This is a different replica than his others. Both the Replica-1 and the MicroKIM-1 use as many original parts as feasible. But for the Altair 8800 micro, he opted for emulation in an Atmel Mega8515. The 8080 used multiple voltages (+5vdc, -5vdc, +12vdc; I wouldn’t be surprised if it also used -12vdc) which would have made replication more expensive.

My kit arrived on Saturday July 24, and when I opened the box, this is what greeted me.

One of his biggest issues was trying to get the Altair 8800 panel as authentic as possible. I think he did a very good job.

All the parts were packed inside the enclosure, which only makes sense.

On the forums, someone asked Vince to serialize the first run. As I said, I received kit #14.

Separating out the package and checking to make sure there is no real damage. The upper board is the main board, and the lower one is the front panel. There are the ICs in two tubes, and appropriate sockets in two more tubes.

Also provided were a buttload of switches, for use on the front panel.

And another bag for the mainboard parts. Another buttload of LEDs for the front panel.

And finally, a CD of documentation, firmware, sample BASIC programs, and other sundries. It’s nice that he spent the extra effort to make a professional looking CD.

So, all looks in order. Time to fire up the soldering iron, separate and verify all the part counts, and get to work. I followed his build directions, which is straight-forward. The mainboard was fairly easy to build, it just took two or three hours, including breaks.

That was the easy part. The front panel is pretty intricate, there are a lot of switches, LEDs and resistors. Since this is the focal point of the whole computer, you really want to make sure you do a good job. Take your time, take breaks often, and double-check your work along the way.

To install the switches, you kind of loosely place the switches in position, attach the front plate to the front panel PCB, and turn the whole thing over. This is so you can align the switches with the front plate. It’s just a matter of wrestling with the switches with one hand, holding the soldering iron with the other, and then applying solder with the third hand. Well, what I did was place a tiny drop of solder on my iron and loosely tack the switch into place, so it wouldn’t fall down. Then I could solder the other lug, and then properly solder the first lug. Then go down the line.

This is how my workspace looked from the top (soldering) side. You can see several of the switches are fully soldered in to place, and one of the switch lugs is poking thru.

Next step was to drop the LEDs in to their positions, and fully assemble the front plate to the board. Then you can turn it over, drop the LEDs into the front plate, and solder the LEDs.

After many hours of work, still including breaks, I had a completed front panel!

Once everything looked OK to the eyeballs, I assembled the case together.

And then buckled it up for the first test spin.

I scrounged up a 9vdc adapter, connected a suitable 2.1mm barrel to it, and proceeded to the first smoke-test. Luckily, the smoke test passed, but the computer wasn’t operational. There was no video coming out either the VGA or the video ports, and the front panel was acting a bit flakey. I could toggle certain switches, but I wasn’t able to operate it to any real degree of success.

Well, these things are bound to happen, so it’s back to the work bench to disassemble the project. Anything like this, the first step you take is to verify all your solder joints. I started on the mainboard; painting all the solder joints with flux and touching them up. Since the front panel was operating to some extent, I figured to focus on the mainboard first.

After retouching all my solder joints, the second smoke test passed, and the machine was operational to a large degree. No video on the VGA, but video on the composite port, and the keyboard LED blinks fast in sync. And I notice that the switches for A5 and A6 are shorted, so it’s back to the work-bench.

I communicated with Vince regarding some of these problems, but I fully expect that all or most of them are operator error. I’ve been able to get the machine to load BASIC, and even flashed a new terminal firmware which Vince recently released, so I’m fairly happy that the machine works as designed. It’s just a matter of finding that little “something”.

Update August 3: Vince suggested to look at the power supply as the culprit. Indeed, it’s a 9vdc adapter, but only 300mA… if I did my math right, 32 LEDs at 20mA is 640mA, so no wonder it was acting flaky. I got a 1.6a adapter which should suffice. /facepalm

Here it is after loading 8k BASIC, at the OK prompt

And the VGA monitor after loading BASIC

To me, this is as much of a reward as a tee-shirt or a personally drawn character sheet. Best of luck, Joe!

Basically five nixie displays and five 7447 in a nice carrier. It just took a little while to beep out the pins, and then I drew this up:

Relatively simple: a Mega32 with a DS32KHZ TCXO. The HV converted is smack in the middle, it takes 12vdc and turns it into 180vdc.

