??? 03/25/07 04:17 Modified: 03/25/07 04:20 Read: times |
#135826 - Use an external oscillator! Responding to: ???'s previous message |
The built-in oscillator of AT89S8253 is proven to be crap! The strange burden caps' recommendation given in the datasheet is an indication for that.
I would suggest to use an external oscillator. If you have a (very expensive) ultra low capacitance FET-probe, then you can check the amplitudes of oscillation at the XTAL1 pin (input) and XTAL2 pin (output). But take note: This is a rather academic approach, because this sort of measurement is really sophisticated. Why? Because you always relevantly change the oscillator characteristics when connecting a whatsoever natured probe to it. You should see a big square wave at XTAL2 and a big sine at XTAL1. DC average level should be at Vcc/2 and amplitudes at about 5Vpp. The level at XTAL1 should be high, but there should not be any clipping. If the amplitudes are too small, then you can try to decrease both burden caps, as Daniel already stated. Decreasing both burden caps by the same amount (!) will increase the amplitudes, while increasing them (by the same amount) will decrease the amplitudes. If the DC average level at XTAL1 is wrong, then you can try to connect a 1M resistor from XTAL1 to GND or from XTAL1 to XTAL2. If, on the other hand, the amplitudes are very high, then you can try to insert a resistance of about 50...200Ohm between the XTAL2 pin an oscillator components. In any case, try to get the best crystal you can. A good crystal offers a very low dampening, means a good crystal shows a low equivalent series resistance. Often, the problems dissapear, when taking a realy good crystal... Kai |
Topic | Author | Date |
AT89S8253 clocking problem | 01/01/70 00:00 | |
The scope probe is the clue!! | 01/01/70 00:00 | |
The tale of a 1000 fingers | 01/01/70 00:00 | |
Alex did what the datasheet demands | 01/01/70 00:00 | |
1M across cap | 01/01/70 00:00 | |
are you sure it is not 1/3? | 01/01/70 00:00 | |
I had the same problem | 01/01/70 00:00 | |
Works!! But is it a solution | 01/01/70 00:00 | |
is it 1/4 or 1/3 | 01/01/70 00:00 | |
Speed unstable | 01/01/70 00:00 | |
May be clock amplitude... | 01/01/70 00:00 | |
Crystal frequency | 01/01/70 00:00 | |
An interesting article... | 01/01/70 00:00 | |
Use an external oscillator! | 01/01/70 00:00 | |
wait a year or more for the errata![]() | 01/01/70 00:00 |