Lufft R2S UMB GmbH Precipitation Sensor
Hi,
I woud like to know if this sensor can be connected to the CR1000 datalogger?
Can i just use its wires which comes from the sensor to connect to the CR1000?
It has Interfaces:RS-485, various RS-485-protocols, analogue output.
With kind regards,
Mark
Do i need to have one of these first to convert rs485 to rs232?
Or rs485 to 4-20mA?
http://www.bb-elec.com/485SD9TB
The Uout1 and Uout2 digital outputs are short-circuit proof high side switches (12V) with integrated
pull-down resistors. The reference potential is the same as for the RS485 interface.
Possible configurations for Uout1 are:
• Tipping bucket simulation with 1mm resolution
• Tipping bucket simulation with 0.1mm resolution
• Tipping bucket simulation with 0.01mm resolution (factory setting)
The length of the output pulse for tipping bucket simulation is typically 50ms.
The type of precipitation is transmitted on output Uout2 in the form of different frequencies. The precipitation types are coded as follows:
Precipitation Type Frequency/Hz
Dry 0
Rain 10
Snow 20
Sleet 30
Freezing Rain 40
Hail 50
If the accumulated precipitation quantity is greater than 0.01mm, the frequency signal is transmitted for 2 minutes. The output of the frequency signal is maintained if a precipitation quantity ≥ 0.01mm is measured within these 2 minutes
Hi,
Can someone shine a light on this?
I found out that:
Uout1 is a switch output like a tipping bucket rain gauge
Uout2 is a pulsed output with the frequency indicating the type of precip
So can i basically wire it like that to my cr1000 pulse ports, one wire as a tipping bucket and the other wire as pulse?
Interface: RS485 semiduplex wire, UMB protocol, pulse and frequency interface
I found this, can i get this sensor to work with cr1000 datalogger?:
IslandMan May 22, 2010 04:46 AM
1. You will have to use a device that converts your RS485 to RS232. I would suggest 485SD9TB from B&B Electronics. I have used them often and they work well.
2. I would connect this device to one of the COM1-4 ports.
'CR1000 Series Datalogger
'Declare Public Variables
Public ModBusData(5)
'Main Program
BeginProg
ModBusSlave (Com1,9600,Pick an Address,ModBusData(),0,0)
Scan (1,Sec,0,0)
ModBusData(1) = Blah1
ModBusData(2) = Blah2
ModBusData(3) = Blah3
ModBusData(4) = Blah4
ModBusData(5) = Blah5
NextScan
EndProg
Can i use these ports on the cr1000 (shown in the link below) to connect the Uout1 (pulse) and the Uout2 (frequency)and the signal ground next to it?
The Uout1 measures the amount of precipitation. Pulse.
The Uout2 the sort of precipitation. Frequency.
Where frequency:
0=dry
10=rain
20=snow
30=sleet
40=freezing rain
50=hail.
The wiring from the sensor is as follows:
1 white, negative power supply
2 brown, positive power supply
3 green, RS485_A
4 yellow, RS485_B
5 grey, not assigned
6 pink, Uout1
7 blue, GND reference potential for the digital outputs
8 red, Uout2
Definitely a P channel on the CR1000 can read either signal. The C channels can read the frequency directly. The pulse probably would directly read on a C channel as well. The C channel would need a pullup resistor if the output is an open collector type.
Hi,
Thanks alot for your reply.
1 What do you mean by "an open collector type" ?
2 i can use the following when i use the pulse port for the pulse wire?
"PulseCount (Rain_mm,1,1,2,0,0.2,0)" this one i use for my current tipping bucket sensor.
But instead of 0.2 above, it has to be 0.01 , 0.1 or 1? (The resolution choosen of the Lufft sensor).
3 am i right in saying that both wires, Uout1=pulse and Uout2=frequency, need signal ground?
So when i use a pulse channel, P1, the GND wire need to be connected next to it, which is ground?
And when i choose to use the C channels, how do i connect both Uout1 (pulse) and Uout2 (frequency) to the C channels ?
Com1 c1=Uout1 and c2=GND?
Com2 c3=Uout2 and c4=GND?
And what will be the crbasic syntax to use when Com1 for pulse and Com2 for frequency are used?
4 theres only one GND wire from the sensor, can i just use an extra wire connected to GND to have 2 GND's to have both wires, pulse and frequency, signal ground?
5 when the cr1000 and the Lufft sensor know eachother , does the cr1000 recognize the Lufft data, or does the crbasic language need customization to have the amount (pulse) and sort of (frequency) precipitation read?
Hi JDavis,
Can you please have a look at my previous post?
Much appreciated.
With kind regards
Mark
An open-collector signal is a sinking pulse. Reading sinking pulses requires a pullup resistor. A P channel in switch closure mode will read it, do to an internal pullup.
The frequency signal is also read with a PulseCount instruction. Use the high frequency option for the mode. The C channels are in the list of usable channels.
Hi JDavis,
Thanks for the reply.
The sensor is in binary data protocol.
Does it influence the way it has to be coded in crbasic?
The sensor being set to UMB protocol would be why Modbus didn't work. You could use your RS485 adapter to connect a PC to the sensor using the Lufft configuration tool. Then, you could set the protocol to Modbus if you want.
I believe the protocol setting does not affect the pulse and frequency signal lines.
thats a relief, so hopefully i can just wire it to my cr1000....
is it possible you can have a look at question nr 3, 4 and 5 in my previous post? regarding the wiring and code?
Currently the R2S is connected to the P1 channel.
The pink wire to P1 and the blue wire to ground.
But im not able to get data from it.
Its currently outside in the rain with a pulsecount instruction.
PulseCount (Rain_mm,1,1,2,0,0.2,0)
This is from the R2S manual:
"The length of the output pulse for tipping bucket simulation is typically 50ms."
Yes, its getting data now.
For the pulse output :(amount of precipitation)
Connected to the P1 channel.
PulseCount (Rain_mm,1,1,2,0,0.01,0)
For the frequency output:(precipitation type)
Connected to the P2 channel
PulseCount (Precipitation_type,1,2,0,1,1,0)
And GND wire connected to ground from P1.
Other question:
The sensor can detect the type of precipitation.
How can i get that to work in a one minute table?
Within a minute the type of precipitation can change ...
These are the codes which correspond with the precepitation type:
Dry 0
Rain 10
Snow 20
Sleet 30
Freezing Rain 40
Hail 50
A histogram is probably the easiest way to handle that. Make an unweighted histogram with bins that have those values in the middle. So, use a lower limit of -5, and an upper limit of 55 with 6 bins. You can use the FieldNames instruction right after the Histogram instruction in your table to change the bin labels.
I would use Form option 001. That gives a ratio, and excludes outliers.
Hi,
Thanks for the reply.
Do you have an example of an instruction on how to acomplish that?
One other thought I just had; you might want to use PeriodAvg instead of PulseCount to read the precipitation type. PulseCount will give you the average frequency since the last time it was called. The situation where that could be a problem is during a change in precipitation type. If you went from dry to hail, PulseCount will report 25.
PeriodAvg takes a snapshot, ie it looks at a short time window. That would reduce your chance of catching a transition, but not eliminate the posibility. The downside of PeriodAvg, is you have to make your scan wait at least 200 milliseconds to read a 10 hz pulse.
Look in the help for the PeriodAvg instruction. It is pretty easy to use. The biggest decisions to make are how many pulses to count and what timeout. Base your timeout on the number of pulses you want to count, times the period of your slowest frequency you need to measure (apart from 0). PeriodAvg uses an analog channel.
Hi,
You mention when it went from dry (0) to hail (50) it will report 25.
But now, in working order, when its dry (0) and start to rain (10), it shows 10, which is good. Sometime when it stop or start raining , i see for very short time other values , like 4 or 7 or something before showing a constant 0 or 10.
Unfortunately it only has rained here, so ive only seen de 10.
Will the 50 (hail) behave the same or will it differ because of the higher number?
With the values of 4 or 7, you have seen what can happen. If you are going to a precipitation type with a higher number, you could end up with an in between number which matches a precip type. So, going up to hail or freezing rain, you could get a count for snow.
Ive seen a similar question in this topic:
https://www.campbellsci.com/forum?forum=1&l=thread&tid=2428
But when use the Select/Case statement , which channel should i use? No pulse i think.
What about PulseCountReset?
When i want to connect the precipitation type wire which is now connected to P2 and ground next to P1, to a SE channel, does the SE channel use ground from the P1 channel or do i have to split the ground wire to have both P1 and SE channel ground?
You don't need to move the ground wire. All the analog grounds are connected together. There would be a very tiny difference in voltage measured, but it is nowhere near impacting your measurement.
Thanks alot for the reply, ill try to rewire and use the PeriodAvg instruction.
Ive re-wired the sensor's wire to SE channel 8.
Ive set to use the PeriodAvg instruction like so:
PeriodAvg (Precipitation_type,1,mV5000,8,0,1,10,10,1,0)
Youre talking about how many pulses to count and what time out.
I have a one second scan and one minute table.
So in order to get the precipitation type in a minute, i need to set the pulse count to 60?
What happens when the precipitation type within a minute changes from rain to hail for example?
And the timeout im not sure how to set that..
I just realize the the output is digital, so SE channel wont work....?
The SE channel works with the PeriodAvg instruction.
Ok, but i get NAN with every setting i change in the PeriodAvg instruction.
i thought SE channels are analog? the output from the sensor is digital...
should i use a com port, which is digital?