[Top] [All Lists]

Re: [Amps] Measuring filament voltage

To: John Lyles <>
Subject: Re: [Amps] Measuring filament voltage
From: Radio WC6W via Amps <>
Reply-to: Radio WC6W <>
Date: Wed, 22 Jun 2022 01:00:29 +0000 (UTC)
List-post: <>
Hi John,
 If you want to measure the voltage at the filament on those follower triodes 
may I suggest the following scheme:

 Use an AD536 to convert to RMS right at the tube.

 Use a LM331 to convert the voltage to frequency.

 Send the frequency out over an optical fiber and read it with a frequency 
counter or alternately convert it back to voltage with a F-V converter.

 Power it with a tiny hi voltage isolation transformer. Only requires milliamps 
to run. 

 Or use a power over fiber cable scheme, though those are rather pricey.

Just a thought,
 Marv WC6W

     On Sunday, June 5, 2022, 10:08:38 AM PDT, John Lyles <> 
 In the commercial RF amplifiers that I have designed, put into 
production or installed and operated, filament voltage is measured all 
the time. For pentodes, tetrodes or triodes with common cathode 
arrangement, it is simple to have two wires going to the socket, 
suitably bypassed for common mode as well differential mode RF noise. A 
cheap DMM won't be accurate enough, depending on the transformer or 
power supply - use true RMS metering. For years this meant taut band 
analog meter movements. All the Broadcast Electronics FM transmitters 
with tubes had these as well as their quality competitors such as 
Collins and Harris. These days one can find a decent digital meter that 
has RMS calibration in case of non-perfect sinewave waveform. I have 
used Newport meters for this. The point is to measure at the socket, not 
the transformer winding.

For common grid circuits where the cathode is carrying common mode RF 
voltage with respect to grid and chassis, it's not so easy. If it is a 
cavity circuit where the structure itself is used to ground the bottom 
end of the resonator (quarter wave cavity for example) then the meter 
circuit is applied there at the ground end of the structure. This is how 
i do it for 2 MW amplifiers at 200 MHz that are grounded grid/screen 
grid configuration. You can tell if there is RF interference, as the 
meter will rapidly change with the RF power comes on. If there is 
appreciable backheating inside the tube (RF and infrared affecting the 
cathode temperature) then it is more complicated and I will leave that 
out of this. Assume that the designer did a good job of bypassing the 
heater carefully for RF.

For HF amplifiers that often use common grid circuit with triodes, it is 
again more difficult to measure at the socket since RF voltage is 
applied to the cathode with respect to grid and to chassis. About the 
best you can do is measure on the transformer secondary (for a center 
tapped filament transformer) just before the bifilar RF choke. You can 
measure with RF off on both sides and create a calibration factor, 
knowing what it is on the cold side of the chokes to estimate what is at 
the tube/socket. Then you know what it is with RF on or off.

I have one amplifier system that is a cathode follower connected 
triodes. The RF voltage is as high as 18 kV peak at 2.8 MHz there. It is 
very difficult to physically measure the filament voltage. The filament 
transformer has low capacitance and RF isolation between windings. In 
that case, can only measure the primary AC voltage, 440 VAC in this 
case. And then create a conversion coefficient for the output voltage, 
measured with a good RMS meter when the RF/HF is locked out.




Amps mailing list
Amps mailing list

<Prev in Thread] Current Thread [Next in Thread>