Oscillator for 2200m

Recently I was frustrated with the amount of time it took to get some simple 2.205 MHz and 2.185 MHz crystals from Futurlec – in the end it took some 8 weeks to only get a small number of what I was after (and only with a bit of poking and prodding). This is not uncommon with Futurlec as you’ll find if you search the ‘net, but the thing is there are very few suppliers of such options and these were needed as I was trying to build something like the VK1SV MEPT transmitter for 2200m. Indeed, to build a crystal based oscillator for 2200m there are very few options, so that made it more frustrating.

Anyway, that frustration led to me to come up with a modified design myself. I saw on element14 that they had an abundance of 22MHz crystals and they’re known for rapid postage – and free at that. Suddenly, the idea seemed feasible for a 137.5 kHz oscillator based on 22MHz crystals – which I found more readily accessible. Only problem was, I only knew the basic idea of using 4000 series CMOS chips as dividers, so I had a lot to learn to do it myself.

Within time though, I’d come up with this design:

Oscillator_schematic

 

As you can see, it’s based primarily around three items:
  • 74HC4060N 14-stage binary ripple counter with internal oscillator
  • 74HC4017N decade counter
  • 22 MHz Crystal

The idea here is simple, the 74HC4060N first provides a 22 MHz Inverter Oscillator with the 22 MHz crystal. This is then used as it’s input for which we derive the ouput on the Q3 pin to achieve a divide by 16. This giving us a frequency of 1.375 MHz. This is then feed into the 74HC4017N so that we can further divide it by 10 which gives us the final frequency of 137.5 kHz. (A total operation of 22 MHz divided by 160.)

The only other bit of the circuit is the power supply. This is contoled with a 78L05 to provide 5V regulated as required by the 74HC chips. This should be noted, as these are not the plain CMOS chips (CD series) as these would not be able to work at the 22MHz clock frequency – normally only workable up to 12 MHz, but in some cases only 1 MHz. As a result, they require a Vcc of 2 to 6 volts.

Going back to the 22 MHz oscillator in the circuit, a couple of notes. The 4 ~ 40pf trimmer capacitor with the 100pF capacitor are to provide the required crystal load capacitance: CL = (C3 x C4) / (C3 + C4) + Cs. With those values – and using an arbitrary stray capacitance (Cs) of 5pF – it will cover a CL range of 8pF to 38pF – that should cover most crystals (mine required 18pF). Also of note in the inverter oscillator circuit is the missing Rd resistor (or Rs depending on your reference) to limit current to the crystal. With most circuits I see online this is ommited, but I think it could be interesting to try and figure out what this value shold be – as there are benefits. Initially I tried with 2k2 and that was far too large and meant no oscillation. But from calculations I guess you could try 75R, but I’m yet to experiment.

The other item to note is that I have observed there is a significant 22 MHz signal on the +ve rail (almost 1.5 Vpp). The above mentioned Rd resistor could help, but more so a decoupling network should ideally be added to the supply rail. Something I might consider if I have time to play with this circuit a bit more.

Anyway, even with these considerations this circuit works. Further, it also seems stable – although I’ve only run it for a few minutes and with no load. Oh, and don’t forget, the output from this is a square wave – so use appropriately. Here’s the output on my scope:

Scope_capture

All the parts you require can be fetched from element14, however if for some reason you struggle with acquiring the 22 MHz crystal there’s good news on that. Another reliable local supplier stocks them: Mini-Kits. And if needs be, I see that Futurlec have a 22 MHz can oscillator for just under $2.

Another option on the crystal front, is to consider using an 11 MHz crystal. But then the circuit is slightly different as you need to divide by 80. You could do this with two 74HC4017N chips (the first could divde by 8, the second by 10), however you’d also need to build the 22 MHz oscillator on it’s own to feed into it. But something to consider.

Anyway, here’s some photos from my resulting prototype:

Vk1is_2200m_osc-001Vk1is_2200m_osc-002Vk1is_2200m_osc-003Vk1is_2200m_osc-004Vk1is_2200m_osc-005Vk1is_2200m_osc-006

Finally, I’d like to send a thanks out to Dimitris (VK1SV). It was only a month or so back I had no idea about how you go about frequency division. I knew basically about frequency multipliers, but hadn’t even heard of frequency division. With his kind introduction to this theory I was able to produce the above and understand how it worked. Further, doing so also made me spend time to better understand inverter oscillators.

Thanks Dimitris! I may just get on 2200m one day soon.

 

 

 

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s