kthompso@uceng.UC.EDU (ken thompson) (08/11/89)
A friend of mine returned his AT clone back to the store for warrenty repair. The owner told him that he would have to buy a new mother board. He said the mother board had smoked because two boards were set on IRQ 4. He said it also destroyed the hard drive. He now refuses to fix or return the PC to him, says he may need it for evidence. I have installed boards with conflicting interrupts before and other than software problems, there was no damage. The interrupt chip is just a TTL input and shouldn't care how many highs are put on it's input. I think my friend is getting the screws. Comments please! -- Ken Thompson University of Cincinnati Dept. of Chemical Eng. kthompso@uceng.uc.edu
karl@ddsw1.MCS.COM (Karl Denninger) (08/17/89)
>Item 5029 (0 resps) by kthompso at uceng.UC.EDU on Wed 16 Aug 89 09:29 >[ken thompson] Subject: Interupt conflicts > >A friend of mine returned his AT clone back to the store for warrenty repair. >The owner told him that he would have to buy a new mother board. He said the >mother board had smoked because two boards were set on IRQ 4. He said it also >destroyed the hard drive. He now refuses to fix or return the PC to him, >says he may need it for evidence. > >I have installed boards with conflicting interrupts before and other than >software problems, there was no damage. The interrupt chip is just a >TTL input and shouldn't care how many highs are put on it's input. I think >my friend is getting the screws. > >Comments please! Comments? Here's some.... First, you CAN damage things by conflicting IRQ settings..... but -- You're most likely to damage the peripheral boards themselves. You see, if one board is driving the pin high, and the other driving it low, well, you get to see how good a fuse some TTL level driver makes... on the peripheral card itself. In this case, you might fry a modem or serial board -- both of which are on IRQ 4. Since the motherboard doesn't drive _anything_ in this situation, it would be slightly difficult to damage it with this type of mistake. As for the disk drive, that would be even harder, as it has no direct connection to the main board at all; it's driven through a number of buffers and other circuitry (the controller card of course). I think your friend is getting the snow job. I'd tell Mr. Clone Shop Owner that you want the machine back, fixed, right now, or you are going to sue. It sounds to me like someone is trying to make a lot of money for doing nothing at all -- this clone shop person could simply fix the jumper settings, charge your friend for a motherboard, and send him on his way -- lots poorer. If you're willing to fly me to wherever the trial is, I'd even be willing to testify in this one. I run the shop here, and have lots and lots of experience -- I bet I qualify as an expert witness :-) == Karl Denninger (karl@ddsw1.MCS.COM, <well-connected>!ddsw1!karl) Public Access Data Line: [+1 312 566-8911], Voice: [+1 312 566-8910] Macro Computer Solutions, Inc. "Quality Solutions at a Fair Price"
gordon@eecea.eece.ksu.edu (Dwight Gordon) (08/17/89)
In article <[24ea313d:5029.1]comp.ibmpc;1@ddsw1.MCS.COM> karl@ddsw1.MCS.COM (Karl Denninger) writes: >>Item 5029 (0 resps) by kthompso at uceng.UC.EDU on Wed 16 Aug 89 09:29 >>[ken thompson] Subject: Interupt conflicts >> >>A friend of mine returned his AT clone back to the store for warrenty repair. >>The owner told him that he would have to buy a new mother board. He said the >>mother board had smoked because two boards were set on IRQ 4. He said it also > >>destroyed the hard drive. He now refuses to fix or return the PC to him, >>says he may need it for evidence. I'm not a lawyer, but it sounds like theft to me. Your state Attourney General probably has some type of consumer protection division. (Kansas even has a toll-free number for us!) They might warrant a phone call. The Better Business Bureau sure does. When he returned the equipment to the store, he probably (better have) got a receipt for the equipment. This receipt has suddenly become important! >> >>I have installed boards with conflicting interrupts before and other than >>software problems, there was no damage. The interrupt chip is just a >>TTL input and shouldn't care how many highs are put on it's input. I think >>my friend is getting the screws. >> >>Comments please! > >Comments? Here's some.... > >First, you CAN damage things by conflicting IRQ settings..... but -- Agreed (as, you say - Highly unlikely for well-designed boards). >You're most likely to damage the peripheral boards themselves. You see, if >one board is driving the pin high, and the other driving it low, well, you >get to see how good a fuse some TTL level driver makes... on the peripheral >card itself. In this case, you might fry a modem or serial board -- both of >which are on IRQ 4. Wait a second, true TTL doesn't source current (at a logic high), it sinks current at a logic low. However, few systems with the C&T chipset are really using TTL inputs. So this point is somewhat mute. More likely the problem is related to what other things are messed-up. INT 4 is used for COM1. If the real problem was two devices connected to COM1 (INT4 and 03F8h Port address), then any device accessing COM1's ports will access more than one device simultaneously. Each of these devices, since they are on a 3-state bus, are not TTL, but 3-state. 3-state devices don't necessarily follow the no-source design rule of TTL. This could blow out the data bus drives. This, in turn, could blow out any other data bus connection (e.g. hard drive controller) which, in turn, could blow out . . . >Since the motherboard doesn't drive _anything_ in this situation, it would >be slightly difficult to damage it with this type of mistake. As for the >disk drive, that would be even harder, as it has no direct connection to the >main board at all; it's driven through a number of buffers and other >circuitry (the controller card of course). Agreed, with the above restrictions. >I think your friend is getting the snow job. I'd tell Mr. Clone Shop Owner >that you want the machine back, fixed, right now, or you are going to sue. >It sounds to me like someone is trying to make a lot of money for doing >nothing at all -- this clone shop person could simply fix the jumper settings, >charge your friend for a motherboard, and send him on his way -- lots poorer. AGREED! Call Clone Shop Owner, and, if no satisfaction is received, call your attourney general's office. (Note - Most times, I've only had to threaten to call the attourney general's office in order to get the ear of a "bad" dealer.) If you do have to go through the attourney general's office, record-keeping can either make or break your case. Make sure to log all calls/contacts with the shop owner. Make copies of all correspondence/receipts/etc. and keep them in a file. I have consulted for almost three years at a company that repairs IBM desktop computers. I've seen a whole lot. The design of the board is not terribly robust. Some times a simple mistake will cause an avalanche of failures. It is possible (although unlikely) that the board is dead from your INT4 mistake. I'm no lawyer, but in my opinion this does not provide an excuse for Clone Shop's owner to TAKE the board(s). At best, he may refuse to provide warranty service, since the unit was tampered-with. However, IBM's open bus architecture was designed for just this such addition of user-purchased boards. Dwight W. Gordon, Ph.D. | 913-532-5600 | gordon@eecea.eece.ksu.edu Electrical & Computer Engineering Department | dwgordon@ksuvm.bitnet Kansas State University - Durland Hall | rutgers!ksuvax1!eecea!gordon Manhattan, KS 66506 | {pyramid,ucsd}!ncr-sd!ncrwic!ksuvax1!eecea!gordon
timothym@tekigm2.MEN.TEK.COM (Timothy D Margeson) (08/17/89)
Hi, Having read some of these discussions of how setting two add-in cards to the same interrupt will cause damage..... Have any of you ever seen electronics??? A TTL gate is designed to take short circuits to ground indefinately (so long as the dissipation spec of the chip is not exceeded). Also, for your info, most PC's with four com ports share interrupts 3 & 4, and somehow these never blow up!!! Summary? Place two add-ins in the same interrupt, what do you get? Some software that won't work! Any other responses about blowing the chips are wrong, and I don't care if on your mothers PC you know of a case where they did 'cause what probably really happened is some non-static knowledgable person got into the PC with pliers or hands and started pulling this part or that part trying to fix the problem he cause to the card he had just installed. Ask yourself this: The last time your worked on your computer, did you wear an anti-static wrist strap connected to your PC and anti-static desk top? If you answered no, then you HAVE damaged your PC, perhaps not in a way that is immediately noticable, but the damage is real non the less (there are ways to minimze this damage using good handling techniques but one minor slip will take out many IC's). -- Tim Margeson (206)253-5240 PO Box 3500 d/s C1-022 @@ 'Who said that?' Vancouver, WA. 98668 e-mail replies to: timothym@tekigm2.MEN.TEK.COM
wek@point.UUCP (Bill Kuykendall) (08/18/89)
>The owner told him that he would have to buy a new mother board. He said the >mother board had smoked because two boards were set on IRQ 4. He said it also >destroyed the hard drive. He now refuses to fix or return the PC to him, I've never seen permanent damage causes by duplicated interupts. I *have* seen this kind of problem caused by misaligning a power connector or installing chips backward. Perhaps the vendor has a legitimate interest in keeping the evidence out of your friend's hands...
