Table 7-1 in the CS215 manual specifies that the White and Black wires should each be connected to a power ground (G) terminal on the CR1000.
My questions are:
1. What is purpose of each of these ground wires, i.e., which respective parts of the circuit are they grounding? This information is missing from the manual.
2. Why aren't these ground wires tied together inside the sensor so that smaller multiconductor cables/connectors can be used?
3. Is there any disadvantage to tying these ground wires to a common ground wire near the sensor and then running the [50 ft] common ground wire back to the G terminal? I have about 25 of these sensors, so I'd really like to reduce the size of my multiconductor cables and connectors.
Both wires are connected to the same ground connection inside the sensor. They are both provided as there is a spare conductor in the cable we use and it is advantageous for long cable runs to use both wires in parallel as it lowers the power ground resistance. The ground resistance is one of the limitations to the maximum cable length you can use for SDI-12 sensors. Using both conductors allows cable runs in excess of 300 metres.
You do not have to use both wires if you do not wish to especially for short distance cables. You can just cut off one of the wires.
Thanks so much, Andrew. That is extremely helpful info!
I have a follow-up question...
I am using the CS215s to measure temp/RH inside buildings. The sensors may be exposed to direct sunlight through the windows, so I am building aspirated radiation shields for them. The shields will include low-flow 12 VDC fans.
I am ordering all the CS215s with 2 meter cables. The distance between each CS215 and the datalogger will be as little as 5 meters to as much as 25 meters. Therefore, I am going to cut multiconductor cables to length in the field and connect them to the CS215 cable with circular plastic connectors or Molex connectors.
I decided not to daisy chain the cables because I may need to move them around. A "home run" cable run from each CS215 back to the datalogger allows for more flexibility in this regard.
Based on Andrew's previous post, I assume that 25 meters is relatively short and does not, therefore, require a "double ground" wire back to the datalogger (especially since I am not daisy chaining the sensors). Therefore, I will only connect my multiconductor connector and cable to one of the CS215's two ground wires.
My question: Can I connect the 12 VDC fan's ground wire to the CS215's power ground wire instead of running it all the way back to the datalogger ground? My motivation for doing this is that I can jump down to much smaller connectors and slightly smaller cables by using a single ground wire for both the CS215 and fan.
25 metres should not be a problem for a single wire for normal SDI-12 communications especially with no daisy chaining.
Adding a fan may or may not work depending on the fan and the extension cabling used.
There are two possible problems with that setup:
1) SDI-12 is a ground referenced communication link. If the voltage drop along the ground wire gets much above 1V then the sensor will not be able to communicate reliably.
2) there is a small risk that electronics in the fan may inject some noise on the ground rail, which even if it does not upset the communications may just get through the internal regulators in the sensor and affect the measurement circuits in the sensor element.
You can check the voltage drop (1) with a voltmeter with the fan on and off. (2) is more difficult to test. You could check for noise with a scope or just do some trials monitoring the readings of the sensor whilst turning the fan on/off.
Just to close the loop on the thread above, I ran some calcs to check whether I should be below the 1 Vdc limit (see below). Then, to test the fan/CS215-L combination, I cut the CS215-L's second ground wire, installed connectors to the end of the provided CS215-L cable and fan wires, spliced the fan ground wire and the CS215-L ground wire together via a t-coupler connector, and ran all the conductors (CS215-L 12V, CS215-L SDI-12, fan tachometer, fan SW12V, and combined ground) back to the datalogger via a shielded 5-conductor cable.
CS215-L
Current: 0.0017 A
Fan
Power: 0.4 W
Operating Voltage: 12 Vdc
Current: 0.4 W / 12 Vdc = 0.0333 A
CS215-L + Fan Current
0.0017 A + 0.0333 A = 0.0350 A
Ground Wire
Max length: 100 ft
Gauge: 22 AWG
Nominal Resistance: 0.0164 Ohm per ft
Total Resistance: 1.614 Ohm
Voltage Drop
0.0350 A * 1.614 Ohm = 0.0565 V
Assuming my calculations are correct, the voltage drop should be about 5% of Andrew's stated max voltage drop of 1 Vdc. Therefore, I think I should be fine with respect to possible problem #1 from the previous post.
Regarding possible problem #2, here is a link to a screenshot that shows no noticeable change in the temp/RH readings when the fan turns off: http://sdrv.ms/1g23MyR
Based on these calcs and the fan test, I think I should be fine. I'll probably also directly measure the voltage drop, but right now I don't have access to bare conductor surfaces to do so.
I hit Submit twice because the Forum was throwing an error when I tried to submit the previous post. I tried to delete the duplicate post, but couldn't.
* Last updated by: mwassmer on 2/13/2014 @ 11:18 AM *
@aps, this thread was a great read. I had very similar question, I wasn’t sure why we need an extra grounding terminal and when I read about the wires running in excess of 300m, I got your point. I’ve been to a lot of electronics forums but this thread can be hailed as an example for other threads, for people who post questions as well as people who help others. I’d spend more time here now since this forum has so many amazing people.
* Last updated by: vedat on 7/18/2014 @ 12:20 PM *