Unilab 10m FM conversion - Receiver VCO and PLL Stability

I noticed that while switching from receive to transmit and back again that sometimes the receiver PLL would drop out of lock. Here then is a brief discussion of how I resolved this issue.

The first thing I did was measure the VCO voltage at the minimum and maximum frequencies I had loaded in the eeprom. To achieve this I had to adjust the inductor and trim cap in the VCO until I had lock with a low control voltage. Otherwise, I couldn't get a lock at the high frequency. This immediately told me I had at least one issue to address.

Frequency         Tuning Voltage
50.74 MHz        0.1V
51.29 MHz        6.7V

Clearly that leaves little margin for things like temperatures change given the maximum tuning voltage under lock is approximately 8V. The control voltage swing needs to be much less than this, but how much?

Looking at the loop filter values left me nonplussed. I'd never seen a loop filter like this before. I have highlighted the relevant components on the schematic below.  I was tempted to re-design this part of the radio but went away and though about this for some time.It eventually dawned on me that there was an embedded lead lag filter, circled on the schematic below. Ignoring everything else I crunched some numbers which suggested that these values would be appropriate if the range in tuning voltage was more like 1V.

Looking at the component values used, below, the most expedient fix would be to change the 2p7 capacitor in series with the varicap.The first capacitor I tried was 15pF which gave a variation in the control voltage of 3.5V - 5.0V. Before I could measure that I had to get the PLL to lock again. I had seen this before with the transmit VCO. Despite the oscillator oscillating, the pre-scaler was giving an erroneous reading due to the input signal being too low. I added a 100 ohm resistor in parallel with R202 which fixed this.

Receiver VCO values before change to 2p7 capacitor

At this point the receiver synthesiser mods are almost complete. After putting the lid on the receiver I noticed the loop went out of lock on one channel. The channel either side was fine. Remove the lid and the channel locked up. Given the control voltage on the test point did not shift significantly for the frequencies above and below this troublesome channel I ruled out the lid modifying the oscillator frequency.

I cut a piece of blank circuit board to the size of the VCO enclosure. As I placed it into the enclosure as a temporary shield and I expected the lock issue to arise. It didn't. But putting the lid on while the temporary shield was in place still unlocked the loop. So the issue wasn't with the VCO.Which ruled out so many weird causes that I was stumped as to what it could be.

After sliding various assortments of metal around the top of the receiver enclosure I confirmed the issue is arising within the section housing the eeprom. I still hadn't got to the bottom of this. The frequency, 50.760 MHz isn't a harmonic of the crystal oscillator or even the 750kHz reference going into the PLL chip. But you can set the frequency 10kHz either side of this with no issue.

However, after another go at solving this I discovered the problem was due to incident light, and not an RF issue at all. It turns out that this one channel in the eeprom changed value if it was in darkness. I couldn't replicate this in my eeprom programmer but it definitely was due to darkness changing the data on the output pins of the eeprom. Anyway, I burned another eeprom and problem solved.

I have put the lid on, tightened the screws and next time I will start writing up the transmitter conversion.

Regards
Richard