Jennings RJ1A Vacuum Relay Measurements

RJ1A Vacuum Relay

After the successful modification of a Heathkit SB-220 for 6 meter operation, an upgrade of the T/R relay to allow QSK operation is now being considered. The 'defacto' relay for high-power QSK at HF is the Jennings RJ1A or equivalent. Unfortunately these excellent relays are only spec'd to 32 MHz ... and those specifications lack information other than voltage and current capability. A copy of the Jennings RJ1A data sheet can be found at RJ1A Specifications.

A call to the tech folks at Jennings yielded some useful information as to the voltage and current specifications higher in frequency. In passing the tech mentioned that they had used the relays in a project at 75 MHz. The tech's best guess was that the 16 and 32 MHz specifications could be 'more or less' linearly extrapolated higher in frequency. Ratings at 64 MHz would therefore be approximately 1.0kV and 5A - more than adequate to handle 1200 watts output power especially considering the lower frequency (50 MHz) of operation. This still left open the question of isolation and return loss for which there doesn't appear to be any data ... at any frequency! This prompted me to generate my own data and it is presented here as others may find it useful.

In most applications, connection to the relays are via copper strap or coaxial cable pigtails. While these methods are fine inside an rf amplifier compartment, a better interface to the test equipment is required if one wants to accurately characterize relay performance. In order to adequately test the relay a fixture was machined from a copper pipe cap. SMA connectors were used for the connections with lead lengths kept as short as possible. Pictures below show the fixture details.

RJ1A Vacuum Relay
RJ1A Vacuum Relay

The first measurement was for relay isolation - from the common connection to the normally open connection with the relay de-energized. A second test from common to the normally closed connection with the relay energized produced results that were indistiguishable from the first test. An HP8595E spectrum analyzer and tracking generator were used for the measurement. All ports were terminated in 50 ohms. The wideband spectrum display can be seen below.

RJ1A Vacuum Relay

Readings were taken at several amateur bands of interest and are presented in the table below. Note that the curve is smooth and 'well behaved' indicating no frequency dependant 'anomolies'.

FREQUENCY ISOLATION
1.8 MHz-64.3 dB
3.5 MHz-59.7 dB
7.0 MHz-55.4 dB
14.0 MHz-49.4 dB
21.0 MHz-46.4 dB
28.0 MHz-43.6 dB
50.0 MHz-38.7 dB
144.0 MHz-29.6 dB

The following impedance/return loss measurement was made using an AIM4170 analyzer. All ports were terminated in 50 ohms. The display shows the frequency range from 1 MHz to 150 MHz.

RJ1A Vacuum Relay

Readings were taken at several amateur bands of interest and are presented in the table below. Note that the curves are smooth and 'well behaved' indicating, again, no frequency dependant 'anomolies'.

MarkerFrequencyRsXsSWRReturn Loss
11.8 MHz50.6490.0841.013143.74
23.5 MHz50.6050.2051.012843.93
37.0 MHz50.6460.3881.015142.51
414.0 MHz50.6400.7281.019540.30
521.0 MHz50.6421.0951.025538.00
628.0 MHz50.6691.4121.031536.19
750.0 MHz50.9022.4771.054231.66
8144.0 MHz52.4796.7981.151623.04

Based on the info from Jennings and the measurements presented here, performance appears to be adequate for operation at 50 MHz. A writeup on the SB-220 modifications for 6 meter operation including the use of the RJ1A relays will be forthcoming ... once the 6-meter sporadic E season winds down!

Back to W1VD WD2XNS WE2XGR/2 home page