HD trouble?
in [Windows XP]
|
From: Bill in Co. on 17 Dec 2009 14:09 Pegasus [MVP] wrote: > "Paul" <nospam(a)needed.com> schrieb im Newsbeitrag > news:hgcvqo$msq$1(a)aioe.org... >> Pegasus [MVP] wrote: >>> >>> >>> The multimeter test might destroy the RAM chip, like the tap on the head >>> with the hammer. I find it difficult to think of a more inappropriate >>> test for a RAM chip. Remember - it contains a hundred million or more >>> transistor gates operating at extremely low currents, less than the most >>> sensitive multimeter can detect! >> >> You can safely test silicon devices, if you use a multimeter with >> "low power ohms" setting. It applies a voltage not intended to >> cause forward conduction in the silicon chips. >> >> If you lost the manual for your multimeter, you can verify this by using >> two multimeters. One multimeter set to "ohms". The second multimeter >> set to "volts". The "volts" multimeter will show you a rough idea >> of the open circuit voltage that the "ohms" meter is putting out. >> (Connect red to red, black to black.) >> >> Test all the "ohms" ranges on the meter, to understand which >> ranges are "high power" and "low power". If an ohms range doesn't >> put out more than 1 volt open circuit, then it should be relatively >> safe to use on just about anything. For the highest resistance >> ranges, the results need some interpretation (since the voltmeter >> has a finite input impedance of its own) >> >> The $100 The $20 >> meter meter >> >> Range Ohmmeter Voltmeter Assumed Confidence >> Reads Reads Power type >> >> Beep Infinity 2.66V High >> Diode infinity 2.67V High High on purpose, for diode >> test >> 200 Infinity 1.01V Low >> 2K infinity 1.01V Low >> 20K Infinity 0.48V Low >> 200K infinity 0.43V Low >> 2M "0.995" 0.23V Low OK, see interpretation >> 20M "0.99" 0.04V Low >> 2000M "001" 0.27V ??? suspicious >> >> (Note - a separate set of tests were done, and the "ohms" multimeter >> never applied more than 1 milliamp of current to the test leads. Typical >> silicon clamp diodes are rated for about 10 milliamps. So there is >> no danger from the level of current flow either.) >> >> In the 2 megohm test case, the "volts" multimeter appears to >> have only a 1 megohm input impedance. Half of the open circuit >> voltage is across the "volts" multimeter, and half is across >> the constant current source inside the meter. We could conclude >> from that, that the actual open circuit voltage applied >> by the multimeter, is 2 * 0.23 volts or 0.46V. And that is suitable >> for low power ohms. So you actually have to stare at the display >> on both devices to understand what is going on. >> >> My suspicion is, the "insulation test" range on my multimeter, is >> actually high power ohms. It is pretty hard to test insulation, >> with a low voltage. Insulation testing would normally be done >> with a "megger". Since I am suspicious of what I see for the >> 2000M range, I likely would not switch to it while working on >> the average PCB. >> >> There are quality voltmeters, with much higher input impedance >> than my $20 "volts" multimeter used in this test. For such a >> device (perhaps a $1000 unit), I likely would not need interpretation >> except for the last test case. And the voltmeter in that case, could >> tell me what the open circuit voltage is, correctly, for the >> rest of the ranges. >> >> The $20 meter is the one I lend to friends :-) >> >> So there are six ranges on my $100 multimeter, that I'd use on a DIMM. >> Without being concerned about any side effects to the silicon. >> >> Some older meters, like my analog Simpson, are a bit meaner. >> They apply nine volts open circuit, and would be unsuited >> for this purpose. I could have run the same set of tests for >> my Simpson, but there is no point, because I simply wouldn't >> use it. My Simpson is good for volts "trend analysis", as digital >> meters aren't perfect for all purposes. There are still >> occasions for reaching for my oldest multimeter. "Ohms" would >> not be one of those reasons. >> >> If a person had concerns about any silicon device they were >> working on, they could look in the datasheet for inspiration. >> I haven't done that in this case for the memory chip. I'm reasonably >> confident, that the six tested ranges on my multimeter, would >> be safe enough. >> >> HTH, >> Paul > > Regardless of the characteristics of any multimeter, it is a totally > unsuitable device for checking out a RAM chip. That is correct.
