Multiple grounds on DC floating battery 1

[thryll]

I currently have a 250VDC battery which has multiple full positive grounds on it. I have pulled all the dc fuses one at a time and the dc potential on the ground still remains. I have been notified that this ground has been present for some time now and in my opinion other grounds have also accumulated. I have been pondering an ac test method. Where I would inject low voltages or currents at a specific frequency (which frequency I am not sure of 25hz ? ) to the battery terminal and check for AC current presence in the individual control circuits while the control circuits are live. If anyone has had any luck with this or could point me in the right direction (voltage with resistor vs current/Frequency suggestions) it would be appreciated.
Thanks

 

[ScottyUK]

I've sucessfully hunted down ground faults using an AC injection set. The unit I used was called the Grouser which looked like it was a very low volume product but it's performance was pretty good. I think it was a CEGB in-house design which has been continued by some enterprising soul after the demise of the CEGB. The detector is the key item - at one point we hooked up a signal generator and a big power amplifier to increase the injected signal when the on-board injection signal from the Grouser ran out of puff.


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

If we learn from our mistakes I'm getting a great education!

 

[thryll]

I was wondering if I could somehow simulate something like the grouser myself?

 

[ScottyUK]

Well, I successfully replaced the signal injection unit with a signal generator driving a big and powerful audio amplifier coupled to the DC bus through a high voltage coupling capacitor. The receiver essentially needs an inductive pickup coil with a narrow band filter tuned to the injection source frequency and a signal strength meter. Keep the frequency low - well below mains frequency - to avoid harmonics and to keep the effects of inductive reactance down.

Let us know how you get on with the receiver design.



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

If we learn from our mistakes I'm getting a great education!

 

[Jebb]

Megger makes device which works just as described called a Battery Ground Fault Tracer which works very well....

 

[rcw retired EE]

I successfuly traced DC ground faults in a large WWII era substation using an isolated AC current source and clamp on ampmeters.

Output of the isolation transformer was connected to ground through a tapped current limiting resistor. The other end was connected to the battery bus lead that showed a ground. A variable voltage transformer fed the isolation unit. We adjusted the variac to get enough AC current to detect.

A timing relay shorted out a portion of the resistor to give a pulsing AC current, jumping from 0.2 A to 2.0 Amps about 1 Hz. The pulsing was easier to see on the ammeter than a straight current.

We clamped on various circuits and followed the wires to the faulted conductors, locating multiple faults in a couple of hours.

 

[raithrovers1]

Scotty UK is right that using a grouser is the best way to go. I have used it on many systems and it works very well. You do have to be careful about what type of systems you can use a grouser on as the signal has been known to cause spurious opeartion of downstream relays etc on instrumentation systems. The best way to prevent this is to follow the instruction manual on how to set up the signal strength to as low as possible.

UPS engineer

http://www.powerups.co.uk

 

[raithrovers1]

Just read your post again. If you have pulled all of the DC fuses one at a time then do you still have a ground fault with all fuses removed?
If so then your ground fault is within the rectifier or battery. Are the batteries Ni-Cad or VRLA? Ni-Cads with metal cases have been known to track down to earth on the battery frame.
Is there DC capacitors within the rectifier? If so sometimes the cases of these can cause earth faults as the negative terminal of the capacitor is connected to the case. If the capacitor is stud mounted on an earthed frame then this can cause a hard earth fault.
Hope this helps!

UPS engineer

http://www.powerups.co.uk

 

[ScottyUK]

If you pulled the load fuses one at a time and sequentially replaced them before pulling the next one and you had continuous indication of a ground on the system then you may have multiple ground faults on different sub-circuits. Pull everything at once and see if it clears, if not then the likely answers are just above here in raithrovers' post.

One further warning about the Grouser and similar instruments: don't use it on the pilot wires of a differential protection scheme when it is in service, otherwise the relays trip and take out a big chunk of the network. Thankfully it was not one of mine - and I'm not sure I'd admit it if it was - but anecdotal from a neighbouring site. Heh-heh.


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

If we learn from our mistakes I'm getting a great education!

 

[thryll]

Engineering didn't like putting a different signal on the dc Circuit and accounting didn’t want to spend the money on a test box so I have been using what in effect is a Simpson meter. I pull the fuses for that circuit and note the resistance between the Pos or Neg lead and ground. If there seems to be a noticeable short I attack that circuit. It is very time consuming but I have been battling it. Just want to say thanks to everyone for your advice and I welcome more.

 

[ScottyUK]

Tell your engineers not to travel too far: they might fall off the edge of the flat earth. Of course if they do go on any long journeys, put them on a bus with the accountants, make sure the the gas tank is full and cross your fingers!

Good luck, sounds like you are being forced to do things the hard way. You could always trip a few critical loads doing it their way, then ask again for the correct tools for the job.


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

If we learn from our mistakes I'm getting a great education!