go for ceramic or Belton types.
I get around these pin capacitance issues by designing for lowish working impedances, but one cannot hide individual tube Miller effcet capacitance paramenters that swamp any capacitance differences between the socket pins. I routinely use high gm tubes and strap the screen to the anode, i.e triode mode; then the penalty becomes the internal tube g2-g1 capacitance though triode working impedances are lower than true pentode. WIth a cold 12BY7 I measure 8pF between g1 and g2 with tube so connected. I come across many circuits that use an extremely high value grid input series resistance which becomes ruinous to the high frequency response.
Le bench baron
.... the "round the circuit bus bar" with thicker copper gauge soldered to the central spigot was the the traditional 1960´s support build method for many amps. I use this today but use a much thicker gauge copper, i.e 12AWG or thicker.
However, with tube bases, on new amp builds I avoid the older McMurdo/phenol bases and go for ceramic or Belton types.
I get around these pin capacitance issues by designing for lowish working impedances, but one cannot hide individual tube Miller effcet capacitance paramenters that swamp any capacitance differences between the socket pins. I routinely use high gm tubes and strap the screen to the anode, i.e triode mode; then the penalty becomes the internal tube g2-g1 capacitance though triode working impedances are lower than true pentode. WIth a cold 12BY7 I measure 8pF between g1 and g2 with tube so connected. I come across many circuits that use an extremely high value grid input series resistance which becomes ruinous to the high frequency response.
Le bench baron