ground wire question

If safety ground is only to short out fault currents, then why were HP LaserJet II printers damaged - hardware damaged on input (centronix) port - due to a problem created by the missing safety (third prong) ground? Read what Hobdbcgv has posted rather than snipping what you don't understand. A missing safety ground can cause a computer to not perform properly. And again, that problem would be unique to that computer's design.

Reply to
w_tom
Loading thread data ...

This wasn't a grounding issue, it was a bonding issue. We had the same problem at IBM with lots of different printers if they had long cables and were located in another room. (particularly in State Farm offices and Kinkos). The grounds checked OK with a variety of testers (Ecos and similar) The final fix from Lexmark was to run a parallel bonding wire between the printer frame and the PC frame. This can happen anytime there is a ground shift, even for microseconds, between the EGC on the circuit the printer is plugged in and the one the PC is plugged in. In your case the printer had a constant ground shift but the problem would have been fixed, simply by bonding the printer to the PC. The real fault is in the sensitivity of "parallel" port hardware. When you read the hardware hacker info sites they constantly remind people how easy it is to blow a parallel port.

Reply to
Greg

What TH does that mean? The statement makes no sense at all - it's like saying "it's not a ground, it's a green wire connected between the neutral bus and an earthing rod".

A bond is a wire is a lead - all ground paths.

(Particularly in printers with the charge buildup from moving paper and moving rubber, when the anti-static brushes are undersized and the ground path is inadequate, and the charge naturally seeks another path.)

Reply to
--

From article 100

Ground. A conducting connection, whether intentional or accidental, between an electrical circuit or equipment and the earth or to some conducting body that serves in place of the earth.

Bonding (Bonded). The permanent joining of metallic parts to form an electrically conductive path that ensures electrical continuity and the capacity to conduct safely any current likely to be imposed.

You can have "bonding" without "grounding".

Earth is always going to be something other than zero volts. Bonding resolves that ground shift in some other way than through connected people or equipment.

When we were setting up LANs it was common to see 20-30 volt spikes between building grounding electrodes in buildings that were less than 100' apart.

Another place where you will see the difference between grounding and bonding is in a swimming pool. The pool is obviously "grounded" as is all the pool equipment via the Equipment Grounding Conductor and grounding electrode. but we still "bond" everything with 8 guage copper to be sure those grounds are the same potential.

I have PCs running on inverters in 2 of my cars as MP3 players. Obviously nothing is grounded but if I don't "bond" the case of the PC to the car chassis I get an ungodly AC hum. That probaby comes from the line filter in the PC supply that establishes ~60v on the PC case. As long as I am bonded that is not a problem.

Reply to
Greg

I think one of you is quoting the NEC and the other is using electrical engineering terms- one goes to physicality, the other to function.

The NEC deals with the (non-OSHA lower voltage end-user power distribution) physical parts - in it, a ground is not a concept that performs a function, it is rather a physical part or group of parts.

Engineers deal with the full spectrum of electrical activity and thus are normally trained to not have a part be considered a function - thus in an experienced engineers mind not speaking in the strict context of the NEC, a metal piece called a bond is a ground path, i.e., a ground.

IMHO, an engineer should/sees a bond as a path used to move non-conductor charges, i.e., current. And If it does not perform the function of a conductor and it carries charges from place to place, it is by definition a ground. I also believe the NEC uses the same two-part distinction - conductor, and non-conductor that is the other stuff that all connects charge flow to earth.

Reply to
Hobdbcgv

The idea of electrically connecting a "common" AKA ground plane is universal. I came from the computer industry and have been in the electrical biz for the last decade. The idea of an equipotential plane is the same, whether you are talking about the DC ground in a computer system or the bonding of a swimming pool. In respect to this printer/PC interface the analogy is still the same. You don't need to ground the two machines to the earth but you certainly need to assure DC ground is the same potential in both machines. Electricians call that "bonding". The combination we used was a straight fat wire between machine frames that drained off the ground shift and we looped ferrites on the signal lines to make them "slower"/less attractive paths for common mode spikes. I suppose we can theorize all day about what the electrons were doing but I know it worked.

Reply to
Greg

An old issue of Datamation Magazine told a classic story. New mainframe was installed but kept failing. At night, the tech would run diagnostics and suffer no failures. But everytime the tech left for coffee, then computer would crash. It got so bad that the tech would loudly announce he was leaving for coffee, stomp out the door, then sneak back to peek in and watch the computer. Still that computer was so smart that it would not crash, until tech actually left for coffee.

Turns out the computer was grounded (bonded) to a same point as the elevator. Only when the tech actually left on the elevator, would the computer crash. Again, if and how a computer is bonded (grounded) can affect reliability. Computers are not haunted. They fail only for good, technical reasons. Some are designed assuming the safety ground exists.

We are probably talking about different printers. Centronix cables were limited to about 15 feet. Printer had to be adjacent to computer. Centronix interfaces to HP LaserJet II printers were damaged by missing safety ground. IOW computer and printer were therefore not bonded to a common point which is what a missing safety ground created. Normally printer did not have a bonding wire to computer because manufacture

*assumed* the safety ground would be provided. Aga>> If safety ground is only to short out fault currents, then
Reply to
w_tom

I have resolved multiple instances where data equipment failure, and data throughput issues were a result of devices with differing ground or chassis potential in relation to earth ground. In most cases they were separated far enough to be served by different branch circuits. In these cases significant current would flow on the data reference ground where each chassis was at a different potential in relation to each other (often just momentarily). In almost every case there were cheap shunt mode surge suppressors involved that were polluting the grounding conductor on one end of the interconnected devices.

I have observed instances where the integrity of data equipment had suffered due to the lack of, or a poor quality grounding conductor. As related above having a quality grounding path present at a piece of data equipment isn't a guarantee of trouble free operation.

Louis

--

********************************************* Remove the two fish in address to respond

Reply to
Louis Bybee

I have resolved multiple instances where data equipment failure, and data throughput issues were a result of devices with differing ground or chassis potential in relation to earth ground. In most cases they were separated far enough to be served by different branch circuits. In these cases significant current would flow on the data reference ground where each chassis was at a different potential in relation to each other (often just momentarily). In almost every case there were cheap shunt mode surge suppressors involved that were polluting the grounding conductor on one end of the interconnected devices.

I have observed instances where the integrity of data equipment had suffered due to the lack of, or a poor quality grounding conductor. As related above having a quality grounding path present at a piece of data equipment isn't a guarantee of trouble free operation.

Louis

--

********************************************* Remove the two fish in address to respond

Reply to
Louis Bybee
040612 2147 - Louis Bybee posted:

There is an old Audio adage: NO GROUND LOOPS

A ground loop has the potential to introduce hum into an audio system. And, as a correlative, the potential to data corruption in a data transfer system. A ground must be fastened at the beginning of the data system, and be continuous throughout, and not connected at any point to another ground throughout. This will reduce the potential for ground loop, and corruption of data.

Reply to
indago

Kinkos had 50' printer cables on a standard Centronix parallel port and it did work. Everyone tried to point out that it wasn't "supported".

Reply to
Greg

That is far from an adage in computers, particularly when you are talking about PC printers. Get a meter and try it. Loog at the DC ground to "3d prong" ground on any old printer with an internal power supply (not a wall wart). Most have the case tied to DC ground. So does the system unit.

In the mainframes the first thing product engineering did when we had ESD problems was start hanging more grounds on the machine. On some old 370s there were 25 or 30 wires you had to disconnect before you did the "baseplate grounding check". They finally removed that procedure since there was no "single point grounding" in the first place.

Reply to
Greg
50' was not supported. But I used twisted cable techniques and shielding to make it work. Then I discovered the handshaking was not controlled by DOS. Some programs would work. Others would not. Eventually had to go with a printer network. Don't something outside of specs does not mean it will fail every time. Just may fail sometime.

Any rate, that pr>> We are probably talking about different printers. Centronix

Reply to
w_tom

"w_tom" Hi, sorry I haven't written back to the group sooner. I was trying to see if I could figure out what was running at the time of the shutdowns. No matter what I have tried the PC has not shut down again. I even left it on for 48 hours with no problems. Strange. I appreciate all the information from you and all that replied. I didn't mean to cause any conflicts over all of this. If I have anymore problems, I will let you all know. Thanks again, Donna J.

Reply to
DJ

Did you review the event logs provided by your operating system as noted in a 6 Jun post? History that the system stores on all detected failures. Did you put a large load (ie electric heater) on that circuit while an incandescent bulb was on? Did the bulb dim and the previously noted 3.5 digit meter show a major voltage drop - as posted 7 Jun? Is it dimming or flickering as was asked on 6 Jun?

These are not just to fix your computer. Do you intend to be alive next year? Intermittent computer failure may have only been a 'canary in the coalmine'. Important experiments that you were expected to perform because experiments were about something more than just a crashing computer. Experiments also eliminated human threats as a reason for computer crash.

Your failures are not strange to an engineering mentality that constantly deals with realities. Many failures start this way before things like, worst case, explosions happen. Those intermittent crashes are consistent even for problems, potentially dangerous, and existing elsewhere. Noted was that you problem could be the precursor to something resulting, worst case, in a building fire. Provided were simple experiments to isolate the problem, at minimum, to something insignificant.

In the meantime as asked in your original post, yes, missing safety ground can cause strange events, including crashes, on some properly built machines and not on others.

Please go back, read those posts, and report back. Do not selectively perform > "w_tom"

Reply to
w_tom

How do you know it happened while off and unplugged?

This sounds more like the amp whent into oscillation (maybe due to speaker wire coils) and burned out the tweeters and itself. Remove SPAMX from email address

Reply to
Jim Michaels

PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.