Does not work on micro:bit V2 hardware

[timleeanieducation]

The dstemp extension does not work on the micro:bit V2. It returns an error code 3 "Read Timeout".

[bsiever]

@timleeanieducation Just for more info, does it work on v1 for you?

Also, where did you get your sensor? There are a variety of copies that have different quirks. So far I've only tested it on sensors that are from a reliable supplier.

I my tests with my sensors, it worked well on both V1 and V2 hardware.

[timleeanieducation]

It works perfectly fine for V1, using the latest stable micro:bit makecode, with the V2 default pull fix (lancaster-university/codal-microbit-v2#17).

You may be right about the sensor supplier. I'll get some DS18B20 for testing from DigiKey, which are manufactured by Maxim Integrated.

I'm also wondering if the V2's sink current capability may have something to do with it. V2 pins (P0 to P2) are not capable of sinking 4mA.

[bsiever]

Thanks for the update and the additional testing! I have tested it on a v2 and it worked...but I'll do more testing as time permits.

I have an alternate version for 2-wire mode, which I've also tested on the v2...but if this (3-wire) doesn't work, I'm pretty skeptical that it will. None the less, it's listed as microbit-dstemp-2wire. The repo shows the wiring for 2-pin mode: https://github.com/bsiever/microbit-dstemp-2wire

[bsiever]

@timleeanieducation I just did some additional testing: I had two sensors (one in 3-wire mode and one in 2-wire mode) and two V2 micro:bits. Both micro:bits worked with each sensor and the respective extensions. The 3-wire setup has a 1M cable and uses a breadboard. The 2-wire version ist just the sensor and alligator clips.

Are you able to share where you got your sensor? I may be able to order one and try it out.

My sensors came from http://yourduino.com/. At the time I ordered them (~1.5 years ago) they had a reputation for supplying authentic DS18B20s. Here's a list of reliable sources: https://github.com/cpetrich/counterfeit_DS18B20#sources

[timleeanieducation]

The cpetrich link is very useful. I downloaded the code to test my sensor. This is what I got:

1. discover_fake_DS18B20.ino
   28-CC-3D-75-D0-01-3C-7D: ROM does not follow expected pattern 28-xx-xx-xx-xx-00-00-crc. Error.
   Scratchpad Register: 90/01/55/05/7F/A5/A5/66/B4
   Info only: Scratchpad bytes 2,3,4 (55/05/7F): not Maxim default values 4B/46/7F.
   Scratchpad byte 5 (0xA5): should have been 0xFF according to datasheet. Error.
   Scratchpad byte 6 (0xA5): unexpected value. Error.
   Scratchpad byte 7 (0x66): should have been 0x10 according to datasheet. Error.
   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 appears to be counterfeit based on 4 deviations.

2. classify_fake_DS18B20.ino
   28-CC-3D-75-D0-01-3C-7D: Family D2 (Clone w/o parasitic power).

The sensor is encpasulated in waterproof shield, purchased from TaoBao, and obviously fake. There are hundreds of shops selling them, and they are really cheap (around 3 to 5 RMB). What I find intriguing is why it worked perfectly on micro:bit V1 but not on V2.

I have ordered a 3-wire and a 2-wire DS18B20 from DigiKey to test things out. Will keep you informed of progress.

[bsiever]

@timleeanieducation Thanks for the extra testing! The sensors require pretty precise timing, which I had to "hand tune" based on the processor speed. So the code used for the v2 is a little different than the v1. It's possible my "tuning" is just a little off for your sensor.

