Long Cable Power Issue 4

[boreholelogger]

I have a 2.565km piece of wire-line 4 conductor cable wound onto a winch that is used in the mineral exploration industry. The cable has an 85 ohm/km resistance which means that there is 218 ohms resistance from one end of the cable to the other, (Where the camera will be connected) and 218 ohms return line back to the power supply. I require 30VDC @ 400mA across the camera system which will be connected on to the end of this cable. I have ran a few tests and the camera system's current changes from 320mA to 400mA constantly which means that the voltage across the camera system fluctuates alot to provide constant power. The issue I have is that the system requires 200VDC at one end of the cable to provide 30VDC @ approx 400mA across a 75 ohm resistor (This resistor sort of simulates the camera system). Also the initial voltage supplied to the camera is 200VDC for a short moment which would blow the camera's regulator to smitherines. The camera's regulator can withstand 9 - 48VDC. Does anyone know how to help me OR anyone that can help me with this issue?. Or does anyone know of a 30 - 300VDC regulator that can be mounted on a PCB?. Also maybe someone might have a circuit that has used a transorb or something to minimise the impact upon the camera system on the initial startup.

 

[boreholelogger]

Yes it was mainly to do with the bandwidth of the cable.

 

[boreholelogger]

Yeah the problem I have with the vicor DC-DC converters is that they are slightly too big for the dimensions of the pipe that they need to go into. I will keep browsing their range though thanks.

 

[boreholelogger]

Can you buy modules that put the DSL comunications over the power line?. I have never seen any, but thats not to say that there isn't. With all of our other equipment we superimpose our pulse type communications over the power wires using flip-flops etc. I would much prefer to be able to put the DSL comms over the power wire, then we can do whatever we like!.

 

[itsmoked]

skogs; the module you are proposing sounds like it needs 175V at the bottom! Furthermore, while I think Vicor is a cool company, I can't imagine them expecting their input sources having 436 Ohms of source impedance.

Still seems to me that using a shunt or even a standard regulator down hole that gets to use all that source impedance to its advantage with the waste heat distributed over a mile as being the way to go.

So far I'm getting ugly numbers like 2W dumped on a zener. ugh!

 

[Warpspeed]

The most obvious and simplest solution seems to be fitting a shunt voltage regulator and suitably large capacitor down the hole as previously suggested.

However, you mention the availability of spare wires in the cable. That raises the possibility of feeding the actual camera voltage back to the surface through a pair of sensing wires which carry no current.

It would then be possible to have a voltage regulator located at the surface that keeps the voltage at the camera constant by tracking load changes. That would eliminate any problems of size or heat dissipation of the equipment down the hole.

 

[boreholelogger]

Warpspeed,
The cable has 4 individual copper conductors encased in a woven steel wire belted casing (Armour). The conductors have 85 ohm/km of resistance and the Armour has 23 ohm/km of resistance. The first configuration I tried was to use 2 conductors for DSL communications and the other two wires for power which made sense to me at the time. The second configuration which I am now using at the moment is two wires for DSL communications and two wires for the positive voltage (Parallel configuration to supply more current and half the resistance) and I am using the Armour as the return path. The problem with the first configuration everyone knows. The problem with the second configuration is that when the cable is spooled onto the winch then the Armour has a 0 ohm resistance (Short circuit). When the cable is fully spooled off the winch then the Armour resistance is about 59 ohms. This means that if I don't have a large voltage input range at the tool I will need to automatically change the voltage at the surface to compensate for the voltage difference.

 

[felixc]

Yup, a problem in your implementation is the ground that swings along with the current peaks. You say DSL communications; is your comm signal electrically isolated, or do you have to live with this floating gound? I guess that your two signal lines are a diff pair. Is there any way to isolate them with a signal transformer, to get rid of the ground fluctuations?

 

[boreholelogger]

The DSL signal communications pair is isolated via a transformer at each end.

 

[VEBill]

I failed to make myself sufficiently clear (it happens all the time...).

!! Parallel the four wires (only 55 ohms one way) !! Use the armour as the return path (0 - 59 ohms). Total resistance only varies from 55 to 114 (much better, eh?).

Then - put the DSL signal onto the power line. This is absolutely trivial since the DSL is probably about 1MHz range (just use ferrite chokes and nF capacitors).

If required - use a 10F SuperCap to smooth out the current (depends on the time frame of the variation). Or, if you need more - add a circuit to adjust the illumination LEDs current inversely to the varying camera. The camera can probably compensate for the slight variation in brightness.

Using the above suggestions almost certainly reduces the problem to a range where there are easier solutions. It wouldn't take much to get the voltage variation into the 9 to 48 volt range (that's a pretty wide range, more than 5:1). Probably don't even need a voltage regulator other than the nice one built into the camera...

BTW - this is a classic case where this should have all been considered at the outset - right at the point where someone first mentions using 2.5km of cable (WHOA!!!!)

