• Please review our updated Terms and Rules here

IBM AT 5170 with battery corrosion (advice needed!)

P

phonophan

Experienced Member
Joined
Nov 10, 2019
Messages
67
Location
New Jersey, USA
Greetings -

Picked up this IBM AT 5170 in the wild... pretty dirty and unfortunately the battery leaked.

So... pretty sure the drive controller is toast... but hoping the motherboard is salvageable.

Would appreciate any advice on what you would do next! Thank you for your time and expertise!







 
Wash the crud off. dry thoroughly--start with plain water. The IDE drive controller's nothing special, but you could try cleaning that and see if it's salvageable.
 
If the generic IDE card doesn't pan out you can grab any of dozens of those types from Ebay. Or you could grab an IBM MFM combo card for around $20-40 if you want to get it back to factory.
 
That's the idea--but before you apply power, make sure it's dry (i.e. no water trapped under components or in sockets).

Some folks use the dishwasher (no detergent, just water), but you probably want to go easy.
 
What are those extra PROM's near the BIOS? Can we have a close up?

Edit: Ignore that. I realised you can click to zoom. Looks like the BIOS is spread across 4 chips. I've read it's an option but never seen it before!

My guess is this is a very early model 1 board!
 
One hint are the TI ZA1250 "piggyback" 128Kx1 DRAM chips. If you need to replace them, they can be a bit hard to find--but I've seen ebay sellers offering pulls.
 
bobba84 said:
What are those extra PROM's near the BIOS? Can we have a close up?
Edit: Ignore that. I realised you can click to zoom. Looks like the BIOS is spread across 4 chips. I've read it's an option but never seen it before!
Click to expand...
Member fs5500 has the same BIOS set, including sticky tape. Pictured at [here].

Chuck(G) said:
One hint are the TI ZA1250 "piggyback" 128Kx1 DRAM chips. If you need to replace them, they can be a bit hard to find--but I've seen ebay sellers offering pulls.
Click to expand...
Noting that as an option, there are at least six alternatives, per [here].
 
Wow, this is very early IBM PC 5170 type 1 motherboard.
(Unfortunately my board was dead, but BIOS dump was succeeded.)
 
One way to dry electronics after washing is compressed air - I've done this for years with new circuit boards that I make after cleaning off the water soluble flux. I usually do two passes, the first air pass is to blast off all the obvious water, the second air pass is to try to get into all the areas that escape visibility. Letting it dry more after that never hurts, but I usually take a pcb right to testing after that.
 
Wash it with the purest water you can find, the water is ok, its the salts that do damage (though with stuff this vintage, there is plenty of meat to go at before tap water will corrode enough). The wife dropped her kindle in the sea 4 years ago, people though i was mad when I started pouring bottled water all over it. Major rinse and a bit of hair dryer and its still working to this day.

I tend to use ultra pure water from work, its the worlds purest water !
 
Thanks all for your advice on this! So after some cleaning and waiting for parts to arrive... i'm back to working at this.

The good news is that the motherboard (seemingly) is working? ...but having some issues still.

SCENARIO 1
Nothing but the graphics card. Loads to PC Basic.



SCENARIO 2
IBM 6480111 16-Bit ISA MFM Hard Disk Controller Card for AT Computer -- A: 3.5 floppy will initiate upon power up... i can hear the disk seemingly is being read but after trying multiple disks, it just loads to PC Basic... so while i hear the audio queues of the floppy working there're no visual. Also, no signs of life from the B: 5.25 floppy. It looks like this controller card may require two floppy cables instead of one piggy backed cable. This card probably needs to go anyway as it's an MFM conroller and I don't have an MFM disk.



SCENARIO 3
This scenario worries me the most because it's essentially how I would like the computer to actually run. Have an XT-IDE card with Compact Flash that seems to be recognized but just hangs at this screen "STARTING MS-DOS".




I'm worried ultimately something is still wrong with the motherboard that it will only boot to PC Basic. Any advice? Thank you for your time and help!
 
Re SCENARIO 2

phonophan said:
Also, no signs of life from the B: 5.25 floppy.
Click to expand...
Normal. The 5170's power-on self test does not attempt any operation on B:

phonophan said:
It looks like this controller card may require two floppy cables instead of one piggy backed cable.
Click to expand...
No. That card has only one connector for floppy drives. See [here].

phonophan said:
A: 3.5 floppy will initiate upon power up... i can hear the disk seemingly is being read but after trying multiple disks, it just loads to PC Basic... so while i hear the audio queues of the floppy working there're no visual.
Click to expand...
The early BIOS that you have does not support 1.44M diskettes, but will read 720K ones. Were you using a 720K one?
 
Re SCENARIO 3

An incompatibility is known to exist between the XTIDE Universal BIOS (XUB) and the 5170 motherboard's IBM BIOS.
See the 'Nil or unstable operation in an IBM 5170 (IBM AT)' entry at [here].
I bet that is what the problem cause is.

Note that if you decide to swap out the 5170 motherboard's BIOS ROM's for the Award ones, in going from four 27128 ROM's to two 27256 ROM's, you will need to change the setting of shunt block U131 on the motherboard (see [here]).
 
modem7 said:
Note that if you decide to swap out the 5170 motherboard's BIOS ROM's for the Award ones, in going from four 27128 ROM's to two 27256 ROM's, you will need to change the setting of shunt block U131 on the motherboard (see [here]).
Click to expand...

OK - I have the new Award Bios in hand... i read what you provided about going from 4 chips to 2 chips and the changes to the shunt block.

...but, please, forgive my ignorance here. I understand that the settings need to be changed but I don't understand what i'm physically doing to change the shunt block setting. I've done switches and jumpers but never this.

Any guidance / advice would be appreciated. Thank you in advance!
 
phonophan said:
...but, please, forgive my ignorance here. I understand that the settings need to be changed but I don't understand what i'm physically doing to change the shunt block setting. I've done switches and jumpers but never this.
Click to expand...
Okay. See [here]. The shunt block can be taken out of its socket. Fitted to the socket in one orientation for '27128' operation, and the other orientation for '27256' operation.
 
modem7 said:
Okay. See [here]. The shunt block can be taken out of its socket. Fitted to the socket in one orientation for '27128' operation, and the other orientation for '27256' operation.
Click to expand...

It's alive! I'm thrilled, thank you!

IvbRq1cl.jpg
 
phonophan said:
What? It's an XT-IDE CF+IDE card... what's wrong with using that in this machine?
Click to expand...

you're getting 8-bit performance on a 16-bit machine, basically halving your throughput

you'd be better served with any standard generic 16-bit IDE card connected to one of the myriad chinese 40-pin IDC to CF adapters and the XUB burned to a ROM socketed on the mainboard if needed
 
You must log in or register to reply here.
Back
Top