(last updated: 10-Feb-2017)
The current configuration is no longer in original 'as acquired' condition.
Memory has been increased from 8MB to 32MB.
Logic board has been replaced.
Keyboard (M0116) s/n 780677
One-button mouse (A9M0331) s/n 592744
| Macintosh SE/30: 68030 CPU w/ 68882 FPU, 32MB RAM | |
|---|---|
| PDS | Farallon PN594-TP V1.0 Ethermac Ethernet Card |
14-Sep-2007
I purchased this from ebay seller '1922clevelander' on September 4, 2007.
Today the the package arrived via USPS Priority Mail in excellent condition. The seller had done a very good job of packaging the unit in discarded styrofoam and crinkled paper. Included were the M5119 main unit itself, a M0116 keyboard, and a A9M0331 one-button mouse, and cables for the mouse, keyboard, and AC. Also included installed is an ethernet adapter already installed. I'll need to research exactly which.
14-Sep-2007
The system is very yellow with age. One side is worse than the other. The underside is almost the original color. The keyboard is funny. The enclosure is very yellow, but the keys (except the space bar), are still their original gray putty color. I wonder if they are a different plastic, or maybe (but I doubt it) the guts have been replaced.
I windexed all of the exterior surfaces, and it cleaned up very nicely. There is some medium burnin on the screen, so I'll be curious about what I see when I power up.
It was dirty inside and I took some photos. Really not too bad, but clearly it had never been cleaned since it was shipped from the factory.
I removed the ethernet board first, starting with the connector board.
Then I removed the harddisk tray and, whew!, I found some kind of liquid trapped between the harddisk tray and the floppy tray. It smells like sardines! Something has dripped in/down from somewhere and it has migrated to this location. I cleaned the exposed surfaces, and then removed the floppy tray.
I could see right away that the floppy drive was filled with dust and lint. Definitely needed cleaning. I removed the floppy from the tray and vacuumed it thoroughly, wiped it down, and remounted it.
I went back to the ethernet connector board and removed from its mounting bracket. There was evidence of some liquid dripping in from there. I cleaned everything and reassembled it. I cleaned the main ethernet board.
I removed the motherboard, and cleaned it up. It had some of that weird liquid on the front edge. A couple of chips show some possible corrosion on the pins, and I cleaned them, but they appear rough.
I cleaned the inside of the cabinet really well, so I hope that I have taken care of this mysterious substance and though it has discolored the floppy and harddisk mounting trays, I'll keep my fingers crossed that the system will bo okay.
15-Sep-2007
It appears that the factory-original 40MB (probably) hard disk has been replaced. A 'MicroNet' sticker is on the drive, plus there is a 'MicroNet' badge applied to the front panel.
To image the drive, I removed it from its bracket and selected SCSI device #4 on JP02 pins 5-6. Then using my VAXstation 3100 with it's external SCSI port, and my cobbled together female-to-female cable I connected the drive and booted the Vs3100. It recognized the drive without any problem, so I allowed the system to boot NetBSD from its internal drive.
The copying took longer than I expected, about 45-50 minutes, and at the end I saw errors that were new to me. I was unable to record them before the scrolled off the screen, but the issue appeared to be that the system was attempting to read beyond the highest logical sector. Since that's all it was, I consider the image to be good.
15-Sep-2007
With everything essential back together, I decided to boot it up. I connected the keyboard and mouse, both to the rear DIN connectors, and attached the AC cord.
At power on, I hear the chime, and watched as the system booted up! I was very happy! After getting a bit familiar with the mac environment and a one-button mouse, I was able to try out a few things. Eventually I found the system info option under the apple logo menu, and saw that I am running Mac OS 7.0.1 and that the system is using almost 2MB, leaving a bit more than 6MB or RAM free. Great!
While the system was just running I was looking with my flashlight for various video adjustements, looking for any that weren't available on the exposed side of the 'analog' board. No additional adjustable items appear, but then I saw a foreign object lodged between the CRT and the analog board. I had noticed this during cleaning but had forgotten about it. I shutdown the system, and carefully extracted the object. It is heavy, metal, and I'm guessing that is a tool to be used when opening the case. I've come across references to a 'case cracker' tool that is helpful for getting the back cover off, and I think this is it. It's really just a short, squat, chubby flat screwdriver blade that I think is inserted in the groove between the front and rear halves of the case, and then twisted to pry them apart. What it was doing *inside* the case I don't know, and it's a wonder that it didn't short something out.
