Author Topic: Fake DS18B20  (Read 2743 times)

0 Members and 1 Guest are viewing this topic.

Online oPossumTopic starter

  • Super Contributor
  • ***
  • Posts: 1472
  • Country: us
  • Very dangerous - may attack at any time
Fake DS18B20
« on: April 19, 2020, 09:29:29 am »
I had suspected many DS18B20 are fake and this confirms it. Lots of great info on genuine and fake DS18B20 temperature sensors. Differences in characteristics and some die shots.

https://github.com/cpetrich/counterfeit_DS18B20
 
The following users thanked this post: kripton2035, felixd, coromonadalix, magic, DrG

Offline magic

  • Super Contributor
  • ***
  • Posts: 7454
  • Country: pl
Re: Fake DS18B20
« Reply #1 on: April 19, 2020, 10:56:55 am »
It's been known for a while (a thread or two on this forum alone) but that's quite an impressive documentation effort.

That being said, over the next few years they will probably fix the most obvious bugs if this information spreads ;)
 

Offline james_s

  • Super Contributor
  • ***
  • Posts: 21611
  • Country: us
Re: Fake DS18B20
« Reply #2 on: April 19, 2020, 05:00:42 pm »
Well if they fix the bugs so the fake parts are indistinguishable from the real ones I'd actually be ok with that. Then I could use cheap clones confident that my projects will work correctly instead of having to spring for overpriced genuine ones.
 

Online coromonadalix

  • Super Contributor
  • ***
  • Posts: 7014
  • Country: ca
Re: Fake DS18B20
« Reply #3 on: April 19, 2020, 05:05:00 pm »
Damn another thing to check ...
 

Offline magic

  • Super Contributor
  • ***
  • Posts: 7454
  • Country: pl
Re: Fake DS18B20
« Reply #4 on: April 19, 2020, 05:49:59 pm »
Well if they fix the bugs so the fake parts are indistinguishable from the real ones I'd actually be ok with that. Then I could use cheap clones confident that my projects will work correctly instead of having to spring for overpriced genuine ones.
I will not be OK with counterfeits even if they are functional. Sell it under your own brand, assholes.

Besides, I did say: the most obvious bugs. By which I mean: things like serial numbers, responses to undocumented commands, whatever other quick check you might try without actually measuring every damn spec guaranteed by Dallas' datasheet. Accuracy is the last thing they care about; the worst of the fakes I have encountered were 2°C off even at room temperature. Wanna bet that this rubbish is actually tested at -55°C? ;D
 

Offline james_s

  • Super Contributor
  • ***
  • Posts: 21611
  • Country: us
Re: Fake DS18B20
« Reply #5 on: April 19, 2020, 07:12:52 pm »
That would certainly be preferable. I'd be happy to buy a DS18B20 with a different brand on it if I evaluated a few samples and found them to perform equally to the Dallas parts. That's true of so many of the fake Chinese parts, the ones that work properly could be sold as compatible parts under their own brand and nobody would feel ripped off.
 

Offline wizard69

  • Super Contributor
  • ***
  • Posts: 1184
  • Country: us
Re: Fake DS18B20
« Reply #6 on: April 19, 2020, 07:16:38 pm »
This is not a good idea at all.  I have nothing against competitive parts and marketed as such but really have massive hate against what are rip off parts -true fakes. 

Well if they fix the bugs so the fake parts are indistinguishable from the real ones I'd actually be ok with that. Then I could use cheap clones confident that my projects will work correctly instead of having to spring for overpriced genuine ones.

If you don’t like to pay for the genuine part design in something different.  Really it is that simple, use an honestly sold competitors product not something marketed to be a rip off. 
 

Offline james_s

  • Super Contributor
  • ***
  • Posts: 21611
  • Country: us
Re: Fake DS18B20
« Reply #7 on: April 19, 2020, 07:21:38 pm »
I'm just talking about hobby projects. If you're designing something into a commercial product then it goes without saying that you should be buying genuine parts from a reputable distributor.

If it's coming direct from China and costs 1/10th the quantity price from the manufacture then just assume it's fake.

I've used quite a few ridiculously cheap Maxim LED drivers and I've never found one that didn't perform exactly like the known genuine samples I got from Maxim years ago. I don't know whether they're fakes or gray market 3rd shift parts or what but they work fine and they are affordable. If I had to pay what Maxim was asking I'd use something else.
 

