The 7501s: The TED Developer Prototype, 116, 116 Portable, Book 116, 232, 264, Canadian Plus/4, 364
Click the Commodore icon to return.
Intending to use pictures or text from this page? Please read this notice. Last modify 21 January 2018.
Lots of this information comes courtesy of one of the original engineers, the one and only Dave Haynie, who graciously took the time to answer all my silly questions. Grateful acknowledgment also paid to Bil Herd, chief of the TED project, who filled in additional detail.
The TED/7501 series have the most colourful (no pun intended) histories of any of Commodore's releases and secret weapons, inspiring many models and many management headaches. The 7501s were most well known in their guises as the ill-fated 16 and Plus/4 series of computers, but were also the primary component of the 264 series. The TED 7360 (the 264's, 16's and Plus/4's I/O chip) project, according to Dave, was originally designed to be a return salvo at the successful Sinclairs, which had Jack Tramiel worried; Tramiel wanted another computer in the "super-low-cost" US$100 range to challenge the ZX-80 series and the Spectrums. In fact, its original price point, according to Bil Herd, was supposed to be just $49.
The confusion is largely blamed on Tramiel's departure, but Dave notes that the technology was very robust and functional, as any owner of a Plus/4 will attest (the irritating built-in software notwithstanding); in fact, six of the seven TED engineers eventually moved on to develop the 128, including Herd himself. If anything, the only sin of the 264s was not being 64 compatible, but this was too great an infraction to be forgiven at the sales counter. If the machines had been cheaper, they probably would have sold better (and as proof, most of the machines sold like hotcakes in liquidation), but as it was most of the engineers (including Dave and Bil) pretty much pronounced the series as stillborn. Later CEO Thomas Rattigan noticed the obvious as well, and finally killed the line as part of his initial financial cleanup in 1985.
Mike Dailly's Plus/4 and 264 Home Page
Annotated Memory Map of the 264/Plus/4 (original by Commodore; additions by Mike Dailly and myself)
Introduced Never officially; board samples were given to developers
Hardware Early hand-bonded 7501 CPU and 7360 TED, otherwise similar to line except for standard 9-pin joyports (see Comments). Appears to have 16K of RAM.
Graphics and Sound Identical to later units. This TED has a defect where the test mode pins wound up higher impedance than expected, requiring a jury-rigged resistor setup. Per Bil Herd, none of the production developer boards had this modification as the defect was presumably fixed.
Eventual Fate Guess.
Bil Herd's original work was on what this unit would become, a development board for TED software production -- as the 264 series was of course not 64 compatible, the Commodore software branch realized they'd better get something on the shelves quick. The 9-pin standard Atari joystick ports were used since most developers wouldn't have had any of the oddball DIN-port sticks, natch. This unit is in Anders' possession; Bil estimates it probably got to Sweden through Sig Hartman's software group (see his original cbm-hackers post). Regretfully, the unit is not working, probably due to bad or degraded passivation in the chips.
Additional Views of the C116 (courtesy Markus Mehring except where noted)
Early Model Keyboard (15K, .jpg, courtesy Marc-Jano Knopp) | Late Model Keyboard (153K, .gif) | Rear Ports (68K, .gif) | Comparison Keyboard Shot (120K, .jpg, courtesy Ray Castaldo)
Views of Haynie NTSC Prototype C116 (courtesy Ray Castaldo)
Portrait (52K) | Mainboard (68K) | Backplate (28K)
Introduced There is some argument here. Some
authorities say never officially, but it was sold eventually in spite
of not having a public rollout (see Eventual Fate).
Markus Mehring has a German sales brochure for the 116 dating from as late
as December 1984 and does remember it being sold in stores; in fact,
my unit is in a German production-grade package.
Hardware 7501 @ 1.76MHz (see What's a 7501, Anyway?), 16K RAM (max 64K), 32K ROM. No user port.
Graphics and Sound Identical to the 16 and Plus/4 (more specifically, the TED chip (see What's a TED, Anyway?)).
Eventual Fate Never released in the US; instead, for American users, the 116 was reincarnated in a 64-like motherboard with a grey keyboard and released as the 16. However, some must have been sold commercially in Germany, at least (Spiro Trikaliotis remembers the 16 and 116 being sold together at the German Allkauf), and On the Edge states they were also sold in Japan. Poor sales with a relatively brief lifespan.
