Gentlemen An isolation xfmr does NOT necessarily reduce leakage current through the ground connector. Most iso xfmrs connect the ground pin, and float line and neutral to the end-use equipment.
If the power supply has components bridging the pri/sec boundary that allow current flow, and depending on if the secondary outputs are ground-referenced, then human (ground-referenced) touch current under some conditions can be very high. If the installed power supply bridges the pri/sec boundary using only galvanic isolation, then almost all human touch current is through the power supply's y-caps that are between primary and ground. And there is no such thing as "medical" y-caps that have low leakage current. The magnitude of current thru the y-caps is simple to calculate; it is the voltage across the cap divided by the cap's reactance. It is important to note that the capacitance for some y-caps can be voltage-dependent. If your empirical measurements are not close to calculated 50/60Hz leakage current values, then it is important to note the quality of the input power, as high crest-factor voltage waveforms will cause the y-caps to "work harder" and thus have more current flowing from the mains input back through ground. Do NOT ever modify components in a certified/recognized component power supply. A slight change in some component values can have profound effects on the performance of the unit under some SFCs and/or transient load conditions. The people that make component power supplies have tested the stability of their units under wide input and output conditions; and if the unit is not stable it will probably not be safe. (CONFLICT OF INTEREST ALERT: my employer makes component power supplies) It is important to use component power supplies with supporting Type Test data that have published leakage currents and indicate test conditions. The OP noted that the test voltage for Touch Current was 264V, so I assume that this is a 1010 or 65-based installation. If the test standard for the component power supply is 60950, then the Touch Current test voltage was probably 254V. You may be using a component outside of its ratings. You must observe the conditions of acceptability for safety-critical components. Please note that the unit's power supply is probably the most safety-critical component in your box. luck, and let's not hurt anybody, Brian -----Original Message----- From: FastWave [ mailto:[email protected]] Sent: Wednesday, November 17, 2004 6:23 AM To: '[email protected]'; 'EMC-PSTC List' Subject: RE: Multiple Leakage/Touch Current Sources Ron, I have conducted hundreds of leakage current tests but never personally redesigned a product to reduce leakage, so here is my "no previous experience" input: My guess is that your leakage current is predominantly line frequency (you may wish to confirm by looking at the leakage current frequency composition). I also guess that the switching power supplies have additional filtering in the front end with capacitors to ground (in addition to the 2 separate filters). This is making for a lot of capacitance to ground which is the source of your leakage current. If these assumptions are correct, the only way to reduce leakage current would be to reduce capacitance to ground (do you really need all those filters). Or, isolate everything behind an isolation transformer (i.e. the medical way). If my presumption that filter capacitance is the cause is correct, it would be additive. Regards, Bill Bisenius, N.C.E. E.D.& D. www.productsafeT.com -----Original Message----- From: Ronald R. Wellman [ mailto:[email protected]] Sent: Wednesday, November 17, 2004 8:26 AM To: EMC-PSTC List Subject: Multiple Leakage/Touch Current Sources Howdy all, I have come across a problem where a product is exceeding 3.5 ma of touch current. The product in question has 2 line filters and 4 Switching Mode Power Supplies. Because of the addition of the fourth power supply the overall touch current can be as much as 4.0 ma. Ideally, we do not want to make the product so it has to be hard-wired to the building mains. Because of the high touch current, does anyone know what the relationships are between different touch current sources? I've heard that touch current is not linear when you add up the different sources and phase shifts between L-G and N-G might be a problem. However, I'm looking for some detailed information on this issue so I can provide ways to investigate the problem and hopefully reduce the overall touch current. BTW, the product is not medical equipment and all touch current tests are done with the product operating at 264 V AC, 50 Hz. Best regards, Ron Wellman ---------------------------------------------------------------- This message is from the IEEE Product Safety Engineering Society emc-pstc discussion list. Website: http://www.ieee-pses.org/ To post a message to the list, send your e-mail to [email protected] Instructions: http://listserv.ieee.org/listserv/request/user-guide.html List rules: http://www.ieee-pses.org/listrules.html For help, send mail to the list administrators: Ron Pickard: [email protected] Scott Douglas [email protected] For policy questions, send mail to: Richard Nute: [email protected] Jim Bacher: [email protected] All emc-pstc postings are archived and searchable on the web at: http://www.ieeecommunities.org/emc-pstc