 

[Skogsgurra]

Smoked

If you look at their site, you will find many models with many different input voltages ranges. And I know that you know that the impedance can be brought to zero with a capacitor. If you choose an AC input model, the capacitor is there for you already. Just behind the rectifier.

Gunnar Englund

http://www.gke.org

 

[itsmoked]

geezz skogs that site is hard to wade through..

I couldn't find one that would do the job. Doesn't mean there isn't one. I just couldn't find one. So I went to other DC-DC makers but always had problems with the desired 30V output.

If he uses a LM137 it would need to dump 12.5W when the camera/lights were doing their 320mA thing and 2W when doing the 400mA thang. Probably means something like 6W average which a pretty moderate heat sink would handle easily. What kind of temps hang out down a mile? I would guess ummmmm 50F?

VE1BLL's idea is a good one. Perhaps mixed with warpspeed's so the two wires running the sense voltage to the surface double as the DSL lines.

BTW: Why are we calling this DSL? Is this really a set of modems?

 

[VEBill]

Re: warpspeed's suggestion - There might be an issue with having 2km+ long sense wires that are also in close parallel with the power wires. By the time that you filter the induced noise out of them, they would probably be too slow to maintain timely control of the distant voltage.


Four Wire Parallel solution - the math...

Using the four wires in parallel with the armour as the return, the total resistance varies from 55 (cable spooled) to 114 ohms (fully out).

Thus if you were to feed in 60 volts at the top end, the voltage limits at the bottom end would be within 14.4 to 42.4 (which is within the camera's limits of 9-48VDC).

Here is the math:

Vin mA R IRdrop Vout
60 320 55 17.6 42.4 (max)
60 400 55 22.0 38.0
60 320 114 36.5 23.5
60 400 114 45.6 14.4 (min)

Note that no voltage regulator is required at the bottom end.

Furthermore, if you add a 10F supercap (or something similar) at the bottom end, you can probably filter out some of the current variations to stabilize the voltage towards the middle of the range.

I'd add a voltage crowbar (perhaps an SCR crowbar using the cable resistance as the voltage dropping load) at the bottom end to protect the circuits. If you need to reset it, then you simply cycle power at the top end.

As already mentioned, it is trivial to insert and extract the ~1MHz DSL signal.

 

[Skogsgurra]

Smoked,

You are a stubborn young boy. The camera can take anything between 9 and 48 volts. What is the problem? Stuck with your darling?

Gunnar Englund

http://www.gke.org

 

[ScottyUK]

Take a look at how the power industry does things: long distance transmission uses high voltage.

If the cable will withstand it, take a supply of 300V or so. The load is 12W, equating to 40mA @ 300V. Assuming that you use a single core each for the positive and negative supply, that is a loop resistance of 436[Ω] and a volt-drop at the load of about 17V, leaving about 283V available to supply the load. A DC/DC converter like the ones Skogs is describing would do the job very well.

Rather than say something is 'too big' (e.g. the Vicor converters), tell us the dimensions you have to work within. It helps because we might be able to suggest something useful if we have all the information.




----------------------------------

One day my ship will come in.
But with my luck, I'll be at the airport!

 

[VEBill]

Higher voltages are 'Good Squared'. The IR drop is reduced because the current is less, AND any remaining IR drops are less of an issue with higher voltages. They're 'Good Squared'.

Size souldn't matter because, although the instrument diameter may be limited, the remaiming third dimension can be as much as about a km or so... ;-)

 

[boreholelogger]

Basically yes this system is an off the shelf modem at the surface and a custom built design at the tool.

 

[boreholelogger]

Yeah the camera board can take 9-48 Volts but the LED power supply board can take 30 - 48V. By using any less than 30VDC at the tool, the voltage fluctuates to much at the tool end due to the increased current draw for constant power. That is why 30VDC is my aim.

 

[boreholelogger]

Yes using high voltages are fine in theory, but when you have people that are not always qualified in the use of electronics, like people using this stuff in the indonesian jungles running off generators etc. I would prefer to keep the voltage as low as possible and still be able to run the equipment successfully (Especially if I am using the armour as a return path). I think the best idea is to put all the 4 wires in parallel and use the armour as a return path. That way everything is safe and we could even use a longer cable.

VE1BLL,
You say putting DSL communcations over the power wire is trivial, then how do you do it?.

 

[boreholelogger]

I appreciate all the feedback I have received it has all be very helpful!. Thankyou to everyone that has contributed and that I am sure will continue to contribute to this matter.

 

[Warpspeed]

Ah ! signal transformers at each end, lovely.

One technique is to have a centre tap in each signal transformer winding. You can then feed dc into one centre tap at one end, and take it out at the centre tap at the other end.

The dc current splits two ways and travels down both sides of the signal circuit in parallel. That will effectively halve the dc cable resistance. The dc current will not effect either transformer because it will be balanced at the centre tap.

The two other available conductors could be paralleled and used as a dc return path.

A suitably large filter cap, and a shunt voltage regulator down the hole should then be all you need.