[ntp:questions] Trying to use Dimension 4 time keeper

unruh unruh at invalid.ca
Tue Sep 24 14:05:29 UTC 2013


On 2013-09-24, Uwe Klein <uwe at klein-habertwedt.de> wrote:
> detha wrote:
>> On Sun, 15 Sep 2013 18:56:26 +0100, David Taylor wrote:
>> 
>> 
>>>On 15/09/2013 18:37, unruh wrote:
>>>[]
>>>
>>>>Yes, I saw it. Not sure what the voltages are on the Sure unit -- the
>>>>direct output of the PPS but I think 5V which is still a bit high. I
>>>>guess one could put a diode in series with the line to bring it down to
>>>>3.3V.
>>>
>>>I would recommend using a restive divider - perhaps 10K and 4.7K to start
>>>with - and monitoring the voltage into the Raspberry Pi.  Using resistors
>>>provides a current limit, and allows the voltage to get down to zero.  If
>>>the voltage is well below 3.3V, the upper 4.7 K resistor could be reduced.
>>> Best to ask on one of the Raspberry Pi support groups as someone has
>>>likely already done the measurements.
>> 
>> 
>> http://jamesreubenknowles.com/level-shifting-stragety-experments-1741
>> might be of interest. For converting serial signals it works fine,
>> for PPS where one worries about being 100ns off maybe not so much.
>> 
> A lot of 3.3V logic comes in 5V tolerant. Look up the datasheet.

The Raspberry Pi specifically states that it is NOT 5V tolerant. 

>
> For a resistive divider ( that includes the diode hacks ) you need
> to compensate the input capacitance i.e. a combined resistive/capacitive divider
> for good rise and fall times.
>
> then use shottky diodes.
>
> one variant is to use a reasonably strong pullup to the input supply voltage
> and a diode ( anode * resistor * input )
> pull down on diode cathode ( usually higher current capability from the driver side) for low.
> High voltage on diode cathode will leave the input to go up by way of the pull up resistor.
>
> uwe



More information about the questions mailing list