If this is your first visit, be sure to check out the FAQ by clicking the link above. You may have to register before you can post: click the register link above to proceed. To start viewing messages, select the forum that you want to visit from the selection below. |
|
|
Thread Tools | Rate Thread | Display Modes |
#16
|
|||
|
|||
SSD and computer casing.
Bob_S wrote:
"Paul" wrote in message news Bob_S wrote: "Paul" wrote in message news Peter Jason wrote: On Sat, 14 Apr 2018 22:25:50 -0400, Paul wrote: Peter Jason wrote: My computer case is of Aluminium and earthed to a domestic copper water pipe. , the casing of the Samsungs SSDs seem to be anodized aluminium. Since my computer case is old it has no provision for 2.5" SSDs I have them resting on an aluminium surface. Is this a problem? I assume the SSDs have internal insulation. I just measured a couple SSDs here, and son of a bitch, there's a hard connection between SSD casing and logic ground! That's not supposed to be good for ESD handling. I was expecting to find a 1 megohm connection between aluminum SSD casing and the logic ground on the SSD. But it's a hard short. ******* The chassis of your computer case, runs at logic ground potential. The casing of your SSD, runs at logic ground potential. This means you can safely rest the SSD on the bottom of the computer case, without any "thru" current flowing. Just don't touch a 3.3V, 5V, or 12V pin from some Molex connector, to the chassis of the SSD. Since it's a hard ground, we can't be carelessly resting a Molex power pin in contact (somehow) with the chassis. While the pins on Molex are flush or slightly recessed, there might be some conceivable way for something to touch. And my finding a hard ground on the SSD chassis, is not a good thing. I'll be extra careful in future, to keep Molex away from it. I was really expecting to find a 1Megohm (high resistance) drain between shiny aluminum SSD casing and the logic ground next to the data pins of the SSD. GND GND GND === logic ground pins are longer tx+ tx- rx+ rx- on the SSD connector. The data pins on the 7 pin data are recessed. Check with the ohmmeter, between GND and chassis, and you'll see the hard short I measured. Summary: 1) Safe to rest SSD on bottom of computer case, no matter what the casing is made of. 2) Don't touch any "hot" Molex pins, to the SSD housing. Just like you would not touch any "hot" Molex pins, to the computer case metal posts. The motherboard has a hard short to the chassis, through the standoffs. Most everything in the computer, has a hard ground. The only poorly designed feature, is "plastic fascia" on the front of the computer, around front USB ports, which is a poor way to handle ESD issues. 3) Don't scuff across the carpet and pick up your SSD by the aluminum casing. Keep unused SSDs in an ESD bag, scuff across the carpet if you must, pick up the SSD by the ESD bag, touch the ESD bag plastic on the inside of the bag, to bring yourself and the bag contents to the same potential. With your other hand, touch the computer chassis. Now, everything is "drained" to the same electrostatic potential, you can remove the SSD from the ESD bag, and connect it to SATA power and SATA data. By using the series resistance of the ESD bag, to limit ESD current flow, you're *safely* bringing the SSD up to the same electrostatic potential as the thing you wish to connect it to. Note: I just changed my handling procedure! as a result of this ohmmeter measurement. I'd assumed SSD chassis were drain connected, not "hard" connected. Now I will have to start using ESD bags for them, just as I do for hard drives. Paul Thanks Paul. I have put my SSDs in plastic sleeves, and rested this combination on cardboard. I thought this might be the cause of my computer restarts. Also I have disabled all "fast startups" via the registry. I wouldn't entirely insulate the thing. Some SSDs have high peak power usage (SandForce compression), and they get a little warm. Put a cardboard underneath if you like. Leave at least one metal surface for cooling. In this review, you can see (and it's mentioned in the text), that thermal pads were used on all the NAND chips. That means both metal surfaces are used for their slight cooling advantage. http://www.thessdreview.com/our-revi...-review-512gb/ Paul Paul, Have no idea why you would expect the SSD to be isolated from case ground. Everything in the computer and the case itself is tied to the grid ground. Mount a motherboard and the mounting screws screw into the case as well as the screw heads tighten down on ground pads on the motherboard. SSD's are also grounded when placed in a mounting tray to case ground. As for the longer pins on power and data connectors, the pins that are designated as ground pins are longer to insure a ground connection is made before any data pins mate to prevent static discharges. This pin design also accommodates "hot swap" capable devices such as hard drives and SSD's. While it's always good practice to use measures to prevent ESD when mounting a device such as a SSD, do not isolate it from the mounting mechanisms such as placing an insulator between it and a metal case. Depending on the SSD (some cases are plastic), the metal case is the heat sink as shown in the link you included in your post. Please note that the logic board is held in-place by a screw in the back that fastens to the SSD case and tightens against a ground pad on the SSD logic board. There are two ground philosophies for protecting assemblies from ESD. Neither was followed in this case. Paul Paul, Not sure I follow your logic on this. The OP said he had his case grounded to a copper pipe. What wasn't stated is how his power supply is grounded or if it was even plugged in. In the US, the 3rd lug on a plug is the green (safety) ground and in a typical home wiring setup, the white wire and green are tied to a single bus bar in the distribution panel so as to prevent ground loops / current loops. So while his setup using a ground from the case to a copper water pipe could potentially create a ground loop if there were other wiring problems, it still is ground as long as there's no interruption to earth ground in that copper pipe and in his wiring and would dissipate any ESD charge. Potential for ESD would be from him picking up the SSD and then plugging it in before he touched the case (grounding himself) and creating a arc between his hand holding the SSD and getting the data or power connectors close enough so the static potential went from him to the ground pins on the cable or the case. That’s' why the manufactures include warnings about using ground straps and/or touching the case before working on it. What people don't realize is that usually, they unplug the power cord from the wall or the power supply and that’s when their case is no longer grounded. But in the OP's situation, he had an additional ground to a water pipe. That may or may not be a good situation depending on his house wiring and I've never seen a power supply plug without a polarized plug. Now if the wall socket (old house wiring) does not have a safety ground, then his water pipe ground is a good idea. I've worked at radar sites and on some of the largest computer installations in the world and worked worked with power engineers that specialize in how "all things" need to be grounded and why. It is a complicated field of study and ESD can create havoc on computer components when you have thousands of volts jumping between you and the computer. So I'm not trying to argue with you but saying "Neither was followed in this case" didn't make sense to me. Perhaps you could expand on the "Neither" a bit. I'm not sure if you are referring to the OP or to your own situation. If it's your own, then I'll shut up and sit down cause I don't know what or how you were doing your measurements. 1) There's no such thing as "ground". The earth is an insulated ball floating in the vacuum of space. The earth is merely a "reference point". It has no special properties as such. "Standing on dirt is not magic." But since we are in frequent contact with dirt though, it's in our best interest to not touch potentials higher or lower than that potential. We could get a shock. 2) To avoid ESD, bring all assemblies to "equipotential". That's the trick. If I unplug a PC from the wall, remove all cabling, I can still *safely* install hardware in it, as long as I follow equipotential rules. This means opening the ESD bag but not removing the new hardware. Touch the inside of the bag while touching the chassis. It helps if you wear short pants and lay the metal-bodied PC flat on your lap while working (that's if you don't own an ESD strap). By bringing the PC, the new hardware item, and you to the same electrostatic potential, there is no danger of ESD damage. And notice in that picture, there is *no* connection to earth ground, safety ground, cold water pipes or anything. You make an "electrostatic island" of yourself, then ensure that all entities are "charged to the same level". I could place my work chair on glass blocks, charge myself to 50kV ESD level, and as long as I followed the rules about touching the inside of the ESD bag to bring the bag to the same potential as the other kit, I can actually install something without damage. The usage of ESD straps (where the ESD strap is fastened to the case via a 10meg series resistor), is for a similar purpose. It's not ground - it's a way to bring the computer chassis and any other objects used in the operation, to the same electrostatic potential. The series resistor inside the strap, performs the same function as the "resistive" surface inside the ESD bag. It's a dissipative path, that limits the ESD peak current flow level. By placing a hard connection between the metal chassis of the SSD and logic ground, now we have to resort to ESD bags again, as there is no dissipative path to logic ground. I can't really afford to scuff across the carpet, and discharge 50kV into the SSD chassis, because the result will now be unknown (I won't know whether I could have damaged it or not, and it could depend on the way the PCB and mount points are designed). If the connection between SSD chassis and PCB was dissipative, then I'd know there was no possibility of a high "peak" current flow during equalization. If the SSD chassis was plastic, I would be using the second philosophy, which says "don't allow static to dissipate into the item at all". Then the advanced ground contacts take the brunt of the equalization current flow. And things like this aren't left to chance in the lab. If I wanted to be sure my decision about the hard ground on the SSD wasn't a mistake, I do this kind of testing. We test for upset level. We test for damage level (which is actually pretty hard to determine, what constitutes "damage"). The bad thing about ESD, is you can "weaken" a piece of silicon, which weeks later, decides to completely fail. And then it's hard to trace back exactly when you "broke it". https://en.wikipedia.org/wiki/Human-body_model These have changed, since the last time I tried this stuff. We had a nasty gadget for doing up to 15kV tests on our products. This is so much more "refined" looking. https://transientspecialists.com/pro...contact-mode-1 There are two ways to damage things with ESD. One is via direct discharge. The other is via "induction". When usage of an improperly designed front USB solution blows out the pad I/O, it's just as likely the transient traveling down the black wire, "induces" a potential in D+ and D-, and blows out the driver pad. When you do ESD analysis, the danger isn't always from the direct path. I will be switching to ESD bags from now on, for my SSDs, and bringing myself to the same potential as the chassis. (The same kind of gymnastics you do, when you don't have an ESD strap handy.) It's not because I know for a fact the SSD will fail. It's because I don't really know, for any individual SSD, whether it has a particular vulnerability or not. If I had a pile of these in the lab, and an HBM-type tester, I could take a shot at finding out. Paul |
Ads |
#17
|
|||
|
|||
SSD and computer casing.