martin@tabbs.UUCP (Martin Walker) (08/18/89)
From article <[1580.2]comp.ibmpc;1@point.UUCP>, by wek@point.UUCP (Bill Kuykendall): >>The owner told him that he would have to buy a new mother board. He said the > >>mother board had smoked because two boards were set on IRQ 4. He said it also > >>destroyed the hard drive. He now refuses to fix or return the PC to him, > The standard design philosophy used in designing interrupt control circuitry is to make all the interrupt line drivers open-collector. This means that the peripheral device can pull an interrupt line low, but cannot drive it high. Thus one can connect as many interrupt lines together as desired. When one of the peripherals wants attention, it merely pulls the interrupt line low. This does not in any way damage any of the other peripherals. This is usually referred to as a "wired-OR" configuration. The device which services the interrupt would then poll the various devices to determine which of them generated the interrupt and act accordingly. On most PC's, however, each peripheral has its own interrupt line, obviating the need for software polling, but the drivers ARE usually still open-collector, for precisely the reasons given above. It prevents damage when people configure their PC's incorrectly! I cannot say for sure if this sensible design philosophy was followed in the PC in question, but I would be rather surprised if open-collector logic hadn't been used. It is after all less complex than fully driven logic.... -- Martin Walker | ..!<smart-host>!ddsw1!olsa99!tabbs!martin @ | "The more people I meet, the more Totally Awesome | I like my dog..." BBS | (:-D D-:)
andyw@vware.MN.ORG (Andy Warner) (08/19/89)
In article <5832@tekigm2.MEN.TEK.COM>, timothym@tekigm2.MEN.TEK.COM (Timothy D Margeson) writes: > [stuff deleted] > Have any of you ever seen electronics??? A TTL gate is designed to take > short circuits to ground indefinately (so long as the dissipation spec of the > chip is not exceeded). I agree, sort of, I thought that most interrupt lines were open drain or open collector, with the line pulled high. So if two boards pull the line low - who cares ? > [stuff deleted] > Summary? Place two add-ins in the same interrupt, what do you get? > Some software that won't work! Correct. > [stuff about static damage deleted] Yes, you may well have blown the card up with static, but if your dealer is as ignorant as he sounds, HE is just as likely to have done it .. Anyway, just for the records, why don't you tell us who he is so we can all avoid him :-) :-) -- andyw Andy Warner US email : andyw@vware.MN.ORG ..or.. VisionWare ..!{amdahl,hpda,uunet!bulus3}!bungia!vware!andyw P.O. Box 3991 Minneapolis MN 55405 USA Voice : (612) 377-3627
karl@ddsw1.MCS.COM (Karl Denninger) (08/19/89)
>Response 4 of 4 (5029) by timothym at tekigm2.MEN.TEK.COM on Fri 18 Aug 89 >14:27 >[Timothy D Margeson] > >Hi, > >Having read some of these discussions of how setting two add-in cards to the >same interrupt will cause damage..... Not will, can. But it's very unlikely (as in I've never actually seen it happen, but have looked at some boards where it is _possible_ to have it occur). >Have any of you ever seen electronics??? A TTL gate is designed to take >short circuits to ground indefinately (so long as the dissipation spec of the >chip is not exceeded). So long as the dissipation spec is not exceeded..... :-) Now, for a few minutes or so, ok. For a number of _hours_, at 100% duty cycle, and with little or no cooling to speak of (check the so-called 'fans' in some of those cheap PCs as well as chip placement) you could have a slight problem. It all depends on the specifics of the situation. Again, I've never seen a board damaged in this fashion. Note that many of the cheap boards are designed with little or no safety margin beyond normal operating parameters. These are the ones which stop working when you do things like this. >Also, for your info, most PC's with four com ports share interrupts 3 & 4, >and somehow these never blow up!!! Right, because most com boards aren't designed _too_ stupidly. But there are some boards, especially the $50 specials you see in a lot of the foreign clones, which aren't carefully put together..... and who said anything about both of the boards which were 'fighting' being COM boards anyways? I've seen more than one card which didn't have buffers on the I/O and/or IRQ lines....saves a few pennies, 'ya know. >Any other responses about blowing the chips are wrong, and I don't care if >on your mothers PC you know of a case where they did 'cause what probably >really happened is some non-static knowledgable person got into the PC with >pliers or hands and started pulling this part or that part trying to fix the >problem he cause to the card he had just installed. Ask yourself this: > > The last time your worked on your computer, did you wear an anti-static > wrist strap connected to your PC and anti-static desk top? > > If you answered no, then you HAVE damaged your PC, perhaps not in a way > that is immediately noticable, but the damage is real non the less (there > are ways to minimze this damage using good handling techniques but one > minor slip will take