I'll try to do a few tests with fake sensors when time permits. Hopefully I can update the extensions so it'll typically work for nearly everyone. (That being said, I think the parasitic power / 2-wire mode will be great for many people and the fake sensors don't seem to support it very well)

[timleeanieducation]

A DS18B20 and DS18B20-PAR arrived today from DigiKey. Tested using discover_fake_DS18B20.ino and micro:bit V1/V2.
bsiever/microbit-dstemp 0.1.21
bsiever/microbit-dstemp-2wire 0.1.6

DS18B20:
28-57-9F-F1-0C-00-00-28: ROM ok.
Scratchpad Register: 88/01/4B/46/7F/FF/08/10/76
Info only: Scratchpad bytes 2,3,4 (4B/46/7F): Maxim default values.
Scratchpad byte 5 (0xFF): ok.
Scratchpad byte 6 (0x08): 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.
Both micro:bit V1 and V2 produced correct temperature.

DS18B20-PAR:
28-E8-CC-FD-0C-00-00-C2: ROM ok.
Scratchpad Register: FF/07/4B/46/7F/FF/01/10/2F
Info only: Scratchpad bytes 2,3,4 (4B/46/7F): Maxim default values.
Scratchpad byte 5 (0xFF): ok.
Scratchpad byte 6 (0x01): 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: Inconclusive. Please change sensor temperature and repeat.
--> DS18S20 counterfeit test not completed, otherwise sensor responded like a genuie Maxim.
micro:bit V1 produced correct temperature. V2 output is -Infinity.

[bsiever]

@timleeanieducation Wow! Thanks for all the testing. It looks like I definitely need to do more testing review with the DS18B20-PAR on the v2....but I still plan to do more testing with "fakes" too. Just to check: did you test either/both sensors with just 3-wire mode (the plain "dstemp" extension) or did you test either/both with the 2-wire mode too (dstemp-2wire)?

Thanks again --- this really helps!

[bsiever]

@timleeanieducation Ah, an update. I didn't realize what the difference was between the PAR and non-PAR versions. The non-PAR version should be able to do either parasitic power or non-parasitic power. Hopefully it will work with either the dstemp or dstemp-2wire extensions if wired correctly.

The PAR version is already internally wired in parasitic power mode, so it would only be expected to work with the dstemp-2wire extension. (See: https://www.maximintegrated.com/en/design/technical-documents/app-notes/2/203.html).

This doesn't explain why the PAR would work on v1 and not v2 --- it may be the difference in the timing (which I can change), or power, or the internal resistance of the v2 (the internal resistance on both the v1 and v2 is technically out-of-spec for the DS18B20).

[timleeanieducation]

Thanks @bsiever for the detailed information. I did not realize you can do a 2-wire conection on a regular DS18B20. I have re-tested all my devices making sure they are wired correctly and using the correct extensions. The results are shown below:

Wiring diagrams
DS18B20 External Powered Circuit

DS18B20 Parasite Powered Circuit

DS18B20-PAR Parasite Powered Circuit

[timleeanieducation]

Just a side note, the batch of sensors I bought from the same shop in TaoBao appear to be three different versions. Under the 3-wire mode and using bsiever/microbit-dstemp 0.1.21 they work for micro:bit V1 but produced a "Read Timeout" error for micro:bit V2. They might be the same chip but with different ROM codes.

TaoBao Unshielded
28-E7-4D-75-D0-01-3C-5E: ROM does not follow expected pattern 28-xx-xx-xx-xx-00-00-crc. Error.
Scratchpad Register: 90/01/55/05/7F/A5/81/66/4E
Info only: Scratchpad bytes 2,3,4 (55/05/7F): not Maxim default values 4B/46/7F.
Scratchpad byte 5 (0xA5): should have been 0xFF according to datasheet. Error.
Scratchpad byte 6 (0x81): unexpected value. Error.
Scratchpad byte 7 (0x66): should have been 0x10 according to datasheet. Error.
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 appears to be counterfeit based on 4 deviations.

