Detection of low current direction?

Status
Not open for further replies.

LMAO

Senior Member
Location
Texas
We have low voltage (24VDC) cubicles and consoles fed off a common low voltage bus. I am trying to find a way to detect a "reverse power" situation where someone might have accidentally connected 24VDC on the load side. We thought about using a Hall Effect sensor to light up an LED but our current (25mA to a few amps) is too small to produce magnetic field large enough to be sensed by Hall Effect sensors.
Any idea how to detect reverse power on low voltage/current DC buses?
thanks
 

rcwilson

Senior Member
Location
Redmond, WA
Add a precison shunt and monitor the millivolts across the shunt. The voltage will go negative on reverse power. A sensitive meter and/or a high ohm shunt is needed to get good resolution.
 

gar

Senior Member
120607-2014 EDT

LMAO:

More information is needed.

Suppose you insert a Schottky diode in series with the hot supply line (non-grounded wire). Will this adversely affect your output voltage? Would the isolated equipment be damaged by this other power source with the diode present?

What would it matter if someone added a 24 V power source to your main bus?
Big problem if it is reverse polarity, otherwise what is the problem?

What is the maximum load current in a cubicle?

How much resistance can you add in series with the source to the cubicle and not create problems?

.
 

LMAO

Senior Member
Location
Texas
120607-2014 EDT

LMAO:

More information is needed.

Suppose you insert a Schottky diode in series with the hot supply line (non-grounded wire). Will this adversely affect your output voltage? Would the isolated equipment be damaged by this other power source with the diode present?

What would it matter if someone added a 24 V power source to your main bus?
Big problem if it is reverse polarity, otherwise what is the problem?

What is the maximum load current in a cubicle?

How much resistance can you add in series with the source to the cubicle and not create problems?

.

regarding Schottky diode, 20mV voltage drop should not be a problem and isolated equipment won't be damaged.

It'd matter to us if there is an external source. Our 24VDC consoles and devices are essential to customer operations and we need to know if there is a problem with main source and some technician is too lazy to identify the problem decides to simply add another 24VDC source to line.

about your other questions, I do not want to introduce a resistance on line. I just wish they made a device like Schottky with constant small voltage drop independent of current and a indicator lighting up when Schottky was on, aka, an LED with a very low voltage and very high current range...
 

gar

Senior Member
120608-1143 EDT

LMAO:

No matter what you do to measure current there will be some impedance added, it could be extremely small. Even if you use a compass needle to measure current it will require some minute amount of energy.

What is your maximum current? What is the minimum back current you want to measure?

Even if you could sense this minimum back current, then that won't prevent someone from adding current internal to the module that does not exceed the load current in the module.

10 turns of wire will increase the sensitivity of a Hall current sensor by 10. And obviously 100 turns is a 100 multiplier. On DC, Hall devices have a zero drift problem in the kind of dynamic range you likely want to work.

Suppose your module load was 100 A and you used an 0.25 V shunt at 100 A, then at 10 MA the voltage drop is 25 microvolts. This is a reasonable trigger voltage with which to work.

.
 

Besoeker

Senior Member
Location
UK
It'd matter to us if there is an external source. Our 24VDC consoles and devices are essential to customer operations and we need to know if there is a problem with main source and some technician is too lazy to identify the problem decides to simply add another 24VDC source to line.
That doesn't sound like "accidentally connected 24VDC on the load side".
Could documentation be circulated to all concerned that such action is not permitted?
Prevent the problem occurring rather than attempting a fix?

But what mode of failures are there with the main source?
If it just gives up and you get no output, couldn't you simply monitor its output voltage and/or current?
It disappears, you register it as a problem, have a latching circuit that drives a fault lamp?
 

LMAO

Senior Member
Location
Texas
120608-1143 EDT

LMAO:

No matter what you do to measure current there will be some impedance added, it could be extremely small. Even if you use a compass needle to measure current it will require some minute amount of energy.

What is your maximum current? What is the minimum back current you want to measure?

Even if you could sense this minimum back current, then that won't prevent someone from adding current internal to the module that does not exceed the load current in the module.

10 turns of wire will increase the sensitivity of a Hall current sensor by 10. And obviously 100 turns is a 100 multiplier. On DC, Hall devices have a zero drift problem in the kind of dynamic range you likely want to work.

Suppose your module load was 100 A and you used an 0.25 V shunt at 100 A, then at 10 MA the voltage drop is 25 microvolts. This is a reasonable trigger voltage with which to work.

.

our current range is 20mA to 20A.
I did consider winding the cable around a solenoid but did not like it because we are looking for something really small to surface-solder and mount inside a terminal block.
 

LMAO

Senior Member
Location
Texas
That doesn't sound like "accidentally connected 24VDC on the load side".
Could documentation be circulated to all concerned that such action is not permitted?
Prevent the problem occurring rather than attempting a fix?

But what mode of failures are there with the main source?
If it just gives up and you get no output, couldn't you simply monitor its output voltage and/or current?
It disappears, you register it as a problem, have a latching circuit that drives a fault lamp?

It is a little complicated; these consoles are used on offshore rigs and multiple power sources are used to maintain a level of redundancy (we use steering diodes). We are also monitoring for out of range voltage (outside 0-28VDC) and want to KNOW if a source is connected downstream to fuses.
 

gar

Senior Member
120608-1541 EDT

LMAO:

Your information is still incomplete. You already have steering diodes and therefore can not back feed your bus, or buses.

What is the magnitude of the back feed current you want to detect? 1 A, 10 MA, 0.1 MA?

MA is milliamperes, I don't change my ways easily.

.
 

Besoeker

Senior Member
Location
UK
It is a little complicated; these consoles are used on offshore rigs and multiple power sources are used to maintain a level of redundancy (we use steering diodes). We are also monitoring for out of range voltage (outside 0-28VDC) and want to KNOW if a source is connected downstream to fuses.
I have some experience with supplying kit to the petrochem sector.
Absolutely everything is documented to the nth degree.
Brackets, nuts, bolts......the whole nine yards.
It should not be permissible for additional power supplies to be installed without a document trail including, as a minimum, written records of why it was done, risk assessments and method statements.
Not to mention approved data sheets for the proposed hardware, certificates of conformity for every part.

I know that doesn't fix your problem. But unauthorised modifications, and that's what you seem to have, are an absolute no no.
 
Status
Not open for further replies.
Top