• Hello Guest! We're hosting a challenge to welcome vintage Intel macs to the MLA during the month of July! See this thread for more information.
  • We've made some quality of life improvements to the Trading Post. More info here.

Adventures with Appleline

thecloud

6502
In the early 1990s, I was handed a slim beige box by a friend, who had gotten it from a person who worked at Apple, and asked if I needed a modem. She thought it was a modem, since it had a RS-232 port which said "modem" on the back panel, but it wasn't actually a modem. We didn't have a 24V AC power supply at the time, nor a serial cable of the appropriate gender, so I put it on a shelf to look at later. Like you, probably, I had no idea what an Appleline was. It has been described as "one of the rarest pieces of commercial hardware Apple has ever produced."

IMG_9877_1.jpg

Back in 1985, Apple wanted to sell its Macintosh Office solution to Business Customers. Macs could be networked with a propietary protocol suite called AppleTalk. However, many large enterprise customers had already invested in an IBM System 360-series mainframe and system network architecture, and were buying PCs to act as IBM 3270-compatible terminals. The Appleline Protocol Converter let you use a Mac as an IBM terminal, in an arguably less expensive way. It could translate between Apple's ASCII-based asynchronous serial communications and the IBM SNA world, which was EBCDIC-based and synchronous. This was no small feat. It needed a small dedicated computer to pull it off.

IMG_9879_1.jpg

But the Appleline had a secret: it wasn't actually made by Apple. It was a rebadged IRMALine, a product made by Digital Communications Associates (DCA), who had acquired it from Technical Analysis Corporation (TAC). Apple simply sprayed a coat of beige paint on the outside. This was confirmed a few years ago by Bill Landolina, one of the IRMALine engineers, as well as by me a few hours ago.

IMG_9878_1.jpg

I held onto the Appleline, with the idea that someday I would be able to hook it up and make it work. Very recently, I was motivated to take it down from the shelf, and to search out a working 24V AC power supply. These are generally unobtainium, but the Appleline box specified the exact model to look for: an Ault 326-4024. I was finally able to find one that was made for an IRMALine.

IMG_9909_1.jpg

There is a reference guide, designed for sales people demonstrating the Appleline, which explains how to set it up and what to expect. In particular, turning on the power is supposed to perform a self-test, causing the red, green, and yellow LEDs on the front panel to blink. I connected the power supply, flipped on the switch (because YOLO), and... nothing happened.

IMG_9910_1.jpg

First I verified with a multimeter that the power supply was actually putting out the expected AC voltage, and the problem was with the Appleline unit. So I removed the four screws on the bottom, opened it up, and had a first look at this board.

IMG_9916_1.jpg

It all looked very clean for 40 year old hardware. As expected, everything on the board says DCA, not Apple. There is a custom socketed Appleline ROM, and interestingly, there are 3 PROMs in sockets at one end which are dated 1987. I got as far as verifying the fuse had not blown and current was flowing to the far side of F1. But then I had a closer look at the capacitors.

IMG_9917_1.jpg

These aren't capacitors. Renco Electronics is known for making coils and transformers, so I assume these fall into that category. The smaller of the two black cylinders looks somewhat deformed and bulgy around the sides. Haven't seen anything that looks quite like these, but I'm not an EE. Do you recognize them? How to proceed?

IMG_9918_1.jpg

The inside of the case shows traces of what the plastic originally looked like (dark grey) before Apple applied the beige paint. Steve's design philosophy where even the inside of a case had to be beautiful clearly did not apply here.

So the adventure remains on hold. Someday, this Appleline will awaken and my Mac 512K, running MacTerminal, will prompt me to Enter Password. I will respond with APPLES, the default password to the Appleline supervisor program, and then I will scream out, "I'M IN!!" But today is not yet that day.
 
The closest thing I've seen to those is the flyback transformer on the analog board of a classic Macintosh, from the same era as the Appleline. But I'm scratching my head. Why would this need a high voltage transformer? It's not driving a CRT. Is it a voltage converter of some kind, e.g. AC to DC?
 
AC to DC is, so far as I know, only and always done with diodes, albeit usually after AC to AC voltage adjustment via transformer. Transformers basically trade current for voltage, or vice versa. Inductors (coils) on the other hand are weird and counterintuitive, and I still haven’t quite wrapped my head around what you would use them for aside from noise filtering. (If anyone more knowledgable wants to chime in on that part, I’d be fascinated.)
I also cannot fathom why this device would have two large-ish cores on the board, like that. The only suggestion I can come up with is that they might have something to do with the hardware connection to the IBM stuff, the design of which would have probably dated from at least twenty years prior. Does anybody know just what _exactly_ this thing was supposed to be hooked up to?
 
Coax points to a SNA network run by a cluster controller. Basically this box replaces the classic 3278 (or similar) terminal with a Macintosh. IBM's terminals required a controller, unlike the relatively simple ASCII or DEC VT100 lines of the time. That controller could be thought of being similar to a modern router. It linked all the terminals and printers at a worksite back to the OS/390 mainframe located elsewhere.

If you are looking for a modern replacement to use this box with, check this out: https://ajk.me/building-an-ibm-3270-terminal-controller
 
Does anybody know just what _exactly_ this thing was supposed to be hooked up to?
The documentation I linked answers this question. The RS232 port labeled "modem" connects to your computer's serial port, where you would run terminal emulation software like LisaTerminal or MacTerminal to talk to it. The BNC connector labeled "controller" is connected via coax cable to an IBM Cluster Controller, either model 3274 (local controller) or 3276 (remote controller). The coax cable connection may be up to 5,000 feet long. Apple also offered its own "Apple Cluster Controller" which could be used instead of an IBM Cluster Controller... these are even more rare than the Appleline boxes.
 
Realistically, I don't see many non-IBM cluster controllers being sold. They were pretty big with the whole vertical integration thing (lavish and expensive support contracts that promised high uptime). The last thing a site needed was downtime from some 3rd party piece of equipment glitching out. Remember, you local motor vehicles office was running these things!
 
I failed to state my question properly. What I was trying to ask was, is there any reason why hooking up to (for example) the 5000' of coax would require the device to have a chunky little transformer core on the board? Or is it probably for some other obscure purpose?
 
I can't imagine the power supply being all that exotic. I'd expect the usual +5V along with -5V since RS-232 is involved. This entire device is so absurd that you can't hate it. Its got both a micro-controller and a 8085 CPU, all to connect to a then decade-plus old mainframe interface standard.
 
inductors also can be used to smooth out 'dirty' ac>dc outputs, aside to (likely not in this particular peripheral but I'll still mention it nevertheless) also being used a a voltage booster whereas 5vdc input gets turned into 7vdc output
 
inductors also can be used to smooth out 'dirty' ac>dc outputs, aside to (likely not in this particular peripheral but I'll still mention it nevertheless) also being used a a voltage booster whereas 5vdc input gets turned into 7vdc output
This was my thought too... it essentially airgaps the output, which might have been a requirement for putting this anywhere near an IBM vertically integrated stack. So those might not be in there for functional engineering reasons, but more as protection against voiding your IBM service contract.
 
Back
Top