Altair 8800c build thread

Discuss construction, troubleshooting, and operation of the Altair 8800c computer

Re: Altair 8800c build thread

Postby Wayne Parham » June 17th, 2022, 8:04 am

AltairClone wrote:Wayne, try these changes on the FDC+ and rev H 2SIOJP in the configuration for running BASIC. See if it changes anything about where you can place the 2SIOJP in the motherboard.

FDC+
S1 Ram - all zeros
S2 Rom - 0100 0000 (PROM disabled, RAM ends at C000h)
S3 Drv - 0000 0<drv>, so 0000 0100 for 330Kb or 0000 0101 for 1.5Mb

On the 2SIOJP, disable both the “status disable” and the “phantom disable” options. After making these changes, let me know if this changes anything about board placement options.

If the above changes make no difference, then move J18 from “8080” to “Z80” and see what that does.

Mike


Will do. I'll be able to try this later this afternoon and will report back.

Not sure if you and Martin discussed this but he emailed me last night and asked that I do the exact same thing.
Wayne Parham
 
Posts: 239
Joined: March 18th, 2022, 3:01 pm

Re: Altair 8800c build thread

Postby AltairClone » June 17th, 2022, 11:36 am

Wayne, yes - Martin and I were talking about this last night. Keep us posted!

Mike
AltairClone
Site Admin
 
Posts: 631
Joined: April 5th, 2013, 10:55 am

Re: Altair 8800c build thread

Postby Wayne Parham » June 24th, 2022, 7:56 pm

More updates:

I have my 9-slot system pretty much completed now.

But I tripped myself up a lot along the way. It's a pretty funny story, so I'll tell you as I go.

Because of that, I really haven't given the rev.H board much attention, as I should. I'll try to find time to do the steps Martin described next week.

First of all, I had several things on my to-do list last week. I had some diskette drives in from eBay sellers that I needed to jumper-configure and test. I also had received my rear chassis panels from Ideal Specialty in Tulsa - I had asked them to cut a DB-37 connector hole for the diskette drives. And lastly, I needed to build the cable assemblies to use the DB-37 connectors.

Here's the panel with the DB-37 cutout. I later drilled a 1/2" hole right beside the power cord for the fuse holder, and a 1/8" hole to mount the FDC+ diskette interface daughter board.

Altair8800c_Rear_Plate_with_DB37_cutout.jpg
Rear panel with DB-37 diskette connector cutout


Ideal Specialty would have been happy to cut those holes for me too - and they could have cut the D-shaped hole for the fuse holder - but that didn't matter to me as much as getting the DB-37 cutout right. I knew I could drill round holes well enough. In fact, if you look closely at the photo of the panel above, you can see that I had already used a punch to mark the place where I was getting ready to drill the 1/2" hole for the fuse holder.

And here's the completed rear panel:

Altair8800c_Rear_Panel_Finished.jpg
Rear panel of completed system


I initially bought stainless screws but later decided to get black ones. I also got some caps for the unused DB-25 holes and for the cassette jacks.

Internal wiring is nice and clean. The ribbon cables fold nicely to run between the slots and through the card cage, out to the connectors. The FDC ribbon cable connects to the 50-pin-to-34-pin daughterboard, which is mounted just below the DB-37. The DB-37 is connected to the 34-pin IDC too, straight through. It could have been done with a ribbon-crimp style connection - which I did on the 34-pin side - but I soldered the 37-pin connector.

Altair8800c_Internal_Cables.jpg
Internal cables


You can't really see it, but I've attached the FDC+ serial connection too. When looking at the system from the rear, the top two DB-25 connectors are the 88-2SIOJP ports. And the DB-25 that's on the third row down is the FDC serial connection. All the rest of the DB-25 holes are plugged.
Wayne Parham
 
Posts: 239
Joined: March 18th, 2022, 3:01 pm

Re: Altair 8800c build thread

Postby Wayne Parham » June 24th, 2022, 8:00 pm

I mounted my diskette drives in a case shown in the "Floppy Drive Enclosure" thread. There are other good options, but this case worked nicely for me.

Altair8800c_Diskette_Drive_Cabinet_Front.jpg
Diskette drive


I considered masking the drives and painting them black to match the case. I could have masked the select-indicator LED and the electronics and applied an adhesion promoter and paint. Might still do it. But I noticed that the early low-density drives were usually black and the later high-density drives were usually tan. I remember that color shift, but didn't really pay attention to the fact that it tracked with the density shift. Not that it really matters, but it pushed me towards leaving the drives alone. Besides, I had bigger fish to fry. The drives looked fine.

Altair8800c_Diskette_Drive_Cabinet_Rear.jpg
Rear panel of drive cabinet


Here's why I really went through all this. I wanted to have DB-37 connectors between the CPU and the diskette cabinet.

