On 03/09/2011 06:40 PM, Doug Collinge wrote: > So, I had a look at the datasheet to see what would happen if the Vdd > lead on a DS18B20 was left floating. All there is to work from is a > block diagram, which doesn't contain enough information. So I asked tech > support at Maxim the following question: > > The block diagram of the DS18B20 shows a block labeled "Power Supply > Sense" that determines whether the chip has Vdd power or not. The > datasheet says to ground Vdd for parasitic operation and it's clear that > the input to Power Supply Sense will see 0V in that case. My question is > what will it see if the Vdd input is allowed to float? If the capacitor > is charged the diode will be a high impedance. What is the impedance of > the Power Supply Sense input? If it is also a high impedance that might > explain why someone is seeing inconsistent behaviour with Vdd floating. > > They promptly replied as follows: > > The Vdd 'sensing circuit' is a high-impedance node, but the Vdd pad is > not. Simply disconnecting the Vdd supply will leave the capacitor > charged and the component would therefore not be powered parasitically. > The capacitor cannot provide enough current to power the component, but > defeats the 0V detection required. > > Floating Vdd is an invalid/undefined condition for the DS18B20. > > We also sell a DS18B20-PAR+ (TO-92 only), which precludes the necessity > for the Vdd pin to be grounded on a board. > > So I think that's a pretty clear explanation: the component expects Vdd > to be wired either to +5 or ground. If it isn't it may think it is > powered when it is not and not function correctly. > > In the specific situation described, outdoors but sheltered from rain, > but high humidity my guess is that the humidity was supplying enough > conductivity to a positive voltage source to keep the Vdd terminal high > enough for the internal sensor to conclude that the chip was powered. It > would therefore not be able to do a conversion correctly on the power > stored in the capacitor and report an error. > > The obvious conclusion is that people should follow the clear > instructions of the manufacturer and ground Vdd for parasitic operation.
Thanks Doug, great information. I guess I read bits and pieces of the DS18B20 datasheet a long time ago and either forgot this particular detail (after things worked for me without following the requirement), or missed it altogether: "When the DS18B20 is used in parasite power mode, the VDD pin must be connected to ground." As you say, it does not get any clearer than that. Their (Maxim tech. support's) response seems to imply that things would not work by leaving Vdd floating. However, it has worked for me for months, and even iButtonLink seem to be shipping sensors with Vdd floatin (the following comes from the manual for iButtonLink's T-Sense sensor [http://www.ibuttonlink.com/pdf/manuals%20for%20t-sense%2005-22-08.pdf]): "NOTE: Important information There is a small chance that the bus could become unstable when using T-Sense units as the only sensor on the bus. This condition comes about as a result of the new design of the T-Sense, using the DS18B20 sensor. The third lead on the DS18B20, (Vdd) is left floating, (to allow design flexibility for sophisticated networks) which leaves the possibility of unpredictable performance. Our engineering department has not been able to create this condition. Experiments with > 30 T-Sensors on the bus, and a variety of interfaces, have shown all devices working properly. In the unlikely event of unstable behavior on the bus, resolution can be accomplished by simply connecting pin 2 (+5 volts) to pin 6 (Aux). Alternately, any MS-xx product on the bus will provide the jumper internally, or the use of a LinkHub as the interface." So I either don't understand what Maxim tech. support said, or I do understand it but then it's mistery why so many DS18B20 seem to work fine with a floating Vdd ;-) In any case, thanks again for the information, and I'll be sure to ground Vdd little by little (but no rush since things are working great for me ;-) ) Your explanation on why high humidity could have caused the 85 degree problem makes sense to me. Cheers, Eloy Paris.- P.S. Regarding iButtonLink's statement that they have not been able to recreate network instability with more than 30 sensors on a bus all with floating Vdd, perhaps they haven't tried in a high humidity scenario like was the case for the one sensor that I had outside? ------------------------------------------------------------------------------ Colocation vs. Managed Hosting A question and answer guide to determining the best fit for your organization - today and in the future. http://p.sf.net/sfu/internap-sfd2d _______________________________________________ Owfs-developers mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/owfs-developers