out many IC's). Well, actually there are several ways to prevent that problem. If you are grounded to the PC's circuit ground physically, and remain so at all times, you cannot damage the componets. It's potential _difference_ across the board/chips/whatever that does the damage -- if you are at the same potential as the chassis, regardless of what that is referenced to "ground", you can't hurt anything (assuming, again, that the chassis is tied to logic ground -- many aren't for safety purposes. In that case you need a jumper or two, and the power darn well better be OFF AND DISCONNECTED). Besides, these days _most_ chips are quite static tolerant. Nothing like the old 4000-series CMOS stuff that would blow up if you looked at it wrong (yes, I worked with that stuff too. Gave ICs a bad name it did....) We generally do use a stat-strap, but not always. I have _never_ damaged a customer or my own system(s) with static discharge, but then again, I am meticulously careful and paranoid (again, those memories of 4000-series CMOS chips do it to 'ya :-) Oh yeah -- do remove your jewelry/watch/whatever before you go poking around in the system, especially if the power is on. Sparks flying don't tend to help things much, and 20+ amp +5V rails can be nasty, and are found in most AT-class systems! Finally, remember that every electronic component is made with an amount of smoke inside, and that is an _essential_ ingredient. Once you let out the smoke, it won't work anymore. Smells bad too. :-) -- Karl Denninger (karl@ddsw1.MCS.COM, <well-connected>!ddsw1!karl) Public Access Data Line: [+1 312 566-8911], Voice: [+1 312 566-8910] Macro Computer Solutions, Inc. "Quality Solutions at a Fair Price"
phil@diablo.amd.com (Phil Ngai) (08/23/89)
In article <176@tabbs.UUCP> martin@tabbs.UUCP (Martin Walker) writes: |The standard design philosophy used in designing interrupt control circuitry |is to make all the interrupt line drivers open-collector. This means that Not in PCs it isn't. |On most PC's, however, each peripheral has |its own interrupt line, obviating the need for software polling, |but the drivers ARE usually still open-collector No, you're wrong. |I cannot say for sure if this sensible design philosophy was |followed in the PC in question, but I would be rather |surprised if open-collector logic hadn't been used. It is |after all less complex than fully driven logic.... If you're so unsure of yourself, how can you make the strong statements I just quoted? In fact, you and many others posting on this subject don't know what you're talking about. Do you just like to hear yourself talk, whether or not you're right? For your information, not that you seem to care, interrupts on the PC bus are active HIGH. -- Phil Ngai, phil@diablo.amd.com {uunet,decwrl,ucbvax}!amdcad!phil "Nonviolence works! India hasn't had to use their nuclear weapons yet."
sv@v7fs1.UUCP (Steve Verity) (08/24/89)
>>is to make all the interrupt line drivers open-collector. This means that >Not in PCs it isn't. >>but the drivers ARE usually still open-collector >No, you're wrong. [flame deleted] >For your information, not that you seem to care, interrupts on the PC >bus are active HIGH. ...So, Do you think that an I/O channel board is allowed to grab the IRQ line, and yank it low, until it needs to generate an interupt, at which point, it drives the line high??? Yes, the interupts on the AT are active high. When a board wants an interrupt serviced, it must first pull the line low *WITH AN OPEN COLLECTOR DEVICE* and then let it *FLOAT* high. (these lines are pulled up on the I/O chanel.) >-- >Phil Ngai, phil@diablo.amd.com {uunet,decwrl,ucbvax}!amdcad!phil >"Nonviolence works! India hasn't had to use their nuclear weapons yet." Steve Verity ...ames!vsi1!v7fs1!sv
phil@diablo.amd.com (Phil Ngai) (08/25/89)
In article <487@v7fs1.UUCP> sv@v7fs1.UUCP (Steve Verity) writes: | ...So, Do you think that an I/O channel board is allowed to |grab the IRQ line, and yank it low, until it needs to generate an |interupt, at which point, it drives the line high??? Looking at the schematics on the IBM serial card, I see that it drives the INT line with a 74125 tri-state driver. The 74125's output enable is controlled by a register set by software. The 74125's input is connected to the serial chip's interrupt line. When an interrupt happens, the 74125 drives the INT line high. So, yes, that does seem to be the way it happens. | Yes, the interupts on the AT are active high. When a board |wants an interrupt serviced, it must first pull the line low *WITH AN |OPEN COLLECTOR DEVICE* and then let it *FLOAT* high. (these lines are |pulled up on the I/O chanel.) Wrong, bozo. The 74125 is not an open collector device. Nor does the INT line float high when the interrupt arrives. It is driven high by the 74125. And in the case that started all this, it is quite possible for two cards which are mistakenly configured for the same interrupt to end up with two INT line drivers fighting each other. You seem to think you know a lot, unfortunately, you don't. -- Phil Ngai, phil@diablo.amd.com {uunet,decwrl,ucbvax}!amdcad!phil "Today surgeons are highly respected but they were once just grave robbers."