Multiple DS18B20 returning ZERO

[PSR B1257]

Hy guys,

my new Arduino-Clone is driving me nuts...
I have been successfully running a temperature-logger with four DS18B20 sensors on a "Sainsmart UNO". Until some idiot  shortet the supply rail of the board...
So I went ahead and got me a nother Arduino-Clone with a logging-shield (https://www.ebay.de/itm/Arduino-Uno-R3-mit-XD-204-Data-logging-shield/281460446703?ssPageName=STRK%3AMEBIDX%3AIT&_trksid=p2057872.m2749.l2648), but lo and behold, the same code is no longer working 
With all four sensors connected, each sensor is replying 0.00. If I connect only one random sensor, everything is fine again. I get three errors and one correct temperatur-reading.

Any ideas are welcome.

P.S.: Merry Christmas 

[2N3055]

Do you run 5V? Do you connect sensor with +VCC (3 wires ) or parasitically (2 wires)? What pull-up resistor do you use?
Do you have access to scope to see waveforms?

P.S. Thanks, Merry Christmas to you too!!

[floobydust]

Maybe your new shield is fighting with the temperature sensors.
The DS18B20's should be on a separate I/O pins - not shared with the SD card or RTC data lines?

Merry Christmas! We got snow and -23C tonight, so a chilly white Christmas.

[PSR B1257]

The 5V supply and such is fine.

Just for curiosity, I've added a 12k resistor in parallel to the existing 4k7 resistor (3-wire connected sensor) which causes the whole mess to work again 

At least for now. A endurance test is running....

[stj]

read the one-wire comms datasheets.
the way you adress multiple devices without a clash is different to talking to a single one.

[floobydust]

Watch long cable runs to the DS18B20's will need a lower (4k7 to 1k) pullup resistor and a 0.1uF decoupling cap across VDD and GND local at each temperature sensor for really good data.

[PSR B1257]

the way you adress multiple devices without a clash is different to talking to a single one.

Yes, but this is not the problem, since I had a already working program on the destroyed Arduino.

Watch long cable runs to the DS18B20's will need a lower (4k7 to 1k) pullup resistor and a 0.1uF decoupling cap

As examined by my little experiment with the 12k resistor in parallel.
A decoupling cap at the sensor however seems like a good idea, to prevent any unexpected issues.

My endurance test went successfully so far. None of the sensors is replying "0.00" or even the error code "-127" since it was started.