On Sun, 15 Apr 2018 22:32:36 -0400, "Bob_S" wrote:
"Paul" wrote in message news Bob_S wrote: "Paul" wrote in message news Peter Jason wrote: On Sat, 14 Apr 2018 22:25:50 -0400, Paul wrote: Peter Jason wrote: My computer case is of Aluminium and earthed to a domestic copper water pipe. , the casing of the Samsungs SSDs seem to be anodized aluminium. Since my computer case is old it has no provision for 2.5" SSDs I have them resting on an aluminium surface. Is this a problem? I assume the SSDs have internal insulation. I just measured a couple SSDs here, and son of a bitch, there's a hard connection between SSD casing and logic ground! That's not supposed to be good for ESD handling. I was expecting to find a 1 megohm connection between aluminum SSD casing and the logic ground on the SSD. But it's a hard short. ******* The chassis of your computer case, runs at logic ground potential. The casing of your SSD, runs at logic ground potential. This means you can safely rest the SSD on the bottom of the computer case, without any "thru" current flowing. Just don't touch a 3.3V, 5V, or 12V pin from some Molex connector, to the chassis of the SSD. Since it's a hard ground, we can't be carelessly resting a Molex power pin in contact (somehow) with the chassis. While the pins on Molex are flush or slightly recessed, there might be some conceivable way for something to touch. And my finding a hard ground on the SSD chassis, is not a good thing. I'll be extra careful in future, to keep Molex away from it. I was really expecting to find a 1Megohm (high resistance) drain between shiny aluminum SSD casing and the logic ground next to the data pins of the SSD. GND GND GND === logic ground pins are longer tx+ tx- rx+ rx- on the SSD connector. The data pins on the 7 pin data are recessed. Check with the ohmmeter, between GND and chassis, and you'll see the hard short I measured. Summary: 1) Safe to rest SSD on bottom of computer case, no matter what the casing is made of. 2) Don't touch any "hot" Molex pins, to the SSD housing. Just like you would not touch any "hot" Molex pins, to the computer case metal posts. The motherboard has a hard short to the chassis, through the standoffs. Most everything in the computer, has a hard ground. The only poorly designed feature, is "plastic fascia" on the front of the computer, around front USB ports, which is a poor way to handle ESD issues. 3) Don't scuff across the carpet and pick up your SSD by the aluminum casing. Keep unused SSDs in an ESD bag, scuff across the carpet if you must, pick up the SSD by the ESD bag, touch the ESD bag plastic on the inside of the bag, to bring yourself and the bag contents to the same potential. With your other hand, touch the computer chassis. Now, everything is "drained" to the same electrostatic potential, you can remove the SSD from the ESD bag, and connect it to SATA power and SATA data. By using the series resistance of the ESD bag, to limit ESD current flow, you're *safely* bringing the SSD up to the same electrostatic potential as the thing you wish to connect it to. Note: I just changed my handling procedure! as a result of this ohmmeter measurement. I'd assumed SSD chassis were drain connected, not "hard" connected. Now I will have to start using ESD bags for them, just as I do for hard drives. Paul Thanks Paul. I have put my SSDs in plastic sleeves, and rested this combination on cardboard. I thought this might be the cause of my computer restarts. Also I have disabled all "fast startups" via the registry. I wouldn't entirely insulate the thing. Some SSDs have high peak power usage (SandForce compression), and they get a little warm. Put a cardboard underneath if you like. Leave at least one metal surface for cooling. In this review, you can see (and it's mentioned in the text), that thermal pads were used on all the NAND chips. That means both metal surfaces are used for their slight cooling advantage. http://www.thessdreview.com/our-revi...-review-512gb/ Paul Paul, Have no idea why you would expect the SSD to be isolated from case ground. Everything in the computer and the case itself is tied to the grid ground. Mount a motherboard and the mounting screws screw into the case as well as the screw heads tighten down on ground pads on the motherboard. SSD's are also grounded when placed in a mounting tray to case ground. As for the longer pins on power and data connectors, the pins that are designated as ground pins are longer to insure a ground connection is made before any data pins mate to prevent static discharges. This pin design also accommodates "hot swap" capable devices such as hard drives and SSD's. While it's always good practice to use measures to prevent ESD when mounting a device such as a SSD, do not isolate it from the mounting mechanisms such as placing an insulator between it and a metal case. Depending on the SSD (some cases are plastic), the metal case is the heat sink as shown in the link you included in your post. Please note that the logic board is held in-place by a screw in the back that fastens to the SSD case and tightens against a ground pad on the SSD logic board. There are two ground philosophies for protecting assemblies from ESD. Neither was followed in this case. Paul Paul, Not sure I follow your logic on this. The OP said he had his case grounded to a copper pipe. What wasn't stated is how his power supply is grounded or if it was even plugged in. In the US, the 3rd lug on a plug is the green (safety) ground and in a typical home wiring setup, the white wire and green are tied to a single bus bar in the distribution panel so as to prevent ground loops / current loops. So while his setup using a ground from the case to a copper water pipe could potentially create a ground loop if there were other wiring problems, it still is ground as long as there's no interruption to earth ground in that copper pipe and in his wiring and would dissipate any ESD charge. Potential for ESD would be from him picking up the SSD and then plugging it in before he touched the case (grounding himself) and creating a arc between his hand holding the SSD and getting the data or power connectors close enough so the static potential went from him to the ground pins on the cable or the case. That’s' why the manufactures include warnings about using ground straps and/or touching the case before working on it. What people don't realize is that usually, they unplug the power cord from the wall or the power supply and that’s when their case is no longer grounded. But in the OP's situation, he had an additional ground to a water pipe. That may or may not be a good situation depending on his house wiring and I've never seen a power supply plug without a polarized plug. Now if the wall socket (old house wiring) does not have a safety ground, then his water pipe ground is a good idea. I've worked at radar sites and on some of the largest computer installations in the world and worked worked with power engineers that specialize in how "all things" need to be grounded and why. It is a complicated field of study and ESD can create havoc on computer components when you have thousands of volts jumping between you and the computer. So I'm not trying to argue with you but saying "Neither was followed in this case" didn't make sense to me. Perhaps you could expand on the "Neither" a bit. I'm not sure if you are referring to the OP or to your own situation. If it's your own, then I'll shut up and sit down cause I don't know what or how you were doing your measurements. The new copper water pipe is near the computer, so I just clipped to it a wire connected to the computer case. Insurance, that's all. |
#18
|
|||
|
|||
SSD and computer casing.
On 4/16/2018 9:00 AM, Wolf K wrote:
On 2018-04-16 02:32, Peter Jason wrote: [...] The new copper water pipe is near the computer, so I just clipped to it a wire connected to the computer case.Â*Â* Insurance, that's all. As others have pointed out, that's like grounding the case after you've unplugged the power cord, which has a ground wire, and grounds the case as ;long as the cord is connected. Nowadays, house wiring is no longer grounded to the water lines: IIRC, the grounding is done at the entry point, at the meter. Thus, the house wiring is grounded via the grid. Since there can be inadvertent grounding, we have ground-fault detector-equipped wall plugs. Code here requires all external and bathroom/kitchen pugs to be ground-fault detectors. These detect a grounded current leak that bypasses the whole-house grounding, and disconnect the plug internally. That's why I'm uncertain about whether your "insurance" is a good idea. I'd like to know. :-) PS: First house we owned grounded the panel to the waterline. But the wiring was two-wire, no ground. Unpolarised, too. Having equipment grounded to differ ground points can be as dangerous as having no ground at all. As stated in previous post, modern house is grounded to a point common to all outlets in the house. With different ground points you can actually have a potential difference between the ground points which can cause currents where the two grounds come together, ie your computer. When we installed the LAN at a company I worked at, we spent a considerable amount of money insuring there was zero potential between the points where the all computers were attached to the LAN. With the shrinking power needs for modern electronics, I have often wonder if the potential difference could be used to run some small electronic devices, and do away with the need for batteries -- 2018: The year we learn to play the great game of Euchre |
#19
|
|||
|
|||
SSD and computer casing.
On 04/15/2018 11:12 PM, Paul wrote:
Bob_S wrote: "Paul"Â* wrote in message news Bob_S wrote: "Paul"Â* wrote in message news Peter Jason wrote: On Sat, 14 Apr 2018 22:25:50 -0400, Paul wrote: Peter Jason wrote: My computer case is of Aluminium and earthed to a domestic copper water pipe. , the casing of the Samsungs SSDs seem to be anodized aluminium.Â* Since my computer case is old it has no provision for 2.5" SSDs I have them resting on an aluminium surface.Â*Â* Is this a problem? I assume the SSDs have internal insulation. I just measured a couple SSDs here, and son of a bitch, there's a hard connection between SSD casing and logic ground! That's not supposed to be good for ESD handling. I was expecting to find a 1 megohm connection between aluminum SSD casing and the logic ground on the SSD. But it's a hard short. ******* The chassis of your computer case, runs at logic ground potential. The casing of your SSD, runs at logic ground potential. This means you can safely rest the SSD on the bottom of the computer case, without any "thru" current flowing. Just don't touch a 3.3V, 5V, or 12V pin from some Molex connector, to the chassis of the SSD. Since it's a hard ground, we can't be carelessly resting a Molex power pin in contact (somehow) with the chassis. While the pins on Molex are flush or slightly recessed, there might be some conceivable way for something to touch. And my finding a hard ground on the SSD chassis, is not a good thing. I'll be extra careful in future, to keep Molex away from it. I was really expecting to find a 1Megohm (high resistance) drain between shiny aluminum SSD casing and the logic ground next to the data pins of the SSD. Â*Â*Â* GNDÂ*Â*Â*Â*Â*Â*Â*Â* GNDÂ*Â*Â*Â*Â*Â*Â*Â* GNDÂ*Â*Â* === logic ground pins are longer Â*Â*Â*Â*Â*Â*Â* tx+ tx-Â*Â*Â*Â* rx+ rx-Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* on the SSD connector. The data Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* pins on the 7 pin data are Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* recessed. Check with the ohmmeter, Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* between GND and chassis, and you'll Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* see the hard short I measured. Summary: 1) Safe to rest SSD on bottom of computer case, no matter Â*Â* what the casing is made of. 2) Don't touch any "hot" Molex pins, to the SSD housing. Â*Â* Just like you would not touch any "hot" Molex pins, Â*Â* to the computer case metal posts. The motherboard has a hard Â*Â* short to the chassis, through the standoffs. Most everything Â*Â* in the computer, has a hard ground. The only poorly designed Â*Â* feature, is "plastic fascia" on the front of the computer, Â*Â* around front USB ports, which is a poor way to handle ESD issues. 3) Don't scuff across the carpet and pick up your Â*Â* SSD by the aluminum casing. Keep unused SSDs in an ESD bag, Â*Â* scuff across the carpet if you must, pick up the Â*Â* SSD by the ESD bag, touch the ESD bag plastic on Â*Â* the inside of the bag, to bring yourself and the bag contents Â*Â* to the same potential. With your other hand, touch Â*Â* the computer chassis. Now, everything is "drained" to Â*Â* the same electrostatic potential, you can remove the Â*Â* SSD from the ESD bag, and connect it to SATA power Â*Â* and SATA data. Â*Â* By using the series resistance of the ESD bag, to limit Â*Â* ESD current flow, you're *safely* bringing the SSD up Â*Â* to the same electrostatic potential as the thing you Â*Â* wish to connect it to. Note: I just changed my handling procedure! as a result Â*Â*Â*Â*Â* of this ohmmeter measurement. I'd assumed SSD chassis Â*Â*Â*Â*Â* were drain connected, not "hard" connected. Now I Â*Â*Â*Â*Â* will have to start using ESD bags for them, just as Â*Â*Â*Â*Â* I do for hard drives. Â*Â* Paul Thanks Paul.Â* I have put my SSDs in plastic sleeves, and rested this combination on cardboard.Â*Â* I thought this might be the cause of my computer restarts.Â*Â* Also I have disabled all "fast startups" via the registry. I wouldn't entirely insulate the thing. Some SSDs have high peak power usage (SandForce compression), and they get a little warm. Put a cardboard underneath if you like. Leave at least one metal surface for cooling. In this review, you can see (and it's mentioned in the text), that thermal pads were used on all the NAND chips. That means both metal surfaces are used for their slight cooling advantage. http://www.thessdreview.com/our-revi...-review-512gb/ Â*Â* Paul Paul, Have no idea why you would expect the SSD to be isolated from case ground. Everything in the computer and the case itself is tied to the grid ground. Mount a motherboard and the mounting screws screw into the case as well as the screw heads tighten down on ground pads on the motherboard.Â* SSD's are also grounded when placed in a mounting tray to case ground. As for the longer pins on power and data connectors, the pins that are designated as ground pins are longer to insure a ground connection is made before any data pins mate to prevent static discharges.Â* This pin design also accommodates "hot swap" capable devices such as hard drives and SSD's. While it's always good practice to use measures to prevent ESD when mounting a device such as a SSD, do not isolate it from the mounting mechanisms such as placing an insulator between it and a metal case. Depending on the SSD (some cases are plastic), the metal case is the heat sink as shown in the link you included in your post.Â* Please note that the logic board is held in-place by a screw in the back that fastens to the SSD case and tightens against a ground pad on the SSD logic board. There are two ground philosophies for protecting assemblies from ESD. Neither was followed in this case. Â*Â* Paul Paul, Not sure I follow your logic on this. The OP said he had his case grounded to a copper pipe.Â* What wasn't stated is how his power supply is grounded or if it was even plugged in.Â* In the US, the 3rd lug on a plug is the green (safety) ground and in a typical home wiring setup, the white wire and green are tied to a single bus bar in the distribution panel so as to prevent ground loops / current loops. So while his setup using a ground from the case to a copper water pipe could potentially create a ground loop if there were other wiring problems, it still is ground as long as there's no interruption to earth ground in that copper pipe and in his wiring and would dissipate any ESD charge. Potential for ESD would be from him picking up the SSD and then plugging it in before he touched the case (grounding himself) and creating a arc between his hand holding the SSD and getting the data or power connectors close enough so the static potential went from him to the ground pins on the cable or the case. That’s' why the manufactures include warnings about using ground straps and/or touching the case before working on it.Â* What people don't realize is that usually, they unplug the power cord from the wall or the power supply and that’s when their case is no longer grounded. But in the OP's situation, he had an additional ground to a water pipe.Â* That may or may not be a good situation depending on his house wiring and I've never seen a power supply plug without a polarized plug.Â*Â* Now if the wall socket (old house wiring) does not have a safety ground, then his water pipe ground is a good idea. I've worked at radar sites and on some of the largest computer installations in the world and worked worked with power engineers that specialize in how "all things" need to be grounded and why.Â* It is a complicated field of study and ESD can create havoc on computer components when you have thousands of volts jumping between you and the computer. So I'm not trying to argue with you but saying "Neither was followed in this case" didn't make sense to me.