Unilab 10m FM conversion - Receiver VCO

Apologies for the delay in posting this part of the project. I couldn't find my notes and my memory was failing me. In fact, when I looked at what I had done I was amazed it worked at all. So I decide to start from scratch. I am breaking this up into a few more parts than I expected. My logic follows and this time I have taken notes!

I assumed the receiver VCO was running on the high side based on a vague reference to this in the technical manual.

1. Standard radio covers 66-88 MHz. Use 77MHz as a starting point.
2. 1st IF frequency is 21.6MHz so Local Oscillator is running 21.6 MHz above 77 MHz, or about 92MHz.
3. For 29.4Mhz the local oscillator needs to be 21.6 MHz above this, or 51MHz.

I know from what I did previously that it wasn't as simple as padding capacitors to lower the frequency. I did a complete rebuild based on material in Solid State Design. I removed the inductor and capacitors from the oscillator board and rebuilt it to match the following circuit:

We need to add the resistor and diode somehow and the tank circuit is no longer tapped on the inductor. I attach a photo of my finished VCO below to help with how I placed the parts.

After soldering all parts except the inductor I used two 33k resistors (value is not critical as long as they are the same) as a divider to apply a voltage on the VCO line. With 8V to the oscillator board I tried a few inductors to see if I had one that was close to 0.44uH. I had trouble getting this circuit to oscillate so I changed C204 from a 68p to a 33p capacitor. Now I could get some oscillation! I held inductors in circuit by hand on the bottom side of the board and looked at the frequency counter. After a few goes I found an adjustable inductor that allowed for oscillations around 51MHz.

I soldered the inductor in place and adjusted the frequency for 51MHz. All going well this suggested the VCO control voltage would be around 4V when the loop locked up. After removing the resistors I plugged the board into the radio and tested.

Lo and behold, no lock. Hmnn....A bit of prodding with the CRO and it seems the output level was too low.
 
The first thing I did was add another 3k3 resistor in parallel with R201 to increase the standing current. I had noticed when I warmed the thermistor with my iron that the output level increased so this addressed the level issue.

But still no lock. Measurement showed the pre-scaler was clocking at twice the rate it should be. I speculate that this was due to insufficient drive, and not the presence of a strong 2nd harmonic, and padding C209 with a 10p surface mount capacitor resolved this.

I then changed the choke in the source to 82uH. At this point the local oscillator is on frequency and the PLL is locked.

If I was doing this again then on reflection I would start by bypassing the existing resistor and thermistor in the source with a 10n capacitor. If the pre-scaler was not clocking correctly I would then change the choke, then pad C209 if required. I'd be interested in hearing how your modification went and what you tried so please leave a comment below.

At this point I still need to check the stability of the loop and I will address this in another post.

73's
Richard

Unilab KL70 10m FM Conversion

The Unilab KL series were rugged FM commercial repeaters very similar, if not electrically identical,  to the Kyodo radio's. Converting KL150's to 2m and KL450's to 70cm is relatively straightforward, giving a 50W near indestructible radio. After successfully converting a KL70 to 6m FM operation I wondered if a 10m FM conversion was possible.

The KL70 was made to work between 66 and 88MHz. So a number of challenges arose or were foreseen:

1. The receiving band pass filters can't simply be re-tuned,
2. The local oscillators need a substantial amount of re-work,
3. The PLL loop filter may need modification, I'm yet to confirm this, and
4. The transmit harmonics will require filtering.

While still relatively straight forward, I've not found time to complete my conversion. However, writing about it is a sure way to elevate the project to a "finished" status so here goes.

In this post I will cover the receiver band pass filters. If you looked at my previous posts you will have noticed I'm a big fan of the Iowa Hills RF Filter Design software. Since the original bandpass filters on the receiver will not tune down to 29MHz I was forced to rebuild them.

Scouring the junkbox I found some nominal 160nH adjustable inductors. A few minutes work with the software and I had the following filter design:

My rejection of the image frequency of 72MHz is likely to be no better than 80dB for the two filters.At this point in time I didn't know about the series tuned filter transformation.  The same inductors used with series LC tanks would yield around 80dB for just one 3 element filter. If you knew of local transmissions on 72MHz then I'd use one filter like that shown, and one with series LC tanks. Let me know if you want more details on this alternative filter.

After etching a board and building two filters with the same pinout as the original filter I removed the filters from the radio, disassembled them, and put my new filter into the filter housings. Prior to putting the filter in the housing they looked like this:

 

With these two re-assembled filters soldered back into the receiver I turned my attention to the other parts of the project. More on that next time.

73's
Richard VK6TT

.

Assembled 20m WSPR transmitter

Here's a picture of the top and bottom side of my 20m synthesised WSPR transmitter. Measured results were 1W pep and reference spurs 42dB below this, or less than 100uW.

Now to put it in a box and really smoke test it.