TaoBao Shielded
28-B1-98-07-D6-01-3C-30: ROM does not follow expected pattern 28-xx-xx-xx-xx-00-00-crc. Error.
Scratchpad Register: 80/01/55/05/7F/A5/81/66/1A
Info only: Scratchpad bytes 2,3,4 (55/05/7F): not Maxim default values 4B/46/7F.
Scratchpad byte 5 (0xA5): should have been 0xFF according to datasheet. Error.
Scratchpad byte 6 (0x81): unexpected value. Error.
Scratchpad byte 7 (0x66): should have been 0x10 according to datasheet. Error.
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 appears to be counterfeit based on 4 deviations.

TaoBao with Pluggable Terminal
28-CC-3D-75-D0-01-3C-7D: ROM does not follow expected pattern 28-xx-xx-xx-xx-00-00-crc. Error.
Scratchpad Register: 90/01/55/05/7F/A5/A5/66/B4
Info only: Scratchpad bytes 2,3,4 (55/05/7F): not Maxim default values 4B/46/7F.
Scratchpad byte 5 (0xA5): should have been 0xFF according to datasheet. Error.
Scratchpad byte 6 (0xA5): unexpected value. Error.
Scratchpad byte 7 (0x66): should have been 0x10 according to datasheet. Error.
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 appears to be counterfeit based on 4 deviations.

[bsiever]

@timleeanieducation Thanks again...I think I spotted a difference between your testing and mine. Did you use an external pullup resistor (the 4.7k) for the 2wire? It shouldn't need it. I enable the internal pull-up resistors on the micro:bit's processor. All of my testing of 2wire omitted the external resistor.

The 2-wire repo has an example of how I wired temperature probes: https://github.com/bsiever/microbit-dstemp-2wire#building-a-sensor-assembly (It's literally two wires and the sensor probe. No other parts at all. That was my actual goal: to minimize parts/cost and make it easier to build a micro:bit compatible probe)

If you get a chance, it'd be interesting to see if removing the resistor has an impact (it would change the net resistance and have some impact on total power and the effort to "pull")

[timleeanieducation]

Interesting... I have that in mind too. Likewise, I would like to minimize parts to save cost. Since I'm going to test for internal pull up, I might as well do the tests where there's no external resistor. Here are the results:

Conclusions:

1. It seems for the DS18B20-PAR sensor, it will work with external, no external and internal pull up.
2. The same is true for DS18B20 in 2-wire configuration.
3. For 3-wire configuration, an external pull up resistor is required.
4. The surprising result is when 3-wire is used with internal pull up, which will work in V2 but not in V1.

[timleeanieducation]

Update: I bought another batch of waterproof DS18B20 sensors from TaoBao because it came with better wiring.

The ROM code is different yet again:

TaoBao with Qhebot Pluggable Terminal
28-60-30-AF-5D-20-01-9E: ROM does not follow expected pattern 28-xx-xx-xx-xx-00-00-crc. Error.
Scratchpad Register: 50/05/4B/46/7F/FF/7F/10/FF
Info only: Scratchpad bytes 2,3,4 (4B/46/7F): Maxim default values.
Scratchpad byte 5 (0xFF): ok.
Scratchpad byte 6 (0x7F): unexpected value. Error.
Scratchpad byte 7 (0x10): ok.
0x4E modifies alarm registers: ok.
0x4E accepts 10 bit resolution: ok.
0x4E preserves reserved bytes: no, got: FF/3F/10. Error.
0x4E accepts 12 bit resolution: ok.
0x4E preserves reserved bytes: no, got: FF/7F/10. Error.
Checking byte 6 upon temperature change: not necessary. Skipped.
--> Sensor appears to be counterfeit based on 4 deviations.

This time, it works for both micro:bit V1 and V2 using the 3-wire bsiever/microbit-dstemp 0.1.21 extension.

[bsiever]

@timleeanieducation Nice module --- it looks like it has the pull-up resistor built in! I wanted to be able to connect the sensor with alligator (or crocodile) clips. Combining that module with these clips (Source of image below and clips: https://www.amazon.com/gp/product/B06XX2JBNC) would be an easy approach: