ds1620 distance signal wiring
Dru
Posts: 10
I'm using a few ·DS1620 temperature sensors around the house.
After a day, the outdoor sensor started returning bad readings occasionally.
The distance is the longest at around 100 feet. I used cat5 cable.· The Data·line seems to be the important one.
I've soldered a 1k ohm resistor at the Data line near the sensor.
Also I've added a 1k ohm resistor at the sending side. That worked reliably for a good 24 hours of straight 1 minute readings.
The other shorter distance devices work fine with a single 1 k ohm resistor on the Data Q line at the sensor only.
As for power in, the·DS1620 needs 5 v· +-· 0.5v.·
Ideas I've found so far:
1. send the·DS1620 a higher power voltage and step it down at the sensor.
2. other combinations of resistors, pull downs... I tried a 10k ohm + pull down on the sending side and that killed everything. Capacitors..
3. use coax cable for the DQ Line.
Any other ideas?
Many thanks,
Dru
Post Edited (Dru) : 3/14/2007 7:42:55 PM GMT
After a day, the outdoor sensor started returning bad readings occasionally.
The distance is the longest at around 100 feet. I used cat5 cable.· The Data·line seems to be the important one.
I've soldered a 1k ohm resistor at the Data line near the sensor.
Also I've added a 1k ohm resistor at the sending side. That worked reliably for a good 24 hours of straight 1 minute readings.
The other shorter distance devices work fine with a single 1 k ohm resistor on the Data Q line at the sensor only.
As for power in, the·DS1620 needs 5 v· +-· 0.5v.·
Ideas I've found so far:
1. send the·DS1620 a higher power voltage and step it down at the sensor.
2. other combinations of resistors, pull downs... I tried a 10k ohm + pull down on the sending side and that killed everything. Capacitors..
3. use coax cable for the DQ Line.
Any other ideas?
Many thanks,
Dru
Post Edited (Dru) : 3/14/2007 7:42:55 PM GMT
Comments
Slightly exotic and fairly pricey, but you could go wireless with the eb500 embeddedlue appmod.
Ray
I've been· reading a bit about cable capacitance , termination and such. I think there is a way I can make this work with capacitors and resistors, or perhaps transistors..
Still looking for ideas! thanks -Dru
Check out my live data graph from C# web application from the first day of data:
Post Edited (Dru) : 3/14/2007 10:11:33 PM GMT
Be sure there is a bypass capacitor across the power supply right next to the DS1620. Something like a 4.7 to 10 microfarad tantalum. The DS1620 is sensitive to power supply noise. In subborn cases it is better to use the single shot mode instead of the continuous mode. But that is probably not necessary if the readings are bad only occasionally. That would be more related to noise pickup on the data line.
The seres resistors would help to protect against damage from miswiring or surges, but they could hurt as much as help with the noise level on the signal wires. They slow down the response, which could help with some kinds of shot noise but is in a fine balance with the cable capacitance. Better to move to a shielded twisted pair. But try the power supply capacitor first, if there is not one there already.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com
So far one simple change seemed to correct the faulty readings. Since I use 5 wires out of the 8 in the cat5 cable, I simply connected the 3 extra unused into the DataQ line-- and only on the microcontroller side! They are not connected to anything at the other end!
For some reason its working solid for the last 12 hours.
Check it out: dru.ackerman.tv