I finished reassembly by attaching the rear cover. Again I powered up and booted. I inserted a spare floppy to check the operation of the drive, and it seemed to work fine, until I ejected the floppy. Apparently something has jammed in the drive, becuase it re-senses a floppy even when there isn't one. So some small plunger switch is probably jammed in the on position now. I'll have to re-open the case and remove the drive to do some more cleaning/lubricating.
Unfortunately while experimenting with the floppy drive and even worse problem has developed. The video flickered, pops were heard from the speaker, and the system locked up. I rebooted and it seemed okay for a few minutes again, and then I saw the same behavior. I'm wondering it it is related to heat buildup now that the case is closed. Or, is it related to the mysterious fluid I found inside? Perhaps some has gotten into the power supply itself (which I didn't remove and clean earlier).
16-Sep-2007
I disassembled the system again, and with the floppy drive removed, I could immediately see that one of the three plunger switches on the front of the drive was still depressed. Using the sharp edge of a knife blade I lifted it back out of its base. I could see a thin shiny film on the white plunger and wiped it off. I believe it is the same smelly substance that I've cleaned from other parts of the interior. I depressed the plunger with my knife blade, and it did not spring back. I wonder if the mystery substance (which has shown some corrosive results elsewhere) has corroded the tiny spring inside the switch so that it no longer resists the plunger. I'll either have to replace the tiny plunger switch, or the drive, I think.
16-Sep-2007
While I had the system disassembled so far, I decided to remove the power supply also, and see if the mystery fluid had gotten inside it. I was hoping to identify the cause of the other, more significant, problem of the system randomly suffering some kind of electrical hiccup and subsequent hang. I carefully removed the power supply, slipping it out from under the CRT neck. I opened it up and looked inside. All was very clean, with hardly a hint of dust even. I reattached the PS cover after a couple of photos.
With the power supply removed, I was able to remove the entire base bracket and further clean the front of the case, near the front ventilation slots. There was a tiny amount of residue still, but not much at all. I also wiped down the entire CRT tube, now that it was all accessible. Well, that's everything.
I reassembled the system, except I left the back cover off. I went ahead and powered up, and was glad to see that the system no longer sensed an inserted floppy. I set the clock and just left the system running to see if it would suffer a hiccup with the back cover removed. Copming back after dinner, I could see that the system had indeed froze at some point, after about an hour according to the clock. Too bad. I'm now back to square one on why this happens. Apprently it is not just heat build up inside the case.
20-Sep-2007
After giving some serious thought to what might be causing the intermittent flashes, burp sounds from the speaker and random freeze-ups, I came to the conclusion that it very well could be remnants of the mystery fluid on the main system board. There were still some unreachable areas (mainly around the SIMM sockets) that I hadn't cleaned. I reasoned that as the system heated up, some of the fluid would start to 'flow', cover adjacent traces or pin pads, producing a short. The short would vaporize the little bit of fluid that cause the short, allowing me to successfully reboot and run, again, until the next occurance.
So, I choose to wash the board in the kitchen dishwasher. I figured that I didn't have a usable system as it was, and thus I couldn't really make things worse. It seemed that there was no reason to try the dishwasher as a method of cleaning the board. Also, lending support to my conclusion, I recall how we (Harry Armstrong) used to clean boards in the early days of Novell. We used a standard kitchen dishwasher!
First, I ran the empty dishwasher through a 10-minute rinse cycle, just to remove any remaining soap and food particles from its last use. Then I put the board in the top rack, component-side down. I figured that the water would drain away better upside-down. I then ran another 10-minute rinse cycle, with no soap or other detergents, just using the hot water. Our hot water comes from the tap at around 120-degrees.
After the cycle was finished I removed the board and blotted most of the water from both sides. Then I put it inside the kitchen baking oven, again component-side down. I had pre-heated the oven to about 130-degrees while the board was in the dishwasher. With the board inside the oven, I switched the oven off, so that the heating elements wouldn't come on again, creating localized hotspots. In about 20 minutes I removed the board, and found no traces of remaining water. The board was absolutely clean of the mystery fluid, and I was anxious to try it.
I reinstalled the board in the system, re-connection the floppy and hard disk. I was a bit anxious, but mostly confident when I switched the system's power on. No need to worry, the system works perfectly! It booted right up, and in the 45 minutes or so that I worked with it, I experienced no hiccups.
I left the system running overnight and most of the next day, for a total time of slightly more than 17 hours. During that period, no problems. I think I fixed it! And the strange 'fishy' smell is gone!
21-Sep-2007
After confirming that my 'dishwashed' main board was working correctly, I re-installed the Farallon ethernet board, and cabled the system to an ethernet hub using the twisted-pair port. I can't find any related software on the system, though, so I don't have the means to confirm the working state of the ethernet board.
After a bit of internet research, I found a couple of helpful sites, one of which pointed out that there is an AppleTalk server package for *nix called 'netatalk'. I installed the package to Aero-6, but it fails to start the atalkd deamon properly. More research needed. But, if this can work, then this would be the way to initially transfer the various pieces of networking software from the web to the macintosh, without too much trouble.
21-Sep-2007
While experimenting with the networking aspects of this system, I wanted to try the Asante Ethernet card (installed in Macintosh SE/30 #2) diagnostics, on the off chance that they would work with the Farallon board. I had put them on a diskette to transfer them over. I fully expected trouble from the floppy drive, but actually it did pretty good. Putting the floppy in isn't the problem, rather it's when the floppy is ejected that the drive still detects a floppy in the drive. The small plunger switch still is not moving its full travel length, but it moves enough tonight to work okay. In any case I can reach through the floppy slot to the plunger switch and lift it lightly with the edge of a knife blade, so any future difficulties won't require removal of the drive, at least. Still, I really should replace the plunger switch when I can find the correct replacement part.
27-Sep-2007
Today I received two batteries from batteryspecialists.com (SLC-based!) for the Real-Time Clock, at $6.95 each, which included free shipping. The old Maxell was obviously original and had rusted out. This battery is some unknown brand, but it is a proper replacement so I installed it into the motherboard after removing it from the system.
After upgrading the memory (next section), I plugged everything back together and booted up. I set the clock in the control panel, and shutdown. I left the machine for a about 60 minutes and then rebooted again. Sure enough, the clock is correct, and the few other settings that I had made (EtherTalk instead of LocalTalk) were still properly set. So, the battery is good!
27-Sep-2007
Also, today I received four sets of four 4MB 30-pin SIMMs from Ebay seller 'XXXXXXXXXX'. I paid just $6 for each set of four SIMMs! I removed the original memory SIMMs which consistes of four 256KB SIMMs and four 1MB SIMMs, and replaced them with the new ones. So I've gone from 5MB to 32MB total RAM in the system. I reinstalled the motherboard (after doing the RTC battery installation) into the system, and powered up. No problems with the boot up and the system sees 32MB! It's not all usable, because of the 'dirty' ROMs, but with MODE32 I'll be able to use it all. Or, if I install NetBSD it should use it all, as I understand it.
27-Sep-2007
After the battery and memory upgrades were complete, I thought that I'd try to determine if the drivers were installed for the Farallon ethernet card. I reasoned that if the card was there in the system, one which appears to have been operational since it's time with KPMG, then the driver is probably on the system. Browsing (ala Finder?) the hard drive contents I went through each folder and subfolder, but I didn't encounter anything that appeared to be the Farallon driver. No files have Farallon or PN594 that was recognizable by me.
On the off chance that the driver really was already on this system, though, I connected the system to the network using a Thick-to-TP transceiver and opened the chooser. I clicked AppleTalk, but, alas, I did not see my Linux box server. Oh, I forgot to select EtherTalk instead of the default-after-battery-failure LocalTalk network in the Control Panel. I went in and choose EtherTalk (which appears to be Phase 2, because of the double arrows) and then went back to the Chooser. I reselected AppleTalk, and, yes!, I see 'protiva', my Linux server. From there, all went smoothly. I managed to mount the shared directory and use it fine.
So, this system already has the driver! It is a good thing that I have imaged the hard disk, since I have had trouble getting my hands on the correct driver.
As a matter of fact, later tonight, I finally found a source for the Farallon driver. It is now at www.proxim.com, and in their 'obsolete' section. Apparently, Farallon was eventually acquired by Proxim. I'm grateful that they still have these old downloads available, but it's too bad that they didn't retain the farallon.com domain. I have downloaded the pertinent pieces of software and will hold them for possible need in the future. Otherwise, I'll continue to just use what is already installed.
28-Sep-2007
Now that I have a working system, with upgraded hardware, my first order of business was to locate and install the MODE32 system extension that will allow this SE/30's 32-bit 'dirty' (24-bit only) system ROM to actually work with 32-bit addresses. I've come across a good number of downloads on many sites, some of which have MODE32, but I had a lot of trouble finding a way to get the MODE32 'BinHex' file over to the Mac. The problem is that the resource and data forks have to be packaged up into a single file for internet travel, but the difficulty was that all of the un-packagers were only available in their own package. So I was in a Catch-22 situation.
I finally came across this site:
http://www.macgeek.org
and snagged this file:
http://www.macgeek.org/downloads/PC_Mac.zip
which had a very welcome solution. Because the Mac SE/30 has a FDHD 'SuperDrive' that can read PC-formatted diskettes, this package can be un-zipped on your PC and the diskette image copied to the floppy. Then the floppy can be moved to the SE/30, which finally allowed me to get a copy of the StuffIt Expander 4.5 over to the Mac, in native format. Now that I have that in place, all the other downloads will be a breeze!
Just a note, here, about versions. It appears that most of the software that I will want to use on this Mac basically requires System 7 or better. I confirmed that MODE32 only works with System 7, and cannot work with System 6. Also, I confirmed that the NetBSD Booter will only work on System 7 or better, with MODE32 enabled. Consideration needs to be given to how this affects my plans for my SE/30 #2, which has only System 6.0.8 installed. One solution there, I guess, would be to add an external hard drive and install System 7 on it, then boot from it. Or maybe use a Iomega Zip drive the same way.
21-Oct-2007
Two or three weeks ago I bought a second-hand Iomega ZIP-100 (SCSI) drive from MAS computers for $15. I had casually tried to use it in the days that followed, and I did dig up the ZIPtools 4.2 driver for it (putting it into the System Folder), but I was unsuccessful at getting it to be recognized by the system. I had been trying to use a ZIP disk that is marked 'BAD', hoping that I could reformat it on the mac, and make use of it. No luck.
Today, I moved the system and ZIP drive to my loft and concentrated a bit more effort on it, which paid off.
Again, I had no luck using the 'BAD' ZIP disk. The drive just wasn't showing on the desktop or being otherwise visibly recognized, though I could see the drives activity light blink when the system was rebooted, so at least the SCSI signals were getting through.
In a bit of desparation, I finally dug out a second ZIP disk that I knew was formatted and working fine in my PC-parallel ZIp drive. I popped it into the drive, and lo!, is spun up and the activity light blinked many times, and I got a popup dialog onscreen asking if I wanted to format the disk! So, this second-hand drive works, and it is recognized by the mac!
Interestingly, on a lark not knowing at all what to expect, I just grabbed the desktop icon for SQUID (my hard disk) and dropped onto the empty ZIPtest icon. The whole hard disk was copied to the ZIP disk in a folder at the root called SQUID, as would be expected. Here's the lark part. I started the Control Panel, then System Disk and chose 'ZIPtest' as my startup disk. Then rebooted. It booted directly from the ZIP disk! This would be a good way to add mass storage to a mac that doesn't have a hard disk, or maybe I can use this method to experiment with other System OSs without having to modify the hard disk. Hmm...
Anyway, after much searching to get my hands on a full, working set of Iomega ZIP tools, at last I found this site:
http://macdrivermuseum.net/disk.shtml
which has the Iomega ZIP tools v4.2.1, which I was able to successfully un-hex and un-stuff. Unfortunately, when I run the Tools application, it errors indicating that the installed driver is a different version (which is true, I'm running v4.2). So, I'll replace the 4.2 driver with 4.2.1 and see if that works better for me.
Well! I discovered something interesting! There IS NO DRIVER already installed! But the system recognizes the drive and the can read and write to it! I found this out while trying to locate the driver (that I thought was already installed) and found none. No wonder the tools application won't work. It wants to talk to a driver that is not there. So I installed the 'Iomega Driver' in the Extensions folder under the System folder. When I reboot, I now see an actual Iomega icon for the drive instead of the generic SCSI disk icon. And the Tools application works, which let's me format disks, and set them to Read-only.
Unless you wanted to write-protect the disk, I don't really see any need for the Tools, or even the driver. The system works just fine with (and can even boot from!) the drive without the driver! Maybe this has something to do with the specific System OS that I'm running (7.0.1), or perhaps there is something special about the SE/30 model.
Anyway, I'm a happy camper!
23-Oct-2007
With the ZIP drive now recognized, and after I learned how to change the System Disk setting in the control panel, I can boot from the ZIP drive if I want. I found this out by copying over the whole hard disk to a ZIP disk and tried booting from it. Sure enough, it worked! Somehow the Mac just looks for the System file wherever it can find it, I guess, because I moved the System Folder into another subfolder and it still works. Must be the nature of the mac.
With the help of the StuffIt Expander, I made diskettes from the disk images from Apple's website for version 7.0 and 7.0.1, installing these systems to a ZIP disk. They works fine, but I'm missing something, since the Ethertalk component was not included in these new installations.
Still, I've learned a bit about putting a mac together, getting the software installed, and testing the different systems.
The floppy drive started to work correctly, too. At first I had to hold the diskette slightly up after insterting it, but as time went on, it just worked.
Unfortunately, throughout this session, I experienced three separate lockups. And I can just smell a bit of the old 'fishy' smell again. I think I should disassemble the system again, and give the motherboard another thorough washing in the dishwasher. I'm also slightly nervous that a capacitor in the power supply has leaked, and is still leaking.
03-Nov-2007
I tried the system again, and immediately had flickering screens, ding-dongs, and various crashes!
In frustration, I removed the logic board again, and carefully cleaned all of the surface corrossion from around the various chips on the board using a toothbrush. I scrubbed everything clean of as much of the corrossion residue as I could possibly reach. I even tried to get under the ICs and memory sockets. When I had done all that I could, I tried re-assembled the system and tried it again.
No luck! Immediately, the same problems!
Okay, desparation had overcome me, so I removed the logic board and swapped it into my SE/30 #2 machine. Though I had checked the various voltages coming from the power supply, I wanted to swap the logic boards to try to at least isolate the fault to the logic board, not any other system component. Well, what do you know? In the SE/30 #2 system, I powered up and saw flashing screens, heard ding dongs and all the rest. I just randomly clicked the little reset button on the left-rear of the board and rebooted over and over. After about 10 tries of this, it worked! It booted right up and was solid! Since the system was on it's side and the logic board was not actually slotted into the frame, I could flex the logic board this way and that, trying to see if that made any difference. It did not.
After letting it run this way for about two hours, I moved the logic board back to it's original SE/30 #1 case, and powered up. It worked! I've let it run now for another three hours, with no problems.
It just occurs to me, now that I'm writing these notes, that maybe the reset or programmer's button(s) were/are flaky. I wonder if some of the mystery fluid has gotten into these switches, and randomly shorts the contacts, like it does in the floppy drive's disk sense plunger. Perhaps my rapid firing of the reset button cleared some of that away, for now anyway.
I've been thinking about how to clean the floppy plunger, and if I come up with something, perhaps I should try to clean the reset button, too. Maybe a syringe of water squirted at high velocity (pressure) would clear out the mystery fluid that's inside the switch/plunger housing, which would then dry with no residue. Hmm. That's an idea worth considering.
03-Nov-2007
With the system appearing to be stable, at present anyway, I got a fresh copy of MacTCP 2.0.6 from Apple's developer website and copied it over to the system. After unstuffing, it appears to be just a control panel item, so I copied it over (onto a previous installation that I guess I did) into the control panel folder. When I ran it, it already had my settings from before somehow remembered. I'm running as 192.168.1.81 for now. I tried to ping the system from my Linux box, but saw no replies. I'm guessing that the MacTCP doesn't implement ICMP packets. So...
I found my previously downloaded copy of MacHTTP, unstuffed it, and scanned the short readme files. Apparently, just running the application is all that it takes. There is no formal installation. So I did.
The system sort of hung there for about 10 or 15 seconds, apparently creating the options file (as I later learned), and then showed a log screen, with memory usage, and session activity. Using my linux box, I browsed to 192.168.1.81 and was presented with a webpage! So it is running! Just like that!
I fiddled a little bit more with it, just exercising it and discovered a good amount of documentation in the form of webpages. Too bad they don't have the installation portion as a text file. It would help a newbie like me. But I'm not a newby anymore!
04-Nov-2007
I disassembled the system, removing the logic board. A careful examination of the Reset and Programmer's switches caused me to rethink trying to clean them on the inside. I don't believe I could get enough water into the switch, or once inside, I don't think I could reliably get it out again. Soooo......
I heated up my soldering iron and removed the two switches from the logic board. I put a small label on each one ('R' and 'P') to indicate which was which, though they appear identical. With them isolated from other components, I tested them for continuity. They both look good, actually.
While I had the soldering iron hot, I also removed the floppy's plunger switch, after removing the drive from it's mounting bracket. With my voltmeter, I confirmed that it is a normally-closed switch.
Interestingly, once free, it worked fine. The plunger moved in and out without trouble, or sluggishness. Hmm... I looked closely and found that the plunger shaft was dry, and not coated in that mystery fluid as it was the last time I had the floppy out of the machine. I manually worked the switch rapidly, many, many times, and it just worked. Well, that being the case, I resoldered it into place, and I lubricated the drive with the Dupont Teflon stuff.
With all of that done, I wanted to see if the switches really were the culprit to my troubles. I reassembled the system enough to get everything together and cranked it up. I started it up at about 2pm and now, at 12:30am, it's still going strong. I'm planning to leave it running overnight.
05-Nov-2007
Well, it crashed...several times. I was still up working when I noticed the system crash at about 3am. It just locked up with the Pyro screen saver on the screen. I rebooted, by cycling the power.
Later this morning, I check it and it crashed with the desktop on the screen, and a bomb dialog. The clock was stopped at 7:05am. I cycled power and rebooted.
Sometime this mid-afternoon I noticed that the system had again frozen up, with just a blank, black screen. I cycled power and rebooted.
At 5:25pm it is still up, and I'm going to just leave it one continuously for days if necessary to see if 'cooking' will help.
05-Nov-2007
I found a page describing how to bless or de-bless a system folder. This allows me to have multiple versions of the System software installed in different folders, and switch between them by de-blessing the old system folder, and blessing the new system folder:
http://developer.apple.com/testing/docs/TNblessdebless.html
Basically, it says that a blessed system folder will have a happy mac icon. To de-bless it, remove either the system file, or the finder file. To re-bless the folder, put the system or finder file back.
That's easy!
19-Sep-2008
During the past several months I have played, on and off, with the system, working to get it running as a webserver. I still experienced random crashes and lockups, but it stays up for fairly long periods now, on the order of hours or days. So, I finally located a copy of MacHTTP and installed it to the system. It's simple to run, just launch it, but I have now discovered one element of the Mac System software that I really wish were'nt so.
Apparently, there is no way to 'auto-launch' a file upon boot, like one can do with MS-DOS's AUTOEXEC.BAT, or Windows' StartUp folder. I'm wishing I had such a facility, so that if I ever do get this system running more or less permanently as a webserver, I'd like to be able to simply cycle the power on it to reboot, and have it start the webserver automatically. I've searched online sources for anything that might perform that function, but can find nothing.
I decided to let the system run 24/7 and see how things would go. For about four weeks I left it on day and night, running the MacHTTP server, and saw crashes and lockups as before. Here is the record that I made of that experience:
BOOT HANG RUNTIME
================ ================ ============
2008-07-31 08:15 2008-08-03 00:03 2d 15h 48m
2008-08-03 00:24 2008-08-04 18:43 1d 18h 19m
2008-08-04 20:50 2008-08-06 08:25 1d 11h 35m
2008-08-06 12:24 2008-08-06 13:35 0d 01h 11m
2008-08-06 13:51 2008-08-06 14:02 0d 00h 11m
(unknown, x4 quick hangs in this period)
2008-08-06 20:35 2008-08-10 03:35 3d 07h 00m
2008-08-10 13:30 2008-08-11 00:52 0d 11h 22m
2008-08-11 15:32 2008-08-16 13:55 4d 22h 23m
2008-08-16 19:55 2008-08-17 10:34 0d 14h 39m
2008-08-17 11:59 2008-08-18 02:52 0d 14h 53m
2008-08-18 09:53 2008-08-20 20:33 2d 10h 30m
2008-08-20 21:28 2008-08-21 15:36 0d 18h 08m
2008-08-21 16:24 2008-08-27 19:31 6d 03h 07m
2008-08-28 15:21 2008-08-27 16:40 0d 01h 06m
So, I had some long periods where the machine worked fine, and others where it was mere hours (or even minutes on August 6) before the machine crashed again.
19-Sep-2008
Considering this recent experience of 24/7 operation, I'm not fairly convinced that a capacitor *inside* the system was the source of the 'mystery fluid', not something externall as I first thought when the system arrived. If this is true, then the capacitor would not be functioning as it should, though it might work sort of. That might explain the random crashes. Now, if I can isolate a bad capacitor inside, that could be the root of all of these difficulties.
03-Jul-2009
Over the last several months I've been thinking about this problem. In the end, after considering that the system initially arrived with a 'fishy' mystery fluid on the exterior and interior of the case, I've become convinced that this was the result of a burst capacitor. Perhaps it popped while the seller set up and tested the system for sale on eBay. Also, I recall trying the motherboard in my Mac SE/30 #2, and saw the same behavior. Thus, the fault seems to lie with the motherboard, not the CRT, or the analog board. Given these things, it would seem that a capacitor is at fault.
Looking closely at the two large axial electrolytic capacitors on the MB (C2 and C11), I can almost detect a small opening, a hole, near the negative lead as it enters the cylinder on C2. This is the 470uF 16V capacitor. I fired up my soldering iron and lifted one lead. Putting it on my multimeter, it reads as infinite resistence. I would have expected a short charging period, but it seems to act like an open circuit. I'm going to replace it.
The other electrolytic (C11) looks okay, but I'm considering replacing it to. It is a 220uF 16V capactitor, again with radial leads.
11-Jul-2009
I realized that I should test the capacitors, to know for sure if they were bad, before I took the trouble to replace them. I lifted one lead from each motherboard capacitor, attached various resistors and checked the RC time constant for each. They look okay. Both capacitors hold their charge very well (at 10V) for a very, very long time when charged and disconnected from a circuit. I conclude that they are just fine and I no longer consider them to be a factor in the random crashes. I resoldered the free leads to the motherboard, putting everything back to normal.
12-Jul-2009
As an 'at last' effort to isolate the trouble with the random crashes, I decided to swap motherboards between this and the SE/30 #2 system hoping that it would tell me if the trouble is in the motherboard itself, or something else like in the analog board or power supply. After making the swap so that this chassis had the #2 motherboard installed, I powered up. Everything looks good and I booted to the hard disk. I fiddled around a bit, launched Word 6.0 and then left the system with the Pyro screen saver showing the current-time clock. If the system crashes, I hope to have the time to record.
03-Aug-2009
About a week ago, after the first crash, I re-installed the ethernet card and re-fired up. The system over the past week or ten days has crashed twice, as it usually does, but these were pretty long stretches. Today, it crashed again, so I called it quits. During this period I have come to the very strong conclusion that the problem still lies with capacitors. Random reading online suggests that spurious and unexplained 'flaky-ness' of electronics is often explained by one of two things: a) bad solder joints, or b) bad capacitors. Since I DID HAVE the fishy smelling fluid (which was also described online), pointing to bad capacitors, and the system is functional otherwise, I've decided to replace all the caps, even though I tested the two larger ones.
There are four different types of aluminum electrolytics on the board:
1- 470uF 16V 85-degC axial (x1 - C2) 2- 220uF 16V 85-degC axial (x1 - C11) 3- 47uF 16V (e1/47/16V) SMT axial (x10 - C1,C3-C5,C7-C10,C12,C13) 4- 1uF 50V (Z1/1/50V) SMT axial (x1 - C6)
I sent an email to my friend Bill asking about the markings on Type 3 and 4. Hopefully the capacitance values will make sense to him.
Got word back from Bill, who says that the SMD caps are 47uF and 1uF as indicated by the second row of markings. I can go for that. Now to get the replacements ordered...
22-Aug-2009
A week or so ago I received from Digi-Key all of the capacitors needed to replace all old ones on the logic board. My friend Bill G. had kindly given me a head gun for surface-mount electronics work, so I really had no excuse for further delay.
Using the gun, I removed all of the old capacitors. It takes a while for them to heat up and become free, since most of the solder lead is underneath the capacitor, but eventually I got them all off, with only one damaged pad. That one, in fact, came completely off the PCB, breaking of from it's trace. Obviously I had applied too much pulling force before the solder had released. I scraped a little of the mask from the end of the trace, and should be able to solder to it without too much trouble.
I noticed that the one cap nearest the front of the PCB (by the ROM SIMM socket) released that 'fishy' smell when I heated it for removal. I bet that particular cap was the culpit that caused me all this trouble.
I ended up using my sharp-pointed soldering iron for installing the new replacement caps, since this seemed to work faster than the heat gun, and kept the heat localized to the caps leads. I found that fluxing the caps leads and pre-tinning them made the work much easier. Each lead extends beyond the case just enough to let me heat it.
I also replaced the two through-hole caps.
With that complete, I reassembled the system (leaving the back cover off, and fired it up. I crashed once after about three hours!
[August 23, 2009] It's now been running for about 26 hours since the one crash last night. I'll keep it going and we'll see what happens.
23-Aug-2009
Well, I finally hit upon a description of how to auto-launch applications when a macintosh is booted:
http://www.essentialmac.com/system/startup2.html
Basically, you just need to put the application (or it's alias) in the Startup Items folder which itself resides in the System folder. That's it!
26-Aug-2009
Well, as it turns out, even after replacing all of the electrolytic capacitors on the main logic board, the system is just as flaky as ever. In addition to the failure on August 22, just after the replacements were completed, here are the results of uptimes until a crash:
2009-08-22 21:08 - 2009-08-24 12:38 2009-08-24 14:35 - 2009-08-24 18:00 2009-08-24 22:05 - 2009-08 (screen blanked, so no clock visible) 2009-08-25 10:05 - 2009-08-25 21:37 2009-08-26 19:32 - 2009-08-26 23:32
It was plain that I had not corrected the problem by replacing the capacitors. It is possible that I didn't get a good connection on all of them, and maybe the location that had a bad cap before, still has a 'bad' cap because of faulty soldering on my part. My next step is to reheat all of the cap connections to make sure that I have reliable solder joints. After that? If the system is still flaky, I'll have to write it off, and look for a new logic board.
30-Nov-2009
Though I had previously swapped logic boards between my two SE/30s, I did it again to confirm that there were no new analog board problems. In the last day or two I have had very poor luck even successfully booting the machine, with the screen staying black, or giving zebra stripes, or bonging and then going black. So I wanted to make sure that it was still the logic board. It was.
After the swap, using SE/30 #2's logic board, this machine ran beautifully for 24 hours. No problems whatsoever.
Considering that this system has the hard disk with System 7.0.1 and MacHTTP on it, I either had to keep the swapped logic board, or swap hard disks instead, thus turning SE/30 #2 into the webserver I've dreamed of.
Once I discovered that the floppy drive is no longer responsive, that settled it. It is easier to swap hard drives and turn my attention to SE/30 #2.
01-Dec-2009
I returned the logic boards to their original chassis. Then I swapped hard disks. Two screws and its done!
20-Dec-2009
Well, I just couldn't leave it alone! Even though I had Mac #2 running with this units hard disk, I just couldn't stand to see a dead Mac around. From eBay, I got a replacement logic board, and today I installed it, and returned the hard disks to their respective machines.
The replacement works except that there is no audio, no sound at all. The small capacitor looks bad, and since I still had some spare caps I replaced it. Still no sound. That's enough, the machine doesn't need sound, though I hate to have a machine less than fully functional. Still, I want this unit to serve as a webserver, so sound isn't needed.
So, except for the sound, both machines are back to original condition.
Copyright © 2006-2025 Jared Blaser. All rights reserved.