Once I had this together, I inserted the nixie carrier, and realized that “untested, as-is” can be risky. But they were cheap enough, and really, that’s why I purchased two of them.

Some testing revealed at least one bad 7447, so I thought to be smart, and just replace them all with socketed 74141s.

But this is an old carrier board, and even though the PCB is double-sided, none of the thru-holes are plated. So the soldering was done on both sides of the 7447, which also meant it was impossible for me to solder the top-side due to the socket.

Luckily for me, I only tried the first 74141 in a socket, because previous experience has shown me for tasks like this, you be very conservative, and test every step of the way.

Out comes the socket, and solder the 7414 into the board.  However, things being as they are, once I had gotten to pulling the third 7447, I had lifted up several lands. Some were “unnecessary”, but some required the application of jumper wires.

At this point, I got pretty frustrated and put the whole thing down. I mean, I got a board for a reason, so I would haven’t to do point-to-point wiring; I might as well just have done this on perfboard.

So I figured, why not just replicate the carrier board? It is handy:

This turned out a lot better. I used BatchPCB, and whoever their fabhouse is, they’ve always produced good quality boards. It did take quite a bit of finagling to solder the leads from the tubes into the board, but eventually everything went well. I was able to go back to my original plan of using socketed 74141s.

Thing is, this is a pretty generic carrier. The nixie leads just wire up into a 6×2 header. All you have to arrange the leads. I didn’t label anything on the board, because I originally considered this just a one-off, but I might go back, relabel everything, and use wider spaced holes for the nixie leads.

This is direct drive, not multiplexed, so it needs at least 2 full PORTs. There is a fifth tube on there, which is wired to the “leftover” bits where the DS32KHZ and INTx buttons attach.

For power, I am using a triple-output switched wall-wart. I found them on Jameco or Digikey for a few dollars. It came with +5vdc, +12vdc, and either -5vdc or -12vdc, which I did not care about.

Again, being conservative, you always test. While trying to determine which was +5vdc, I found that this was wired for black to be +5vdc, red was ground. Luckily, everything inside the wallwart was labeled on it’s own PCB, so I just removed the negative voltage, and wired red to +5vdc, yellow to +12vdc and black to gnd. No need for pictures :p

IRL, I’m a big fan of animation. Cartoons, actually. I’d rather watch a good cartoon then watch CSPAN. And actually, my TV is more often tuned to Boomerang than anything else.

I don’t watch Cartoon Network much anymore, they had their salad days with things like Camp Lazlo, Billy and Mandy, and Squirrel Boy. But CN decided to go with “kids reality programming”, and appearing to forsake their eponymous name.

Nick was also once a bastion of cutting edge animation, in the form of Zim, Rocko and The Angry Beavers (none of which would probably ever make it past nanny state censorship today), but they have chosen to make Spongebob their bread-n-butter, and import commercial products which have almost no originality.

Then let’s not forget the older classics like The Pink Panther, and even older still, Popeye.

Quite a bit of television animation is just so much crap, where the producers and networks have chosen profit over originality. Risk mitigation negates creativity. That’s their prerogative. They are, after all, a for-profit entity.

There is quite a bit of web-based animation, but those still seem to be fairly slim. I haven’t looked at things like Homestar or Foamy for quite a long time. And there’s Explosm, altho they’re more of a web-comic, and their animated shorts are fairly irregular.

Anyways. Joe Murray, the guy who created both Rocko and Lazlo (mentioned above), is working on a project called KaboingTV, and is looking for “startup capital”. You’ll have to read the full details, but the basic idea is to have a source of independent animation, not controlled by big TV, and to hopefully draw other animators.

So why, in a tanking economy, would you donate to an entertainment venture? Well, because this is my blog :) And I’m just posting in the hopes of helping Joe get his project going.

This is an analog clock face using a silicon wafer.

I was randomly browsing ebay when I suddenly stumbled on several wafer listings. On a whim I did a “Buy now”, and this is what I got:

It’s an 8″ round wafer. The pattern is interesting. Not knowing a single thing about silicon wafers, I like to think this is a production test wafer, cause it almost likes like those varying shades of gray you see for copier or photograph tests.