Â* Perhaps you could expand on the "Neither" a bit.Â* I'm not sure if you are referring to the OP or to your own situation. If it's your own, then I'll shut up and sit down cause I don't know what or how you were doing your measurements. 1) There's no such thing as "ground". The earth is an insulated Â*Â* ball floating in the vacuum of space. The earth is merely Â*Â* a "reference point". It has no special properties as such. Â*Â* "Standing on dirt is not magic." But since we are in frequent Â*Â* contact with dirt though, it's in our best interest to not Â*Â* touch potentials higher or lower than that potential. We could Â*Â* get a shock. 2) To avoid ESD, bring all assemblies to "equipotential". Â*Â* That's the trick. If I unplug a PC from the wall, remove all cabling, I can still *safely* install hardware in it, as long as I follow equipotential rules. This means opening the ESD bag but not removing the new hardware. Touch the inside of the bag while touching the chassis. It helps if you wear short pants and lay the metal-bodied PC flat on your lap while working (that's if you don't own an ESD strap). By bringing the PC, the new hardware item, and you to the same electrostatic potential, there is no danger of ESD damage. And notice in that picture, there is *no* connection to earth ground, safety ground, cold water pipes or anything. You make an "electrostatic island" of yourself, then ensure that all entities are "charged to the same level". I could place my work chair on glass blocks, charge myself to 50kV ESD level, and as long as I followed the rules about touching the inside of the ESD bag to bring the bag to the same potential as the other kit, I can actually install something without damage. The usage of ESD straps (where the ESD strap is fastened to the case via a 10meg series resistor), is for a similar purpose. It's not ground - it's a way to bring the computer chassis and any other objects used in the operation, to the same electrostatic potential. The series resistor inside the strap, performs the same function as the "resistive" surface inside the ESD bag. It's a dissipative path, that limits the ESD peak current flow level. By placing a hard connection between the metal chassis of the SSD and logic ground, now we have to resort to ESD bags again, as there is no dissipative path to logic ground. I can't really afford to scuff across the carpet, and discharge 50kV into the SSD chassis, because the result will now be unknown (I won't know whether I could have damaged it or not, and it could depend on the way the PCB and mount points are designed). If the connection between SSD chassis and PCB was dissipative, then I'd know there was no possibility of a high "peak" current flow during equalization. If the SSD chassis was plastic, I would be using the second philosophy, which says "don't allow static to dissipate into the item at all". Then the advanced ground contacts take the brunt of the equalization current flow. And things like this aren't left to chance in the lab. If I wanted to be sure my decision about the hard ground on the SSD wasn't a mistake, I do this kind of testing. We test for upset level. We test for damage level (which is actually pretty hard to determine, what constitutes "damage"). The bad thing about ESD, is you can "weaken" a piece of silicon, which weeks later, decides to completely fail. And then it's hard to trace back exactly when you "broke it". https://en.wikipedia.org/wiki/Human-body_model These have changed, since the last time I tried this stuff. We had a nasty gadget for doing up to 15kV tests on our products. This is so much more "refined" looking. https://transientspecialists.com/pro...contact-mode-1 There are two ways to damage things with ESD. One is via direct discharge. The other is via "induction". When usage of an improperly designed front USB solution blows out the pad I/O, it's just as likely the transient traveling down the black wire, "induces" a potential in D+ and D-, and blows out the driver pad. When you do ESD analysis, the danger isn't always from the direct path. I will be switching to ESD bags from now on, for my SSDs, and bringing myself to the same potential as the chassis. (The same kind of gymnastics you do, when you don't have an ESD strap handy.) It's not because I know for a fact the SSD will fail. It's because I don't really know, for any individual SSD, whether it has a particular vulnerability or not. If I had a pile of these in the lab, and an HBM-type tester, I could take a shot at finding out. Â*Â* Paul Here's an extreme example. Rene |
#20
|
|||
|
|||
SSD and computer casing.
Wolf K wrote:
On 2018-04-16 02:32, Peter Jason wrote: [...] The new copper water pipe is near the computer, so I just clipped to it a wire connected to the computer case. Insurance, that's all. As others have pointed out, that's like grounding the case after you've unplugged the power cord, which has a ground wire, and grounds the case as ;long as the cord is connected. Nowadays, house wiring is no longer grounded to the water lines: IIRC, the grounding is done at the entry point, at the meter. Thus, the house wiring is grounded via the grid. Since there can be inadvertent grounding, we have ground-fault detector-equipped wall plugs. Code here requires all external and bathroom/kitchen pugs to be ground-fault detectors. These detect a grounded current leak that bypasses the whole-house grounding, and disconnect the plug internally. That's why I'm uncertain about whether your "insurance" is a good idea. I'd like to know. :-) PS: First house we owned grounded the panel to the waterline. But the wiring was two-wire, no ground. Unpolarised, too. I'm not an electrician, but I don't think Safety Ground comes from the grid. In the diagram here, the House near the guys name "Gene", you can see a ground rod just outside the house. I think that's Safety Ground, which is not exactly the same thing as Neutral. Neutral and Safety Ground should be close in potential (~1V say), but aren't always exactly the same thing. And under fault conditions, Neutral could end up at just about any potential. That's why you don't necessarily want to use Neutral as a "cheap chassis ground". http://indexxit.com/wp-content/uploa...ru-970x912.jpg And some distribution systems, don't use the Neutral idea. They actually have the whole return current flow through Earth. What happens to your Safety Ground, on a system like this ? Must depend on where exactly, you drive the rod. https://en.wikipedia.org/wiki/Single-wire_earth_return And when it comes to "rods", the practice isn't the same all over the continent. Some areas have poor soil conditions, and multiple rods are required. Only a local electrician will know the "local rules". And nobody actually wants grounding rods that require "maintenance", because nobody would tell the owner about that. Who would suspect they're expected to look after their ground ? Paul |
#21
|
|||
|
|||
SSD and computer casing.
On 04/16/2018 9:57 AM, Paul wrote:
Wolf K wrote: On 2018-04-16 02:32, Peter Jason wrote: [...] The new copper water pipe is near the computer, so I just clipped to it a wire connected to the computer case.Â*Â* Insurance, that's all. As others have pointed out, that's like grounding the case after you've unplugged the power cord, which has a ground wire, and grounds the case as ;long as the cord is connected. Nowadays, house wiring is no longer grounded to the water lines: IIRC, the grounding is done at the entry point, at the meter. Thus, the house wiring is grounded via the grid. Since there can be inadvertent grounding, we have ground-fault detector-equipped wall plugs. Code here requires all external and bathroom/kitchen pugs to be ground-fault detectors. These detect a grounded current leak that bypasses the whole-house grounding, and disconnect the plug internally. That's why I'm uncertain about whether your "insurance" is a good idea. I'd like to know. :-) PS: First house we owned grounded the panel to the waterline. But the wiring was two-wire, no ground. Unpolarised, too. I'm not an electrician, but I don't think Safety Ground comes from the grid. In the diagram here, the House near the guys name "Gene", you can see a ground rod just outside the house. I think that's Safety Ground, which is not exactly the same thing as Neutral. Neutral and Safety Ground should be close in potential (~1V say), but aren't always exactly the same thing. And under fault conditions, Neutral could end up at just about any potential. That's why you don't necessarily want to use Neutral as a "cheap chassis ground". http://indexxit.com/wp-content/uploa...ru-970x912.jpg And some distribution systems, don't use the Neutral idea. They actually have the whole return current flow through Earth. What happens to your Safety Ground, on a system like this ? Must depend on where exactly, you drive the rod. https://en.wikipedia.org/wiki/Single-wire_earth_return And when it comes to "rods", the practice isn't the same all over the continent. Some areas have poor soil conditions, and multiple rods are required. Only a local electrician will know the "local rules". And nobody actually wants grounding rods that require "maintenance", because nobody would tell the owner about that. Who would suspect they're expected to look after their ground ? Â*Â* Paul Where I was brought up on my dads farm in Saskatchewan about 1952, Electric power was installed to the local farms with a single conductor to the yard pole and 2 copper rods about ten feet long where driven into the ground at the pole to use the earth as the neutral conductor. Rene |
#22
|
|||
|
|||
SSD and computer casing.
Rene Lamontagne wrote:
On 04/16/2018 9:57 AM, Paul wrote: Wolf K wrote: On 2018-04-16 02:32, Peter Jason wrote: [...] The new copper water pipe is near the computer, so I just clipped to it a wire connected to the computer case. Insurance, that's all. As others have pointed out, that's like grounding the case after you've unplugged the power cord, which has a ground wire, and grounds the case as ;long as the cord is connected. Nowadays, house wiring is no longer grounded to the water lines: IIRC, the grounding is done at the entry point, at the meter. Thus, the house wiring is grounded via the grid. Since there can be inadvertent grounding, we have ground-fault detector-equipped wall plugs. Code here requires all external and bathroom/kitchen pugs to be ground-fault detectors. These detect a grounded current leak that bypasses the whole-house grounding, and disconnect the plug internally. That's why I'm uncertain about whether your "insurance" is a good idea. I'd like to know. :-) PS: First house we owned grounded the panel to the waterline. But the wiring was two-wire, no ground. Unpolarised, too. I'm not an electrician, but I don't think Safety Ground comes from the grid. In the diagram here, the House near the guys name "Gene", you can see a ground rod just outside the house. I think that's Safety Ground, which is not exactly the same thing as Neutral. Neutral and Safety Ground should be close in potential (~1V say), but aren't always exactly the same thing. And under fault conditions, Neutral could end up at just about any potential. That's why you don't necessarily want to use Neutral as a "cheap chassis ground". http://indexxit.com/wp-content/uploa...ru-970x912.jpg And some distribution systems, don't use the Neutral idea. They actually have the whole return current flow through Earth. What happens to your Safety Ground, on a system like this ? Must depend on where exactly, you drive the rod. https://en.wikipedia.org/wiki/Single-wire_earth_return And when it comes to "rods", the practice isn't the same all over the continent. Some areas have poor soil conditions, and multiple rods are required. Only a local electrician will know the "local rules". And nobody actually wants grounding rods that require "maintenance", because nobody would tell the owner about that. Who would suspect they're expected to look after their ground ? Paul Where I was brought up on my dads farm in Saskatchewan about 1952, Electric power was installed to the local farms with a single conductor to the yard pole and 2 copper rods about ten feet long where driven into the ground at the pole to use the earth as the neutral conductor. Rene Were there ever any surprised ? Unhappy cows in the milking parlor ? Apparently cows don't like receiving an electric shock through the udder, from the milking machine :-) And apparently a tiny potential difference is enough to do it (1V level). Paul |
#23
|
|||
|
|||
SSD and computer casing.
On 04/16/2018 10:26 AM, Paul wrote:
Rene Lamontagne wrote: On 04/16/2018 9:57 AM, Paul wrote: Wolf K wrote: On 2018-04-16 02:32, Peter Jason wrote: [...] The new copper water pipe is near the computer, so I just clipped to it a wire connected to the computer case.Â*Â* Insurance, that's all. As others have pointed out, that's like grounding the case after you've unplugged the power cord, which has a ground wire, and grounds the case as ;long as the cord is connected. Nowadays, house wiring is no longer grounded to the water lines: IIRC, the grounding is done at the entry point, at the meter. Thus, the house wiring is grounded via the grid. Since there can be inadvertent grounding, we have ground-fault detector-equipped wall plugs. Code here requires all external and bathroom/kitchen pugs to be ground-fault detectors. These detect a grounded current leak that bypasses the whole-house grounding, and disconnect the plug internally. That's why I'm uncertain about whether your "insurance" is a good idea. I'd like to know. :-) PS: First house we owned grounded the panel to the waterline. But the wiring was two-wire, no ground. Unpolarised, too. I'm not an electrician, but I don't think Safety Ground comes from the grid. In the diagram here, the House near the guys name "Gene", you can see a ground rod just outside the house. I think that's Safety Ground, which is not exactly the same thing as Neutral. Neutral and Safety Ground should be close in potential (~1V say), but aren't always exactly the same thing. And under fault conditions, Neutral could end up at just about any potential. That's why you don't necessarily want to use Neutral as a "cheap chassis ground". http://indexxit.com/wp-content/uploa...ru-970x912.jpg And some distribution systems, don't use the Neutral idea. They actually have the whole return current flow through Earth. What happens to your Safety Ground, on a system like this ? Must depend on where exactly, you drive the rod. https://en.wikipedia.org/wiki/Single-wire_earth_return And when it comes to "rods", the practice isn't the same all over the continent. Some areas have poor soil conditions, and multiple rods are required. Only a local electrician will know the "local rules". And nobody actually wants grounding rods that require "maintenance", because nobody would tell the owner about that. Who would suspect they're expected to look after their ground ? Â*Â*Â* Paul Where I was brought up on my dads farm in Saskatchewan about 1952, Electric power was installed to the local farms with a single conductor to the yard pole and 2 copper rods about ten feet long where driven into the ground at the pole to use the earth as the neutral conductor. Rene Were there ever any surprised ? Unhappy cows in the milking parlor ? Apparently cows don't like receiving an electric shock through the udder, from the milking machine :-) And apparently a tiny potential difference is enough to do it (1V level). Â*Â* Paul No, no surprises, everything worked as expected. The cows were all happy as we hand milked them, no electric milking machines like nowadays. to this day I remember squirting milk into the cats faces and ****ing them off... :-) Rene |
#24
|
|||
|
|||
SSD and computer casing.
On Sun, 15 Apr 2018 10:50:46 -0600, ken1943
wrote: On Sun, 15 Apr 2018 08:32:03 -0700, Ken Blake wrote: On Sun, 15 Apr 2018 09:25:01 -0600, ken1943 wrote: Go to Amazon.com and buy an SSD tray or mounting bracket for each SSD you have. They are only a few dollars each and will let you install the SSDs in 3.5" drive bays. I wrapped one in bubble wrap and stuck in a 3.5 bay. Do they get hot? I don't know for sure, but if they do, what you did could start a fire. They run cool. Check specs. on one, they only draw under 2-4 watts for the size I use. Probably depends on how many chips/capacity one has. I would still worry about fires. Even if they run cool, something else near them may run hot. As far as I'm concerned, putting anything flammable inside the case is a bad idea. |
#25
|
|||
|
|||
SSD and computer casing.
"Paul" wrote in message news
Bob_S wrote: "Paul" wrote in message news Bob_S wrote: "Paul" wrote in message news Peter Jason wrote: On Sat, 14 Apr 2018 22:25:50 -0400, Paul wrote: Peter Jason wrote: My computer case is of Aluminium and earthed to a domestic copper water pipe. , the casing of the Samsungs SSDs seem to be anodized aluminium. Since my computer case is old it has no provision for 2.5" SSDs I have them resting on an aluminium surface. Is this a problem? I assume the SSDs have internal insulation. I just measured a couple SSDs here, and son of a bitch, there's a hard connection between SSD casing and logic ground! That's not supposed to be good for ESD handling. I was expecting to find a 1 megohm connection between aluminum SSD casing and the logic ground on the SSD. But it's a hard short. ******* The chassis of your computer case, runs at logic ground potential. The casing of your SSD, runs at logic ground potential. This means you can safely rest the SSD on the bottom of the computer case, without any "thru" current flowing. Just don't touch a 3.3V, 5V, or 12V pin from some Molex connector, to the chassis of the SSD. Since it's a hard ground, we can't be carelessly resting a Molex power pin in contact (somehow) with the chassis. While the pins on Molex are flush or slightly recessed, there might be some conceivable way for something to touch. And my finding a hard ground on the SSD chassis, is not a good thing. I'll be extra careful in future, to keep Molex away from it. I was really expecting to find a 1Megohm (high resistance) drain between shiny aluminum SSD casing and the logic ground next to the data pins of the SSD. GND GND GND === logic ground pins are longer tx+ tx- rx+ rx- on the SSD connector. The data pins on the 7 pin data are recessed. Check with the ohmmeter, between GND and chassis, and you'll see the hard short I measured. Summary: 1) Safe to rest SSD on bottom of computer case, no matter what the casing is made of. 2) Don't touch any "hot" Molex pins, to the SSD housing. Just like you would not touch any "hot" Molex pins, to the computer case metal posts. The motherboard has a hard short to the chassis, through the standoffs. Most everything in the computer, has a hard ground. The only poorly designed feature, is "plastic fascia" on the front of the computer, around front USB ports, which is a poor way to handle ESD issues. 3) Don't scuff across the carpet and pick up your SSD by the aluminum casing. Keep unused SSDs in an ESD bag, scuff across the carpet if you must, pick up the SSD by the ESD bag, touch the ESD bag plastic on the inside of the bag, to bring yourself and the bag contents to the same potential. With your other hand, touch the computer chassis. Now, everything is "drained" to the same electrostatic potential, you can remove the SSD from the ESD bag, and connect it to SATA power and SATA data. By using the series resistance of the ESD bag, to limit ESD current flow, you're *safely* bringing the SSD up to the same electrostatic potential as the thing you wish to connect it to. Note: I just changed my handling procedure! as a result of this ohmmeter measurement. I'd assumed SSD chassis were drain connected, not "hard" connected. Now I will have to start using ESD bags for them, just as I do for hard drives. Paul Thanks Paul. I have put my SSDs in plastic sleeves, and rested this combination on cardboard. I thought this might be the cause of my computer restarts. Also I have disabled all "fast startups" via the registry. I wouldn't entirely insulate the thing. Some SSDs have high peak power usage (SandForce compression), and they get a little warm. Put a cardboard underneath if you like. Leave at least one metal surface for cooling. In this review, you can see (and it's mentioned in the text), that thermal pads were used on all the NAND chips. That means both metal surfaces are used for their slight cooling advantage. http://www.thessdreview.com/our-revi...-review-512gb/ Paul Paul, Have no idea why you would expect the SSD to be isolated from case ground. Everything in the computer and the case itself is tied to the grid ground. Mount a motherboard and the mounting screws screw into the case as well as the screw heads tighten down on ground pads on the motherboard. SSD's are also grounded when placed in a mounting tray to case ground. As for the longer pins on power and data connectors, the pins that are designated as ground pins are longer to insure a ground connection is made before any data pins mate to prevent static discharges. This pin design also accommodates "hot swap" capable devices such as hard drives and SSD's. While it's always good practice to use measures to prevent ESD when mounting a device such as a SSD, do not isolate it from the mounting mechanisms such as placing an insulator between it and a metal case. Depending on the SSD (some cases are plastic), the metal case is the heat sink as shown in the link you included in your post. Please note that the logic board is held in-place by a screw in the back that fastens to the SSD case and tightens against a ground pad on the SSD logic board. There are two ground philosophies for protecting assemblies from ESD. Neither was followed in this case. Paul Paul, Not sure I follow your logic on this. The OP said he had his case grounded to a copper pipe. What wasn't stated is how his power supply is grounded or if it was even plugged in. In the US, the 3rd lug on a plug is the green (safety) ground and in a typical home wiring setup, the white wire and green are tied to a single bus bar in the distribution panel so as to prevent ground loops / current loops. So while his setup using a ground from the case to a copper water pipe could potentially create a ground loop if there were other wiring problems, it still is ground as long as there's no interruption to earth ground in that copper pipe and in his wiring and would dissipate any ESD charge. Potential for ESD would be from him picking up the SSD and then plugging it in before he touched the case (grounding himself) and creating a arc between his hand holding the SSD and getting the data or power connectors close enough so the static potential went from him to the ground pins on the cable or the case. That’s' why the manufactures include warnings about using ground straps and/or touching the case before working on it. What people don't realize is that usually, they unplug the power cord from the wall or the power supply and that’s when their case is no longer grounded. But in the OP's situation, he had an additional ground to a water pipe. That may or may not be a good situation depending on his house wiring and I've never seen a power supply plug without a polarized plug. Now if the wall socket (old house wiring) does not have a safety ground, then his water pipe ground is a good idea. I've worked at radar sites and on some of the largest computer installations in the world and worked worked with power engineers that specialize in how "all things" need to be grounded and why. It is a complicated field of study and ESD can create havoc on computer components when you have thousands of volts jumping between you and the computer. So I'm not trying to argue with you but saying "Neither was followed in this case" didn't make sense to me. Perhaps you could expand on the "Neither" a bit. I'm not sure if you are referring to the OP or to your own situation. If it's your own, then I'll shut up and sit down cause I don't know what or how you were doing your measurements. 1) There's no such thing as "ground". The earth is an insulated ball floating in the vacuum of space. The earth is merely a "reference point". It has no special properties as such. "Standing on dirt is not magic." But since we are in frequent contact with dirt though, it's in our best interest to not touch potentials higher or lower than that potential. We could get a shock. 2) To avoid ESD, bring all assemblies to "equipotential". That's the trick. If I unplug a PC from the wall, remove all cabling, I can still *safely* install hardware in it, as long as I follow equipotential rules. This means opening the ESD bag but not removing the new hardware. Touch the inside of the bag while touching the chassis. It helps if you wear short pants and lay the metal-bodied PC flat on your lap while working (that's if you don't own an ESD strap). By bringing the PC, the new hardware item, and you to the same electrostatic potential, there is no danger of ESD damage. And notice in that picture, there is *no* connection to earth ground, safety ground, cold water pipes or anything. You make an "electrostatic island" of yourself, then ensure that all entities are "charged to the same level". I could place my work chair on glass blocks, charge myself to 50kV ESD level, and as long as I followed the rules about touching the inside of the ESD bag to bring the bag to the same potential as the other kit, I can actually install something without damage. The usage of ESD straps (where the ESD strap is fastened to the case via a 10meg series resistor), is for a similar purpose. It's not ground - it's a way to bring the computer chassis and any other objects used in the operation, to the same electrostatic potential. The series resistor inside the strap, performs the same function as the "resistive" surface inside the ESD bag. It's a dissipative path, that limits the ESD peak current flow level. By placing a hard connection between the metal chassis of the SSD and logic ground, now we have to resort to ESD bags again, as there is no dissipative path to logic ground. I can't really afford to scuff across the carpet, and discharge 50kV into the SSD chassis, because the result will now be unknown (I won't know whether I could have damaged it or not, and it could depend on the way the PCB and mount points are designed). If the connection between SSD chassis and PCB was dissipative, then I'd know there was no possibility of a high "peak" current flow during equalization. If the SSD chassis was plastic, I would be using the second philosophy, which says "don't allow static to dissipate into the item at all". Then the advanced ground contacts take the brunt of the equalization current flow. And things like this aren't left to chance in the lab. If I wanted to be sure my decision about the hard ground on the SSD wasn't a mistake, I do this kind of testing. We test for upset level. We test for damage level (which is actually pretty hard to determine, what constitutes "damage"). The bad thing about ESD, is you can "weaken" a piece of silicon, which weeks later, decides to completely fail. And then it's hard to trace back exactly when you "broke it". https://en.wikipedia.org/wiki/Human-body_model These have changed, since the last time I tried this stuff. We had a nasty gadget for doing up to 15kV tests on our products. This is so much more "refined" looking. https://transientspecialists.com/pro...contact-mode-1 There are two ways to damage things with ESD. One is via direct discharge. The other is via "induction". When usage of an improperly designed front USB solution blows out the pad I/O, it's just as likely the transient traveling down the black wire, "induces" a potential in D+ and D-, and blows out the driver pad. When you do ESD analysis, the danger isn't always from the direct path. I will be switching to ESD bags from now on, for my SSDs, and bringing myself to the same potential as the chassis. (The same kind of gymnastics you do, when you don't have an ESD strap handy.) It's not because I know for a fact the SSD will fail. It's because I don't really know, for any individual SSD, whether it has a particular vulnerability or not. If I had a pile of these in the lab, and an HBM-type tester, I could take a shot at finding out. Paul Paul - now you're being an ass who thinks he knows everything. Here's a clue for you - you don't, so stop acting like it. We come here to share knowledge - not to boost your friggin ego. -- Bob S. |
#26
|
|||
|
|||
SSD and computer casing.