Offline dgtl

  • Regular Contributor
  • *
  • Posts: 183
  • Country: ee
Re: Fake DS18B20
« Reply #8 on: April 19, 2020, 08:24:11 pm »
I've had the same struggle with DS18B20-based sensors. As I needed ready-made sensors with wires attached, it was very difficult to find good ones. The whole world is full of the same stuff as sold on aliexpress, each having different kind of counterfeit chips in them. Even if they are kind-of-okay at 20C, then at 80C the readings are totally bogus, not even +/-5C. They have issues with parasite powering and are terribly bad at long lines with some noise.
Finally, I found one type of sensors, that so far have been precise enough and are digitally indistinguishable from real DS18B20 to believe that they actually contain the real chips (sensors with red-blue-white wires and red tag with black writing at wire end, about 4x more expensive).

One good indication of a real DS18B20 is scratchpad byte 6 (Reserved, non-constant). It seems, that this register is legacy from DS18S20 and it should read the same value for the same temperature readout and different values for different temperatures. So far, I havent found a clone, that would implement this register.
 

Offline DrG

  • Super Contributor
  • ***
  • !
  • Posts: 1199
  • Country: us
Re: Fake DS18B20
« Reply #9 on: April 22, 2020, 02:23:18 pm »
I am impressed with the effort of Chris Petrich https://github.com/cpetrich/counterfeit_DS18B20 and decided to test out the ones I had on hand. Thought I would share the results (only one real surprise).

Ordered from Mouser:



8-D6-A1-E8-05-00-00-F6: ROM ok.
  Scratchpad Register: BE/01/4B/46/7F/FF/02/10/FE
  Info only: Scratchpad bytes 2,3,4 (4B/46/7F):  Maxim default values.
  Scratchpad byte 5 (0xFF):  ok.
  Scratchpad byte 6 (0x02):  ok.
  Scratchpad byte 7 (0x10):  ok.
  0x4E modifies alarm registers:  ok.
  0x4E accepts 10 bit resolution:  ok.
  0x4E preserves reserved bytes:  ok.
  0x4E accepts 12 bit resolution:  ok.
  0x4E preserves reserved bytes:  ok.
  Checking byte 6 upon temperature change: not necessary. Skipped.
  --> Sensor responded like a genuie Maxim.

Ordered from Mouser:



28-B8-47-0B-06-00-00-45: ROM ok.
  Scratchpad Register: C6/01/4B/46/7F/FF/0A/10/17
  Info only: Scratchpad bytes 2,3,4 (4B/46/7F):  Maxim default values.
  Scratchpad byte 5 (0xFF):  ok.
  Scratchpad byte 6 (0x0A):  ok.
  Scratchpad byte 7 (0x10):  ok.
  0x4E modifies alarm registers:  ok.
  0x4E accepts 10 bit resolution:  ok.
  0x4E preserves reserved bytes:  ok.
  0x4E accepts 12 bit resolution:  ok.
  0x4E preserves reserved bytes:  ok.
  Checking byte 6 upon temperature change: not necessary. Skipped.
  --> Sensor responded like a genuie Maxim.

Ordered from Banggood (this was the surprise)



28-8A-8B-C7-06-00-00-EE: ROM ok.
  Scratchpad Register: A3/01/4B/46/7F/FF/0D/10/CE
  Info only: Scratchpad bytes 2,3,4 (4B/46/7F):  Maxim default values.
  Scratchpad byte 5 (0xFF):  ok.
  Scratchpad byte 6 (0x0D):  ok.
  Scratchpad byte 7 (0x10):  ok.
  0x4E modifies alarm registers:  ok.
  0x4E accepts 10 bit resolution:  ok.
  0x4E preserves reserved bytes:  ok.
  0x4E accepts 12 bit resolution:  ok.
  0x4E preserves reserved bytes:  ok.
  Checking byte 6 upon temperature change: not necessary. Skipped.
  --> Sensor responded like a genuie Maxim.

All the ones below (from various export houses, Bangood, DIY, off of Amazon) were all:
--> Sensor appears to be counterfeit based on 4 deviations.








- Invest in science - it pays big dividends. -
 
The following users thanked this post: oPossum, mycroft


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf