# [Amps] Question about separate power supply/RF section

Jim Garland 4cx250b at miamioh.edu
Sat Jan 19 16:47:03 EST 2013

```No, the B- is only grounded at the RF deck, at the same spot that the
cathodes are grounded. To steer the cathode current properly, it is
important that the B- lead NOT be grounded at the power supply. If it were
grounded, then you are correct that it would merely be a parallel connection
with the other ground wires. However, a single point ground is what is
required.

There will be a negligible voltage difference between the power supply/rf
deck chassis and the B- lead, because of the single point. If the power
supply should become completely disconnected from the RF deck, then the 220
ohm resistor, mentioned in my earlier email, will keep the B- at the same
potential as the power supply ground, within a volt or two.

I don't want to push the analogy too far, but in your house wiring there is
a single point ground between neutral and the bare ground, which is at the
utility entrance. The single point ground keeps the Neutral line at ground
potential, but also keeps any return currents from flowing in the bare
ground wire.  It the same concept we're talking about here.

73,

Jim W8ZR

_____

From: Vic, K2VCO [mailto:k2vco.vic at gmail.com]
Sent: Saturday, January 19, 2013 2:16 PM
To: Jim Garland
Cc: amps at contesting.com
Subject: Re: [Amps] Question about separate power supply/RF section

The top of the swamping resistor is connected to the grid, but the bottom is
bypassed for RF and connected to the bias and protective circuits. You are
correct that it is to provide a load for the exciter.

In the circuit you describe, is your B- lead grounded at the power supply?
If not then there would be a voltage between the two chassis or (most
likely) they would be connected together by ground wires, so in essence the
B- lead becomes an extra ground wire.

On 1/19/13 1:00 PM, Jim Garland wrote:

Hi Vic,

Ahh, now I see what you're doing, though I still don't quite see how you're
providing operating and cutoff bias to the control grid of the 4cx1000a. I'm
guessing the "swamping resistor") is to provide a 50 ohm load for your
exciter and is DC isolated from the grid, but of course I could be wrong.

It's a matter of personal taste, but I prefer to avoid using the chassis and
power supply grounds as the return path for plate current. In my own tetrode
amp (a current project), I run a B- lead from the external power supply to
the RF deck, and anchor it at the point where the cathodes (three GU74Bs)
are grounded to the chassis. That way the cathode current works its way back
to the power supply via the B- lead, and not through the chassis or ground
wires. (A similar approach is used in your house wiring, where electrical
codes require that return currents on the AC line return through the neutral
wire and not the ground wire.) With my approach a protection diode in the
power supply is also desirable (anode grounded) to protect against flashover
in the power supply. I measure plate current with a 2A ammeter in the B-

73,

Jim W8ZR

_____

From: Vic, K2VCO [mailto:k2vco.vic at gmail.com]
Sent: Saturday, January 19, 2013 1:19 PM
To: Jim Garland
Cc: amps at contesting.com
Subject: Re: [Amps] Question about separate power supply/RF section

I think my initial message was misleading. I think my brain was foggy, too.
I was conflating two different designs that I was thinking about!

The tube is a 4CX1000A, AB1 with a swamping resistor. The grid current (if
any) is measured in series with this resistor. The cathode is grounded. I
have put a 25-watt 0.25 ohm resistor from the negative side of the power
supply to ground, essentially a meter shunt. The plate meter is a
milliammeter with a series resistor that reads the voltage across this
shunt. The meter is in the RF section, so the wire from the negative side of
the power supply is not really the B- connection, it is just the metering
voltage.

I think all I need is a diode in the RF section to protect the meter in case
of an arc.

The power supply section is finished except for the cover, which I hope to
do this weekend. Then I can start the fun part.

On 1/19/13 10:23 AM, Jim Garland wrote:

Vic,

I'm a little puzzled by your proposed schematic. You didn't mention what
tube you'll be using in your new amplifier, but I'm guessing you'll be using
a grounded grid triode(s) design. The customary practice is to measure the
plate current with an ammeter in the B- lead of the supply. The cathode of
the tube is raised above ground by the operating bias, so that the voltage
developed across a low-value resistor to ground from the B- lead is used to
measure the grid current, not the plate current.

There are basically three ways an arc can take place in the standard design.
First, there can be an arc to ground in the RF deck. Second, there can be an
arc to ground in the power supply. And third, there can be an internal arc
in the tube. In the first two cases, the arc current flows from ground up
into the B- lead and back into the power supply. Each of these cases is
usually addressed by means of a diode in both the RF deck and power supply,
each of which has it's anode grounded and its cathode on the B-minus lead.
Although there has been endless debate on the subject, I've never seen any
reason to use back-to-back diodes, or multiple diodes in parallel. Since the
grid current flows from the B- lead to ground through a resistor, a diode
with it's anode connected to the B- line will clamp the voltage to 0.7V and
seriously distort grid current readings. Normally, one wants a 5-10 ohm
resistor between B- to ground for grid current measuring purposes. BTW, a
single 6A10 diode has a surge rating of 250A, so there's really no need to
parallel more than one.

There's one other precaution one should make, and that's to put a 1 W
resistor (say 220 ohms) between B- and ground in the power supply. This
resistor has high enough value that it won't distort grid current
measurements significantly. It's purpose is to keep the B- from soaring
above ground if the power supply is turned on with the HV cable to the
amplifier disconnected. It's basically just to provide an anchor for the B-
side of the power supply. Normally, a negligigle current flows through it. I
checked my most recent power supply and discovered that without the
resistor, the B- lead drifted up to several hundred volts above ground, when
the RF deck was disconnected from the power supply.

Aside from these precautions, and possibly placing protection diodes across
the plate meter terminals, I don't believe any other arc suppression
measures are needed or desirable.

73,

Jim W8ZR

--
Vic, K2VCO
Fresno CA
http://www.qsl.net/k2vco/

--
Vic, K2VCO
Fresno CA
http://www.qsl.net/k2vco/
```