PDA

View Full Version : Help. I shorted out my 5155's power supply.



ibmapc
July 6th, 2010, 08:57 PM
I attempted to install a turbo 8088 motherboard in my 5155. The board powered up fine when hooked up externally, but I must have shorted something to ground when I mounted it in the system because the power supply fan did not start. So then I disconnected all the power leads (motherboard, disk drives, even the monitor) and flipped the switch on. Still nothing. So I then pulled out my multi-meter and found zero volts on all leads. I took the cover off and found the 3 amp fuse near the switch and removed it to test it (hoping for an easy fix) but found the fuse tested good. I can't see anything that looks burn't inside. When turned on it makes a single click that sounds like a small relay but no fan and no output. Can someone help? I'd be willing to ship it both ways if someone would be willing to look at it. Any sugestions?

Chuck(G)
July 6th, 2010, 09:04 PM
Maybe you already know this, but most SMPSUs won't work with no load at all. Use a hard disk drive as a dummy load (IDE is fine) and see if you get some action that way.

ibmapc
July 6th, 2010, 09:14 PM
Maybe you already know this, but most SMPSUs won't work with no load at all. Use a hard disk drive as a dummy load (IDE is fine) and see if you get some action that way.

OK, I tried that with two different IDE hard disks. It would not spin either one, but the one with the led blinked momentarily(less than one second).

Chuck(G)
July 6th, 2010, 09:43 PM
Sounds like you've got some digging to do in your PSU. It might not have been a motherboard short after all.

I'd start by checking capacitors for shorts.

ibmapc
July 6th, 2010, 09:58 PM
Sounds like you've got some digging to do in your PSU. It might not have been a motherboard short after all.

I'd start by checking capacitors for shorts.

I'm a little chicken around those big capacitors, especialy the ones with the "SHOCK HAZARD WARNING". Chuck, I see you're in the "Pacific Northwest". I'm in Albany, OR. How far away are you? Would you be willing to look it over for me?

modem7
July 7th, 2010, 02:39 AM
The 5155 PSU will start up with only it's motherboard (which is a 5160 motherboard) connected.
I know from measurements done a few years ago that a 5160 motherboard draws about 15 Watts.

I tried only an IDE drive connected to my 5155 PSU - no start (fan turns momentarily).
I tried two IDE drives - no start (fan turns momentarily).
I tried three IDE drives - no start (fan turns momentarily, and for longer than with just two drives connected).
I tried four IDE drives - PSU runs.

So your PSU may be good. Try connecting the original motherboard (or four IDE drives) and see how that goes.

MikeS
July 7th, 2010, 07:49 AM
Maybe even remove the MB from the case and put everything back the way it was when it worked, in case there's a short at one of the mounting screws. And of course make sure that the connectors have the black wires in the middle, i.e. next to each other. It's pretty hard to catastrophically "short out" a PC PS.

ibmapc
July 7th, 2010, 03:30 PM
You guy's are great! When I connected the original MB and flipped the switch, the fan came on and appears to run normal, so I turned it back off and connected the HDD with the LED and flipped the switch again and the fan came on, the LED lit and the drive spun up normally. So it would seam the PSU is fine. Just my trouble shooting methods were faulty. Glad I didn't buy that refurbished power supply from CODEMICRO!! Thanks everyone!!

Greg

Chuck(G)
July 7th, 2010, 04:44 PM
The 5155 PSU will start up with only it's motherboard (which is a 5160 motherboard) connected.
I know from measurements done a few years ago that a 5160 motherboard draws about 15 Watts.

I tried only an IDE drive connected to my 5155 PSU - no start (fan turns momentarily).
I tried two IDE drives - no start (fan turns momentarily).
I tried three IDE drives - no start (fan turns momentarily, and for longer than with just two drives connected).
I tried four IDE drives - PSU runs.

So your PSU may be good. Try connecting the original motherboard (or four IDE drives) and see how that goes.

I'll put it in my file that a big 5.25" hard drive may be required. :) (Many consume upwards of 30W)

modem7
July 7th, 2010, 05:55 PM
I'll put it in my file that a big 5.25" hard drive may be required. :) (Many consume upwards of 30W)
I didn't think one IDE drive would be enough, but it was certainly a surprise to me that four were required.
I've just done the same experiment on a 5160 PSU, and it required three IDE drives.
I guess that reflects:
1. These old switch mode PSUs require more load (minimum load) than modern switch mode PSUs, and
2. IDE drives draw less power than I thought.