From: Bill in Co. on 17 Dec 2009 14:18 Paul wrote: > Pegasus [MVP] wrote: >> >> >> "Paul" <nospam(a)needed.com> schrieb im Newsbeitrag >> news:hgcvqo$msq$1(a)aioe.org... >>> Pegasus [MVP] wrote: >>>> >>>> >>>> The multimeter test might destroy the RAM chip, like the tap on the >>>> head with the hammer. I find it difficult to think of a more >>>> inappropriate test for a RAM chip. Remember - it contains a hundred >>>> million or more transistor gates operating at extremely low currents, >>>> less than the most sensitive multimeter can detect! >>> >>> You can safely test silicon devices, if you use a multimeter with >>> "low power ohms" setting. It applies a voltage not intended to >>> cause forward conduction in the silicon chips. >>> >>> If you lost the manual for your multimeter, you can verify this by using >>> two multimeters. One multimeter set to "ohms". The second multimeter >>> set to "volts". The "volts" multimeter will show you a rough idea >>> of the open circuit voltage that the "ohms" meter is putting out. >>> (Connect red to red, black to black.) >>> >>> Test all the "ohms" ranges on the meter, to understand which >>> ranges are "high power" and "low power". If an ohms range doesn't >>> put out more than 1 volt open circuit, then it should be relatively >>> safe to use on just about anything. For the highest resistance >>> ranges, the results need some interpretation (since the voltmeter >>> has a finite input impedance of its own) >>> >>> The $100 The $20 >>> meter meter >>> >>> Range Ohmmeter Voltmeter Assumed Confidence >>> Reads Reads Power type >>> >>> Beep Infinity 2.66V High >>> Diode infinity 2.67V High High on purpose, for >>> diode test >>> 200 Infinity 1.01V Low >>> 2K infinity 1.01V Low >>> 20K Infinity 0.48V Low >>> 200K infinity 0.43V Low >>> 2M "0.995" 0.23V Low OK, see interpretation >>> 20M "0.99" 0.04V Low >>> 2000M "001" 0.27V ??? suspicious >>> >>> (Note - a separate set of tests were done, and the "ohms" multimeter >>> never applied more than 1 milliamp of current to the test leads. Typical >>> silicon clamp diodes are rated for about 10 milliamps. So there is >>> no danger from the level of current flow either.) >>> >>> In the 2 megohm test case, the "volts" multimeter appears to >>> have only a 1 megohm input impedance. Half of the open circuit >>> voltage is across the "volts" multimeter, and half is across >>> the constant current source inside the meter. We could conclude >>> from that, that the actual open circuit voltage applied >>> by the multimeter, is 2 * 0.23 volts or 0.46V. And that is suitable >>> for low power ohms. So you actually have to stare at the display >>> on both devices to understand what is going on. >>> >>> My suspicion is, the "insulation test" range on my multimeter, is >>> actually high power ohms. It is pretty hard to test insulation, >>> with a low voltage. Insulation testing would normally be done >>> with a "megger". Since I am suspicious of what I see for the >>> 2000M range, I likely would not switch to it while working on >>> the average PCB. >>> >>> There are quality voltmeters, with much higher input impedance >>> than my $20 "volts" multimeter used in this test. For such a >>> device (perhaps a $1000 unit), I likely would not need interpretation >>> except for the last test case. And the voltmeter in that case, could >>> tell me what the open circuit voltage is, correctly, for the >>> rest of the ranges. >>> >>> The $20 meter is the one I lend to friends :-) >>> >>> So there are six ranges on my $100 multimeter, that I'd use on a DIMM. >>> Without being concerned about any side effects to the silicon. >>> >>> Some older meters, like my analog Simpson, are a bit meaner. >>> They apply nine volts open circuit, and would be unsuited >>> for this purpose. I could have run the same set of tests for >>> my Simpson, but there is no point, because I simply wouldn't >>> use it. My Simpson is good for volts "trend analysis", as digital >>> meters aren't