Transformer grounding resistor to reduce circulating currents

[CRINGEYBOY]

Trying to figure out if adding a resistor to the core ground-to-tank connection will help with circulating currents. What's the best way to determine the right resistance and wattage?

 

[theduality78]

You may need to make iterative calculations or even some trial and error with different resistor values. And monitor the current levels and resistor temperature carefully. In my opinion it's often best to start with a lower resistance and slowly increase it while observing the current.

I'm no expert so let's see what the others might say

 

[Shazzzam]

Measure the current first, then calculate resistance using Ohm's law, choose a resistor with wattage rating at least double the calculated power dissipation for safety

 

[Mr_electricals]

The trick is balancing enough resistance to break the current path without floating the core too much and picking up noise.
Start by measuring the existing current between core and tank, clamp meter with a low scale milliamp setting helps. Once you've got a ballpark value, you can size the resistor to drop that current to something negligible, you'd usually end up in the 10–100 ohm range with 10W+ resistors depending on the situation, whatever you use should be non inductive and well insulated

Drafted this from a big textbook I've not been able to finish

 

[OhmMy]

Absolutely, you nailed it! The trick is all about finding that perfect balance with the resistance and keeping everything non-inductive. And I love that you thought to check the current first,that's a smart move

 

[CRINGEYBOY]

Thanks all for the input! The consensus is to start by measuring the current before picking a resistor, which makes sense. I'll grab a clamp meter with a low milliamp setting and see what numbers I get. The 10–100 ohm range with 10W+ resistors seems like a good ballpark but I'll keep an eye on temperature and noise levels too. Has anyone ever run into unexpected interference after adding resistance? Would be good to know what to watch for.

 

[OhmQu3st]

Everyone's got some solid advice. I had a similar experience a few years ago while setting up a grounding system for a tube amp chassis. I threw a resistor between the chassis ground and the signal ground to eliminate a hum loop, but I forgot to check for any interference. Turns out I picked up all sorts of noise from a nearby fluorescent ballast

What I took away from that is that non-inductive resistors and decent insulation are really important. It's great that you're already considering potential interference

 

[CRINGEYBOY]

Thanks everyone for the excellent guidance. We measured about 150mA of circulating current, and after installing a 50-ohm non-inductive resistor, it has dropped to a negligible level.

@OhmQu3st your story about interference was a good reminder. We have not observed any new noise on the circuit.

 

[Zakimasood1]

Adding a resistor between the core ground and tank usually isn't the best first step—circulating currents are most often caused by having multiple grounding paths, so the priority is to ensure you've got a single-point ground for the core. If you confirm a resistor is really needed, the right way to size it is by measuring the induced voltage between the isolated core and tank, then deciding on a safe target leakage current (often just a few milliamps). From there, you calculate R = V/I and use P = V²/R to get the minimum wattage, always choosing a resistor with plenty of thermal margin. That said, it's generally safer to follow OEM or IEEE/IEC guidance and use proper core-ground monitoring rather than inserting resistance on your own.

 

[CRINGEYBOY]

This provides a very different and more fundamental approach than what was discussed earlier.

The method of measuring open circuit voltage instead of circulating current is a more rigorous engineering practice than the one I followed. It seems I have implemented a solution that treats the symptom, while your recommendation to ensure a true single point ground would correct the root cause. Completely isolating and re-running the ground connections in this particular installation would be a substantial undertaking. Given that the 50 ohm resistor is already installed and has reduced the current, are there any specific long term operational risks or failure modes I should monitor for with this kind of fix in place?