Send a sensor output (+5v to -5v) to a remote destination

Thread Starter

KG6DOQ

Joined Jul 9, 2019
4
I inherited an earthquake sensor (installed on a mountaintop) that has an analog output. Basically, when idle it produces 0v, but when there's motion it will swing between a max of +5v and -5v. The problem is getting the voltage info back to our classroom.

The system this came out of used a drum recorder in the classroom end. It worked over an analog dedicated 2 wire phone line, which I also have access to. The original system had a circuit (which is long gone) that would produce a 1000hz tone on the phone line when the sensor voltage was 0v, drop the tone to around 775hz on -5v and bring the tone up to around 1270hz on +5v.

At the receiving end, there was another circuit that did the opposite. ie. the audio tone would be converted back to +5v/-5v voltage and drive the drum recorder.

I am seeking advice as what would be the best way to duplicate the function of this circuit or even if that is the best choice. I also have a network connection to the mountaintop available, so I could probably put this into an A-D converter of some type.
 

Sensacell

Joined Jun 19, 2012
3,011
How far away is the sender Vs receiver?

You could convert the voltage to a current, send this over the phone line, convert it back to a voltage at the other end.
 

danadak

Joined Mar 10, 2018
4,057
V to F PSOC.JPG Two possibilities.

Network connection : Use ESP8266 or ESP32 to stream data into network.
This would be wireless to your network endpoint top of mountain. Data logger

https://tttapa.github.io/ESP8266/Chap16 - Data Logging.html

https://randomnerdtutorials.com/esp8266-wireless-weather-station-with-data-logging-to-excel/

Google "ESP8266 data logger", lots of projects.



Or stay with tone via FSK approach.

The other approach retain FSK (V to F) interface. Image above.

Driving long cable length and issues of lightning strikes, direct or indirect, and cable
capacitance, might need external low Z analog buffer with protection coverage (like
diodes, zeners, MOVs...) on it.

Note most of chip resources unused, other features could be added. Also SAR shown
configed as differential in, 4 channels. You can change that to single ended or more/less
channels. Trivial to do. Dev board for this ~ $10. Chip has fairly accurate clock (+/- 2%)
and very accurate (+/- .1%) onchip reference for SAR.

Maybe ~ 10 lines of code to do this.

One other issue, your sensor has a bipolar output. To work with either of the above
solutions you need to offset that into the common mode range of the chip A/D, so
my calling this a single chip design not accurate, an opamp to do the offset will be
required. Note there is a technique of doing offsets with a R divider to a + rail that
translates the sensor bipolar into unipolar range. I can supply that info if you decide
to go that route.


Regards, Dana.
 
Last edited:

MrChips

Joined Oct 2, 2009
26,517
Is there power available where the sensor is installed?
If not, can you installed a solar power battery system?

One solution I may suggest is to digitize the ±5V analog signal and send the data wirelessly either direct RF or to a web server.
 

danadak

Joined Mar 10, 2018
4,057
You could always modulate a low power laser if there is a direct obstruction
free path down the mountain.


Regards, Dana.
 

danadak

Joined Mar 10, 2018
4,057
No, just that whatever the application you come up with a way of doing it with an ESP...
You need to do some homework, I recommend Cypress, Atmel, Freescale,
the ESP parts, ATTINY.....and different languages depending on what makes
sense to me and I think might be of interest to the customer.

Whats the problem Albert ?


Regards, Dana.
 

Thread Starter

KG6DOQ

Joined Jul 9, 2019
4
How far away is the sender Vs receiver?

You could convert the voltage to a current, send this over the phone line, convert it back to a voltage at the other end.
The sender is about 20 miles away. The 2-wire phone line is really a phone line, which can only pass audio in the 300hz - 3000hz range. It's not a twisted pair that I can put voltage on.
 

Thread Starter

KG6DOQ

Joined Jul 9, 2019
4
Is there power available where the sensor is installed?
If not, can you installed a solar power battery system?

One solution I may suggest is to digitize the ±5V analog signal and send the data wirelessly either direct RF or to a web server.
There is power available where the sensor is installed. There is also a wired ethernet connection available.
 

danadak

Joined Mar 10, 2018
4,057
If you have a wired 10/100 connection I would recommend that over a telephone
line/analog approach.

Is this a turn key design, no development, no code desired approach, that you are
seeking ? If so a LabView collection at the end point for graphing/data collection
might be easiest. Use a DAQ with an Ethernet connection on the mountain to
generate the data stream. These guys have been around a long time -

https://www.dataq.com/data-acquisit...MIuNf74uWq4wIVlFcNCh1NPghgEAAYASAAEgLs_PD_BwE

https://labjack.com/wifi-daq?gclid=EAIaIQobChMIu77wsuaq4wIV445bCh2t4wN1EAAYAyAAEgJZa_D_BwE




Regards, Dana.
 

djsfantasi

Joined Apr 11, 2010
8,394
I haven’t noted the degree of accuracy desired. Are there three values to be transmitted (5, 0, -5)? Or does the application require values from -5 to 5V? Does that range need to be continuous? Or would a set of discrete values representing the range be sufficient? For example, -5V to 5V in increments of 0.005V. Or any other increment.

I am reminded of old technology. I was thinking of old dial-up modem technology, but realized there may be a simpler approach. Using DTMF, one could transmit 16 discrete values over the phone line. Using two DTMF times, you could transmit 256 discrete values; three tones give you 4096 discrete values. Or steps of 0.0025V.

So, using a 12V rail to rail op amp in a summing configuration gives a 0-10V input for a 2 bit ADC. Input that to a micro controller with a DTMF library and with three tones, send the data over the phone line.

A local micro controller decodes the DTMF triplets into a float value of the voltage. And does whatever is necessary.
 

marcf

Joined Dec 29, 2014
267
Just curious, but do you know what signal levels the original system used?

20 miles is a very long distance to send DTMF, or any signal in the old 'POTS' system's pass band.

5km (3.1miles) is (was) the maximum distance in the days of POTS for AWG 24.

If the existing wire is 2 conductor AWG 24, sending and receiving over the 20 mile (32Km) distance will not be a trivial problem.

Also I would be very careful to provide lighting protection and provide proper grounding.

Of course a balanced 600 ohm isolation transformer with appropriate circuitry would be a very good idea as well.
 

Thread Starter

KG6DOQ

Joined Jul 9, 2019
4
I haven’t noted the degree of accuracy desired. Are there three values to be transmitted (5, 0, -5)? Or does the application require values from -5 to 5V? Does that range need to be continuous? Or would a set of discrete values representing the range be sufficient? For example, -5V to 5V in increments of 0.005V. Or any other increment.

I am reminded of old technology. I was thinking of old dial-up modem technology, but realized there may be a simpler approach. Using DTMF, one could transmit 16 discrete values over the phone line. Using two DTMF times, you could transmit 256 discrete values; three tones give you 4096 discrete values. Or steps of 0.0025V.

So, using a 12V rail to rail op amp in a summing configuration gives a 0-10V input for a 2 bit ADC. Input that to a micro controller with a DTMF library and with three tones, send the data over the phone line.

A local micro controller decodes the DTMF triplets into a float value of the voltage. And does whatever is necessary.
 
Top