Buying custom crystals...

[Sparkey55]

Re: Buying custom crystals...

Crystal swap meets. Ham, SWL, Scanner freaks, etc. Ask around, look online.

[Sparkey55]

Re: Buying custom crystals...

It would be awesome if I could tune the receiver all the way from 88MHz to 162MHz so I could at least hear FM Broadcast and NOAA weather as well...

Scanners that cover those freqs are a dime-a-dozen. Why bother?

[eccerr0r]

Re: Buying custom crystals...

Because it's a no cost addition if it works with my homemade PLL!
Extra bonus for the homemade PLL!

[Sparkey55]

Re: Buying custom crystals...

Because it's a no cost addition if it works with my homemade PLL!
Extra bonus for the homemade PLL!

Knock yourself out then.

[eccerr0r]

Re: Buying custom crystals...

Well, why wouldn't one want to add this feature in? I guess it's not absolutely "no cost" because it would force the PLL to run in the fringe (the VCO would need to have a larger frequency spread) but if all that was needed was to change the multiplier to "overclock" or "underclock" the superheterodyne... get it?

[Agent24]

Re: Buying custom crystals...

I'm not entirely sure what it is you are trying to do, but my general experience has been that if you are trying to design something which requires a very specific and impossible part, then you might just be overlooking a much simpler way to do the same thing.

But, as I said, I don't really know what you're doing, so I could well be completely wrong...

[eccerr0r]

Re: Buying custom crystals...

Well, the simpler way is to buy a Baofeng (or rip up the radio...)

To be clear:

Goal: Make my Regency HR-2B able to receive and transmit on all amateur frequencies without needing to get all crystal combinations
Constraint: Keep HR-2B as original as possible (i.e. do not desolder anything if at all possible, only connect to serviceable points.)
Stretch goal: Make the Regency HR-2B into a scanner as well.

As in Sparkey55's point on that webpage selling $15/ea crystals, the constraint is part of the radio itself: it uses triplers, which makes life difficult for getting specific crystals or PLLs to work. The idea is to compensate for them and still be able to come up with the correct frequencies needed to transmit/receive. The receive side uses one tripler, the transmitter uses two stages.

If it had been using all doublers, this would make the PLL implementation _MUCH_ easier.

Ideally I don't resort to DDS, though another solution is a 1Hz resolution DDS without jitter (asking for 277.7777Hz resolution is probably asking for something weird, 1Hz would at least cover all needed frequencies within 0.7Hz). I really don't want to get an off the shelf DDS that costs more than a BaoFeng, that's another problem. Come to think of it, how would one create merely a square wave digitally controllable oscillator with 1Hz resolution that can scale from 48.000MHz to 49.333MHz with 20PPM accuracy?

[eccerr0r]

Re: Buying custom crystals...

EUREKA!

I just figured out my solution...

I just realized if I had a 10KHz signal (which is easily gotten from a 4MHz clock) and I simply divide this by a non-power of 2 or 10: 9... THERE'S MY 1111.1111Hz refclock!!! Why didn't I think of this before!!! Now I can get rid of this crazy 277.7777 divider and get rid of a bit of jitter that would cause!

Granted, now I need to do some wacky math in the microcontroller... floating point even...

[eccerr0r]

Re: Buying custom crystals...

Oh no... the eureka moment dies... then again my original solution was flawed.
My original solution ended up actually requiring a 1Hz refclock - while possible, this would really make lock time take forever! so I have to throw that idea out the window

So I have to use the new 1111.1111 Hz refclock... but this makes things run too fast, the 5KHz channel step separation goal dies, have to go to 10KHz.

Back to the drawing board.

[goontron]

Re: Buying custom crystals...

^ Why are you so opposed to cracking it open and gutting the oscillator section may i ask? One thing you will learn over the years once you are licensed is that second hand radios are going to have modifications done to them 99% of the time.

[eccerr0r]

Re: Buying custom crystals...

Tread lightly, leave things as much as they were - ability to revert back for nostalgic purposes or if the project turns to a dead end.

Hacking on a CTCSS board is quite a bit different than pulling out the analog multipliers... and FRS is already in the 70cm band - only requires tuning/PLL multiplier changes. Heck that's what I sort of did with the Kenwood, had to tune out of the high 70cm range and down into the ham bands, though the next step is that I have to figure out how the controller communicates to the PLL and force the PLL to run in the ham bands because the current microcontroller blocks transmitting there -- I'll need to hack up a new microcontroller for it. Odd that receiving seems to work fine out of its designed band though that needed its own reverse engineering effort.

