Dead LC PDS ethernet?

[blturner]

I’m attempting to setup a Farallon LC PDS card with an RJ-45 connector I had lying around in a Performa 637CD I recently obtained. I originally picked up the card for an LCII that’s in storage.

The card installs fine, and powers on with a green LED, but so far I have not been able to get a network connection using System 7.5.3 with Classic Networking or Open Transport. I tried configuring the network manually and dynamically.

I installed this driver from the macintosh garden: https://macintoshgarden.org/apps/farallon-lc-pds-ethernet-driver-setup

Running the diagnostics from that software package does offer some clues, but I’m not sure if it’s indicating a hardware failure or configuration error. Screenshots attached. It looks like the card is attempting to send packets, but all fail.

 

[Glenn Anderson]

Is this one of those cards that has a big ASIC (Farallon 3170828-00-01), a ROM, and 2 SRAM chips? Loopback failing in the controller sounds like that ASIC has failed, particularly with RAM test passing and register access passing. Are you definitely running the test with the card not active for AppleTalk or TCP/IP? That might interfere with the loopback test.

Before writing the card off, I'd try cleaning the slot connector with a deoxidizer, and check the ASIC for anything stuck between the pins, or pins that are loose due to bad solder joints.

One other idea, if the 20MHz crystal on the card failed, it wouldn't surprise me if that resulted in the symptoms you are seeing, where accessing the RAM and controller works, but nothing communications related works, including loopback tests.

 

[blturner]

Is this one of those cards that has a big ASIC (Farallon 3170828-00-01), a ROM, and 2 SRAM chips? Loopback failing in the controller sounds like that ASIC has failed, particularly with RAM test passing and register access passing. Are you definitely running the test with the card not active for AppleTalk or TCP/IP? That might interfere with the loopback test.

Before writing the card off, I'd try cleaning the slot connector with a deoxidizer, and check the ASIC for anything stuck between the pins, or pins that are loose due to bad solder joints.

One other idea, if the 20MHz crystal on the card failed, it wouldn't surprise me if that resulted in the symptoms you are seeing, where accessing the RAM and controller works, but nothing communications related works, including loopback tests.

It's got a different chip number, 5000112-00-01. I blew some compressed air on the slot connector and the pins tonight, there was a sneaky cobweb in there, tried it again but no change. I'll give the oxidizer a try when I have a chance to pick some up. As far as I can see from a visual inspection, there's nothing obvious that looks out of place.

Here's a picture of the card:

 

[Glenn Anderson]

It's got a different chip number, 5000112-00-01. I blew some compressed air on the slot connector and the pins tonight, there was a sneaky cobweb in there, tried it again but no change. I'll give the oxidizer a try when I have a chance to pick some up. As far as I can see from a visual inspection, there's nothing obvious that looks out of place.

Here's a picture of the card:

That chip is the ROM, it is the one with S9406AN on it that I would check the pins on. I think S9406AN is a date code (probably 1994 week 6), under the label it will have "DP83902AV" and "ST-NIC™", that is the ethernet chip. The label sticking over the edges might be obscuring something.

The ECLIPTEK 20MHz oscillator is the other part that I'm wondering about. Oscillators sometime fail or become unstable due to mechanical shock, but there is no way to visually detect that.