crystal, does it work without grounding? - Electronic Circuits Projects Diagrams Free

skyrock · 8th May 2008, 01:08 PM

Hello everybody,

As with the little crystal problem i've had last few days, i finally found out why it didnt work but I dont understand it..

It happened that the crystal I used cannot be grounded in order to work. I can only make it work like the circuit in the attachment. Is it normal?

I'm using Nigel's main board and Led board for testing using PIC16f628

http://www.winpicprog.co.uk/pic_tutorial_main_board.htm

I attached the 2 legs from the crystal to RA6 & 7, and then my board turns on the LED on portb blinking at 1 second interval. But this is without any capacitor and grounding... (SPOOKY..) Does anybody knows why?

EDIT: it doesnt work already after i tried it again. would it because there's charge left inside the crystal? It's just plain wierd..

**Nigel Goodwin** · 8th May 2008, 02:49 PM

Connect the capacitors, they are an essential part of the oscillator circuit - without them it 'may' work some of the time, due to stray capacitance, but it's unlikely to be reliable.

joshua17ss2 · 8th May 2008, 03:42 PM

Is it possible its running on the internal oscillator and just ignoring the connection to the external crystal?

picbits · 8th May 2008, 05:26 PM

More than likely the stray capacitance of the PCB or prototype board he's using is enabling the crystal to oscillate as Nigel said above.

colin mac · 8th May 2008, 07:58 PM

Same thing happened me the other day .I was wondering what was wrong until I removed the caps. It's still working away now.
I don't have the part no. for the crystal to find out what the story was.

speakerguy79 · 8th May 2008, 08:55 PM

Breadboards have a lot of parasitic capacitance. That said, I've always used xtals with caps on breadboards.

eblc1388 · 9th May 2008, 08:02 AM

I never have problem with crystals not working on breadboard.

The technique is simple.

Place the crystal directly across the MCU pins. Do not use any extra connection wire or links to connect the crystal and its two capacitors.

Most projects I saw on the Net placed the crystal some distance away from the MCU pins, then connect them using links so that it looks neat and tidy.

Don't do that.

skyrock · 9th May 2008, 08:44 AM

Quote:

Originally Posted by eblc1388

Place the crystal directly across the MCU pins. Do not use any extra connection wire or links to connect the crystal and its two capacitors.

Most projects I saw on the Net placed the crystal some distance away from the MCU pins, then connect them using links so that it looks neat and tidy.

Don't do that.

can you explain further? what is the reason behind this?

regards.

ericgibbs · 9th May 2008, 09:03 AM

Quote:

Originally Posted by eblc1388

I never have problem with crystals not working on breadboard.

The technique is simple.

Place the crystal directly across the MCU pins. Do not use any extra connection wire or links to connect the crystal and its two capacitors.

Most projects I saw on the Net placed the crystal some distance away from the MCU pins, then connect them using links so that it looks neat and tidy.

Don't do that.

hi,
I agree, doing this way I never have a problem with the osc..

Kryten · 9th May 2008, 09:30 AM

I agree with ericgibbs and eblc1388

The shorterway from the Crystal to the cnotroller the bether. This way you get the most correct freq. as you can control the stray capacitance (witch is on all PCBs more or less).

There are some ways to control the stray capacitance as shown on the pic below (snapped from DS1302 Datasheet). For Crystals that are used with PICs the caps are used for stability and to keep the crystals BandWith very tight. (I mean to have read somewhere that its like 3-4 Hz)

skyrock · 10th May 2008, 12:40 PM

I've finally found a way to make the crystal to work. After trying to abandon the 16f628 by using 16f84a, i noticed a statement from this spec, saying that it is recommended to use HS mode for crystal higher than 3.5mhz rather than XT mode. so this took me a week to finally solve.

http://www.biltek.tubitak.gov.tr/gel.../pic16f84A.pdf

Page:22, (or 24 of acrobat)

gramo · 11th May 2008, 01:35 AM

Quote:

Originally Posted by skyrock

I've finally found a way to make the crystal to work. After trying to abandon the 16f628 by using 16f84a, i noticed a statement from this spec, saying that it is recommended to use HS mode for crystal higher than 3.5mhz rather than XT mode. so this took me a week to finally solve.

http://www.biltek.tubitak.gov.tr/gel.../pic16f84A.pdf

Page:22, (or 24 of acrobat)

Dont abandon the 16F628! The 16F84 is not nearly as good a tool

8th May 2008, 02:49 PM	#2 (permalink)
Nigel Goodwin Super Moderator Join Date: Nov 2003 Location: Derbyshire, UK Posts: 22,787	Connect the capacitors, they are an essential part of the oscillator circuit - without them it 'may' work some of the time, due to stray capacitance, but it's unlikely to be reliable. __________________ PIC programmer software, and PIC Tutorials at: http://www.winpicprog.co.uk

9th May 2008, 08:02 AM	#7 (permalink)
eblc1388 Experienced Member Join Date: Jan 2005 Location: UK Posts: 1,966	I never have problem with crystals not working on breadboard. The technique is simple. Place the crystal directly across the MCU pins. Do not use any extra connection wire or links to connect the crystal and its two capacitors. Most projects I saw on the Net placed the crystal some distance away from the MCU pins, then connect them using links so that it looks neat and tidy. Don't do that. __________________ L.Chung

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

Similar Threads
Thread	Thread Starter	Forum	Replies	Latest
Will this work?	MagnumForce	General Electronics Chat	7	11th November 2007 12:12 PM
DIY Temperature Compensated Crystal Oscillators TCXO	oliverb	Electronic Projects Design/Ideas/Reviews	2	6th November 2005 09:16 PM
Crystal questions	zachtheterrible	General Electronics Chat	7	9th April 2005 02:19 PM
Crystal oscillator questions	Johnson777717	Micro Controllers	0	19th August 2004 10:41 PM
Need help grounding some exposed metal.	RisingFire	Electronic Projects Design/Ideas/Reviews	6	30th January 2003 06:17 AM

8th May 2008, 03:42 PM	#3 (permalink)
joshua17ss2 Experienced Member Join Date: Mar 2008 Posts: 55	Is it possible its running on the internal oscillator and just ignoring the connection to the external crystal?

8th May 2008, 05:26 PM	#4 (permalink)
picbits Experienced Member Join Date: Jan 2007 Location: Northants, United Kingdom Posts: 467	More than likely the stray capacitance of the PCB or prototype board he's using is enabling the crystal to oscillate as Nigel said above.

8th May 2008, 07:58 PM	#5 (permalink)
colin mac Experienced Member Join Date: Jun 2007 Location: Ireland Posts: 46	Same thing happened me the other day .I was wondering what was wrong until I removed the caps. It's still working away now. I don't have the part no. for the crystal to find out what the story was.

8th May 2008, 08:55 PM	#6 (permalink)
speakerguy79 Experienced Member Join Date: May 2007 Location: USA Posts: 496	Breadboards have a lot of parasitic capacitance. That said, I've always used xtals with caps on breadboards.

10th May 2008, 12:40 PM	#11 (permalink)
skyrock Experienced Member Join Date: Apr 2008 Posts: 40	I've finally found a way to make the crystal to work. After trying to abandon the 16f628 by using 16f84a, i noticed a statement from this spec, saying that it is recommended to use HS mode for crystal higher than 3.5mhz rather than XT mode. so this took me a week to finally solve. http://www.biltek.tubitak.gov.tr/gel.../pic16f84A.pdf Page:22, (or 24 of acrobat)