Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Looking for expert on cable telemetry

Status
Not open for further replies.

barryn56

New Member
Hi,
I have developed some equipment that uses a monocable (single wire with external braided armour) CAN bus telemetry which works up to 1000m, but I need something that will go to 3000 metres cable length. Speeds are not so high (9600 baud is fine), but an advantage would be to be able to send some power as well as send/receive data over the same line - can anyone provide me with a commercial off-the-shelf product, offer to design, build and test ($$) or help point me in the right direction?
Thanks,
Barry
 
Last edited:
I don't think you're going to get there. Not with CAN, unless you can find a better transceiver than the 82C251 or a cable with minimal DC resistance. How many nodes BTW?

Do the following thought experiment. Pick a cable resistance. Then imagine a 24V Class II power supply. Now ask yourself how much voltage at what current will be available at the end of 3000 meters. It's hen scratchings.
 
Last edited:
Thanks for comments

Thanks for the note, but I understand from the CAN design that, at 10kbps the distance quoted is 6000 metres (though this is twisted pair, we operate in a very low noise environment). Plus there are other companies with other types of tools (logging) that routinely use higher speed CAN bus at lengths of more than 5000 metres. There are others that use other systems (FSK) too - and I'm not married to CAN bus, it is just widely available and seems to have the right capabilities.

Barry
 
Your problem is not related to choosing the CAN bit timing parameters that trade cable distance for bit rate. Your problem with using the cable for power is related to the DC resistance of the cable. You must account for the IR drop on the power lead from the power supply to the node that requires the power. Then you need to account for the IR drop on the return line back to the power supply.

Assuming you can get the power where you want it, you then have the issue of transmit levels from the sending node to the receiving node. Assuming the transmit half of the transceiver is beefy enough to charge and discharge that much cable you then have the problem of receiver thresholds.

That people have done things in a way that convinces you that it is easy is not a surprise. However when you start to combine requirements then things quickly become more difficult.

I think that your requirements are going to be difficult to meet. Your best chance to get the data over that distance is to forget the power on the cable and stick to differential drivers and receivers without the headache of a common ground.

As I mentioned before the number of nodes on this network is crucial. In the best case there will be one transmitter and one receiver. If you start sprinkling nodes along the trunkline your life will get much more complicated.
 
Thoughts

Thanks for the new comments, Papabravo. We do only have a surface and downhole system, so no intermediate sensors, etc. For the resistance, the cable provides abut 280 ohms for 3000 metres, but that is with the cable wound on the drum, so the armour is shorted out. As you run into the hole, the resistance increases. I was thinking I can measure the voltage arriving at the tool, and send this back up the telemetry to surface, so we can control the surface voltage to keep the downhole voltage correct. We have several ADCs available on our downhole processor board.
 
Thanks for the new comments, Papabravo. We do only have a surface and downhole system, so no intermediate sensors, etc. For the resistance, the cable provides abut 280 ohms for 3000 metres, but that is with the cable wound on the drum, so the armour is shorted out. As you run into the hole, the resistance increases. I was thinking I can measure the voltage arriving at the tool, and send this back up the telemetry to surface, so we can control the surface voltage to keep the downhole voltage correct. We have several ADCs available on our downhole processor board.
Hmmm What you have is a more interesting problem than the one of a static network cable that runs 3000 meters. Wound on the drum you measure the resistance of just the center conductor.

Returning to my power supply thought experiment. Start with a 100VA Class II 24V power supply. The maximum available current would be in the neighborhood of 86 mA. If your snesor only draws say 25 mA then you would drop 7 volts in the cable and you would have available (24-7) or 17V @ 25 mA. If the sensor requires say 50 mA then you would be dropping twice that or 14 volts in the cable leaving only 10V @ 25 mA for the sensor.

You say that as the cable unwinds the resistance will go up because you are adding resistance in the jacket. This will make the problem get worse by limiting the available current and reducing the voltage available at the sensor.

Let us assume that with the cable unwound the resistance increases by 20% to 336 Ohms. Now the maximum current available is only 71.4 mA and our 50 mA sensor now has only 7.2 Volts @ 50 mA because 16.8 volts is lost in the cable. Some 5V regulators will start to behave erratically as the voltage drops below 7V which means you have to start thinking about LDO regulators for a bit of extra headroom.

This was just the quickie analysis I did in my head as a result of long experience with CAN in factory automation. Power along the network cable is a tricky proposition. Also if your power supply output voltage has a tolerance your 24 volts might actually be only 23 volts, tying another limb behind your back.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top