Raven
July 7th, 2010, 09:09 PM
I got a 5170 recently with lots of documentation including that from an installed HDD upgrade. It mentions something called a "load resistor" which was placed in the casing of a 5170 without a HDD to create that minimum load, otherwise the system wouldn't function right. This resistor had to be removed when installing a HDD, which is why it was mentioned in the documentation.

I suspect that the same applies for older systems, load resistor or no, and that seems to be what you guys have come up with here.

Just sharing my two cents - a shame I didn't come across this thread sooner, as I might have saved some hassle, heh.

modem7
July 7th, 2010, 11:56 PM
I got a 5170 recently with lots of documentation including that from an installed HDD upgrade. It mentions something called a "load resistor" which was placed in the casing of a 5170 without a HDD to create that minimum load, otherwise the system wouldn't function right. This resistor had to be removed when installing a HDD, which is why it was mentioned in the documentation.
The load resistor you write of is mentioned (and there's a link to a picture) at the 5170 entry at our Wiki (VCWiki): http://wiki.vintage-computer.com/index.php/IBM_PC_AT_%285170%29

I have a 5170 model 068 (no hard drive) in my house presently. I'm unsure why the load resistor is required, because the PSU fires up okay when I disconnect it.


I suspect that the same applies for older systems, load resistor or no, and that seems to be what you guys have come up with here.
Just sharing my two cents - a shame I didn't come across this thread sooner, as I might have saved some hassle, heh.
The issue wasn't about whether a load was present or not - it was about the amount of load required.

QuantumII
July 8th, 2010, 12:34 AM
That dummy load was a weird animal. Why the large metal cage? Maybe it acts as a heat sink.

modem7
July 8th, 2010, 02:43 AM
That dummy load was a weird animal. Why the large metal cage? Maybe it acts as a heat sink.
The resistor is positioned in the cage middle, sitting on two thin metal legs (each with a cross section of about 1mm x 5mm). And so heat transfer from resistor to cage does not appear to be a major consideration.

The top plate looks to me to be a wonderful piece of over-engineering.

Raven
July 8th, 2010, 03:35 AM
The issue wasn't about whether a load was present or not - it was about the amount of load required.

I know - I figured that a load resistor provided X load. I realize your point, though.

wrljet
July 8th, 2010, 04:19 AM
Pic of Compaq Deskpro 386 load resistor...

3874

3875

QuantumII
July 8th, 2010, 04:32 AM
Hehe, that was a cool looking dummy HDD. It's of the kind that can fool a seller into thinking "This PC has a HDD but it does not work". It could also be used as a decoration on a newer PC to give it a bit of a vintage look :-)

Raven
July 8th, 2010, 02:16 PM
I have several 3.5" to 5.25" hdd adapters that are like that but designed to hold a HDD - like using them in modern boxes.. I even found a beige one by some streak of luck that matches my latest case's beigeness. They even have little light holes, a few of them even having a proper LED and plug for the rare IDE disk that had a header for a HDD light right on it, or you could run it to the motherboard. Most of them just had a clear colored plastic piece for a light mounted on the odd IDE disk to shine through, though.

One of them tricked me though, the first time I saw it.. It was a beige desktop case, and it looked like it had a 5.25" MFM disk. I expected some sort of cool XT or AT clone, but when I opened it it was a 486 and the HDD turned out to be an EMPTY adapter bracket, heh. The case was awesome on that machine, but unfortuantely had many broken clips making it fall apart.. I kept it so I can look out for another.

ibmapc
July 9th, 2010, 09:25 PM
OK, I got it working with the turbo 8088 MB. Thing is when I ran FALCON AT, I heard the what sounded like an arc and saw smoke so I powered it off immediatly and started taking things apart to see if the source of arcing could be found. I could not find any thing obvious, but the big transromer in the power supply looks funny.(see picture). Am I overloading the PSU with this 10mhz Turbo 8088 MB? Does that brown stuff on the transfomer look normal or has it been overheated? Am I risking PSU failure if I continue to use this MB?

