DS18B20 - marked as obsolete(!)


Senior Member
Hopefully there's been a terrible mistake somewhere! I'm just gearing up for a little project involving over 10 DS18B20 temperature sensors, but I'm struggling to find any sources. I started on the flea market and the big-forest-place and the bad reviews/reports of fakes put me off, so I moved on to the professional electronics suppliers. Most seem to be out of stock; then (shock horror) Digikey have them listed as 'obsolete'!
The manufacturer - Maxim - have the details in their parametric search, but all links to the page seem to be broken at the moment.
I of course remembered the PICAXE shop and came here; but I'm now concerned that the PICAXE store description says 'new die revision'. I'm aware there are plenty of variants of the DS18B20 - but there seems to be little information on which ones are directly compatible with our friendly ReadTemp12 command.

So, does anyone know which variants are safe to use? Are there any later alternatives? Can hippy/Technical reveal exactly which die revision is stocked in the store?
I'm actually worried about not using the so familiar DS18B20!!



Futurlec has the SOIC8 outline chip

Newark has the TO92 format

also see:



Senior Member
It appears Digikey has obsoleted the DS18B20 in favour of the DS18B20+ which can be found here -


The "+" indicates a lead-free package and is presumably for RoHS compliance. As far I can tell there should be no functional differences between a plus and non-plus version. The READTEMP commands should work with any DS18B20.

Digikey seems to have differentiated between the plus and non-plus versions of the DS18B20 whereas ourselves, and near everyone else, have simply continued to refer to the plus version by its non-plus name. Even Digikey do that in their URL reference. It would not surprise me if all are now plus versions no matter how they are named or where purchased.

As to our "new die" description; I expect that meant new for the time, and has actually been there since the description went up some years ago.

I believe the "new die" was the one which reduced the temperature conversion time down from a fixed, about one second, time to however long it takes. That happened quite a while ago. I believe the M2 and X2 PICAXE chips poll for the temperature conversion to complete so they handle both old and new, whether it takes one second or less. That may actually be the case for all PICAXE chips.

Whatever die version we have is likely whatever die version is supplied when we buy the parts in. There shouldn't be any functional differences no matter what version of die there is.


Senior Member
Thanks techEldar and hippy - that's very helpful information.
It doesn't help that the Maxim page is down for 'DS18B20' and there is no mention of the variants (i.e. +) elsewhere on their site!

Good news also about the temperature conversion time - I'm polling 8 sensors in one circuit, and I was thinking that an 8 odd second delay might cause problems for higher speed monitoring. On to breadboarding and hopefully all will be well.

Thanks again.


Senior Member
an 8 odd second delay might cause problems for higher speed monitoring
Note that if it gets to be a problem, on the X2s you can use the one-wire commands to initiate the temperature read on all 8 DS18B20s, and then after a second, read them all.
owout C.0,%1001,($CC,$44) ' initiate one-wire temperature read
owout C.1,%1001,($CC,$44) ' initiate one-wire temperature read
pause 1000
owout C.0,%0001,($CC,$BE)
owout C.1,%0001,($CC,$BE)
owin C.0,%0000,(b46,b47) ; read in result
owin C.1,%0000,(b48,b49) ; read in result


Senior Member
Thanks Ibenson, appreciated. I'm currently on 14M2s, but I think I've got some 20X2s hiding somewhere.
So hard to get a hobby back up and running when life is so busy!