Come to think of it, I may need the new microcontroller anyway, as I was planning to also add microcontroller support to replace the DIP switches on the CTCSS board of the Kenwood. That is also a problem for this radio - it has no way of storing the CTCSS codes for a particular programming, nor does it have a way of saving different tones for transmit and receive. As it's only a 10W radio, I'm not sure when I'll start this - would like to see 2m working first since I know my 35W icom will at least try to transmit in the 70cm ham bands, it appears. Besides, building a PLL from scratch seems more interesting than microcontroller work, I've already hacked up multiple microcontroller projects in the past, not much new...

Besides, at 50-60MHz it's already starting to get into problematic logic speeds for the PLL, so the analog multipliers "help" to an extent. Trying to get a digital circuit with off the shelf through hole parts to run at 144MHz is a problem in itself if I want to directly synthesize the LO, not to mention the digital logic is now emanating RF in the 2m band...

[goontron]

Re: Buying custom crystals...

^Are you sure its the micro and not a discrete bandpass filter on that kenwood? They are known for that.

Many radios use RF sources in the carrier frequency range, its all about shielding, but that's more complexity, so i think i see why you want to make a PLL and use the multipliers for that reason.

[eccerr0r]

Re: Buying custom crystals...

Yes it's the microcontroller, at least it seems to be from these observations: I programmed a bunch of frequencies to try transmitting into the dummy load. The frequencies that seem to "work" I see a large increase in power consumption. The frequencies it does not like, the power does not increase while transmitting into the dummy load as if it didn't even bother transmitting. It's not even a gradual drop, it just stayed at nominal power consumption of the receiver plus a few milliamps for the antenna relay.

And right, shielding is a solution, but shielding does not help trying to force logic devices that take 25ns to propagate through and trying to feed a 144MHz signal into it. The signal will change before it finished getting through the chip!

[eccerr0r]

Re: Buying custom crystals...

Hmm... perhaps I should see what the PLLs in the kenwood are actually trying to do when I try to transmit... but it seems really discouraging when it doesn't draw power when trying to transmit. Perhaps there is a BPF somewhere, but where...

The receive circuitry, however, yes it has a BPF that I hacked to get it to work...

[eccerr0r]

Re: Buying custom crystals...

Well, learned something from the Kenwood which I will need to apply to the homemade PLL. Varactor tuning is going to be a beach. I'll be lucky if I can go from 144 to 164MHz, going down to 88MHz will likely not be possible unless I add more varactors in parallel, perhaps, and hope they don't disable the high end frequencies(!)

Still don't have a viable solution for the PLL however, as right now it looks like the Atmel's timer counter isn't sufficiently fast enough to be the primary/only divider. The only thing that will work now:

* I will need 10KHz/9 reference clock
* My goal of 5KHz channel step is gone. 10KHz is best I can do now.
* Need a Low (for transmitter) and High (for receiver) range control.
- High range for the receiver will enable a divide by 3 which will allow an AVR to be able to count.
- Low range for the transmitter is still broken with no solution. Currently stuck with either a 20KHz channel separation (no bueno!) or perhaps go to a 555.55555555 Hz reference clock, which will then allow 10KHz transmit channel separation. On the bright side, with 555.5555555 Hz it will allow a 5KHz receive channel separation. It will also slow down lock time to probably a half second or so (just a guess) which will make scanning very slow Plus pressing PTT one will need to wait a second for lock, then transmitting would be possible.
The other possibility is I use two different reference clocks, one for receive and one for transmit ... Doesn't help channel separation any, but will speed up scanning.

[redwire]

Re: Buying custom crystals...

Si5351 looks interesting. I2C-Programmable Any-Frequency CMOS Clock Generator + VCXO.

https://hackaday.com/tag/digitally-c...led-oscillator

[eccerr0r]

Re: Buying custom crystals...

Thanks, yeah I was thinking about how I could use a clock chip from a motherboard though stepping by 1KHz is not enough, I need to step by ... oh these numbers are coming up again and again...
277.7777777 Hz (for 5KHz step on transmit, or 2.5KHz step on receive). If I don't get these fractional Hz accurate, I'll end up with erroneous carrier center frequencies due to the multipliers. However if the step is small enough, I'm not sure what happens if you transmit or received with say 147052066 Hz for 147.052MHz because that's what the multipliers ended up with. Would that be okay?

[Andyy]

Re: Buying custom crystals...

You need this

http://g4oep.atspace.com/channels/channels.htm

[eccerr0r]

Re: Buying custom crystals...

That lone Si570 chip is what I need.

Alas getting that ONE chip is around $18 a pop (in quantities of 1)... ugh. Not pleasant.

[Andyy]

Re: Buying custom crystals...

Have you tried getting a free sample from silabs?