Hello everyone!

I'm working on my frequency counter, and now I need to create a 1 Hertz signal, and I'm doing that by using a 4521 chip (a 24-stage frequency divider).

But I have a problem, when I power the circuit, the chip only starts "oscillating", after a while (sometimes), or after I touch (yes, literally touch something, like the crystal, or sometimes when I plug a 9V battery, it only starts counting after I stop touching with my hands the insulator of the battery connector, or other times when I touch some capacitor...). This is really odd :/

- Chip's Vdd and Vdd1 pins are connected to V+ (5V)
- Chip's Vss and Vss1 pins are connected to Ground
- Chip's master reset pin is connected to ground
- A1 pin is connected to ground
- Y2 pin (pin 4) is connected to one of the pins of the crystal
- A2 pin (pin 6) is connected to the other pin of the crystal
- Y1 pin is floating, like I even saw in some schematics.
- Both the pins of the crystal are connected to the ground throught 22pF caps (one for each pin)
- The Vdd pin is even decoupled to ground throught one 100 nF cap (a late addition).
- The crystal is 4.194304Mhz
- All the outut pins are left floating - except for the Q23 pin that is connected to a resistor and a LED - (that should be blinking).
- The circuit is powered by an AC adaptor or by a battery and the voltage is regulated to 5V by a 7805...

When I power up the circuit, it should start "oscillating", but that only happens sometimes or after waiting a random amount of time...

Guys what is wrong?
Is it static or something like that? If so, how to solve this problem?

http://www.nxp.com/documents/data_sheet/HEF4521B.pdf

Thanks a lot for reading!
Miguel

 

clearly your oscillator has trouble starting up. why don't you use schematic from datasheet? you connected Y2 directly to crystal while datasheet shows large resistor value Ro (tens to hundreds of kOhm...). also datasheet also shows large pullup resistor on A2 (1.8M) which you don't have.

 

"Both the pins of the crystal are connected to the ground throught 22pF caps (one for each pin)"

Shouldn't one be 22pF and the other 82pF or so? I thought it was the difference in value of these capacitors that "kicked" the crystal into oscillating?

 

Hello and thank you for your answers. I've been trying to fix this and took your replies in consideration but I still cannot get it working. :/

@panic mode and @paulktreg
By putting those resistors and caps I cannot get it working either. :/ In fact it is even worse - the oscillator wont even start.

Now I just went back to the "old circuit", I just removed both my caps to replace them... And it started... Right now the oscillator is working without any caps.

Update: Lol? Now I don't have a single capacitor and my oscillator is starting everytime I powerup!

This isn't supposed... or is it?... :/ I'm soldering my frequency counter on a veroboard, but this oscillator I decided to first try out in a cheap breadboard... If I'm not wrong breadboards may have capacitance between "tracks"... So I don't know if it's such thing that is starting the circuit... And perhaps it wont work when I solder it...

Do you guys have any more ideas or possible explanations? :/

Edit 2: With 1 pin at 22pF and the other at 10pF it is also starting easily (no resistors connected).

Edit 3: It seems that the and 22pF/10pF solution is better than my initial 22pf/22pF BUT it is still not as stable as putting no capacitors at all... When I leave the circuit turned off for a while, even de 22pF/10pF has a little trouble to start even tought it seems to be better than de original one. Higher values in capacitors do not seem to be working... I'll try more capacitor configurations this morning. Now it's very late here and I'll get some sleep.

Thank you for your anwsers!
Miguel

 

Get rid of the breadboard because its capacitance and inductance are far too much for your high crystal frequency.

I wish I was given $1.00 for every problem reported here caused by a breadboard.

 

but this oscillator I decided to first try out in a cheap breadboard

Solderless breadboards are great.... at suitably low frequencies. I'm willing to push them into the 1MHz range for analog circuits, with careful routing, but I prefer to stay below a few hundred kHz. I have a friend that has made a beautifully working function generator on a breadboard that goes to 20MHz, but he really knows his stuff and spent a lot of time doing it -- it wasn't for any practical purpose, he just wanted to prove to himself that he could do it.

 

Both HCF4521B from STM, and CD4521B from TI, only guarantee the maximum clock frequency to 2MHz. You might get more consistent results if you use a higher supply voltage.
I also agree with Audioguru's comments about solderless breadboards.

 

Hmmm all right guys thank you! I will solder the oscillator this afternoon and post the results here. I'll use a trim cap in one of the crystal pins as I really don't feel like desoldering components (which is really not my expertise at all...).

 

I just finished soldering it in my veroboard, I used a 10pF cap and a trimmer cap that goes till 45pF and it seems to be working fine for now. So thank you guys!

I still need to solder a lot of other chips on the board so I don't know how that will affect the oscillator, if I get any more trouble on this, I'll post it here.

Thank you again.

Miguel