Simple question about ceramic oscillators

John Whitfield

I got the oscillator to function correctly on the bread board, but when I soldered it onto a regular board it doesn't work. Is it possible that got the oscillator to hot during soldering? I don't think I kept it hot to long, but I did use a normal size iron, not one of the smaller ones.

Thanks,
-John Whitfield

cbmeeks

It is totally possible. I definately recommend a heat sink when working with components. The iron shouldn't touch the pins for more than about 1-2 seconds.

cbmeeks

Guenther Daubach

John,

in the meantime, I have hand-soldered hundreds of these Murata ceramic resonators without killing one. I found one defective on a couple of hundred boards that I give away for machine-soldering. So, just one defective resonator out of about 800 is not a bad ratio. But you never know, maybe, you got the one with the 1/800 probability right at the beginning of your SX career . Nevertheless, it is a good idea using a small iron, and applying heat for a short time only, and let the solder joint cool for a while before soldering the next pin.

Another idea is to use a small 3-pole socket for the resonator, like on the SXTech board. With this socket installed, you can remove the resonator when you want to do on-board debugging without the need for a jumper to open the OSC1 line to the resonator when the SX-Key is in action. For just programming the SX you don't need to pull out the resonator - just for debugging.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Greetings from Germany,

G

DaveG

John,

On your regular board, do you have a resistor across the resonator(OSC1 to OSC2)? Like a 10K or 1 meg.
Sometimes the resonator has difficulty starting without this resistor. See Appendix E, Table 35 of the SX Key
Development System Manual for the correct value.

Dave G

william chan

John,

I think your resonator is too far away from the SX.
Move it closer.....


William

PJMonty

William,

Since John didn't post a picture of his PCB layout, what makes you think the resonator is "too far away"?

Thanks, PeterM

Miner_with_a_PIC

I made the mistake of locating my 50 Mhz resonator about 2 inches away from my SX chip during my PCB design. I found it necessary to use a 75K resistor in series with the OSC1 and OSC2 pins to get stable and robust performance. Fortunately I wasn't dead in the water, but it did detract from performance. Just a thought, there is considerable parasitic capacitance on bread boards, as such, the PCB is not the same system that you tested on. In any case, do solder a resistor in parallel as this is a good general practice. The value depends on the resonator, I typically start at a high resistance value like 1M and move down, all the time checking to ensure good start up and operational robustness; I have never gone below 75K.

JT

Post Edited (Miner_with_a_PIC) : 6/30/2005 11:41:13 PM GMT

Guenther Daubach

Oops,

I always follow the specs shown in the Ubicom SX datasheets. With a 50 MHz Murata ceramic resonator, I parallel 10 kOhms (as specified), and never had problems with this configuration. The lengths of the OSC1/2 leads between the SX pins, the resonator, and the resistor, usually are less than some millimeters here.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Greetings from Germany,

G

Miner_with_a_PIC

Guenther , I agree that following specifications is typically what one should do, but my experience with my "poor" PCB resonator placement seems to indicate that the oscillator position can be extended beyond a few millimeters. For optimum performance in the future I will definately bring the oscillator much closer to the SX but I now have a sense of the upper limit on distance and won't worry if I place an oscillator 3 or 15 mm away; they will both work. It is important to also note that minimizing the osc1/osc2 circuit loop area will also minimize inductance of the oscillator circuit. I did not run my lines in such a manner as to minimize this area and the circuit still works. Bottom line here is those oscillator circuits take a licking and keep on ticking. Use some caution when preparing you pcb layout and all will be turn out just fine.

JT

John Whitfield

I burned it up, because I removed the oscillator and replaced it with a machine pin set and the new oscillator worked.

Thanks,

-John