problem with MicroMac accelerator in SE

[retr01]

There you go, @KennyPowers! Good job!

Try to get magnification. Got a phone with an excellent camera and a flexible stand to hold the phone that clips on the table? It could help. You got this!

Cheers,

 

[JDW]

Looks like a fun job without magnification.

Not sure one can do that kind of fix right without proper magnification. Honestly, I wish I could afford one of those amazing dual optic microscopes that Bruce and Mac84 use. Those are amazing, but so is the price tag! I don't do teensy tiny SMD work for a living, so I cannot justify that expense. What I have to make do with is the poor man's solution...

https://www.amazon.com/Microscope-Bysameyee-Magnifier-Rechargeable-Collection/dp/B07Q357P49/

But even that price isn't cheap enough if you only plan to repair a single trace on one PCB, in which case you may need to opt for a pair of magnifying glasses instead. Or if you have an iPhone mount, you can try @retr01 's suggestion.

 

[retr01]

What I have to make do with is the poor man's solution...

As with some Amazon sellers, I caught another misspelling and grammar issue, as seen on that screenshot: "WiFi Off/NO." "No" is palindromic of "on," but I wonder why the seller uses that word?

Nevertheless, it is indeed a nifty device!

But even that price isn't cheap enough if you only plan to repair a single trace on one PCB, in which case you may need to opt for a pair of magnifying glasses instead. Or if you have an iPhone mount, you can try @retr01 's suggestion.

Yup, this is what I use:

https://amzn.to/3nSqRPL

 

[KennyPowers]

Well, I punted on repairing this accelerator and put this cleaned and recapped SE back together without it. The logic board was pristine and retrobrighting went well...I think it came out nice:



I did order one of these for when I get around to attempting to repair that accelerator board, but I've got too many other systems and hardware in this lot to go through to get hung up on one thing. I think I've got enough Apple IIs here to put one in every room of my house .

 

[retr01]

Well, I punted on repairing this accelerator and put this cleaned and recapped SE back together without it. The logic board was pristine and retrobrighting went well...I think it came out nice:

Looks beautiful!

I did order one of these for when I get around to attempting to repair that accelerator board,

And the price of that item went down. Cool!

I think I've got enough Apple IIs here to put one in every room of my house .

Awesome! You can post those in the Apple II thread here on TD. We need more Apple II folks coming on here.

Cheers!

 

[retr01]

Hey @KennyPowers, I see that your SE has a 1.44 MB Superdrive. Does that mean BOTH of those 3.5" disk drives are 1.44 MB?

 

[KennyPowers]

Hey @KennyPowers, I see that your SE has a 1.44 MB Superdrive. Does that mean BOTH of those 3.5" disk drives are 1.44 MB?

Yes, they are. I tore both drives down, cleaned, lubed, and replaced that ejection gear that always strips in these (a process I've now performed on five various Apple 3.5" drives). This SE also had a working 40MB Quantum HD in it, but I replaced that with a BlueSCSI. It was getting cramped in there with two floppies and a fat HD squeezed under the CRT, and the BlueSCSI is just more convenient. I routed the SD card out the back on a ribbon cable for easy access:

 

[retr01]

Yes, they are. I tore both drives down, cleaned, lubed, and replaced that ejection gear that always strips in these (a process I've now performed on five various Apple 3.5" drives). This SE also had a working 40MB Quantum HD in it, but I replaced that with a BlueSCSI.

Cool. That is one thing I need to do with the SuperDrive in my SE/30. I cleaned it and used disk cleaner. It works fine now, but I need to do the lubing procedure.

Cheers!

 

[KennyPowers]

Now that I have better equipment and some trace repair experience under my belt, I think I'm going to tackle getting this accelerator working again. The consensus a year ago was there's a continuity break somewhere (probably due to leaky cap damage) that's keeping it from seeing a couple of the SIMMs.

Pretty much all of the vias under each capacitor look corroded. However, they all appear to go to either the power or ground planes (4-layer board). So, I checked continuity from each capacitor's solder pads to either the power or ground pin on one of the logic ICs and all had good continuity.

So that hopefully narrows it down to one or some of the corroded-looking traces around the board:



There are more like that, but you get the idea. So, time to start scraping solder mask and checking traces I guess

 

[YMK]

What's the DRAM part number on your SIMMs?

 

[KennyPowers]

What's the DRAM part number on your SIMMs?

Here you go:

 

[retr01]

Using Preview app, Apple helps us to get the numbers down.

AAA4М204J-06A
NPNX 9322 D6

 

[KennyPowers]

So I checked every iffy-looking trace and via, and they all have continuity across the iffy-looking parts. Also looked around the board under the microscope for bridged pins, etc and didn't see anything like that. I also ruled out a bad SIMM a year ago (swapped SIMMs with the SE logic board), and I verified continuity of every SIMM contact to the socket. So I'm thinking either cracked solder joint(s) somewhere or I did a crappy job soldering (and resoldering) the new caps a year ago. I have some new 25V 22uF polymer tantalum caps for this board (original electrolytics were 16V 22uF). I guess the next thing to do would be to install those and give it a try?

 

[YMK]

I also ruled out a bad SIMM a year ago (swapped SIMMs with the SE logic board)

I couldn't find anything on those DRAM parts, but it's possible they're fast enough for the SE, but not for the accelerator.

I'd try several sets of RAM before reworking the board any further.

 

[KennyPowers]

I couldn't find anything on those DRAM parts, but it's possible they're fast enough for the SE, but not for the accelerator.

I'd try several sets of RAM before reworking the board any further.

I know the previous owner of the SE/accelerator, and he was using it with those SIMMs (over 20 years ago). I do have a big box of 30pin SIMMs though, so I could try a few different sets anyways. I have to install these new caps anyways though because there aren't any on the board right now

 

[YMK]

The /OE, /CAS, /RAS, /WE and address lines are typically tied together between slots.

You can check continuity between those.

Depending on the design of the board, the RAM data pins may be tied to the 030 data pins.

If they are, finding breaks is fairly simple.

 

[KennyPowers]

The /OE, /CAS, /RAS, /WE and address lines are typically tied together between slots.

You can check continuity between those.

Depending on the design of the board, the RAM data pins may be tied to the 030 data pins.

If they are, finding breaks is fairly simple.

Thank you! The /CAS, /RAS, and /WE lines don't appear to be tied together between the slots on this card. The address lines appear to be, but I didn't check all combinations of all 12 address lines because your message made me have a "duh" moment. When I last tried to use this card, the error message on boot was:



So, I started checking continuity between each SIMM's data pins and the data pins on the 030. They all had continuity except for two...D7 on one of the rear SIMMs wasn't connected to D7 on the CPU, and D5 on the other rear SIMM wasn't connected to D13 on the CPU, which jives with the error message. I just didn't understand what that error was telling me back then I started working back from each SIMM slot to the CPU and eventually found a bad via on each of those two data lines. I've repaired them and now have good continuity from the SIMM back to the CPU. Going to wait on reinforcing the repairs with solder mask until I see it work though...fingers crossed!

 

[YMK]

The data lines between the SIMMs and 030 may not be ordered.

For example, D0 of a SIMM might be mapped to something other than D0, D8, D16 or D24 of the 030 for ease of routing.

The error messages suggest they're grouped by octet, but they could be scrambled within that grouping.

Or the designers may have kept the data lines fully ordered.

An address line failure could appear as a data error, since data might be written to locations other than the intended. The address space "mirrors" or folds over itself.

 

[KennyPowers]

The data lines between the SIMMs and 030 may not be ordered.

For example, D0 of a SIMM might be mapped to something other than D0, D8, D16 or D24 of the 030 for ease of routing.

The error messages suggest they're grouped by octet, but they could be scrambled within that grouping.

Or the designers may have kept the data lines fully ordered.

An address line failure could appear as a data error, since data might be written to locations other than the intended. The address space "mirrors" or folds over itself.

All of the other data lines were organized by octet and accounted for (I toned-out each one):

SIMM 1: D0-D7
SIMM 2: D8-D15
SIMM 3: D16 -D23
SIMM 4: D24-D31

The only two data lines that didn't have continuity to a pin on the CPU were the two I mentioned before, and I was able to visually follow those traces from the SIMM slot back to the CPU.

Ya, there could be an address line failure too, but there was an obvious problem with those two data lines.

 

[KennyPowers]

It's alive!