Anyways, all I did was get two peices of 9″ unfinished MDF, glue them together (to make it thicker), and pick up a clock movement at Michael’s. The movement came with gold hands, but I didn’t think that would contrast very well with the shiny surface of the wafer, so I got black hour and minute hands, instead.

Drilling a hole thru the wafer was a bit daunting. I actually cracked it, but thankfully it’s not very visible. I need to polish it now, after my fingers have been all over it. If you ever handle a wafer, make sure you really do use clean cotton gloves, something that won’t scratch it. Not only will you prevent smudges, but you might even actually scratch the surface, it’s extremely delicate. There are several smaller scratches, but they don’t seem to appear, but even so…

But damn if that isn’t sexy.

I found this while browsing a “new” website, Your Nixie Clock Headquarters. I saw reference to a VFD Round Clock, which ultimately led me to Kosbo.com. As soon as I saw it, I just had to buy one. It was a bit on the expensive side, $80, but Kosbo isn’t really overrating themselves when they call this “unique”.

It’s fully assembled, sans power supply and enclosure. I have a 12v wall-wart I used for power. They will eventually offer an enclosure, but it didn’t grab me, so I’ll skip on theirs and try to figure out something to do.

I’m kind of curious where they got the VFD from. I’ve never seen an analog clock face like that, but that doesn’t mean much.

I don’t know who the original source for these are, but if you search for “nix180″, you’ll find several. They’re all relatively cheap, $25. At least, cheaper than trying to unsuccessfully design my own step-up converter.

In any event, I decided to build an Eagle library for it, if I ever do get off my fat ass and finish this. This is the first time I’ve tried to build a library, so I’m not sure how good it is, if it even works.

This is the package:

I just used “regular” pins, so when you add this to a schematic, it’ll be like a header. You can move the pins closer to where they’ll need to be, and that way, the schematic will look neater (IMO).

Anyways, here is the library. I don’t make any claim whatsoever. Feel free to use it as you will. If you do use it, perhaps drop me a line and let me know? And perhaps someone else did a better library than I did.

This is a small cigar box, almost just the perfect size to fit the PCB. The board was about 0.5″ too big on the width, so I had to decide what to do: shave off the PCB, or shave the inside of the box? It would have been easier to shave off a quarter inch of the PCB on each side, but for some reason, that gave me nightmares. So I used a dremmel to shave out the inside of the box.

Maybe you can see it, maybe not. But on each top and botton, I shaved off the inside to about half-way down, so I could sit the board in. Just screw it in to place. I had to stack four 14-pin sockets into each display, so it would lift up enough to clear the lid.

Yea, maybe not the most professional job, but hey.

A single 3.5″ floppy with room for a small something-or-other.

And from the back:

Lots of ports: 10baseT, 10base2, AUI, serial, parallel, PS/2 keyboard and mouse, a 50-pin SCSI, and what looks like two PCMCIA slots.

The guts aren’t that glamorous.

There’s no obvious model number on it, so I’ll have to dig around for some specs, etc.

I used to collect DEC equipment, mostly old iron, like VAXen. I had a couple of 11/730s, some rackmounted PDP-11 (one was an 11/23, the other was an 11/73, and I think one other).

Over time, I’ve had to pass up on offers for 11/750s and an 11/780, although the 11/780 required 3-phase, so it would have been a display only.

Unfortunately, old iron is expensive in terms of real-estate. It’s also pretty heavy on the power requirements. I had to follow a process to boot the VAX 730: power on the RA80, and let that spin for about 5 minutes, and then hit RUN on the processor, and cross my fingers that I don’t pop a fuse somewhere.

After four moves, I no longer have any of my old stuff, which is a shame. I don’t even remember where it all ended up. Some of the other non-DEC stuff was pretty cool, but I draw a blank at most of it. The only outstanding thing was a hand-wired S100 system with two displays: one was a console, the other ran a debugger.

Oh well.

I did not try to plug it in. The power cord is very frayed and the insulation is cracking, falling off. The best estimate is that this was in an attic for the last 30-40 years.

Unfortunately, it was unprofessionally refinished at some point in its life, so any original value is lost, but that’s Ok.

I’ll see if I can remove the chassis and do what I can, at the very least, replace the power cord. And since the original finish is already shot, I can do that myself, too.