[sdiy] Error on the SoundBite board!

[mikko.helin at jippii.fi]

Hi,

First, the hex inverter on SoundBite is 74AC04 which is buffered vs. 
74HCU04 which is unbuffered. Anyway, it may be that there are other 
reasons for oscillator problems than the component values used (unless 
they have been verified). In this thread is similar case, here the 
problems are relaxation oscillations (unfortunately the problem was not 
solved, anyway the same 100k feedback resistor was used here as well):

http://www.burnyourbonus.info/sci.electronics.design/thread717.html

There the problem seems to be the hysteresis that keeps up the 
relaxation mode oscillations.

Suggestions to decrease relaxation oscillations in this Intel app note 
for 8048 (!!):

ftp://download.intel.nl/design/mcs51/applnots/23065901.pdf

I think simply adding a 0.1 uF bypass cap directly to pins 7-14 would 
propably help, maybe the 1 Mohm to Vcc as well. There are propably 
reasons why so low Rf was choosen in the first place (cause it's cheap 
demo board they propably wouldn't like to use an extra HCU gate).

Noticed just the Eric's earlier post, maybe the reason the mod did work 
was because the board was cleaned and not the the component swap itself?

-Mikko
---- Paul S. wrote ---------------------

> R20 was 100K change to 1M

>>

> R21 was 1K, change to 100ohms.

Not having the schematics here, what were these resistors?

These are resistors used with the crystal and a 74HCU04 hex inverter to 
form the crystal oscillator. Look at any HCU04 datasheet, the circuit is 
there. If you are a REALLY old fart, then the original circuit is in the 
Motorola McMOS Handbook, c.1977.

Paul S.



__
Wippies-vallankumous on täällä! Varmista paikkasi vallankumouksen eturintamassa ja liity Wippiesiin heti!
http://www.wippies.com/