5110A Synthesizer Driver & 5100A Frequency Synthesizer

Brooke Clarke 2005

HP 5100A over 5110


This is a synthesizer based on add, subtract, multiply NOT on PLL methods.  It is big and heavy, but has very fast coherent switching and good phase noise.  This model was used by BR communications to develop their chirp sounder .  For remote control on the back there is a wire for each of the buttons on the front. That's a lot of wires!
HP Memory Project: The First Commercial Application of Frequency Synthesis -
HP Journal May 1964:  A 0-50 Mc Frequency Synthesizer with Excellent Stability, Fast Switching, and Fine Resolution
App Note 96: Frequency Synthesizers Jan 1969
HP Catalog 1965 (pdf pg 95) - "Synthesizer design and construction make possible output signals whose spurious content is 90 db or more below the selected frequency. Signal-to-noise ratio in a 3 kc band centered on the selected signal is more than 60 db. Particular care in Model 5100A-5110A design results in a very clean output signal over the full 50 mc range. The high order of spectral purity permits accurate doppler measurements, microwave spectroscopy, narrow-band telemetry, stable local oscillator for a transmitter and/or receiver, automatic testing of crystal filter response and many other applications."

What Goes Wrong

The 5100A A1A4 assembly takes in 1 MHz and after amplifying the signal with Q1 drives Step Recovery Diode CR6.  In order to get the diode to deliver more output power it has been built into an assembly (E1) consisting of the SRD, a heater and a temperature sensor.  Q2 drives the heater.  Running the Step Recovery Diode hot increases its lifetime and hence the output.  Running electronic components hot causes their lifetime to be shortened and in this case the SRD fails.  When that happens the 10 outputs die and the whole 5100 Synthesizer no longer works.

Top of HP 5100 A4A1 PCB
HP 5100 A4A1 trace side optically flipped
Comp side of A1A4 PCB
Trace Side Optically Flipped
The A1A4 Printed Circuit Board holds the Step Recovery Diode in the lower right (off white circle). 

One fix is to find a stud mount varactor diode and wire it in place of the SRD and leave the heater disconnected.

Another possible fix is to use the circuit from the PRC-25 A15  / PRC-77 A45 module.  This module has a 1 MHz crystal oscillator and using a one transistor circuit generates a lot of harmonics.  In the PRC-25 the transistor is a PNP 2N2273 and in the newer PRC-77 it's an NPN  2N918.  A 10 pF capacitor (typo in the PRC-77 manual) feeds the 1 MHz signal to the base.  The base has a 10k resistor to ground.  The emitter has a parallel circuit consisting of a 1k2 resistor and a 0.05 uF cap to ground.  The collector is transformer coupled to the output with the top of the transformer primary connected to +9.5 volts and decoupled to ground with 0.05 uF.  Positive supply isolating by means of a 200 uH choke.  The transistor is off during the negative input half cycle and turns on only during the positive half cycle generating a narrow pulse rich in harmonics.  The output of the module consists of harmonics at 1 to 12 Mhz and so is not directly suitable for use in the 5110 which puts out 30 to 40 MHz.

Brooke's Home, Time & Frequency, Electronics Page
[an error occurred while processing this directive] page created Oct 15, 2005.