perfect for all purposes. There are still >>> occasions for reaching for my oldest multimeter. "Ohms" would >>> not be one of those reasons. >>> >>> If a person had concerns about any silicon device they were >>> working on, they could look in the datasheet for inspiration. >>> I haven't done that in this case for the memory chip. I'm reasonably >>> confident, that the six tested ranges on my multimeter, would >>> be safe enough. >>> >>> HTH, >>> Paul >> >> Regardless of the characteristics of any multimeter, it is a totally >> unsuitable device for checking out a RAM chip. > > Doing an ohms test on the chip itself has no meaning. (Like measuring > the resistance from DQ0 to DQ7. That wouldn't mean anything in > particular.) Exactly. It can be fairly useful for some discrete devices like diodes and transistors, however, but is completely inadequate for integrated circuits. > If you're checking continuity on a series damping resistor, > that could be used as a check that the wiring is OK. For > example, doing this, shows the resistor pack is soldered > down OK. > > contact ---- resistor ----- Memory_chip > ^ ^ > | | > +- ohm from here to here -+ > > There aren't a lot of other tests you can do. And that's a pretty limited test. > If you buy "generic" DIMMs by the barrel full, one good test > to run, is measure the rail to rail resistance. Some failed > cheap bypass caps are a dead short, and can cause the motherboard > socket to get burned. (I've seen reports of this on Newegg.) Again, that is a pretty limited test. Extremely limited!! > If you were doing incoming inspection, a quick check between > rails might be a good test. (We actually used to do that > for large circuit packs as well, as an optional test before > applying power. I was surprised, while recording the results > for this test, how well correlated it was. There was little > unit to unit variation, in the ones I recorded. If something > was out of the ordinary, I wouldn't plug in the unit, until > determining why.) > > But I can't see a lot of reasons to be using a multimeter > on a DIMM. Because there really aren't any, for all practical purposes. Simply checking whether or not a capacitor on the board is shorted is a very limited test of the IC (actually, it's not even testing the IC itself).
From: Bill Cunningham on 17 Dec 2009 17:21 "Bill in Co." <not_really_here(a)earthlink.net> wrote in message news:e4hskx0fKHA.5608(a)TK2MSFTNGP05.phx.gbl... > Paul, this only applies to checking *discrete* devices, like diodes and > transistors, it does not apply to integrated circuits. So I couldn't use a multimeter to check my machine's transformer that steps down 120 V powersupply or the system buses in the motherboard? Bill
From: Bill in Co. on 17 Dec 2009 19:01 Bill Cunningham wrote: > "Bill in Co." <not_really_here(a)earthlink.net> wrote in message > news:e4hskx0fKHA.5608(a)TK2MSFTNGP05.phx.gbl... > >> Paul, this only applies to checking *discrete* devices, like diodes and >> transistors, it does not apply to integrated circuits. > > So I couldn't use a multimeter to check my machine's transformer that > steps down 120 V powersupply or the system buses in the motherboard? > > Bill Of course you could. We were talking about semiconductor devices (like ICs = integrated circuits) in this discussion, not transformers, resistors, etc. But just FYI, a transformer is also a discrete (i.e. not integrated) device. :-)
From: Bill Cunningham on 17 Dec 2009 20:22
"Bill in Co." <not_really_here(a)earthlink.net> wrote in message news:u3NS9U3fKHA.1112(a)TK2MSFTNGP04.phx.gbl... > Of course you could. We were talking about semiconductor devices (like > ICs = integrated circuits) in this discussion, not transformers, > resistors, etc. But just FYI, a transformer is also a discrete (i.e. not > integrated) device. :-) Isn't the motherboard, system bus, address bus and so on made of ICs? Can a person test memory with a multimeter via the motherboard and its bridges? Bill |