Solution Finder Get Tech Support

Issues with Line Spikes.

Posted by: StephanL on

We have a mature design of a meter that utilizes a TinySwitch TNY266.  The input is 85/265VAC and the outputs are isolated 5V and 24V. Suppression circuitry includes X and Y safety capacitors.  The meter utilizes signal wire to connect to a remotely mounted transducer.  This transducer couples the signal wires to earth.

The issue is that lightening strikes in the area can traverse the AC power lines across the Y cap, through tthe circuitry, through the signal wires, and through the transducer to earth.  Along the way it damages the meter. 

Any advise as how to minimise damage from this?

Thanks.

Files

Comments

Submitted by PI-YY on 04/12/2017

Hello StephanL,

 

Thank you for interested in using Power Integrations product.

 

From the schematic you show, i didn't see how the meter and transducer are connected. If you can provide more information, it will be more help to dig in more.

 

I think the first thing to figure out is where is the lightening surge comes from. Is it from the AC power line or from the transducer itself.

 

If the Ycap see the lightening strikes and it cannot stand, something should damage at the primary side of the power supply. Is there anything damaged at the power supply side. if not, the power supply should still work and the output should still maintain 24V and 5V. Then something related to the meter and transducer will cause the failure.

 

 

Best Regards

 

 

 

 

Submitted by StephanL on 05/08/2017

<p>I've sketched in additional circuitry to illustrate what's going on.&nbsp;&nbsp; The earthing points can be located fairly far apart.&nbsp;</p>

Attachment Size
Additional circuit information. (82.05 KB) 82.05 KB
Submitted by PI-YY on 05/18/2017

Hello StephanL,

 

Thank you for the detailed information about the connection.

 

There are two connection from the power supply to the meter, one 5V and one 24V. If the damage is from power supply side which means through the input line and the power supply and then damage the meter, it should be either the 5V output and 24V output. i noticed there is a zener at 24V output, i am wondering for those units which damages the meter, is the zener diode D1 still work normally or it is also damaged. If you add another 5V zener at the 5V output, will this help proetect the meter.

 

As this is lighting surge which we can simulate on bench. i am wondering if you can duplicate the issue by doing a standard lighting surge test. If possible, we can monitor the input of the meter which means the voltage and current of the 5V output and 24V output. In this way, we know which signal damage the meter.

 

best Regards