Bob_S wrote:
Paul - now you're being an ass who thinks he knows everything. Here's a clue for you - you don't, so stop acting like it. We come here to share knowledge - not to boost your friggin ego. But there is a lesson here. For ESD protection, there is the concept of equipotential. If two assemblies have exactly the same potential, no damaging currents flow into semiconductors. This is how and why ESD-protected labs are designed the way they are. There are high-resistance paths everything in the lab, so that everything is brought to the same potential safelt. If I lay my SSD on the ESD mat on the benchtop, or on the ESD flooring, they're all charged to the same potential. I snap on my ESD strap, to bring myself to the same potential as the mat, before I start work on something. Ground is just a reference point. Nothing more. It's used as a return path for power systems (a conductor-of-convenience). Similar in a way, to how the metalwork in your car, forms the second wire for DC power distribution inside the car. And I gave a practical example. To work safely, wear short pants with exposed leg skin. Put the desktop PC on its side on your lap. Now, touch the inside of the ESD bag first, then pull out an item, install. It's like your own ESD-proof lab, in miniature. You can do a whole build that way. Paul |
#27
|
|||
|
|||
SSD and computer casing.
On 04/16/2018 9:39 PM, Paul wrote:
Bob_S wrote: Paul - now you're being an ass who thinks he knows everything. Here's a clue for you - you don't, so stop acting like it. We come here to share knowledge - not to boost your friggin ego. But there is a lesson here. For ESD protection, there is the concept of equipotential. If two assemblies have exactly the same potential, no damaging currents flow into semiconductors. This is how and why ESD-protected labs are designed the way they are. There are high-resistance paths everything in the lab, so that everything is brought to the same potential safelt. If I lay my SSD on the ESD mat on the benchtop, or on the ESD flooring, they're all charged to the same potential. I snap on my ESD strap, to bring myself to the same potential as the mat, before I start work on something. Ground is just a reference point. Nothing more. It's used as a return path for power systems (a conductor-of-convenience). Similar in a way, to how the metalwork in your car, forms the second wire for DC power distribution inside the car. And I gave a practical example. To work safely, wear short pants with exposed leg skin. Put the desktop PC on its side on your lap. Now, touch the inside of the ESD bag first, then pull out an item, install. It's like your own ESD-proof lab, in miniature. You can do a whole build that way. Â*Â* Paul Thanks Paul for the Great ESD tips, My earlier post was screwed up with the extreme example, I somehow lost the link I was trying to include. I'll try again. https://www.youtube.com/watch?v=x94BH9TUiHM hope that one hangs in there. Rene |
#28
|
|||
|
|||
SSD and computer casing.
Rene Lamontagne wrote:
On 04/16/2018 9:39 PM, Paul wrote: Bob_S wrote: Paul - now you're being an ass who thinks he knows everything. Here's a clue for you - you don't, so stop acting like it. We come here to share knowledge - not to boost your friggin ego. But there is a lesson here. For ESD protection, there is the concept of equipotential. If two assemblies have exactly the same potential, no damaging currents flow into semiconductors. This is how and why ESD-protected labs are designed the way they are. There are high-resistance paths everything in the lab, so that everything is brought to the same potential safelt. If I lay my SSD on the ESD mat on the benchtop, or on the ESD flooring, they're all charged to the same potential. I snap on my ESD strap, to bring myself to the same potential as the mat, before I start work on something. Ground is just a reference point. Nothing more. It's used as a return path for power systems (a conductor-of-convenience). Similar in a way, to how the metalwork in your car, forms the second wire for DC power distribution inside the car. And I gave a practical example. To work safely, wear short pants with exposed leg skin. Put the desktop PC on its side on your lap. Now, touch the inside of the ESD bag first, then pull out an item, install. It's like your own ESD-proof lab, in miniature. You can do a whole build that way. Paul Thanks Paul for the Great ESD tips, My earlier post was screwed up with the extreme example, I somehow lost the link I was trying to include. I'll try again. https://www.youtube.com/watch?v=x94BH9TUiHM hope that one hangs in there. Rene That looks like fun. I hope it pays well :-) And in that movie, "every day is a sunny day". Must be fun if there's a little weather. Sooner or later, the boss is going to make you work, when you shouldn't be working. Paul |
#29
|
|||
|
|||
SSD and computer casing.
On 04/16/2018 8:38 PM, Bob_S wrote:
"Paul"Â* wrote in message news Bob_S wrote: "Paul"Â* wrote in message news Bob_S wrote: "Paul"Â* wrote in message news Peter Jason wrote: On Sat, 14 Apr 2018 22:25:50 -0400, Paul wrote: Peter Jason wrote: My computer case is of Aluminium and earthed to a domestic copper water pipe. , the casing of the Samsungs SSDs seem to be anodized aluminium.Â* Since my computer case is old it has no provision for 2.5" SSDs I have them resting on an aluminium surface.Â*Â* Is this a problem? I assume the SSDs have internal insulation. I just measured a couple SSDs here, and son of a bitch, there's a hard connection between SSD casing and logic ground! That's not supposed to be good for ESD handling. I was expecting to find a 1 megohm connection between aluminum SSD casing and the logic ground on the SSD. But it's a hard short. ******* The chassis of your computer case, runs at logic ground potential. The casing of your SSD, runs at logic ground potential. This means you can safely rest the SSD on the bottom of the computer case, without any "thru" current flowing. Just don't touch a 3.3V, 5V, or 12V pin from some Molex connector, to the chassis of the SSD. Since it's a hard ground, we can't be carelessly resting a Molex power pin in contact (somehow) with the chassis. While the pins on Molex are flush or slightly recessed, there might be some conceivable way for something to touch. And my finding a hard ground on the SSD chassis, is not a good thing. I'll be extra careful in future, to keep Molex away from it. I was really expecting to find a 1Megohm (high resistance) drain between shiny aluminum SSD casing and the logic ground next to the data pins of the SSD. Â*Â*Â* GNDÂ*Â*Â*Â*Â*Â*Â*Â* GNDÂ*Â*Â*Â*Â*Â*Â*Â* GNDÂ*Â*Â* === logic ground pins are longer Â*Â*Â*Â*Â*Â*Â* tx+ tx-Â*Â*Â*Â* rx+ rx-Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* on the SSD connector. The data Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* pins on the 7 pin data are Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* recessed. Check with the ohmmeter, Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* between GND and chassis, and you'll Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â*Â* see the hard short I measured. Summary: 1) Safe to rest SSD on bottom of computer case, no matter Â*Â* what the casing is made of. 2) Don't touch any "hot" Molex pins, to the SSD housing. Â*Â* Just like you would not touch any "hot" Molex pins, Â*Â* to the computer case metal posts. The motherboard has a hard Â*Â* short to the chassis, through the standoffs. Most everything Â*Â* in the computer, has a hard ground. The only poorly designed Â*Â* feature, is "plastic fascia" on the front of the computer, Â*Â* around front USB ports, which is a poor way to handle ESD issues. 3) Don't scuff across the carpet and pick up your Â*Â* SSD by the aluminum casing. Keep unused SSDs in an ESD bag, Â*Â* scuff across the carpet if you must, pick up the Â*Â* SSD by the ESD bag, touch the ESD bag plastic on Â*Â* the inside of the bag, to bring yourself and the bag contents Â*Â* to the same potential. With your other hand, touch Â*Â* the computer chassis. Now, everything is "drained" to Â*Â* the same electrostatic potential, you can remove the Â*Â* SSD from the ESD bag, and connect it to SATA power Â*Â* and SATA data. Â*Â* By using the series resistance of the ESD bag, to limit Â*Â* ESD current flow, you're *safely* bringing the SSD up Â*Â* to the same electrostatic potential as the thing you Â*Â* wish to connect it to. Note: I just changed my handling procedure! as a result Â*Â*Â*Â*Â* of this ohmmeter measurement. I'd assumed SSD chassis Â*Â*Â*Â*Â* were drain connected, not "hard" connected. Now I Â*Â*Â*Â*Â* will have to start using ESD bags for them, just as Â*Â*Â*Â*Â* I do for hard drives. Â*Â* Paul Thanks Paul.Â* I have put my SSDs in plastic sleeves, and rested this combination on cardboard.Â*Â* I thought this might be the cause of my computer restarts.Â*Â* Also I have disabled all "fast startups" via the registry. I wouldn't entirely insulate the thing. Some SSDs have high peak power usage (SandForce compression), and they get a little warm. Put a cardboard underneath if you like. Leave at least one metal surface for cooling. In this review, you can see (and it's mentioned in the text), that thermal pads were used on all the NAND chips. That means both metal surfaces are used for their slight cooling advantage. http://www.thessdreview.com/our-revi...