i was come to a wierd machine condition. it will power up "properly" sometimes, sometimes, it just doing anything wrong. such as moving the motors down etc. i suspecting the main board reset circuit was somehow have a problem. when i put a reset switch (there is a dedicated space for it), i can do it when on, and all is played normally. the chip who control it is MC34064. it generate /reset line for the rest. i had try to put a cap in addition to pull up resistor, but no luck. i think i am gonna replace it with a new one.

yet, the problem is, i am not sure it is sold in local shop here. so, i think i must replace it with some discrete component. i just know RC combination for active high reset line, can someone told me for the active low one?
or any suggestion?


thanks
Posted on 2004-01-09 05:24:39 by dion
yes probably the cause is in that component, to be sure, could you try to simulate the power on reset (gnd) by hand some until seconds after the power on ?

:)

Yeh maybe you can try to put a high capacitor instade, just for test.

B7
Posted on 2004-01-09 11:16:23 by Bit7
The first step of desiging a RC POC circuit is choosing the R value. This value should make a good pull-up R, (forcing a logical 1) or good pull-down (forcing a logic 0) for the type of logic you are using. This is easier with MOS logic now then it used to be with TTL, as TTL needs a non-negligable ammount of current to be pulled down.

OK, having the R you pick a C to give you a timeout as long as you need... the worst case power up time for the voltages to be stable, plus a fudge factor. Make RC at least this time, or 10 times as much, and don't forget large cheap caps tend to be +/- 20% in value.

Just make it a BIG cap and don't worry.
Posted on 2004-01-10 14:53:47 by Ernie
Thanks all...

i make a pull up resistor and cap connect each other to gnd, while i take the line out of the middle to the reset pin, is it okay?
Posted on 2004-01-13 05:58:13 by dion
It should be pointed out that the MC34064 and similar circuits do a much better job than the simple RC circuits. It provides not only power-on reset, but also brown-out protection.

Therefore, before going for the RC circuit, I would test the MC34064 first. Remove it from the circuit, connect a 10k resistor from the /RESET output to Vin and apply a variable voltage ( 4.0 to 5.0V) between Vin and GND.

If the output switches between HIGH and LOW as the input voltage is varied around the threshold (approx. 4.6V), then the device is OK in all probability and the rest of the circuit shouldbe checked:
1. is the pullup in the original circuit OK?
2. is there a capacitor from /RESET to GND that could be leaky? (usually a cap is used with this device to create a min. reset pulse width)
3. is there another device connected to the /RESET output that is somehow forcing it to LOW, or some intermediate state?

This is what I would do.
A better part, that guarantees a minimum reset pulse width, is the DS1233 from Maxim (formerly Dallas Semiconductor) which could provide a replacement, but the pinout is different.

If you are set to go with the RC, just connect the C to GND and the R to +5V. Their junction point should go to the /RESET input of the micro. I generally like using Schmitt trigger gates between the RC and the actual RESET input to the micro. Just pay attention to the active state of the reset signal the micro needs. You may need two gates to get the right active state.
Posted on 2004-01-13 11:50:14 by VVV
Hi dion,

If you'r still keep searching the mc34064 reset chip, take a look at the dead old maxtor harddisk (??? meg. size).
Posted on 2004-02-10 02:26:35 by andaman
thanks andaman, i cant believe that i just bought that thing yesterday, and still not test the board yet until now. i'll let you all know about it :)
Posted on 2004-02-10 21:16:07 by dion