The AM Forum

A recent post on the contesting.com "amps" forum points out that it is dangerous to connect power supply b minus to chassis ground.

I built my grounded-grid 813 amp with B minus separated from chassis by a 25 ohm resistor and a .01 6kv capacitor as per advice at the time.

Yet, in my current AM rig build, I have the plate supply B minus at chassis (earth) ground.  Am I misinterpreting the K1JJ/Tesla rig diagrams?

The plate supply being lifted a few ohms from chassis is for grid and plate current metering in a grounded grid amp.

If you don't lift the B- then you can only measure the cathode and plate current in the fil center tap or from the cathode directly.

It's not to do with safety directly, but more to do with safety of the metering circuits.

Does your rig have any metering in direct contact with the HV?  Is it a grounded grid class C amp?

--Shane
KD6VXI

I was pointed here by other forum members:

https://www.w8ji.com/metering_amplifier.htm (https://www.w8ji.com/metering_amplifier.htm)

The concept is also that if the B+ shorts to chassis, the stored energy in the filter caps will blow the meter shunt up, leaving the chassis hot with B+...so create a path that doesn't affect the metering but will pass the fault current from chassis back to the B- on the power supply, hence power diodes and low voltage caps that fail shorted.

Ed

Every piece of equipment in my station, with the exception of a couple modems, has a lug on the back and a piece of braid connecting to a copper ground buss.  This goes to ground via a wide copper strap.  That ground braid is the first thing I connect and the last thing I disconnect.  That way, if high voltage goes astray it has someplace more desirable than I am towards which to make its way.
Having spent my formative years in the broadcast business, I have more than once tripped over 4 inch wide copper strap that wasn't affixed to the floor as tightly as it could have been.  But it was always in evidence.  Recently, I have seen installations where it was lacking, and I wasn't impressed. 
Whether we choose to float the  B-minus or strap it to ground is a circuit choice and both ways have their place, depending on components and design.  But a good connection between equipment and Ground will buy you a few more birthdays.
73 de Norm W1ITT

The W8JI setup works great. If you're running seperate chassis' for the RF deck and power supply be sure to run a connection between the two chassis. Without it, if the shunt resistors open it creates a shock hazard between the two. I know of one ham who didn't and is now a SK because of it.

Whenever I make a "2 piece' amplifier I always make the rf deck a 3 prong plug in the wall for itself.

I then wack a hole in the power supply and install a single gang 20A receptacle.  The receptacle has hot, neutral and ground on it.  This way if the end user uses the power supply receptacle or a house circuit, it should still have a bond on the chassis.

I also stick a 1/4 20 bolt out the back of each unit to ensure they are grounded.

It's super easy to also use a molex connector between units. Have B-, ground and whatever else is needed through the molex.

I make a 4 switch rf deck.
1. Bkower
2. Filament
3. High Voltage
4. Standby (for indirectly heated tubes)

These are all done through a molex.  Hv switching is done via neutral.  In this case, it's impossible to have HV without the molex in.

--Shane
KD6VXI

One thing I did on the two piece amp I just finished building/refurbishing, besides a ground braid and three-wire AC connectors...

I used SHV connectors for B+. If you do not know these, they are similar to a BNC, but the center pin/socket is both recessed and reversed.  Build up a cable using appropriate coax and these connectors.  This means, that if the B+ is connected, so is the shield, which I firmly ground to chassis at both ends...A redundant path even if both chassis get plugged into a cheap plastic two wire extension cord and the braid between chassis is missing and neither is screwed into the rack. In my case, the coaxial relays are also bolted to the metal rack which is tied to station and AC ground directly by a #4 cable, as well as by the metal power distribution (fancy power strips) inside it....

Ed

With a B+  to  grounded grid arc inside a triode. the  fault path is through the chassis, up through your  25 ohm resistor... to B-..then back to the B- of HV  filter caps..completing the loop.   ur  25 ohm resistor will  explode.   These days,  6A10/ 10A10  diodes are wired from chassis  to  B-.   That keeps the B-  within   +/-  .7 V   from chassis potential.    Sometimes  2 x diodes in series are required.  ( +/- 1.4 V).   

IF  the RF deck and  HV supply are in separate  box's,   safety diodes as described above are used in both the  RF deck and the HV supply.    And both cabinets bonded together.   Then  bond  each cab to ur spg ( single point main station ground). 

Diodes are also paralleled to increase  surge capacity in case of a B+ to grid / chassis  arc.    B+  fuse used in B+ leg, then into a  50 ohm  glitch resistor.   IF B+  say arced to grid on a triode, the 50 ohm resistor LIMITS the fault current to a safe value, then < 2 msecs later, the HV  fuse  INTERUPTS the fault current...event over..and nothing blows up.   I also install a 2 nd  HV fuse, between plate  xfmr sec..and input to  FWB / FWD /  FWCT.   No  more shorted diodes in the HV  supply.  Cro bar anything u want, nothing happens..except a blown HV fuse.

Jim   VE7RF