|
Chuck Harris
Hi Raymond,
toggle quoted message
Show quoted text
I was curious, do you have any direct experience with Dallas NVRAM's losing data from desoldering grounded boards? The reason I ask is that the Dallas 1225AD NVRAM contains a special control chip that is continuously powered by an internal lithium cell. The builtin control chip's power is isolated from the Vcc pin, so grounding out that pin will not prevent it from protecting the internal CMOS RAM's data. The builtin control chip disables all external control signals (OE, WE, CE) when the Vcc supply is outside of the manufacturer specified range of 4.5 to 5.5V (DS1225AD). The builtin control chip further requires that external control signals be greater than 2.2V and less than 0.8V to be recognized. And, the control chip requires that Vcc be valid for a minimum of 125ms before it will end its write protection lock, once again allowing the CMOS RAM's contents to be altered. Protection further requires that Vcc come up from invalid to valid faster than 300us; Vcc just meandering its way into the valid range won't release the lock. Manufacturer's specs are just a document, and are sometimes wrong, which is why I am curious if you have any direct experience with Dallas's NVRAM specifications being wrong. Good electronics practice requires that both the circuit board, and the soldering/desoldering equipment be grounded in normal use. Most of us are a little casual about grounding the circuit boards we work on, but are rock solid on grounding our soldering/desoldering equipment, as that comes at no extra effort when using UL soldering/desoldering stations. If you do have any direct experience with Dallas being wrong, I am going to have to rethink what I was taught some nearly 40 years ago by a NASA contractor that I did assembly work for. -Chuck Harris Raymond Domp Frank wrote: On Sun, Aug 11, 2019 at 11:49 AM, gjm45janssen wrote:If you have to desolder the NVRAM, make sure that the board carrying it isn't grounded while desoldering, as would be the case if the 'scope is switched off but still plugged in: The earthing of your soldering iron might provide a logic low level to a write (enable) pin, possibly destroying the contents of the chip. In general, beware of transients. Depending on the chip (I haven't looked up the specs), it might be wise to temporarily connect the Vcc pin to the potential that the tip of your soldering iron is at (usually safety earth) so only a logic high level could reach the chip via your soldering iron. Most chips need a logic low level to enable writing. |