Chuck(G)
July 9th, 2010, 11:00 PM
Nothing obvious that I can see. I doubt that a 10MHz 8088 board would make a big difference (are you running an 80C88 or V20 to keep the power down?). But aging components can and do make smoke. Power it up again and watch carefully. Often the component giving trouble will give itself away by the smell.

modem7
July 9th, 2010, 11:05 PM
The brown stuff is lacquer (what the wire in the transformer/filter gets coated with). It's not unusual to see lacquer there. It's quite a bit, but I wouldn't consider that to be an definite indicator of overheating.

The 5155 PSU should be able to handle the 10mhz Turbo 8088 motherboard.

The symptoms you describe could be caused by a ruptured line suppression capacitor. See the 'Line Suppression Capacitor' section of http://members.dodo.com.au/~slappanel555/failure.htm
In your photo, the two blue things are line suppression capacitors. They look okay in the photo, but take a close look at them. In my 5155 PSU, there's also one wired directly to the contacts on the power switch.

Of course, there are other components that can arc/smoke.

One way of detecting the location of arcing is to operate the equipment in a dark room, being extra careful due to the greater risk of shock (e.g. difficulty seeing opened PSU).

Chuck(G)
July 10th, 2010, 07:54 AM
I'll add that if what you heard was a loud "snap" then just a bit of smoke, but the system kept going, it's likely a popped bypass or decoupling capacitor somewhere on one of the boards, not in the power supply. This happens more than you'd think on old gear. If it's a bypass (value around 0.1uF, don't worry about it--just make a note to replace it when you have time. If it's a decoupling (usually tantalum), move it a bit higher on your priority list, but don't sweat it.

QuantumII
July 10th, 2010, 09:04 AM
I'll add that if what you heard was a loud "snap" then just a bit of smoke, but the system kept going, it's likely a popped bypass or decoupling capacitor somewhere on one of the boards, not in the power supply.

I second this. I have experienced the same multiple times. One snap, flying debris and some smoke. My S-100 backplane did this 3 times, one cap each time power was applied. The first "explosion" severed a PCB trace as well :-P Look around the motherboard for signs of a broken small capacitor. Usually they are blue or light brown.

Might as well show us a large picture of the motherboard, so that we can help you look at it :-)

ibmapc
July 10th, 2010, 01:34 PM
Might as well show us a large picture of the motherboard, so that we can help you look at it :-)

Here's a couple of pics of the 10 mhz MB. I can't see anything here but maybe a more trained eye might see something. Also, on a side note, anyone have info on this board?
I have not been able to figure jumper settings or how to switch between turbo and slow speed. I've tried "cntrl+alt +/-", and "cntrl+alt+pgup/pgdwn" and a bunch of other keyboard manuvers with no luck, it remains in turbo mode.

modem7
July 10th, 2010, 02:33 PM
It sounds to me like your PSU/motherboard combination is working, but that you are trying to locate the component that arced/smoked. Is that the case?


Also, on a side note, anyone have info on this board?
Your motherboard looks very much like a PIM-TB10. The manual for that is at http://members.dodo.com.au/~slappanel555/manuals.htm


I have not been able to figure jumper settings or how to switch between turbo and slow speed. I've tried "cntrl+alt +/-", and "cntrl+alt+pgup/pgdwn" and a bunch of other keyboard manuvers with no luck, it remains in turbo mode.
When you tried "cntrl+alt +/-", did you try the "-" on the keypad (i.e. not the "-" on the main section of the keyboard).

ibmapc
July 10th, 2010, 03:19 PM
It sounds to me like your PSU/motherboard combination is working, but that you are trying to locate the component that arced/smoked. Is that the case?


Your motherboard looks very much like a PIM-TB10. The manual for that is at http://members.dodo.com.au/~slappanel555/manuals.htm


When you tried "cntrl+alt +/-", did you try the "-" on the keypad (i.e. not the "-" on the main section of the keyboard).

Yes, the combo seems to work but when Falcon AT is run, the video goes to EGA Mode(I'm using an ATI EGA Wonder) and since this game was meant to run on a 286, it pushes the system pretty hard and that's when I hear the arcing and see the smoke (happened on two separate occasions).
For now I've put the original MB back because the arcing noise made me nervous, but I'll probably try this MB agian soon. I've tried both "-" keys with cntrl+alt.

