Click here to see it done: Scelbi Layout in 14 Seconds
Archive for February, 2012
The memory chip replacement I mentioned in the previous post was routine and got the motherboard up and smiling.
I figured that since I had the machine apart, I should clean and grease the floppy drive. This turned out to be an adventure that took a good part of three days.
Since I had never successfully restored a floppy drive to working condition in a couple of previous attempts on more modern macs, I was a bit concerned. First I found a couple of web sites that described what to do, or so I thought. I followed the directions cleaning some of the gunk off of the worm gear and the mechanism and then lightly greased it. However when I tried to boot the machine, it barely worked. I was able to boot the machine only two times between countless failures. Once booted, it seemed stable, so I was fairly certain that the problem was in the drive, not the motherboard. At this point I disassembled and cleaned the drive, some more. This repeat effort also resulted in a drive that barely worked.
At this point, I really studied the mechanism and determined that a clean worm gear is critical to proper head positioning. A dirty worm gear or worm gear follower could really mess positioning up. As it rotated, any grime or dried grease could shift the follower forward or backward a bit, ruining head positioning. I went back and cleaned more gunk out of the gear and removed piece of grime that had attached itself to the follower.
I also examined the circuit board to see what kind of debugging or calibration was possible. The parts all seem to be off the shelf parts, that should still be attainable, and their data sheets are on line, so I felt comforted by that. There are only two two calibration points. One is the read amplifier gain, and the other is the differential input phase adjust. There is no adjustment for the motor speed control, which is somehow controlled by software, anyway. For now, I left both adjustments alone. I did measure resistance of the adjust pots, so I could put it back to factory settings, in case I decided to change either setting later on.
I also partially disassembled the sliding mechanism and really cleaned it out, which you can’t do without disassembly. The red arrows in the image denote slip rings and a spring that need to be removed in order to do this. Don’t do this in a carpeted area, since you may never find a dropped slip ring in a carpet. At this point the mechanism worked as smooth as new and I thought I had a chance for success.
This time, rather than reassembling into the carrier housing, I just attached via the ribbon cable and used some paper to insulate the drive from the chassis. Once again, the drive didn’t work. Somewhere around this time, I placed the drive upside down outside of the chassis in order to get a better view of things. Lo and behold, it booted several times in a row. I played with the machine for a while before going to bed.
The next day, I moved the drive back into the chassis, just prior to reassembling and it started behaiving badly again. Moving it back out of the chassis, it started working again. Maybe it was sensitive to position, since it was on it’s back when outside the chassis and in it’s normal position when inside the chassis. Moving the drive right side up, and setting it carefully into the chassis, I immediately encountered more errors. Trying different positions only revealed that it didn’t work when in it’s normal position inside the chassis.
I was stumped. This drive is very well built, I didn’t see anyway in which gravity could affect operation. In fact, it is so well built that I thought that it would be very unlikely that it could ever get out of mechanical adjustment. Do to this, I resisted the great temptation to fool with mechanical adjustments. I know there is a website that suggests raising the head could help things. Based on what I saw, I figured that that was unlikely to significantly help matters, especially when the problem for that person really could have been a dirty worm gear or follower throwing the head positioning out of kilter.
I even got out my scope and checked the phase adjust test points. The phase relationship between the two signals was dead on. I thought it was unlikely that there was a problem with the circuit, though gain could still be off. I had no real way to know what gain should look like, so I left well enough alone.
I put things back together and tried some more experiments with positioning the unit, only to find that putting it in the chassis brought bad performance.
Finally, it struck me, the CRT was right next to the floppy when I had the floppy in the enclosure. In addition, I had removed the shielding/mounting bracket for my testing. EMI from the high voltage CRT was screwing up the drive read and write capability! Loosely putting the mounting bracket over the drive to shield it from the CRT, quickly changed performance from awful to normal! I had solved the issue.
A couple of lessons learned about Mac 400K drives should be shared here.
The good news about this, is that I believe that these mechanisms are extremely well built and should last a long time with a bit of careful maintenance.
I really don`t need another retro project, but I managed to pick up an all original 128K mac this weekend. I got it from a local source for a fair price. It does need a DRAM repair, but I have original mac 128K drams left over from a 512K upgrade that I did on my original unit back in 1984! I`ve really been wanting to get one of these computers. The original unit I bought back in 1984 was so upgraded, hacked on and otherwise beat on, it was beyond reasonable restoration. I had to turn it into a macquarium. On the other hand, the unit I just picked up, appears almost completely original and hardly used. I still have my original 1984 imagewriter, so this is going to be quite a retro trip. I already found a couple of books I have saved, including Doug Capp`s “Macintosh”
It’s been a while since I found an Apple 1 program that didn’t run on the Brain Board/Wozanium, but there is a new program out for the Apple 1 that uses the ACI for audio out.
Copied here, is my post on the Applefritter forum, describing the fix.
This program will not work as is, with versions Brain Board/Wozanium pack firmware less than version 5.2. There is an issue with the the gosubs 750s in the 900 range (at the end of the program). Those gosubs call directly into the original Cassette interface driver to generate sound. This driver is still present in the Wozanium PROM, but is incompatible with Apple II hardware and will crash the system.
There are several fixes available.
1) remove the call 750s at the end of the program. LInes 900 and 910.
2) Change the pokes in line 999 to call the Wozanium A2 cassette Driver at 0xd0d7:
999 POKE 750,169:POKE 752,32:POKE 753,215:POKE 754,208:POKE 755,96:RETURN
3) If you have a PROM burner, fix the Wozanium cassette driver to work with the A2 hardware by erasing a bit in the PROM – change location 0x01EA in PROM from 0xBC to 0xAC.
4) If you don’t have a PROM burner, send your board or prom back to me and I’ll erase the bit for you.
With fixes 2, 3 or 4 in place, you can attach a preamp to the cassette output of the Apple 2 and get the same sounds as if you were running on an actual Apple 1. It’s kind of funky because the A2 has a built in speaker, which is bypassed, but it works.
If you are running a Wozanium PROM image in an Apple II emulator either:
1) remove the call 750s at the end of the program. Lines 900 and 910.
2) change line 999 to call the Wozanium ACI driver at 0xd1d7:
999 POKE 750,169:POKE 752,32:POKE 753,215:POKE 754,209:POKE 755,96:RETURN
Finally if you don’t have a preamp for the cassette output, there is an inexpensive design on my website that will work for this application.
I’ve done several online searches for a complete MM5740 data sheet over the last several years. There is a version that can be found on many of the data sheet archive search engines, but it is missing at least one page, containing crucial information. I knew it was part of National’s MOS/LSI data book from the mid ’70s, but I couldn’t locate an online version of that book, and I really didn’t want to buy a physical copy. One of the key questions that couldn’t be answered with the existing online data sheet, is what exactly is the difference between the AAC/AAD/AAE/AAF versions.
Finally while exploring the possibility of using an AVR as a replacement decoder for the MM5740, I did another search and the data book turned up as a PDF at bitsavers.org. The last page of the data sheet show that the AAE version has N-key rollover and the AAF version has 2-key rollover. Otherwise they are the same. This means that the AAF version should be usable in the Apple/Datanetics keyboards, only with the loss of the n-key rollover feature. Also, in this data book are some app notes, that provide information on the AAC and AAD versions. Those versions decode the matrix differently, in order to make using an external PROM decoder simpler. They will not be compatible with Apple/Datanetics keyboards.
Once I figure out the best way to cut it out of the complete data book, I’ll put a copy of the complete MM5740 data sheet and associated app notes up on one of my web pages. For now, you can try to download the complete National MOS/LSI data book from bitsavers.org. Recently, I’ve have trouble accessing that site, but if you are persistant, you should be able to get through.
This is NOT a rev 0 system. Is there really a chance Dan could sell this?
This small section looks half-decent after the first pass, but there is a lot of tedious tweaking to follow. The rest of the board, isn’t even close to this standard.
The CPU board has been a lot of work. The next SCELBI board I’m going to layout will probably be one of the easiest, which is the front panel board.
If anyone ever said that these early micros were simple, they were wrong.
This board, alone, has 27 ICs, compared to the Apple 1’s 57 (including 16 DRAM chips on the Apple). However the complexity is really revealed by the number of vias. The SCELBI CPU board has about 240 versus the Apple 1’s 296.
For each replica PCB board layout, I go through 4 main phases. Placing the major components is phase 1. Connecting the traces is phase 2. Next phase of this board will be aligning traces with an image of an original board. The last phase is an exhaustive checking phase. This is only the first of at least five boards I need to layout before I can power up a replica SCELBI.
An image of an original SCELBI stock certificate was sent to me about a week ago, by early SCELBI employee, Frank Zawacki. More about Frank and the image can be found on my SCELBI page.
Wendell Sander sent me an incredible scan of an original photo used in an early Apple 1 ad. A lower res image of this scan has been added to my registry page.
Also, earlier today, I was lucky enough to have a brief chat with Jon Titus of Mark-8 fame about an Apple ][ project I have had in the works for a while. Jon seems like a great guy and was very accommodating. Hopefully I'll have time to make more headway on this Apple ][ project, so I can share the details with readers.