Altair8800c_Diskette_Drive_Connected.jpg
CPU and diskette drive connected


The connector cable is a three-foot DB-37 straight-through cable. They are still readily available, as are five-foot variants.
Wayne Parham
 
Posts: 239
Joined: March 18th, 2022, 3:01 pm

Re: Altair 8800c build thread

Postby Wayne Parham » June 24th, 2022, 8:04 pm

Here's the internal data cabling of the diskette drive cabinet:

Altair8800c_Diskette_Drive_Cabinet_Open.jpg
Diskette drive cabinet open, showing wiring


The cable is a straight connection - none of that PC-era twisted cable stuff - and it runs to a 34-pin IDC connector. That connects to a short 34-pin-to-DB37 cable. Could have been done with a single cable too.

Altair8800c_Diskette_Drive_Internal_Data_Cable.jpg
Close-up of the internal data cable


Altair8800c_Diskette_Drive_DB37.jpg
Close-up of the DB-37 connector
Wayne Parham
 
Posts: 239
Joined: March 18th, 2022, 3:01 pm

Re: Altair 8800c build thread

Postby Wayne Parham » June 24th, 2022, 8:19 pm

So now for the "messy" part of the story. The last few posts show the end-result, but this time I had to work for it. I initially connected the CPU to the diskette drives several weeks ago, and had enjoyed running lots of stuff on the system. I have a dozen diskettes made already, with all kinds of fun on them. But it was all done with a one-foot 50-pin ribbon cable from FDC+ board to daughterboard, and a one-foot 34-pin ribbon cable from daughterboard to diskettes.

All good stuff.

Then it came time to start testing the DB-37 cables. And I also had those diskette drives to test. I had only actually run a couple of drives, and I have two systems which I planned to have two drives each, so I still had some drives to test. And I was in kind of a hurry.

Not really a good idea to be in a hurry.

My first try at the 37-pin cables worked great. But I had a couple of those cables, and I wanted to test each of them. That way I'd have some "known good" units to set in the "tested" pile. The second cable I tried didn't work. So I put the first one on and it stopped working too.

In my mind, this meant that second cable did a bad thing. And I should add that I was trying both three-foot and five-foot cables. The TEAC interface specifies a maximum interface cable length of five-feet if 1KΩ pull-up terminator resistors are used, and a maximum of ten-feet if 330Ω terminator resistors are used. Considering I had a couple feet of internal ribbon cables, the three-foot cable was probably going to work no matter what, but the five-foot cable was questionable, so I wanted to test it.

So my first thought was the five-foot cable screwed me up. And since going back to the three-foot cable didn't fix it, I started worrying that I blew a board. I actually kind of assumed that. But since it was the long cable that did it, I thought to myself - this is why having spares makes troubleshooting easier - and I installed my other FDC+ along with the shorter cable that worked before. I powered up, toggled stop/reset and run and fully expected to boot. It didn't.

Neither FDC+ board worked. None of the cables worked. I guessed that I'd shorted something, hit it with noise, who knows what. I even traced the signal path from the diskette interface to the nearest chips on the FDC+ boards, solder-wicked 'em off, installed sockets and fresh chips. They're U17, U18 and U19. That'll fix it.

Nothing. Nada. No boot. I had completely trashed my pretty little Altair.

So I wrote to Mike, and described my situation. I assumed I had blown the boards and I asked him to sell me another. Once I was back in black, I'd troubleshoot the bad boards. He didn't think it was the FDC+, but sent me out another one anyway. I told him I'd put a scope on the drive select and motor start signals and see what they were doing. I hadn't done that yet, and it was clearly the next step. But it was nice to know a new board was on the way.

Then it occurred to me, since the signals are all open-collector, it would be pretty hard to blow anything by shorting to ground. By "pretty hard" what I mean is impossible. To blow this board, you pretty much need a lighting strike. I wrote to Mike and said that, and he agreed. He must have been totally cracking up at this point.

And the new board showed up and did the same thing. So I knew I must have overlooked something obvious.

I pulled out all the cables. I'd show you a photo, but I've hit my attachment quota. But what it was is the edge connector for one of the drives had a finger that was mangled, and touching an adjacent finger. That was the whole problem, all this time!

How in the world had I missed that?!!

Made me think of the guys on the show "The IT Crowd" saying, "Did you turn it off and on again?"

This was way too long an explanation to end so stupidly, but there it is. So funny!

After throwing that cable away, everything else worked great. The original chips are back in their places - socketed now - and they work just fine. All the other cables work fine. And Mike offered to allow me to return the FDC+ board he sent me. That's support, man.

But I'm keepin' it.

Last thing to share in this already way too long post is a suggestion to check your diskette drives if you get 'em off eBay. And where else would you get them?

In spite of my cable charade, there are some other fairly high failure points. In fact, it's part of what made me even step onto that thin ice. I'd say only about 75% of the used drives on eBay work. The sellers probably connect them and see if they turn on, if even that. They may be listed as "tested" or even "refurbished" but many of them just don't work.

I have found the sellers will send a replacement if you tell them a drive doesn't work, but obviously you need to do this in a timely fashion, before the trust-rating system has expired. So don't buy a drive when you start your build and wait to test it 'til you get everything done and ready. Wait until you're nearly ready for the drive before buying it.

I had a couple drives that would seek and appear to work, but wouldn't read. One would read but wouldn't write. Then there was this one - sent as a replacement for a bad drive - which was also bad but easily repaired. It had a pretty large scrape across the top board, cutting the traces for the index pulse sensor. It was an obvious-enough problem, I felt it was worth attempting to repair. If it didn't work, I could always still send it back. But after jumpering the traces, it worked.

So my point is to definitely have a healthy suspicion of drives coming from eBay. Most work, but a large minority don't. My experience is one out of every three to four have been DOA.
Wayne Parham
 
Posts: 239
Joined: March 18th, 2022, 3:01 pm

Re: Altair 8800c build thread

Postby AltairClone » June 25th, 2022, 9:35 am

Despite the troubles along the way, the final build is looking great!

Mike
AltairClone
Site Admin
 
Posts: 631
Joined: April 5th, 2013, 10:55 am

Re: Altair 8800c build thread

Postby BillO » June 25th, 2022, 12:17 pm

Nice clean build Wayne!
BillO
 
Posts: 136
Joined: November 11th, 2020, 6:29 am

Re: Altair 8800c build thread

Postby Wayne Parham » August 7th, 2022, 3:43 pm

Updates:

I have fully assembled and tested both my "red" 18-slot Altair 8800c and my "blue" 9-slot system. Photos of each can be found earlier in this thread.

They both worked pretty much right away, but each had its own unique idiosyncrasy that I had to work through. None were design problems or anything of that sort, just little "hitches" that I thought might be worthwhile to share.

First was the disk drive interface cable problem I mentioned above. Easy enough to resolve once I discovered this:

Bad_Floppy_Cable.jpg
Damaged edge connector

Somehow I had mangled that cable end, which obviously made the drive non-functional. But after I replaced it, the system was good as gold. In the end, I had about five feet of cable run between the FDC+ card and the diskette drives. Works just fine at that distance.

Second was the issue I mentioned earlier in the thread with ROM BASIC on my Rev.H 88-2SOIJP board. It was intermittently loading BASIC, no matter how I set the sense switches. I could always boot from diskette, but I couldn't always start ROM BASIC from the front-panel.

I initially thought it was slot-sensitivity issue, that the SIO board worked right if it was placed close to the CPU. But that was just coincidence. It wasn't a slot sensitivity problem.

Both Martin Eberhard and Mike Douglas helped me out with this. They each have enormous patience, or if they didn't, they at least didn't let it show.

Martin gave me several troubleshooting suggestions, including even some small test programs to enter from the front panel. He even let me send the board back to verify. Worked for him, so we continued to look at my system. Martin quickly came to the conclusion that the problem was in the front-panel.

Mike also gave me lots of pointers, including a willingness to help me diagnose the interface board with a scope. I have two interface boards, so once we determined the problem was there, I knew it would be pretty easy to find. When I dove in a little closer, I realized that one of my interface boards worked properly but one didn't. Even the one that didn't work would boot diskette and do pretty much everything except STOP. After IPL, the stop toggle would no longer stop the system.

Once I began to focus on the STOP problem, I ended up swapping chips until I found the defective part. It was the 74LS02 NOR chip in location "L." Swapped it out and the STOP toggle was able to halt the system.

The last thing I encountered was the card guides in the 18-slot system. I had ordered 36 guides from JM Precision but they are actually a little too tall and the slots don't go deep enough. So they prevented the cards from penetrating the edge connectors enough for reliable operation. I ground the tops 1/4" to get the right dimension for the ejectors, but the slots were still wrong. They didn't go deep enough to allow the boards to seat properly. I ended up disposing of all of them and having new ones made from the STL file in the following archive:


The files in the ZIP archive include the left and right 9-slot card guides and the single card guide as is used on the 18-slot board. Of course, you'll need 36 of those to provide for all slots on an 18-slot motherboard. If you have a 9-slotter, you just need the left/right 9-slot pair. They're different setups.

I might add that the 18-slot motherboard from JM is excellent. It's thick enough to support an engine block. And it could be that some edge connectors are taller than others. I used the connectors sold at JM and advertised as being "manufactured by the original company to the original design for a true authentic result." If you get those, be sure to download the STL file above and have card guides made from it.

I already listed the hardware for mounting the 9-slot motherboard earlier in this thread. It's actually in a link to the 3D Printing Service thread. The hardware needed for the 18-slotter follows:

(72) #6 x 3/8" Truss head Phillips wood screw - These are used to mount 36 card guides, a pair for each of the 18 slots.
(36) M3 x 10mm female threaded stand-off - These mount the board to the chassis.
(36) M3 x 12mm screw - These mount the board to the standoffs, going through the edge connectors.
(36) M3 x 6mm screw - These mount the standoffs to the bottom plate of the chassis.

You'll need to drill the chassis to mount the 18-slot motherboard. You can use the motherboard to fabricate a drilling template out of cardboard.
Wayne Parham
 
Posts: 239
Joined: March 18th, 2022, 3:01 pm

Re: Altair 8800c build thread

Postby BillO » August 7th, 2022, 9:01 pm

Helpful information. Thanks for taking the time Wayne.
BillO
 
Posts: 136
Joined: November 11th, 2020, 6:29 am

PreviousNext

Return to Altair 8800c

Who is online

Users browsing this forum: No registered users and 2 guests