Chuck(G)
July 10th, 2010, 03:52 PM
Looks like you've got something very close to this turbo mainboard (http://www.artofhacking.com/th99/m/U-Z/31632.htm). If the jumpers line up, you may want to fool with jumpering/un-jumpering JP6 to alter the speed.

As far as software-controllable turbo mode goes, try Ctrl-Alt-keypad anything.

ibmapc
July 10th, 2010, 04:08 PM
Looks like you've got something very close to this turbo mainboard (http://www.artofhacking.com/th99/m/U-Z/31632.htm). If the jumpers line up, you may want to fool with jumpering/un-jumpering JP6 to alter the speed.

As far as software-controllable turbo mode goes, try Ctrl-Alt-keypad anything.

You're right Chuck, that board looks very close. The only difference I can see is my board doesn't have JP9 for parity enable/disable. I wonder if that means that it doesn't use parity. Or, maybe it does and can not be dissabled. How can I determine if parity is used? Sorry for wondering off topic but my mind tends to do that a lot.

ps The board that Modem7 mentioned in his post also looks very close. It's manual also specifies JP6 for turbo.

modem7
July 10th, 2010, 04:31 PM
According to Mueller's Upgrading & Repairing PCs book, the 5155 PSU is rated at 114W. Compared to a standard 5155, you're running a faster motherboard and you've probably installed a hard drive as well. Maybe you're right. Maybe Falcon AT use is indeed pushing the PSU loading to it's limits (but not enough to invoke the PSU's overload protection).

Re the component that smoked. Have you tried disconnecting the 5155 from the mains for a while (10 minutes should be adequate), then using your nose to home in on the component.

ibmapc
July 10th, 2010, 05:13 PM
According to Mueller's Upgrading & Repairing PCs book, the 5155 PSU is rated at 114W. Compared to a standard 5155, you're running a faster motherboard and you've probably installed a hard drive as well. Maybe you're right. Maybe Falcon AT use is indeed pushing the PSU loading to it's limits (but not enough to invoke the PSU's overload protection).

Re the component that smoked. Have you tried disconnecting the 5155 from the mains for a while (10 minutes should be adequate), then using your nose to home in on the component.

Yes, I've tried that but my nose can't find the source. And you're comment about pushing the PSU to it's limits is exactly what I'm afraid of. Maybe I'll put the Turbo board back in and see if I can kill it. At least at the moment I can still find refurbished PSU's. In the future, I'm sure they will be harder to find.

RE the hard drive. I have only a CF card connected to a Siliconvalley Computers IDE adapter.

Chuck(G)
July 10th, 2010, 06:03 PM
Does your 5155 have a hard disk installed, or just floppies? If it's just a floppy system, you've got enough +5 for your board. Remember that the original 5150 PSU was only 63.5W. If you have a hard disk installed, try disconnecting it.

Your motherboard uses parity, BTW. (4 banks of 8 data bit + 1 parity).

Your motherboard is one of many variations of the basic ERSO XT board. I've got two variations here myself--an 8 and a 10 MHz one.

modem7
July 10th, 2010, 08:18 PM
I've just calculated an approximate power consumption of the 5155's CRT unit. It's only about 15W. That's based on the fact that the unit is powered only by a dedicated 12Vdc wire pair from the PSU, the wire pair rated at 1.5A.

So if we make some approximations:
1. 5155 CRT unit = 15W
2. 10MHz motherboard = 20W (based on fact that a 5160 motherboard draws about 15W)
3. Various cards = 20W
4. CF card = 2W (a wild guess)

So Chuck is right. There's certainly enough power available.

Maybe a dud PSU - defective component that only starts to fail above a certain current level.


Yes, the combo seems to work but when Falcon AT is run, the video goes to EGA Mode(I'm using an ATI EGA Wonder)
I should have picked this up before.
The CRT unit will be connected to the composite video connector on your ATI EGA Wonder.
What's the go there? It's set for CGA mode?
If the composite video out is switching from CGA to EGA, I bet the CRT unit is being highly stressed.

ibmapc
July 10th, 2010, 09:21 PM
I should have picked this up before.
The CRT unit will be connected to the composite video connector on your ATI EGA Wonder.
What's the go there? It's set for CGA mode?
If the composite video out is switching from CGA to EGA, I bet the CRT unit is being highly stressed.

Yes the video on the internal CRT flickers alot in EGA mode, but this has not been a problem when using the original MB at 4.77mhz. Does the faster clock speed(10mhz) effect the composit out of the EGA Wonder in EGA mode? Anyone care to speculate what if any damage might be occuring with the internal monitor when the EGA Wonder switches to EGA mode? I believe it goes interlaced causing the flickering display. I was under the impression that the EGA Wonder was designed to operate in EGA mode without harming the 5155's monitor.

Chuck(G)
July 10th, 2010, 09:41 PM
I think you're okay there. A composite monitor simply wouldn't sync, unlike a TTL mono monitor which most likely be turned into a crispy mess.

QuantumII
July 11th, 2010, 02:27 AM
Here's a couple of pics of the 10 mhz MB. I can't see anything here but maybe a more trained eye might see something.

Could be one of the blue capacitors. They are scattered everywhere on the board, and the one that made the pop (If one did, that is..) could just have a small crack in it making it hard to detect. On the other hand, as have been mentioned in this thread, if it works fine you should not worry too much :-)

ibmapc
July 11th, 2010, 04:57 PM
Could be one of the blue capacitors. They are scattered everywhere on the board, and the one that made the pop (If one did, that is..) could just have a small crack in it making it hard to detect. On the other hand, as have been mentioned in this thread, if it works fine you should not worry too much :-)

Check out this close-up of a blue capacitor. I saw the small cavity in the top of it when I inspectected the board under magnification. This capacitor is right next to the slot where the EGA Wonder was plugged in. So, my guess is that when the video board was switched to EGA mode, this capacitor didn't handle it well and started arcing ind smoking.

Now, I need to find a replacement. Printed on the side is the number 104. Anyone have any sugestions on a source for replacement.

Chuck(G)
July 11th, 2010, 05:00 PM
It's what I said earlier--a 0.1uF bypass capacitor. Just about any type of this capacitor is suitable. I buy them by the 100s--they're cheap. You can use a film, ceramic...etc. type. It doesn't matter in this application.

ibmapc
July 11th, 2010, 05:11 PM
It's what I said earlier--a 0.1uF bypass capacitor. Just about any type of this capacitor is suitable. I buy them by the 100s--they're cheap. You can use a film, ceramic...etc. type. It doesn't matter in this application.

will one of these work?

Chuck(G)
July 11th, 2010, 05:29 PM
Probably; I'm assuming that its voltage rating is at least 10V.

ibmapc
July 11th, 2010, 08:47 PM
It's rated for 50V

Chuck(G)
July 11th, 2010, 09:30 PM
You're fine, then.

ibmapc
July 12th, 2010, 04:03 PM
I would like to thank everyone who helped me through this minor crisis!! You guys are really great!! I managed to un-solder the offending capacictor and re-solder the new one(note, I only burnt two fingers and now there's a funny brown spot on the dining room table). But, the system is running perfectly with no arcs or pops and no more smoke, even when worked hard in EGA mode running Falcon AT!! Now, if someone could help me find a V20 that will run at 10mhz and maybe an 8087 to go with it, that would be really cool.

QuantumII
July 12th, 2010, 10:15 PM
Cool! It's great that you got it repaired :-)

wouldrichest
July 14th, 2010, 04:43 PM
Maybe even remove the MB from the case and put everything back the way it was when it worked, in case there's a short at one of the mounting screws. And of course make sure that the connectors have the black wires in the middle, i.e. next to each other. It's pretty hard to catastrophically "short out" a PC PS.

You guy's are great! When I connected the original MB and flipped the switch, the fan came on and appears to run normal, so I turned it back off and connected the HDD with the LED and flipped the switch again and the fan came on, the LED lit and the drive spun up normally. So it would seam the PSU is fine. Just my trouble shooting methods were faulty. Glad I didn't buy that refurbished power supply from CODEMICRO!! Thanks everyone!!


__________________
watch free movies online (http://moviesonlinefree.biz/)