PDA

View Full Version : 5160 wont start, mobo problem?



rtwo
September 13th, 2010, 02:50 AM
Got myself a free 5160 together with a 5154 monitor.
When I switch power on, I can hear the harddisc tick once and the powerfan spins up, nothing more. Have been testing for a while, when I disconnect the mobo connectors, the harddisc spins up normaly. Have removed all cards, but still the same problem. Have tried a power supply from my 5170, when connected to them mobo, only the fan spins up an one tick in de harddisc, when I remove the mobo connectors, the harddisc spins up.
My conclusion, a faulty motherboard??, or can it be something else.
Oh yeah when I connect only P1 on the mobo, harddisc spins up, when I connect P2, the harddisc ticks once.

If it is a faulty motherboard, can I replace that one with a mobo from a 5155?, I have read that those mobo's are the same.

per
September 13th, 2010, 04:19 AM
Got myself a free 5160 together with a 5154 monitor.
When I switch power on, I can hear the harddisc tick once and the powerfan spins up, nothing more. Have been testing for a while, when I disconnect the mobo connectors, the harddisc spins up normaly. Have removed all cards, but still the same problem. Have tried a power supply from my 5170, when connected to them mobo, only the fan spins up an one tick in de harddisc, when I remove the mobo connectors, the harddisc spins up.
My conclusion, a faulty motherboard??, or can it be something else.
Oh yeah when I connect only P1 on the mobo, harddisc spins up, when I connect P2, the harddisc ticks once.

If it is a faulty motherboard, can I replace that one with a mobo from a 5155?, I have read that those mobo's are the same.

Step 1, check that all motherboard switches are properly set. Step 2, make sure the video card is properly installed. Step 3, start the machine, and wait a coupple of minutes. If nothing happens durning step 3, check the motherboard for faulty/shorting capacitors. If no broken capacitors are found, you will need to use an osclioscope to check for any activity on the motherboard (A0, A1 and D0-D7 is of particular interest).

rtwo
September 13th, 2010, 09:07 AM
Step 1, check that all motherboard switches are properly set. Step 2, make sure the video card is properly installed. Step 3, start the machine, and wait a coupple of minutes. If nothing happens durning step 3, check the motherboard for faulty/shorting capacitors. If no broken capacitors are found, you will need to use an osclioscope to check for any activity on the motherboard (A0, A1 and D0-D7 is of particular interest).

Switches are set properly, all card are installed properly, severall times, strated the machine, waited a long time, nothing happens, no beeps, nothing. Checked the motherboard visual, didn't notice any broken capacitors. Don't have a oscilioscope, so I guess I am out of options.
Can I change the mobo with one from my 5155???

per
September 13th, 2010, 09:46 AM
Switches are set properly, all card are installed properly, severall times, strated the machine, waited a long time, nothing happens, no beeps, nothing. Checked the motherboard visual, didn't notice any broken capacitors. Don't have a oscilioscope, so I guess I am out of options.
Can I change the mobo with one from my 5155???

Yes, you can swap the motherboard with the 5155. However, then you end up with an 5155 with a broken motherboard.

You will need a multimeter to check the capacitors, though. However, you should check the power-good signal first after you have located a multimeter.

One last thing you could try to do is to swap the RAM chips in bank 1 with the RAM chips in bank 2.

Chuck(G)
September 13th, 2010, 09:51 AM
You know, there seems to be a train of thought among some collectors of "If I can't fix it, then I'll junk it". I'm not saying that that's the case here.

A moderate suggestion to the forum in general--if you have something that doesn't work and you can't fix, consider offering it to someone who might be able to do something with it.

rtwo
September 13th, 2010, 10:38 AM
@per:
I will try bank switching first, I am not an electronics expert so testing stuff will be hard for me to do, all I can do is some visual checking and some voltages checking.
I have two 5155 pc's, one in top conditioning and one more for spare parts, harddisc, floppydrive butcasing and keyboard are crap, but the monitor is still working, so is the powersupply, so I don't mind using its motherboard in the 5160.
@Chuck(G):
I don't mind giving away parts that don't work, even parts that still work, but I live in the Netherlands and its hard to find information about vintage IBM computers overhere, thats one of the reasons I visit this forum. So I don't know other collectors living nearby. I don't mind sending stuff to you! or anyone else overthere ;)

Chuck(G)
September 13th, 2010, 11:28 AM
#rtwo,

Complete IBM technical reference material has been posted for the 5160 here and on the retrocomputing site, including complete schematics and BIOS source. That's all anyone really needs to get a non-working system going.

framer
September 13th, 2010, 04:55 PM
Really, have pulled the HD and tried booting from a floppy? Does it ever beep during bootup? I really think you getting ahead of yourself.

