Can you supply a link to NEY Contact Manual: Electrical Contacts for Low Energy Uses"? I can't find a copy. Below are my notes on dry switching. This is pulled from our Engineering Wiki, so links have been lost, but I hope content is useful. In one of our applications, we have a mercury cell driving a tamper protect circuit (holding private keys) and tamper switches. The total current draw is 13 uA, so the tamper switches are maybe 20% of that. The switches have a minimum current rating that can't be met with current from the mercury cell. We have a low leakage diode and series resistor to main power (+3.3V) on each switch so a higher current is run through the switch whenever main power is available. The current then backs down to a few uA when main power is not present. The mercury cell was in a battery holder. The mercury cell and the battery holder had nickle plated contacts. In the field, we kept losing keys with a flag indicating we'd lost battery power. The cell and holder manufacturer (the same company) said they'd never had contact problems. But, we moved to a soldered in cell, and that problem went away. Low current switching or just maintaining a contact is tricky! Harold Dry Switching Tyco paper on contact systems - Mostly a discussion of relays, but applies to switches also. Points of interest include: pdf page 4 - Contact surface contamination depends on contact material, ambient atmosphere, temperature, and time. pdf page 4 - Contacts are cleaned during closure by wiping, mechanical pressure, and bounce. pdf page 5 - Contacts can also be electrically cleaned through a process called "fritting" where the increase in contact resistance results in very thin conducting areas between contacts and current through these conducting areas heats the area enough to destroy the insulating film on the contact. "If the layers have not been mechanically destroyed by the closing of the contacts, or if the contacts have been closed for a long time without conducting sufficient current, the electrical effect of fritting will contribute towards establishing a metallic contact, despite layers on the effective contact area. The term fritting describes the electrical breakdown of the oxide/foreign layer when a sufficiently high voltage (fritting voltage) is applied across a closed contact. Due to the applied voltage and the very short distance (the thickness of the layers) between the two potentials an extremely high electric field is generated. The low conductive layer will break down and a small current (a few nA) is forced through very thin channels in the layer. The resulting local high current density heats the conducting channels up quickly, destroying the layers, until finally (within a few ms) a metal to metal bridge is established, electrically linking the two surfaces. The value of fritting voltage depends on the contact material, composition and thickness of the layers, conductivity and composition of the contact surface. Voltages in the range of only a few volts up to some hundreds of volts may be necessary for fritting to occur." pdf page 7 "The term dry circuit describes applications with extremely low loads (e.g. LED's) or circuits which are switched with the electrical load having been previously disconnected, e.g. by electronic means.In these cases the current is too low to establish an electro-thermal cleaning effect and the voltage is below the fritting voltage. The nonconductive oxide layers on the contact surface will not therefore be electrically destroyed. The only remaining cleaning effect is the mechanical destruction of the layers which is sometimes insufficient (e.g. low switching frequency) to give a reliable contact or to keep the contact resistance within specification limits. The correct choice of contact materials is critical in such cases for reliability. As described above, the contact resistance increases with the formation of layers on the contact surface. The thickness and speed of growth depends on contact material, ambient atmosphere and temperature. This is especially important when the contact resistance is tested after a prolonged period of storage." fig4.9.png IEEE Paper - "Measurements made during and after exposure show that application of 20-mA sealing current approximately limits the contact resistance to less than 21 ohms. Smaller currents allow correspondingly higher contact resistances and fluctuations." Abstract here IEEE Paper - "The authors examine the sealing current issue, and discuss the way in which splices degrade through corrosive attack in the service environment. They also discuss the mechanism by which sealing current inhibits this degradation, the use of this understanding in applying sealing current to a subscriber loop, and a low-energy sealing current supply implementation for ISDN (integrated services digital network). While the focus is on the application of sealing current to subscriber loops, the discussion is equally valid in addressing special services circuits and interoffice cables." Abstract PICLIST discussion of dry switching: "... My first experience with this problem was a gold-on-gold contact in a hermetically sealed switch. The current was about 1mA, and after sitting closed for hours and hours, the contacts would noisly go open for a few seconds, then re close (usually) Boosting the current by 10x solved the problem, and I fed that back to the manufacturer, who put that into later production versions of the product." Full Thread Though no documentation related to switches has been found, an intermittent "sealing current" of a few mA is expected to keep contact resistance low over time due to electromigration. Contact Fretting of Electronic Connectors, IEICE TRANS. ELECTRON., VOL.E82{C, NO.1 JANUARY 1999. --=20 FCC Rules Updated Daily at http://www.hallikainen.com Not sent from an iPhone. --=20 http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist .