jmacz journey

[jmacz]

Right, but there's a switch inside the jack that disconnects the speaker when headphones are plugged in. I don't think it's impossible for that switch to get intermittent and cause symptoms similar to what you're describing.

What I'm going on is that when that jack is removed, there will be no sound from the speaker.

So played with the headphone jack and it seems to be functioning properly. With headphones plugged in, the chime is routed to the headphones, with the headphones not plugged in, the chime is routed to the speakers. Both however exhibit the same problem which is that the chime is staticky, low volume, and fades to no sound and then after that, no more sound until the next reboot. With the headphones however, it's clear that there is power as you get a slight hiss but again, no sound.

But, I had an interesting observation... after boot up, again there's no sound, I'm using the Sound control panel and as you change the volume, there's no sound coming from the speaker or headphones. BUT, there's something weird. Normally when playing with the Sound control panel, you get the beep sound (at the sound level you choose) but if you choose 0 (ie. mute), the menu bar blinks instead as an alert. So choose sound levels 1-7, you get the sound at the proper volume, but choose the lowest option, you get mute and the menu bar blinks instead. That's the normal expected behavior. BUT on my SE/30, it doesn't matter what sound level you choose, the menu bar always blinks. That tells me even the OS is detecting something wrong. The System doesn't do this menu bar blinking based on the speaker being disconnected or not (I just tested this on another computer). It's supposed to only blink when at no volume. But on this SE/30, it's blinking on any volume setting.

RE: Floppy vs SCSI - put System 6.0.8 on a SCSI device and booted and that worked fine. So it's not the SCSI subsystem but something in System 7+ is exposing a hardware problem on boot whereas System 6.0.8 boots fine and doesn't expose the issue. (I had a typo in my message earlier in this thread -- not sure why I typed 6.1.8).

 

[jmacz]

Moved the SE/30 problem discussion to its own thread to avoid cluttering this one. https://68kmla.org/bb/index.php?threads/se-30-sound-and-boot-freeze-issues.46398/

 

[jmacz]

Finished my second Design 68 keyboard:



Went with white keycaps again.. different style. Was looking for black keycaps with white writing, and a narrow font like the above, but couldn't find one. This particular keycap set didn't have a good choice for the power button keycap so went with the little dog next to the cmd key.

 

[jmacz]

As described in the other thread, my SE/30 is fixed and working. There was a broken data trace that I finally found. It also had a physical issue, cracked corner on the front case. I had to do some body work on that corner:



Basically auto body repair work ... fill it, sand it, paint it, etc. The white chunks were missing. Here is the corner after refinishing:



And here's the final completed SE/30:



I'm happy with how it turned out. I need to look into the memory gotchas on the SE/30. I put 16MB in 4 sticks into it and I'm running Mode 32. But need to understand what the differences are with mode 32 and replacing the ROM with a 32bit clean one.

 

[jmacz]

I've been slowly working on the Brainstorm 512K I got a month back. It ran into a shipping issue when the previous owner got it, and the end result was a damaged case as I posted earlier:



Inside, the machine had two obvious upgrades done to it: 1.) Brainstorm memory upgrade and 2.) SuperMac SCSI upgrade. After looking at these further, the memory upgrade takes the machine from 512K to 2MB. The SCSI upgrade comes with Mac Plus ROMs and provides an internal 50 pin connector. Both upgrades are working. The memory is present and I tested the SCSI using a ZuluSCSI (MacPlus mode enabled) and it worked great. The floppy wasn't working so I went to clean it (and replace the usual gear) and found that the floppy drive was also upgraded from the usual single side 400K drive to an 800K drive. After cleaning, replacing the gear, cleaning the head, and tuning the zero track alignment, the floppy is working great. I left the analog board capacitors alone for now but will circle back and replace those at some point.

Now getting back to the case, the plastic is pretty thin where it shattered so there wasn't much original material for any new filler to adhere to. So I had to put more plastic weld, even on the front face (for more strength) and hope to blend it in. Here's the original glob of plastic weld with some fine sawing to get the basic shape:



Not much to look at yet but with some more grinding with a dremel, sanding, some progress:



You can see the cracks near the monitor also filled in. And then some more sanding:



There was even more sanding after the above... but the final result of that corner:



and the rear:



And then there was this HD20 that I received along with my Mac Plus that had been painted black:



That was also refinished to match my 512K. With the combined system looking like this:



Kinda not sure what the point is.. as @bigmessowires pointed out, this 512K with the ROM upgrade, SCSI upgrade, memory upgrade, floppy upgrade, it's basically a Mac Plus (which I already have). I'm not even using the SCSI upgrade as I've got this HD20 plugged into it via the floppy drive. Does this mean I'm still looking for a vanilla 512K? maybe? Don't know. Not sure how many of these Brainstorm/SuperMac upgraded units are out there in the wild still, but I guess it's something different and so I'll keep it.

 

[jmacz]

Found some time to start debugging the Applied Engineering TransWarp Mac SE II Accelerator TW1340 I was lucky to find inside a Macintosh SE I got a few weeks back.



I mentioned earlier the board was very clean, and it really is. I put it under the microscope and I could not find a single thing wrong with it visually. I toned out the main SE logic board to ensure that all the pins to the PDS connector were good. They were. Then I started toning out the connections on the accelerator board and at least the first few things I checked looked ok. I had mentioned I put a working FPU on the board as well as it was missing. That didn't help either.

As I continued toning things out, I was thinking about the memory test I performed. I had replaced the SIMMs on the accelerator with known working ones and it didn't help... but did I fully check them. So I took the 4 SIMMs that came on the accelerator and put them into my IIci to see what I would find. All 4 of them seemed good, passed memory tests, but.... they were 2MB simms (for a total of 8MB). I remembered not all systems can use 2MB simms. I went back to the "known working" SIMMs I tried in the accelerator (which didn't work)... also 2MB simms.

So I grabbed a set of 4MB simms (16MB total) I had lying around and placed those on the accelerator. Works!!!

The accelerator is working and the SE is recognizing the memory on the accelerator as expected.

Now I'm not sure what I'm going to do with this as I'm in a similar dilemma as I have with the Brainstorm 512K. My intent was a normal SE, because I already have an SE/30. With this accelerator, it's more like having another SE/30. I guess it's a good problem to have.

 

[jmacz]

Actually scratch that... it couldn't be that simple and it wasn't. Reproduced the issue with the 4MB simms. And then got the accelerator working with the original 2MB simms. So not a memory thing. Most likely a solder joint somewhere as nothing else is changing except changing memory modules so clearly some flex is exposing a bad solder joint somewhere on the accelerator. Same issue happens on another SE I have so most likely the issue is on the accelerator board. Fun stuff.

 

[mdeverhart]

Might also be worth checking those socketed parts, see if the contacts need to be cleaned or de-oxidized. Seems like flexing could cause them to seat in a way that does or doesn’t make contact properly, and I know others have had problems with sockets on other boards.

Edit: Or the SIMM sockets on the accelerator themselves

 

[jmacz]

Might also be worth checking those socketed parts, see if the contacts need to be cleaned or de-oxidized. Seems like flexing could cause them to seat in a way that does or doesn’t make contact properly, and I know others have had problems with sockets on other boards.

Edit: Or the SIMM sockets on the accelerator themselves

That is a great point. I just used deoxit on both the simm sockets as well as the PDS connector. Can't repro the issue now. Keeping my fingers crossed.

 

[jmacz]

Some progress, after boot I'm able to muck around in the Finder, copy files, etc... but every so often when I run a program, it crashes with a "Memory Manager Error". I'm guessing my issue is with the memory. Will need to test the simms some more, check the sockets for cracked solder joints, and tone out all the memory lines on the accelerator card. Again, absolutely zero issues with the accelerator removed.

On another note, running TattleTech shows a 40Mhz 68030 and it's recognizing the FPU I installed. 40Mhz!

But I did find this post on Apple Fritter from back in 2015:

I souped it up with an Applied Engineering TransWarp '030 board running at 40 MHz. It was like turning it into a IIfx! More interesting, in retrospect, is what it did for my RAM situation. The SE is natively limited to 4 MB. System 7 had been out long enough that a lot of new apps were too RAM-hungry for a 4 MB box to thrive and make use of multitasking. I stuck with System 6 (in mono-Finder mode) but more and more apps required 7 so that wasn't a good long-term strategy either. The AE TransWarp board had its own separate array of 4 30-pin SIMMS slots, each of which could take up to 4 MB, at least in theory.

In practice, when the board first shipped, Applied Engineering warned me that until they ironed out some bugs and released a patch, I'd be limited to 4 MB. I bought it anyway, AND optimistically shelled out for the full 16 MB, which was more expensive than the accelerator board itself! I'm not sure that's a good comment on my judgment, but I did it.

It took something like 18 months but eventually they sent out an email about getting the update to be able to use the full 16. It wasn't a software patch; I had to remove the accelerator board and mail it to them and they'd alter it physically in some unspecified fashion and mail it back.

It worked! Not only did I have 16 MB of RAM on a Mac SE, I had 16 MB of RAM even under System 6! The largest contiguous chunk of RAM was 4 MB (or lower if I launched enough apps to eat more than 12 MB). That gives some clue as to how they did it. But I had a blisteringly fast little toasterMac with enough RAM to thrive under 6.0.8 (with MultiFinder, finally) and, 7.1.

I don't if my board had this change or not... whether a board without this change would only show 4MB even though more had been installed? or whether you could install it but the system wouldn't work or be stable if you went over 4MB. And looks like this patch required a hardware change. I guess I can try putting 1MB simms in there and see if it's any more stable?

If it isn't any more stable with 4MB of memory, then it sounds like I may have an upgraded board with some fault in it (again, probably memory related).

 

[olePigeon]

@jmacz Holy cow those restorations turned out great. I don't think I would have even attempted the putty approach. Can I ask what paint you used? I have a beige plus with such color discoloration, that no amount of Retr0brite would fix it. It needs a new coat.

Edit: Also, something I read about that you could try in the future if you get another broken Mac and need to putty some stuff: Take some clay and press it against the case to pick up the texture. Then after the final sand, apply one more thin coat of epoxy putty, then use the clay to apply the original case texture.

 

[jmacz]

The color is way too bright, I needed to find something more beige (more of a brown tone) if I were to do it again, so just wanted to throw that warning out there but for the 512K it was a Krylon Satin Ivory color that I used.

 

[jmacz]

@jmacz Holy cow those restorations turned out great. I don't think I would have even attempted the putty approach. Can I ask what paint you used? I have a beige plus with such color discoloration, that no amount of Retr0brite would fix it. It needs a new coat.

Edit: Also, something I read about that you could try in the future if you get another broken Mac and need to putty some stuff: Take some clay and press it against the case to pick up the texture. Then after the final sand, apply one more thin coat of epoxy putty, then use the clay to apply the original case texture.

Ooh, that's a good idea!

 

[jmacz]

More on the SE/TransWarp accelerator.

When booted, the memory total matches what I have on the accelerator board. For fun, I decided to remove the memory on the main logic board. That did not work. The system will not boot at all without any logic board memory. It's still used in some capacity at startup.

I decided to remove the accelerator and fully test the logic board memory simms using MacTest SE. That didn't go so well because I don't have the SCSI loopback device -- the test won't proceed very far without it. So I tried MacTest Pro Emergency as well as Snooper 2.0. MTP Emergency crashes during the RAM test with the accelerator removed. Possibly something wrong with the main logic board memory simms. So removed those and replaced with another set. Now it's stable running the RAM tests. I left it looping for a while and it's doing fine.

So I had faulty main logic board memory. I need to rerun memory tests with the accelerator installed and new main logic board memory.

As of now, the system is stable with the accelerator but only if 4MB is used. With 8MB or 16MB, I get all sorts of error including "Memory Manager Error", random crashes, unimplemented trap errors, MacsBug reports crashes during memory move instructions. Looks like I am limited to 4MB.

Summarizing the changes so far:

• Main Board Memory - replaced as it won't pass MacTest Pro ram checks with the original memory.
• Dexoit + Cleaning on SIMM/PDS Connectors - no longer hanging on startup with garbage on the screen.
• Limit to 4MB of Accelerator Board Memory - the card came with 8MB on it (2MB simms) and that was clearly unstable; 16MB (4MB simms) also was not stable. 4MB (1MB simms) is stable.

That quote I posted earlier suggests the card needed physical changes to support more than 4MB and also had bugs so doesn't look that really worked or was done to this particular card even though it does recognize more than 4MB.

 

[jmacz]

SE with TransWarp Accelerator is working great now. Doesn't look like the card supports more than 4MB. It recognizes 8 and 16MB, but hits issues when using that much. At 4MB, everything's great. Had it loop through hardware/memory tests for hours today and at 4MB, no issues.

68030 with FPU @ 40MHz. It's fast. Limited to 4MB, but that's fine.

Remaining projects in the queue:

• SuperMac Spectrum 24/V repair, has a broken inductor and a broken nubus connector. If I get it working, it will add to my stable of SuperMac Spectrum cards (24/III, 24/IV, 24 PDQ).
• Apple Multiple Scan 17" Trinitron Monitor repair, screen isn't stable (keeps changing size on every boot), missing front button door. Will try to figure out what's going on and while in there, probably recap the thing.
• SuperMac Spectrum 24 PDQ lockup running SuperMac 1024x768 @ 75Hz mode only. Need to get back to this.
• 2.3GB Fujitsu Magneto Optical Drive unable to read 2.3GB disks.
• Apple 400K External Floppy Drive has read issues and needs usual maintenance.

Almost there. Gotta keep from growing the queue again.

 

[jmacz]

Ok hmm... I had previously posted in this thread about my Mac 512K with Brainstorm memory upgrade and SuperMac SCSI interface. At the time I had mentioned I was searching online for more information and only found 3 other references to 512Ks with the Brainstorm memory upgrade.

Here's the picture of my board after cleaning:



Here's one of the 3 that I had found online, and I had found this picture in a classified listing from the UK (from years ago):



I obviously did not look at this image closely enough. I was just looking back at these pictures and realized, hmm, that board really does look just like mine... wait... the serial number... it's the same board!!!

Unfortunately, I cannot find the original listing anymore. It's not coming up in searches. But looks like the machine I now have in my possession came from the UK! How interesting.

 

[MicroCoreLabs]

I've been slowly working on the Brainstorm 512K I got a month back. It ran into a shipping issue when the previous owner got it, and the end result was a damaged case as I posted earlier:

View attachment 68161

Inside, the machine had two obvious upgrades done to it: 1.) Brainstorm memory upgrade and 2.) SuperMac SCSI upgrade. After looking at these further, the memory upgrade takes the machine from 512K to 2MB. The SCSI upgrade comes with Mac Plus ROMs and provides an internal 50 pin connector. Both upgrades are working. The memory is present and I tested the SCSI using a ZuluSCSI (MacPlus mode enabled) and it worked great. The floppy wasn't working so I went to clean it (and replace the usual gear) and found that the floppy drive was also upgraded from the usual single side 400K drive to an 800K drive. After cleaning, replacing the gear, cleaning the head, and tuning the zero track alignment, the floppy is working great. I left the analog board capacitors alone for now but will circle back and replace those at some point.

Now getting back to the case, the plastic is pretty thin where it shattered so there wasn't much original material for any new filler to adhere to. So I had to put more plastic weld, even on the front face (for more strength) and hope to blend it in. Here's the original glob of plastic weld with some fine sawing to get the basic shape:

View attachment 68162

Not much to look at yet but with some more grinding with a dremel, sanding, some progress:

View attachment 68164

You can see the cracks near the monitor also filled in. And then some more sanding:

View attachment 68163

There was even more sanding after the above... but the final result of that corner:

View attachment 68166

and the rear:

View attachment 68167

And then there was this HD20 that I received along with my Mac Plus that had been painted black:

View attachment 68168

That was also refinished to match my 512K. With the combined system looking like this:

View attachment 68169

Kinda not sure what the point is.. as @bigmessowires pointed out, this 512K with the ROM upgrade, SCSI upgrade, memory upgrade, floppy upgrade, it's basically a Mac Plus (which I already have). I'm not even using the SCSI upgrade as I've got this HD20 plugged into it via the floppy drive. Does this mean I'm still looking for a vanilla 512K? maybe? Don't know. Not sure how many of these Brainstorm/SuperMac upgraded units are out there in the wild still, but I guess it's something different and so I'll keep it.

That hard drive cleaned up nice! How easy was it to get the black paint off?

 

[jmacz]

That hard drive cleaned up nice! How easy was it to get the black paint off?

Used IPA to get it off the cable. On the case, used thinner to get as much off as I could, then reprimed it and painted it.

 

[olePigeon]

68030 with FPU @ 40MHz. It's fast. Limited to 4MB, but that's fine.

I think you need Compact Virtual for that (3.0.2 version available on the Garden.) It'll convert any RAM in excess of 4MBs into a RAM disk, then use the RAM disk for virtual memory, thereby cleverly bypassing the 4MB RAM ceiling.

Only problem is that I don't know if the commercial version is the same as the version that shipped with the myriad of accelerators.

 

[jmacz]

Interesting, thanks for the tip!