framer

rtwo
September 14th, 2010, 12:56 AM
I did try to boot without HD, but same problem.
Found the following tread and it looks like my problem: http://www.vintage-computer.com/vcforum/showthread.php?21443-5160-weird-power-problem&highlight=5160
With P8 connected the 5160 wont boot up, disconnected at least the HD spins up, so I removed -12 from the P8 connector, started the 5160, HD spins up, 2 beeps (POST error), but nothing on screen, no memory count, just a cursor in the left upper corner, thats it, but its more than before ;)
Tried booting from floppy, it reads quickly, than 2 beeps,etc.
Anything else I can try or do?

per
September 14th, 2010, 01:12 AM
I did try to boot without HD, but same problem.
Found the following tread and it looks like my problem: http://www.vintage-computer.com/vcforum/showthread.php?21443-5160-weird-power-problem&highlight=5160
With P8 connected the 5160 wont boot up, disconnected at least the HD spins up, so I removed -12 from the P8 connector, started the 5160, HD spins up, 2 beeps (POST error), but nothing on screen, no memory count, just a cursor in the left upper corner, thats it, but its more than before ;)
Tried booting from floppy, it reads quickly, than 2 beeps,etc.
Anything else I can try or do?

That indicates that one of the filter caps on the -12v line is toast.

I've asked before, but are you sure Sw1-nr5/6 are both in the "ON" position AND that the graphics card has it's own BIOS?

You could eventually try to use a different graphics card.

rtwo
September 14th, 2010, 02:48 AM
Yes they are, and yes it has, tried the graphics card from my 5155 but still no luck.
Tried everything I could, guess I will have to stock this one as is.
Thanks however for your help.

per
September 14th, 2010, 04:14 AM
Sounds like the BIOS extension on your graphics card (if its an EGA or VGA) may be troubling, if it has a BIOS extension at all.


tried the graphics card from my 5155 but still no luck.

You did set the switches accordingly to that one too (thar's problably CGA, which is "On/Off" or "Off/On")?

Anyways, pictures of the graphics cards may help.

james1095
September 14th, 2010, 11:06 AM
I would bet money that you have a shorted tantalum capacitor on the motherboard. These are small often yellow blob-like parts with either 2 or 3 legs. I had this exact problem with mine, tracking it down turned out to be easy because after I power cycled the machine a few times listening for the hard drive to make a noise, the offending capacitor exploded! Took about 10 minutes to replace and is working perfectly now.

rtwo
September 20th, 2010, 11:53 AM
Ok, I finally got me a digital pocktsize multimeter, borrowed it from a friend, have been testing it using the function "measuring resistance and capacitance", what should I exactly look for?
If I get a reading of 46,5 in display, does that mean the capacitor is ok? and if there is no reading at all, it is broken?

Chuck(G)
September 20th, 2010, 11:57 AM
Check capacitors by using the lowest resistance range (i.e. the continuity testing range). A zero- or near-zero ohm reading indicates a short circuit--and it's the shorted caps that will keep your system from starting.

Unfortunately, you'll have to unsolder suspect capacitors from the PCB, as testing for shorted caps in-circuit is very difficult to localize.

rtwo
September 20th, 2010, 12:21 PM
I can switch it from: kΩ .0l / kΩ 0.l / kΩ ol. / MΩ .0l / MΩ 0.l / Ω 0l.
which one is lowest resistance?

Chuck(G)
September 20th, 2010, 12:24 PM
Ω 0l.

(I take that as a range measuring in ohms.) Some DMMs have a buzzer also for a continuity test (makes things easy because you can use your ears rather than your eyes).

james1095
September 20th, 2010, 01:38 PM
The capacitors are all in parallel too so it can be really tough to narrow down which is the bad one unless you remove them from the circuit. That's easier said than done without good tools. It's really easy to damage the through-hole plating on a motherboard and ruin the board. I use a Hakko 808, but a Radio Shack desoldering iron with a fresh tip can get the job done with some patience. I got lucky and the bad capacitor exploded on mine after a few power cycles, that made it easy.

Chuck(G)
September 20th, 2010, 02:40 PM
Physical inspection for bulging, discoloration, etc. should always be the first step. If it looks bad, it probably is.

I've always done my desoldering with my Weller WTCP and the big Soldapullt. Works very well with through-hole stuff.

channelmaniac
September 20th, 2010, 04:15 PM
Everyone that troubleshoots shorted capacitors should have one of these:

http://www.anatekcorp.com/blueesr.htm

The ESR meter is nothing more than a high frequency milliohm meter that will go down to a hundredth of an ohm: .01 ohm. By doing this you can measure the caps to find the one with the lowest resistance to find the one that is shorted.

It takes copper traces to go from one cap to the next and those traces have very tiny, but measurable resistance values. When you get to the shorted cap you'll find a lower resistance value. Remove that cap and you should have the short circuit gone.

RJ

tezza
September 20th, 2010, 04:41 PM
It takes copper traces to go from one cap to the next and those traces have very tiny, but measurable resistance values. When you get to the shorted cap you'll find a lower resistance value. Remove that cap and you should have the short circuit gone.

RJ

Yes, it was a shorted cap on my 5170's disk controller that prevented any sign of life. With that replaced, volia! It rose Lazarus-like from the dead. See
http://www.classic-computers.org.nz/blog/2009-08-17-reviving%20an-IBM-AT--diagnosing%20a%20PSU%20non-start.html

Tez

channelmaniac
September 20th, 2010, 05:10 PM
Yes, it was a shorted cap on my 5170's disk controller that prevented any sign of life. With that replaced, volia! It rose Lazarus-like from the dead. See
http://www.classic-computers.org.nz/blog/2009-08-17-reviving%20an-IBM-AT--diagnosing%20a%20PSU%20non-start.html

Tez

Exactly!

But if you have dozens of the caps it's tough to find the bad one without desoldering an end off each until you find the bad one. If you use the ESR meter, you'll find an area of caps that have lower resistance than others (.05 ohms vs .01 ohms).. when you find the lowest with that ESR meter then desolder it and you'll have the bad one quickly isolated. ;)

Chuck(G)
September 20th, 2010, 07:31 PM
If you're handy, you can build one of the many do-it-yourself milliohmmeters on the web. Most are 4-wire (Kelvin bridge) setups using op amps and simple to build. Rich Cappels has a rather clever circuit using 1KHz bursts of low-level AC.

H-A-L-9000
September 21st, 2010, 10:29 AM
IMO it's not hard to find the broken one - give it an ampere or two with a regulated power supply, and the capacitor will budge. Safety glasses recommended though.

channelmaniac
September 21st, 2010, 10:47 AM
IMO it's not hard to find the broken one - give it an ampere or two with a regulated power supply, and the capacitor will budge. Safety glasses recommended though.

Depends on what is stronger... the traces or the cap... You might end up with a blown trace.

H-A-L-9000
September 21st, 2010, 11:47 AM
That's why I said regulated power supply. The tracks can take quite some current: http://www.basista.de/download/strombelastbarkeit_leiterbahnen.pdf

Power traces are usually wider than 0.25mm so 2A are fine. Also, you won't have it on for hours.

james1095
September 21st, 2010, 01:11 PM
Everyone that troubleshoots shorted capacitors should have one of these:

http://www.anatekcorp.com/blueesr.htm

The ESR meter is nothing more than a high frequency milliohm meter that will go down to a hundredth of an ohm: .01 ohm. By doing this you can measure the caps to find the one with the lowest resistance to find the one that is shorted.

It takes copper traces to go from one cap to the next and those traces have very tiny, but measurable resistance values. When you get to the shorted cap you'll find a lower resistance value. Remove that cap and you should have the short circuit gone.

RJ

I have a Capacitor Wizard which is a similar device. The problem with using it for this is that tantalum capacitors have very low ESR, that's one of their advantages, so a good one will often appear shorted to an ESR meter.

Most multimeters are not precise enough to measure very low resistances. A milliohm meter with good quality probes can do it but even then it's not 100%. Even a shorted component will still have *some* resistance, and if the adjacent parts are nearby and on big fat traces or power/ground planes it can still be tough to locate. Worth a shot though.

Chuck(G)
September 21st, 2010, 01:56 PM
If you've got a well-stocked hellbox, a simple milliohmmeter can be constructed with an LM317 voltage regulator and a few odds and ends.

But if it were my board, I'd just start with the capacitors closest to the power connector and work out from there.

rtwo
October 13th, 2010, 02:00 AM
Ok, have been busy for a while, but have started testing the mobo, could not find the problem so I switched the mobo with a working one from a 5155.
Computer starts up, single beep, floppy seek, harddisc spins up and starts reading, but nothing on screen, screens stays black. Switched the 5154 for a 5151 screen, adjusted the brightnes to full, and all I can see are horizontal lines. plugged in the CGA adapter from my 5155 (is working), the same, only horizontal lines, tried the graphics adapter from my 5170 with the monitor from my 5170, again horizontal lines, but this time with a blinking cursor. Tried all possible switch settings (5 and 6) but no luck. Switched Bios roms with latest version, same problem. Mobo is working, so are the adapters from my 5155 and 5170, the computers boots up, with single beep, so no errors, floppy seek, HD seek. What else can there be wrong, or what else can I try?

dongfeng
October 18th, 2010, 04:39 PM
Did you adjust the dip switches on the motherboard to correspond with the new graphics card?

Be careful, you might damage the card or monitor by plugging the incorrect one into it..

http://www.howard81.co.uk/upload/vcf/xt/xtdips.gif

rtwo
October 20th, 2010, 12:08 AM
Like I said in my post, I tried all combination but no luck, btw, the cards I tried from other systemes work fine.