Page 2 of 2

Re: Wake My Hardware Up Before You Go-go

Posted: Sun Dec 18, 2011 10:30 pm
by Dan
74LS05 looks ok to me? shall skype you sometime soonish to clarify this and a few other things.

Re: Wake My Hardware Up Before You Go-go

Posted: Sun Dec 18, 2011 10:36 pm
by Fred
OK, I'm online and ready when you are :-)

Re: Wake My Hardware Up Before You Go-go

Posted: Sun Dec 18, 2011 11:14 pm
by Dan
umm sorry, i meant when i am not at work. LOL. unless you feel like calling a land line from your end?

Re: Wake My Hardware Up Before You Go-go

Posted: Sun Dec 18, 2011 11:23 pm
by Fred
Hmm, no worries, so tomorrow morning my time, then? Suits me :-) I removed your number from the post btw, but kept a record of it for me.

Re: Wake My Hardware Up Before You Go-go

Posted: Sun Dec 18, 2011 11:32 pm
by Dan
cool! thats my direct work number! :-)

Re: Wake My Hardware Up Before You Go-go

Posted: Sun Dec 18, 2011 11:42 pm
by Fred
Re this stuff, I think I killed two birds!

If our resistors can stomach only 1/4 watt, then we're talking 1k above the zener and 500 ohm (or 470) across the zener to pull the input down quickly and effectively. If they can stomach less than that, then we need other values. Given our choice of 470 on ADC inputs, we should probably have fairly good resistors anyway in order for them to survive 14.4V connection.

The same should therefore apply to the digital input circuits. The zener is not the limiting factor, resistor heat is. The digital inputs won't have the resistor across them though, only the pull up, or will they? I guess you want them to have a definitive connected and disconnected state too.

About the zeners, we currently have them at 5.6V however that means we're loading the schottky diodes by DEFAULT in the digital inputs and wakeup, which I don't like. Better to reduce to 5.1V and put up with a slightly below 5V signal level. This is OK because the switch point on the CPU and on the logic chip is much lower than 5V. However, thinking about this some more, the pull down across the zener would be better as a higher value such that the zener was definitely loaded and pulled high properly. maybe 1k to signal and 1k across zener makes the most sense and keeps the BOM simple.

Thoughts?

Fred.

Re: Wake My Hardware Up Before You Go-go

Posted: Sun Dec 18, 2011 11:48 pm
by Fred
Hmmm, thinking again, during cranking voltages could be very low, and we don't want accidental triggering, so perhaps even 1k series limit and 2.4k pull down would be better? That way the input gets the majority of the voltage and can go as low as 6 or 7 V without issue and still look 'on'.

Re: Wake My Hardware Up Before You Go-go

Posted: Sun Dec 18, 2011 11:51 pm
by Dan
agreed RE: cranking voltages.

I like the 1K series limit and 2.4K pull down concept! :-)

Re: Wake My Hardware Up Before You Go-go

Posted: Mon Dec 19, 2011 12:30 am
by Fred
OK, I won't add it to the list yet, I'll do the tests first and try to get some basic sleep/wake code happening reliably first.