Have you seen this chip before? (Datasheet wanted)

[FixyMacFix]

Hi,

I'm reviving an old IBM PS/2 20 MB hard drive from the late 80s / early 90s.

It's mechanically fine, but the PCB and various components are badly corroded due to leaking caps. Long story, but I've fixed a lot and made some progress. However it would be easier with some documentation...

So two questions:

1) I was wondering if anyone recognises the chip in the picture. It'd be immensely helpful with the datasheet as this is the chip where all address and data lines from the main connector go to.

2) In case anyone on this forum would actually happen to have the schematics for this thing I'd be forever grateful. I know IBM published some stuff but I haven't found any for this particular hard drive. Or fixed disk drive as they called them...

Thanks

P.S. Of course I could fetch a broken drive and transfer the logic board from that etc... but where's the fun in that? 

[fzabkar]

Could we see a detailed photo of both sides of the PCB?

"612X117" appears to be an IBM internal part number. I expect that the IC would be an ASIC.

You can see the other "612" part numbers on the label and in these photos:

http://knoppix.net/store/img/g/cxkAAOSwxaBi63ej/s-l1600/Vintage-IBM-6128285-WDL-320-20MB-PS-2-Model-30-50-.jpg

https://external-preview.redd.it/iLwA0rVq3GtTGEQbk8TUVLnzJcZ_JNwDWoFVa3l1dL0.jpg?width=640&crop=smart&auto=webp&s=d2182bf9f3372d2d0dc3fe0e4fac9aff2c7d23db

https://www.researchgate.net/profile/Ahmet-Erdogan-2/publication/4245046/figure/fig3/AS:394715372507139@1471118886218/Reference-images-of-different-types-of-PCB.png

Some research suggests that your drive was used in the Model 30. A compatible Seagate model was the ST-125L:

https://www.ebay.com/itm/324800023347

A photo of the Seagate PCB ...

https://i.ebayimg.com/images/g/GU0AAOSwfSthTCe3/s-l1600.jpg

... suggests that it matches the pinout of the fixed disk in the PS2 Model 25:

https://mirrorservice.org/sites/www.bitsavers.org/pdf/ibm/pc/ps2/PS2_Model_25_Technical_Reference_Jun87.pdf (see page 119)

This is the other side of the WDL-320 PCB:

https://m.media-amazon.com/images/I/81X0foGyDtL.jpg

The Data 0-7 pins appear to be grouped together in the expected place, and they appear to be bussed to pins 1-7 of U19 (the mystery IC).

Therefore it would appear that U19 is some kind of Micro Channel interface IC. I expect that the drive would plug into the PS2's MCA slot via a dumb riser card.

[FixyMacFix]

Thanks!

Photos are attached (note, at the time the first photo was taken there was a cap missing on C6 (marking hidden behind the chip - it's near the J1 marking) and caps C10 and C11 should've switched position).

What you're saying makes a lot of sense.

When I first started on this caps had leaked and there was really bad corrosion. Still is, probably doesn't show well in the pictures (and I've cleaned a LOT).

Anyway, drive wasn't spinning up, the LED was always on. I found pin 14 and 15 on U19 were not connected so fixed that. Drive now spins up, I can hear the heads move a little shortly after power-up and machine recognises there is a drive. However it can't identify it.

As such I'm wondering if the problem is it can't read the ROM. The ROM (which I believe is U11 - chip closest to the connector) however seems to be (mostly) connected to U21 (the IBM chip next to the mysterious U19/612x117). I've verified continuity  for all the pins on the U11 chip, so my hypothesis is some signal is not making its way from U21 through U19 and back to the connector.

The board appears to have at least one hidden layer however, because there are some pins which are connected, but I can't see a connection. Have tried a strong light but can't shine through the PCB.

Hence the wish for a circuit diagram and/or datasheet for the U19 chip.

Re your comment about Microchannel: The model 30 has an ISA bus. MCA drives have a wider connector and (as you rightly say) connect directly to the riser card. In this machine however there's a dedicated drive connector on the planar (IBM naming convention for motherboard...). Rumours say these drives are mostly MFM but with some IBM proprietary stuff added to it (inline power etc - not sure what else differentiates it from a stock MFM drive - but stock MFM drives do not work in this machine). They might not be MFM... anyway:

If you draw a line straight across the board from Y3 to C3 ish, that's where the corrosion was. The rest of the board was not affected. The worst area was between U19 and C10 and C11, which leads me to believe there is some pin which lacks a connection to somewhere. Or a short...

I probably should give up, but it would be so damned cool to get this working again...

I guess I could get a crashed drive and compare logic boards.

Thanks again for your help and patience!

[fzabkar]

This drive may have MFM recording technology, but there is an HDD controller on the same PCB. The original MFM drives had no on-board controller. The interface looks like a precursor to IDE. (Of course you are right, there is no MCA bus. I was confused because several online sources claim that your drive was available in the Model 50.)

As for the drive failing to detect, it could be that it is unable to read a hidden firmware area on the platters (where the defect list would be stored), and the drive then locks out access to the data area. At least that's how modern HDDs behave.

Edit:

FWIW, the IBM WDS-2120 120MB SCSI HDD of that era had an S80C196KW microcontroller for the SCSI interface. It had the same QFP-64 package as U19.

Your drive has an MFM/RLL read data processor IC.

32P541-CH, Silicon Systems, Read data processor, 5V and 12V, PLCC-28:
https://www.ardent-tool.com/8573/SSI_32P541-CH_Datasheet.pdf

Edit #2:

This PCB appears to be using the same chip, but with a different IBM part number:

http://www.ibmfiles.com/pages/ps2capacitors.htm

64F8574 = 612X117 ??