-review-512gb/ Â*Â* Paul Paul, Have no idea why you would expect the SSD to be isolated from case ground. Everything in the computer and the case itself is tied to the grid ground. Mount a motherboard and the mounting screws screw into the case as well as the screw heads tighten down on ground pads on the motherboard.Â* SSD's are also grounded when placed in a mounting tray to case ground. As for the longer pins on power and data connectors, the pins that are designated as ground pins are longer to insure a ground connection is made before any data pins mate to prevent static discharges.Â* This pin design also accommodates "hot swap" capable devices such as hard drives and SSD's. While it's always good practice to use measures to prevent ESD when mounting a device such as a SSD, do not isolate it from the mounting mechanisms such as placing an insulator between it and a metal case. Depending on the SSD (some cases are plastic), the metal case is the heat sink as shown in the link you included in your post.Â* Please note that the logic board is held in-place by a screw in the back that fastens to the SSD case and tightens against a ground pad on the SSD logic board. There are two ground philosophies for protecting assemblies from ESD. Neither was followed in this case. Â*Â* Paul Paul, Not sure I follow your logic on this. The OP said he had his case grounded to a copper pipe.Â* What wasn't stated is how his power supply is grounded or if it was even plugged in.Â* In the US, the 3rd lug on a plug is the green (safety) ground and in a typical home wiring setup, the white wire and green are tied to a single bus bar in the distribution panel so as to prevent ground loops / current loops. So while his setup using a ground from the case to a copper water pipe could potentially create a ground loop if there were other wiring problems, it still is ground as long as there's no interruption to earth ground in that copper pipe and in his wiring and would dissipate any ESD charge. Potential for ESD would be from him picking up the SSD and then plugging it in before he touched the case (grounding himself) and creating a arc between his hand holding the SSD and getting the data or power connectors close enough so the static potential went from him to the ground pins on the cable or the case. That’s' why the manufactures include warnings about using ground straps and/or touching the case before working on it.Â* What people don't realize is that usually, they unplug the power cord from the wall or the power supply and that’s when their case is no longer grounded. But in the OP's situation, he had an additional ground to a water pipe.Â* That may or may not be a good situation depending on his house wiring and I've never seen a power supply plug without a polarized plug.Â*Â* Now if the wall socket (old house wiring) does not have a safety ground, then his water pipe ground is a good idea. I've worked at radar sites and on some of the largest computer installations in the world and worked worked with power engineers that specialize in how "all things" need to be grounded and why.Â* It is a complicated field of study and ESD can create havoc on computer components when you have thousands of volts jumping between you and the computer. So I'm not trying to argue with you but saying "Neither was followed in this case" didn't make sense to me.Â* Perhaps you could expand on the "Neither" a bit.Â* I'm not sure if you are referring to the OP or to your own situation. If it's your own, then I'll shut up and sit down cause I don't know what or how you were doing your measurements. 1) There's no such thing as "ground". The earth is an insulated Â*Â* ball floating in the vacuum of space. The earth is merely Â*Â* a "reference point". It has no special properties as such. Â*Â* "Standing on dirt is not magic." But since we are in frequent Â*Â* contact with dirt though, it's in our best interest to not Â*Â* touch potentials higher or lower than that potential. We could Â*Â* get a shock. 2) To avoid ESD, bring all assemblies to "equipotential". Â*Â* That's the trick. If I unplug a PC from the wall, remove all cabling, I can still *safely* install hardware in it, as long as I follow equipotential rules. This means opening the ESD bag but not removing the new hardware. Touch the inside of the bag while touching the chassis. It helps if you wear short pants and lay the metal-bodied PC flat on your lap while working (that's if you don't own an ESD strap). By bringing the PC, the new hardware item, and you to the same electrostatic potential, there is no danger of ESD damage. And notice in that picture, there is *no* connection to earth ground, safety ground, cold water pipes or anything. You make an "electrostatic island" of yourself, then ensure that all entities are "charged to the same level". I could place my work chair on glass blocks, charge myself to 50kV ESD level, and as long as I followed the rules about touching the inside of the ESD bag to bring the bag to the same potential as the other kit, I can actually install something without damage. The usage of ESD straps (where the ESD strap is fastened to the case via a 10meg series resistor), is for a similar purpose. It's not ground - it's a way to bring the computer chassis and any other objects used in the operation, to the same electrostatic potential. The series resistor inside the strap, performs the same function as the "resistive" surface inside the ESD bag. It's a dissipative path, that limits the ESD peak current flow level. By placing a hard connection between the metal chassis of the SSD and logic ground, now we have to resort to ESD bags again, as there is no dissipative path to logic ground. I can't really afford to scuff across the carpet, and discharge 50kV into the SSD chassis, because the result will now be unknown (I won't know whether I could have damaged it or not, and it could depend on the way the PCB and mount points are designed). If the connection between SSD chassis and PCB was dissipative, then I'd know there was no possibility of a high "peak" current flow during equalization. If the SSD chassis was plastic, I would be using the second philosophy, which says "don't allow static to dissipate into the item at all". Then the advanced ground contacts take the brunt of the equalization current flow. And things like this aren't left to chance in the lab. If I wanted to be sure my decision about the hard ground on the SSD wasn't a mistake, I do this kind of testing. We test for upset level. We test for damage level (which is actually pretty hard to determine, what constitutes "damage"). The bad thing about ESD, is you can "weaken" a piece of silicon, which weeks later, decides to completely fail. And then it's hard to trace back exactly when you "broke it". https://en.wikipedia.org/wiki/Human-body_model These have changed, since the last time I tried this stuff. We had a nasty gadget for doing up to 15kV tests on our products. This is so much more "refined" looking. https://transientspecialists.com/pro...contact-mode-1 There are two ways to damage things with ESD. One is via direct discharge. The other is via "induction". When usage of an improperly designed front USB solution blows out the pad I/O, it's just as likely the transient traveling down the black wire, "induces" a potential in D+ and D-, and blows out the driver pad. When you do ESD analysis, the danger isn't always from the direct path. I will be switching to ESD bags from now on, for my SSDs, and bringing myself to the same potential as the chassis. (The same kind of gymnastics you do, when you don't have an ESD strap handy.) It's not because I know for a fact the SSD will fail. It's because I don't really know, for any individual SSD, whether it has a particular vulnerability or not. If I had a pile of these in the lab, and an HBM-type tester, I could take a shot at finding out. Â*Â* Paul Paul - now you're being an ass who thinks he knows everything. Here's a clue for you - you don't, so stop acting like it. We come here to share knowledge - not to boost your friggin ego. Bob_S, Better cool your irons before you start branding people. :-( Rene |
#30
|
|||
|
|||
SSD and computer casing.
Bob_S, Better cool your irons before you start branding people. :-( Rene The comment was for Paul and he can certainly speak for himself. This is for you. When someone insults by speaking down to others, I believe they need to be called out for it. I question how an individual can answer all questions about computer hardware and software in such infinite detail. I find that level of knowledge absolutely amazing. I think he is a great researcher but comes up shy on the fix actions he's proposed. My 40 years of experience and credentials in the telecommunications and computer science fields, makes me question those that have all the answers all of the time. And no, I don't have to cool my irons. |
Thread Tools | |
Display Modes | Rate This Thread |
|
|