Problem setting the frequency of CD4046 PLL / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Problem setting the frequency of CD4046 PLL

1 2 next

Move Thread

LAN_403

Dr. Shark

Mon Mar 06 2006, 06:38PM

Registered Member #75 Joined: Thu Feb 09 2006, 09:30AM
Location: Montana, USA
Posts: 711

I am constructing a Class E amplifier, and for the drive signal I intend to use a 4046 PLL chip (the CMOS version that runs off 5V max), since that's what I have at hand to generate a square wave at something like 4MHz frequency. Now the trouble is that I cant seem to get the frequency above about 1MHz. Reducing the size of the timing capacitor below 100pF or so does not have any effect at all.
The circuit I am using is

from Steve, apart from the timing capacitor I have also replaced the 47k and 33k R's by 15k and 10k respectively. I have made no changes to the low-pass filter on the feedback section.

Obviously I am missing a great deal of PLL basics, but google did not help here. Can enyone enlighten me what exactly I am missing here?

Marko

Mon Mar 06 2006, 07:21PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

If you are talking about 74HC4046 then you can go for 20mhz max, CDs/HEFs are 2,7mhz max.
VCO is controlled by two resistors and capacitor, and you can control its frequency range by changing their values.
Here is a datasheet so you can compare basic schematics:

Dr. Shark

Mon Mar 06 2006, 08:04PM

Registered Member #75 Joined: Thu Feb 09 2006, 09:30AM
Location: Montana, USA
Posts: 711

It is in fact the CD4046, but still it should be good for more than 1MHz. The weird thing is, even when I remove the capacitor entirely, and cut the resistor down to 20Ohm, the frequency does not go up!

I have got the data sheet printed out, but I find it rather confusing and not very informative, it does not even give a formula for calculating f_0, let alone the frequency offset etc.

I think digital stuff is not for me, I better stick with old-fashioned, non-IC circuits.

Marko

Mon Mar 06 2006, 08:39PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

CD4046, HEF4046... cannot work at more than 2,7Mhz and that must be at 15V.
With 5V supply maximum frequency (as datasheet says) is 1Mhz.
YOu hit the maximum and you must increase supply voltage for more.
TO operate at 4Mhz you must use 74HC4046.

Dr. Shark

Tue Mar 07 2006, 08:44AM

Registered Member #75 Joined: Thu Feb 09 2006, 09:30AM
Location: Montana, USA
Posts: 711

Thanks for pointing out the obvious to me. So the reason why the datasheet does not give a formula for the frequency in terms of C and R is because it depends on voltage. Well, no surprise with a VCO, thinking about it. Damn.

Marko

Tue Mar 07 2006, 09:07AM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

Dont get mad now, im just trying to help.
And 74HC4046 wil obivously solve all your problems (supply voltage and wanted frequency).

Dr. Shark

Tue Mar 07 2006, 10:40AM

Registered Member #75 Joined: Thu Feb 09 2006, 09:30AM
Location: Montana, USA
Posts: 711

Don't get me wrong, I was sincerely happy that I finally know what was going wrong. I was not being sarcastic in my last comment, even if it sounded like I was.
Actually I am much more comfortable using 12V for supply, the only reason I used 5V was that I though I had ordered the fast version of the chip - which aparently I haven't.

Dr. Shark

Sat Mar 18 2006, 09:46AM

Registered Member #75 Joined: Thu Feb 09 2006, 09:30AM
Location: Montana, USA
Posts: 711

More PLL trouble ahead.
I think I have killed the pahse comparators in both PLL driver I have build. The reason I think this is the case: I used to be able to change the VCO frequency slightly by turning the 10k pot (still referring to Steves Mark I PLL here). Now this does not work anymore. Is this a typical failure mode for the CD4046, or am I just overlooking something (again)?

ragnar

Sat Mar 18 2006, 11:33AM

Registered Member #63 Joined: Thu Feb 09 2006, 06:18AM
Location:
Posts: 1425

Twiddling the 10K pot will adjust the lock.

I recommend you replace the resistors (on pins 11 and 12) with potentiometers, and hook a digital multimeter up to the output.

Now twiddle the pots and get a feel for things. You'll find that the pot on pin 11 (IIRC, might be vice versa) sets the centre frequency, and the pot on pin 12 will set your side-side offset. If you've got the feedback hooked up, the PLL will try and lock... without feedback, twiddling the pot would just give you fine frequency adjust.

I'd imagine the 4046 died because you took out the capacitor and shorted pin 11 to ground through a 20-ohm resistor

Put a 10-ohm resistor in series with the positive supply - this will limit the current, meaning nothing should cook too badly if it wants to die.

So, starting fresh:
Hook up to 15V, regulated, smoothed, with 10-ohm in series with positive.
Put in a 100pF or 150pF (start with 150pF, preferably) timing capacitor.
Twiddle those pots until you have a feeling for how it affects the frequency.
Set your centre frequency with the first pot. Set your offset with the other pot.
Now fine tune with the 10K lockpot.

I'm sleepy. Bits of what I said are probably wrong, the god of PLL may smite me in my dreams, but that's what I'd try considering you don't quite know whats going on. Look up some PLL tutorials on google. There's at least two good ones out there that explain what goes on pretty well.

ragnar

Fri Mar 24 2006, 01:17PM

Registered Member #63 Joined: Thu Feb 09 2006, 06:18AM
Location:
Posts: 1425

Joe, if you're still reading this thread, I actually got around to putting together a circuit for you.

The CD4046 generally won't go above 1.4MHz with Vin +10V.. BUT:

My particular version (CD4046BE) and probably yours too will have a "Max Vin" of 15V. Then if you read further on, there's a "MAX MAX Vin" of 18V. In small writing at the bottom of my datasheet, it also says the input is zener clamped to "+20V". Ah, HA!

SO, if you can get a very precise 20V reference, you can run the chip at max.
I used a 15-ohm resistor on pin 16 (positive) to limit any current in case the zeners were clamping early. Because I didn't have any regulators handy, I strung some old, half-dead AAs together until I got ~20.5V. I then ran the AAs through a resistor for a few seconds, a few times, to bring the volts down to 19.9.

If you're only using the PLL as an oscillator:
- connect pin3 to pin4.
- disconnect any aerial.
- get rid of the centre-frequency-offset resistor, (on pin11 IIRC)

Now, on the datasheet they claim that 50pF is the minimum timing capacitor you can use. And they're not telling lies - if you go smaller (e.g. 30pF), then the max frequency you can reach is reduced.

If you go above, (100pF), the max frequency is also reduced. Perhaps you can use a variable AM capacitor to find out where the ideal ideal max range is available... might be between 40-60pF.

I didn't meter the resistor on pin12. Put a 50K pot there (preferably something like a 25-turn cermet) in place of the resistor. As you move from 50K-ohm to the smaller end of the scale, the frequency will start to rise, peak, then fall off again. Go back to this peak and tune around it.

With all these minor optimizations, I was able to run a CD4046 at 4.4MHz - 20V input, 50pF timer.
My scope showed a smooth and steady squarewave with fairly slopey edges. There was nothing connected to the output - a UCC or two might make the wave even worse, BUT it should trigger nicely enough.

You might be able to tweak that even more if you get rid of the 6V divider and that 10nF capacitor... err, perhaps tie that pin to positive or negative, depending on if it swings the frequency one way or another. You won't need to connect that 10K divider pot either. Just use the VCO section of the chip.

Good luck!

1 2 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob