• Updated 2023-07-12: Hello, Guest! Welcome back, and be sure to check out this follow-up post about our outage a week or so ago.

zombie Quadra 610 mostly boots...sometimes

ChadVDR

Active member
I had to join to consult the 68k experts so I can help somebody relive some old memories.

This Quadra last ran just fine in 2015. It was fully functional then and was left in the garage.

Last week, I opened the Mac and about six surface-mount electrolytic capacitors were rolling around. I thoroughly washed with alcohol and scrubbed away any trace of the goo. The original Lithium battery was intact and looked new! I pitched it anyway.

I desoldered the remaining capacitors and cleaned all of the pads thoroughly and re-tinned them. I soldered tantalum capacitors in and I've checked their orientation probably ten times now. Most importantly, I checked that each capacitor leg was indeed making contact with the vias on the board. All of them check out.

I thoroughly removed the flux, rinsed thoroughly with alcohol, blew every drop of liquid out, and let it sit overnight and through the next day. Once I pressed power, I heard the bong and got a "?". Turns out that no matter what I do, the SCSI drives won't do anything more than wake up and then standby. They're not receiving any seek instructions. HDD, CD, even an external SCSI with internal termination wouldn't boot or seek.

I created a bootable floppy of the System 7 disk tools and hey! It boots! As soon as I put the floppy in, it gives me a happy Mac and boots into the utility. The graphics seem odd but I don't know if I can attribute that to the 10-switch adapter to HD-15 VGA monitor. I'm running at 66hz RGB 480 but it is very green and seems misaligned vertically. What sync mode and resolution etc. should be best to use when testing with this adapter? My monitor is a ViewSonic CRT from around 2001.

Pressing reset does nothing in any scenario. Pressing Interrupt within the disk tools bootup causes the crash but the crash doesn't load completely with any crash code etc. if that makes sense. Just an empty white window that appears.

Things I've tried:

With and without the new PRAM battery installed.
ADB keyboard with every thinkable key combination on start.
With and without any SCSI devices plugged in.
With and without any RAM installed and with each RAM stick individually and tested in each slot.

Does anybody know if a PSU could cause this? Could somebody with a similar PSU tell me what Ohms you get from all of the power pins to the PSU case (ground)? The pinout is like red, red, red, black, black, black, black, yellow blue. When it is running, all the voltages seem OK except that the black (I assume ground, right? They Ohm out to about 0ohm to ground with my meter) gives me about 2-5microvolts when the computer is running.

Here's a clue. If I run it for a while, it will eventually get less and less successful with the boot process. It will go from booting, to just reading the floppy long enough to get a happy Mac, to not reading the floppy at all nor giving a question mark, to finally not giving a cursor. If I let it sit for 15-20 minutes, it will succeed on the first try. When it is successful, artifacts or weird random pixels will appear as glitches here and there. Sometimes it will actually just freeze completely.

Pretty at a loss here. I'll try to get pics of my screen and board but hopefully this sounds like something somebody has experienced. I have never really troubleshooted with 68K so please forgive my utter ignorance with the terminology and conventions.

Thanks for reading and here's hoping I get this together in time for a very happy Christmas.
 

foetoid

Well-known member
It just so happens that I have both a fully functional 610 and a 610 that powers but won't chime/boot. Been meaning to do some diagnostics on the PSU. Will bust out the multimeter and let you know what I find.
 

foetoid

Well-known member
Here's the Ohm results, measured from the header on the MB.

Bad Machine (ASTEC PSU):
Pin1: 107.6
Pin2: 107.8
Pin3: 107.7
Pin4: 0
Pin5: 0.2
Pin6: 0.3
Pin7: 0.1
Pin8: 0.4
Pin9: 2.395

Good Machine (TDK PSU):
Pin1: 102.3
Pin2: 102.1
Pin3: 102.2
Pin4: 0.2
Pin5: 0.2
Pin6: 0.6
Pin7: 0.4
Pin8: 304.8
Pin9: 253.9
 

ChadVDR

Active member
Hey! Thanks for jumping right in with a very helpful post.

I'm slightly suspecting my PSU less at this point and I'll explain. For now, let's talk numbers. I have an ASTEC and my black wires all seem to run to the frame of the PSU. They all equal about 0.4ohm to frame and 0.4-0.5 between black wires. If you truly have a 0.00ohm reading on one of your pins, this may indicate a direct short bypassing some component. Are you reading Ohm or kOhm or ? because my reading are almost all in the kOhm range and rise as though charging capacitors.

Here's why I doubt the PSU is a major problem at this point and may have jumped the gun.

I pulled the board and figured it was time to prod each and every leg of every IC! Yuck! Just go ahead and try to count how many micro legs are on one bus controller. Well, I started in a corner close to a cap and...well I found several corroded legs that were making no contact at all. The chip is a TI NAND controller. I pulled it up and checked under it. The pads all cleaned up OK, though I did lift one in the process. I checked continuity on each of the pads and found they were just fine. Issue is the lifted pad. It doesn't seam to have any trace on the board. If a pad has no trace at all (leading to no VIA or other component) then is it just a "blank" that serves no purpose? Or (and I shudder here) are they direct to a ground plane and I have to find a way to ground the pin that is now without a pad?

Check the bottom left capacitor and to the right of that you will find U46.

Do you mind (I know I'm asking a lot here so I understand) tracing pin 2, 3 and 4 on that chip (assuming #1 is the white stripe on the chip) to see if they go anywhere at all? All three of these pins go to pads that seem to go nowhere and I pray they don't actually go to ground.

I know I'm on to something and will update as this develops. I keep seeing this topic with similar symptoms pop up so I hope this can bring closure to a lot of people and maybe they'll find a way to fix their IC legs.

quadra_610_logic.jpg
 

robin-fo

Well-known member
Last week, I opened the Mac and about six surface-mount electrolytic capacitors were rolling around.

I know from my 660AV that the solder from other chips close to the capacitors can also crumble, resulting in loose chips. Best check all neighbouring solder joints.
 

foetoid

Well-known member
If you truly have a 0.00ohm reading on one of your pins, this may indicate a direct short bypassing some component. Are you reading Ohm or kOhm or ? because my reading are almost all in the kOhm range and rise as though charging capacitors.

Sorry, it looks like I may have gotten a bad read off Pin 4 (see picture)

As far as pulling the board, that's a for tomorrow project. :) Will get back to you on that.
 

Attachments

  • PXL_20221213_062540323.jpg
    PXL_20221213_062540323.jpg
    2.4 MB · Views: 26

ChadVDR

Active member
I know from my 660AV that the solder from other chips close to the capacitors can also crumble, resulting in loose chips. Best check all neighbouring solder joints.
And just like that I have already checked every pin I can reach on the board and the total death toll is….five. The “IOSB” chip next to the cap just right of the NUBUS slot looks like crap but only five legs came up when tugged. I’ll have to pick up some baking soda etc. this week and then have a crack at micro-soldering. Shoot me.
Sorry, it looks like I may have gotten a bad read off Pin 4 (see picture)

As far as pulling the board, that's a for tomorrow project. :) Will get back to you on that.
No need to pull the board for this one. Just slide the hard drive forward once you unplug it.

Wonder if my issue will give you clues to your Mac. So far I could boot but would eventually hang and the problem appears to be one NAND and one what I assume is a SCSI or I/O controller.

This is quickly becoming no longer a mystery. More just a nightmare. I see bodge wires in my future.
 

lobust

Well-known member
If a pad has no trace and no via then it is not connected and is only there for mechanical purposes (something to fix the pin to so that it's not dangling in thin air).

I have a Q800 board, which shares common system architecture with the Q610, with dead scsi that I have not been able to diagnose. I was unable to find any broken trace anywhere on the scsi subsystem or upstream of the scsi controller. I replaced the scsi controller itself with no change, and also noted no change in behaviour with the scsi controller removed altogether. If you can get a copy of SCSI Probe on bootable floppy and run it you might find that it reports the bus is not terminated, as that's what mine does.

I have a spare Q800 board, and another one en-route that I bought recently on ebay, but don't currently have a functional PSU to boot them. For a long time I have intended to get a logic analyser and see if I can find some functional difference between the good board and the bad one, but I haven't had time for such projects thus far.
 

ChadVDR

Active member
The first part is good. The second part I don't like so much.

I suppose there'd be no harm in just clipping the parts of the NAND chip that go nowhere, eh?

Does anybody know what the U46 NAND does or what a NAND IC like this normally would do?

Do you think "IOSB" has to do with SCSI?

Assuming the SCSI controllers were removed completely, would I expect the crashing and graphical glitches even when booted to floppy?

Since you mention it, I will try the SCSI probe if I get it running reliably. It will boot from floppy but has to be stable.

I know it's a lot of questions but I'm on a bit of a time crunch and hoping for the best. I'll wash with baking soda tonight, solder the bits and then ultrasonic clean it as soon as I can get to my cleaner.
 

Attachments

  • disk tool.jpeg
    disk tool.jpeg
    116.5 KB · Views: 11
  • Happy.jpeg
    Happy.jpeg
    151.4 KB · Views: 18

lobust

Well-known member
The first part is good. The second part I don't like so much.

I suppose there'd be no harm in just clipping the parts of the NAND chip that go nowhere, eh?

Does anybody know what the U46 NAND does or what a NAND IC like this normally would do?

Do you think "IOSB" has to do with SCSI?

Assuming the SCSI controllers were removed completely, would I expect the crashing and graphical glitches even when booted to floppy?

Since you mention it, I will try the SCSI probe if I get it running reliably. It will boot from floppy but has to be stable.

I know it's a lot of questions but I'm on a bit of a time crunch and hoping for the best. I'll wash with baking soda tonight, solder the bits and then ultrasonic clean it as soon as I can get to my cleaner.

I don't know what U46 does on the 610, it's a simple logic gate ic, it should be fairly trivial to see what it's connected to and figure out it's function. I don't have a 610 unfortunately.

The scsi controller is the 53C96. IOSB is a multipurpose IO controller. A faulty IOSB could potentially prevent the onboard scsi from working, but I'd expect there to be multiple other IO subsystems also not working in that case...

Page 15 of the 610/650/800 developer note (http://www.retro.co.za/ccc/mac/Quadra650/Mac Centris Quadra Devnote.pdf) shows the system architecture block diagram.
 

joshc

Well-known member
Do you think "IOSB" has to do with SCSI?

From the dev notes for the 610:

IOSB IC

Like the MEMC IC, the IOSB IC combines functions performed by several ICs in
previous Macintosh designs. The IOSB includes:

- I/O data bus buffers
- a SWIM2 floppy disk controller
- VIA1 and VIA2
- address decoding for I/O devices
- sound circuits, including separate sound buffers for input and output

The IOSB IC provides the data bus services of the MC68030 that the MC68040 does not
provide. Those services are byte steering, which allows 8-bit and 16-bit devices to be
connected to a fixed byte lane, and dynamic bus sizing, which allows software to read
and write longwords to 8-bit and 16-bit devices. Those services allow the Macintosh
Centris 610, Centris 650, and Macintosh Quadra 800 computers to work with existing I/O software.

Here is the link to get the dev notes:

 

joshc

Well-known member
Cap goo can affect the legs of some ICs and it's possible that the SCSI controller (53C96) is faulty. A scope would help a lot with diagnosing what's going on with the SCSI bus on your board.
 

ChadVDR

Active member
Well I learned a lot from you guys today. Reading that post kind of clears up that the IOSB is important. It's no wonder none of the SCSI ports try to access data at all. It is also helpful to find that there are automatic terminators so you can run with no internal cable connected and boot externally and no need for an external terminator if no device is connected externally.

Still cleaning up the pads etc. It's going to be a very very very long and painful ride. No need to meter your U46, foetoid, since I've learned that the blank pads really do go nowhere at all.

I may have some of the chips completely restored by tomorrow. While I can't do as much work as I'd like due to the holiday being my deadline, I am encouraged that all but maybe five total legs (on the IOSB) and one chip (U46) need attention right now. The other legs, no matter how corroded, should at least be resilient enough to last until I can provide long and patient repair work on a then-known good computer. At that point, I will be desoldering many chips and doing proper permanent fixes. For now it's magnetic wire and "at least it has continuity".
 

ChadVDR

Active member
VICTORY!

Not only did it run just fine, it even booted from the old HDD. Thing was stuck and needed a bump against the desk to unstick it but it was enough to get it spinning. I was utterly shocked when I saw the OS 8 boot screen. Hasn't booted in seven years. It was also able to boot from floppy as usual and even from my raSCSI!

The IOSB had about five lifted legs and the U46 was completely removed, cleaned and put back.

Everything is working excellently except the persistent display issues. Does anybody have a clue what the next step may be? Has this error occured before? 90% of the display is OK but it has black pixels peppering everything and the green alignment is way off. I am using a DB15 to HD15 (VGA so they call it) connector with 10 dip switches. I've tried every mode of RGB and VGA and it only works with Composite Sync 1 and Separate Sync. Here is a pic of the issue and the weird dancing/squiggly menu bar.

Guys thanks for getting me here and I owe 610 users a pinout of the U46 chip and my other findings that I will upload.

When I do get to doing the long-term repair of this board, I plan to remove any and all corroded chips and clean and resolder them. When I take them off the board I will do my best to document and provide diagrams to help others in tracing etc.

Let me know what you guys think!

desktop.jpegjiggle.jpeg
 

AwkwardPotato

Well-known member
Do you have another monitor you can test with? Judging by how the area of the screen outside the raster is also too green, I wonder if the green drive/cutoff controls on the monitor are out of whack. Not sure if you can adjust those in the on-screen menu on a relatively late CRT like that though. Some Macs are picky about the sync-on-green setting on the VGA adapter, but I'm fairly certain that the 610 isn't one of those.

As for the bad pixels: could possibly be rotten traces between VRAM (U41-44) and the RAMDAC (U10), or it could be that the RAMDAC itself is faulty. A few months ago I had a Q700 with practically the exact same fault and it was indeed the RAMDAC that was gone. At that point you're pretty much SOL unless you can find a new motherboard, or a donor LC475/575/Q650/800 mobo to steal the chip from.

If you think it would be useful in finding any bad traces, and I can find it, I wrote down the pinout of the RAMDAC somewhere while troubleshooting that 700. Not the exact same part as what's on these boards but the pinout & board layout seem to be mostly the same between the two.
 

ChadVDR

Active member
If you happen to have the pinout, it can only help at this point.

I do have other monitors but they will be LCD and I know those are picky about resolution and sync but I will try. If this CRT works fine with all my other computers, does that narrow it to the motherboard as the issue? This monitor has an attached video cable so that’s not the issue as I get a perfect image with all other Macs. Then again, would I even know what type of sync those are using and would it make a difference? I mean, if it has this issue when sync is set to Separate, do you think it would go away if I had a monitor that could Sync on Green? Or if the issue appears on Separate sync but a modern Mac syncs on V or H that would explain why those are fine and this one isn’t?

Just to be clear, I hope nobody thinks the image I posted is my board. That is from the recap a Mac site. Is that what lead to your judgement about the VRAM circuit?

Edit: I just found the manufacturer spec for the chip which includes the pin out for the chip itself, but if you can tell me how it connects to the VRAM then that is still more helpful.

The specs say it’s a SCSI controller. Forgive me for being ignorant of 68k macs but isn’t the video done on CPU and then what path does that take to get to the video out?
 
Last edited:

ChadVDR

Active member
Correction: the chip NEXT TO the RAMDAC is a SCSI controller heh my mistake. So I look forward to any info on the RAMDAC. I did a quick tracing and the DAC and the small IC's below it seem to at least be attached but I fear what you said about it possibly being borked internally.

This repair is 99% complete. I've been playing a few games on the old HDD and it hasn't caused any issues at all. I'm thinking it makes more sense to start a new thread about this final issue and I'll come back to this one with a detailed write-up of the symptoms and solutions with diagrams.
 

ChadVDR

Active member
You guys did it!
68kchristmas.JPG

I was able to give somebody a 1994 Christmas today. The Quadra was recapped, a few IC's repaired, the shell was bleached in hot peroxide and brand new feet attached. The only flaw is a bit of snowy pixels that will be fixed thanks to lobust and AwkwardPotato pointing out how the VRAM functions.

In Gray mode, the machine performs absolutely flawlessly. All three internal drives plus the BlueSCSI perform without a single glitch. Audio is great too. ADB devices work perfectly. The only remaining issue is the VRAM IC rot that will be re-soldered after cleaning. I'm actually surprised just how quietly this machine works. I can only barely hear the PSU fan over the drive and the drive sounds just as I have always remembered big old spinning platters.

I can't tell you how good it feels to see this setup and I still owe it to you to post my findings and circuit traces since no diagrams of this board seem to exist in the public.

Merry Christmas!
 
Top