linrad support: The local oscillator for 70 to 10.7MHz conversion.
(Dec 18 2002)

A series resonance X-tal oscillator

The oscillator is the same as the oscillator used for conversion from 10.7 to 2.5 MHz. but there is no frequency divider at the output. The schematic diagram is given in fig. 1.

An oscillator can be seen as an amplifier which has feedback through a filter. The amplifier of a series resonance oscillator must have both a low input impedance and a low output impedance, preferrably well below the series resonance impedance of the X-tal, to preserve the selectivity associated with the high Q of the X-tal itself.

The frequency is adjusted with a trimmer in series with the X-tal.


Fig.1. Schematic diagram for the 59MHz X-tal oscillator.

Selecting different X-tals

The complete circuit diagram is shown in fig. 1. Only one of the MPSH10 transistors is conducting, the others are blocked by having +0.5 V at the emitter.

The frequency is adjusted with a trimmer in series with the X-tal.

Selecting X-tal

The oscillator is selected by a simple low speed serial interface that is adopted for all Linrad hardware. Each unit has four communication pin's.

Select(input)
Clock (input)
Data (input)
Status (output, open collector)

These pins are intended to be connected directly to the parallel port of the PC for small systems, one separate data pin for each select input with all the other pin's in parallel. Larger systems need a decoder to convert 6 pin's to one of 64 pin's.

The frequency control is level controlled and all wires are well decoupled to avoid sending interference into the oscillator. The user program is responsible for setting the d-type flip-flops right so only one of the 74HC4053 outputs is at minus 5 volts while the other four are at plus 5 volts.


Fig.2. Frequency control.

Frequency stability of RX70

Fig. 3 shows the frequency drift of two RX70 units when the units are switched on and the temperature starts rising from 20 degrees. The final temperature is 38 degrees and the final frequencies are 70.011534 and 70.011541. Assuming a temperature independent temperature coefficient one woulf find a value of about 20Hz/deg or 0.3ppm/deg for the unit that drifts fastest. The temperature coefficient varies with the temperature however and one can guess a value of about 10Hz/deg for one unit and close to zero for the other at the final temperature. It is obvious that the crystal oscillators have to be phase locked if accurate calibration is desired. Fig. 3 shows typical behaviour of crystal oscillators without temperature control.


Fig.3. A stable signal is fed to two RX70 units and the frequency is monitored while the temperature of the units rises from 20 degrees.
Fig.4 shows the frequency drift during about 40 minutes for the two first RX70 prototypes using the oscillator of fig.1. Three signal generators, two HP 8657A and one HP ESG1000A were combined and the summed signals were then split and fed into channel 1 of the two different RX70 units. The outputs of the RX70 units were routed to channel 1 and channel 2 respectively of a RX10700 unit, then to a Delta44(mod) in a PC running Linrad.

The long term stability is a thermal problem, the short term stability is affected by several factors. Since so many signals are present simultaneously in fig. 4 it is possible to separate the different factors to some extent. The ESG1000 is a more stable generator than the 8657A. It has a slightly higher amplitude that makes the traces white while the 8657A traces are red.

The difference between the two ESG1000A traces represents the difference in frequency drift between the two RX70 local oscillators. Since one unit has about 10Hz/deg and the othe a small temperature coefficient one can find that the temperature during the 40 minutes varied by about 0.2 degrees. Fig 4 is the best of about 10 measurements, the reason for the temperature changes is mainly variations in outdoor wind speed. Wind speed was very low during this measurement.

The 8657A generators are sensitive to the mains voltage. They respond in a similar fashion to mains voltage variations although one is a little better than the other.

The short term instabilities peak at about 0.2Hz/min as one can see in the AFC graph. It is obvious that some contribution is from the ESG1000A generator but it is hard to quantify it well. The corresponding value for RX10700 is 0.02Hz/min which leads to an expected value of about 0.1 Hz/min for the RX70 unit considering the 4.5 ratio between the LO frequencies.

Here is some information about state of the art VHF crystal oscillators


Fig.4. Frequency stability. Three different 70.0115 MHz signals derived from two HP8657A signal generators and one HP ESG1000A are fed to the inputs of two RX70 units. Temperature and mains voltage variations are the main reasons for frequency variations. Flicker noise is low despite the fact that the oscillators are not run in class A.