One of two machines designed by Commodore Japan (and with little feedback from the home offices in West Chester) -- the other being the Ultimax -- the 116 is unique for its bizarre rubber "Chiclet" keyboard, and was only known released in Europe, according to Nicolas Welte; it has a motherboard very similar to the Plus/4, but totally unlike the Commodore 16 (which had a newly manufactured design). Its sales were weak at best; after its apparent market failure, it seems that Commodore Deutschland GmbH, uncomfortable about all that stock sitting around, sold them off fire-sale as educational systems and some of these are also still in use. The picture of the box, taken of my own unit, says (freely translated): "16K Memory, Super BASIC, 121 Colours, Help Key; high resolution graphics, 75 program commands, sound and music generator, eight programmable function keys, extensive peripherals [for] connection."
The keyboard on the 116 is awful, just terrible. Both my units (the one I acquired from Ray Castaldo as well as the unit built from the casing Martijn van Buul sent me), has a keyboard that is absolutely mush, bad even amongst Chiclet keyboards which are a bad lot themselves. Worse is the very early 116, which had only one SHIFT key in the lower right (the left SHIFT was now an Escape key -- your guess is as good as mine!), and no shift lock, moving the INST/DEL key there instead! Later models, including my case, have the INST/DEL key up next to the solitary CLEAR/HOME key (top right, in the function key row), an improvement, but not by much, because they never did put the SHIFT/LOCK back (it's now the English pound key). Ray's image above shows the details. Despite the terrible keyboard, the design still manages to be very attractive and compact and was in fact designed by Ira Velinsky, then working for Commodore Japan and Sam Tramiel, who was also the designer of the B/P series cases.
One interesting thing is the cursor controls; on the 116, they are linked together into a angular diamond that works like a disc controller, very neat in theory but positively maddening in practice.
By the way, it is definitely possible to build an NTSC 116. Because the TED chip is not intrinsically PAL or NTSC, simply replacing the crystal and Kernal ROM is sufficient. (Where you would actually get an NTSC 116 Kernal ROM is a more pressing issue, but an NTSC 16 Kernal ROM will do just fine.) This also applies to any of the 264s, including the 16 and Plus/4. In fact, Ray Castaldo was able to create an NTSC machine doing just that, using a 14.x MHz crystal; this is the machine that yours truly presently owns now and exhibited at Vintage Computer Festival 7. Only Dave Haynie's original prototype, which Ray also owns and is displayed above, was truly manufactured as an NTSC system; his unit, obviously, was purely for testing and had no FCC clearance. Interestingly, Bill Degnan notes that there are unused holes on the 116 motherboard which closely fit a standard C64 NTSC crystal.
Despite its obvious flaws and annoyances, according to Dave Haynie, the 116 is the real deal in being the closest machine to the original TED concept.
This briefcase-sized portable made its only known appearance at CES 1984 before disappearing completely. In a possibly apocryphal mention in RUN's April 1984 issue, the magazine notes that Commodore would not divulge any further information about it until its release, though they did show it off to show attendees, but the Portable never did escape development: the 116 was introduced instead.
Jim Butterfield notes that this tiny incarnation of the 116 was made for Commodore Deutschland GmbH (he saw it at Commodore Fachausstellung, Frankfurt); it had the diamond rocker pad, but was shrunk to an extremely diminutive size (easy to do, given that the bulk of the 264s is given over to the keyboard, not the circuitry, thanks to TED). The Book 116, as its name implies, was intended for BASIC programming texts and was basically built into or attached to the back cover. The power supply, naturally, apparently didn't fit so well in the book binding.
Complete Views of the Commodore 232 (.jpg, courtesy Hans Karsten)
Portrait (27K) | Badge/Keyboard Closeup (28K) | Rear Ports (17K) | Serial Number/Rear Plate (21K) | Board Number (35K) | Board Overview (40K)
Additional Views of the Commodore 232 (.jpg, courtesy Dan Benson)
Portrait (74K) | Back (37K) | Backplate (17K) | Screenshot (51K)
Introduced Never officially, but models apparently leaked out.
Hardware Same as the 264 but with 32K, although the 232 lacks a user port and ACIA.
Graphics and Sound Identical to the 16 and Plus/4.
Eventual Fate Scrapped prototype; however, the unit is known to have made it into production and a smattering are known to exist today thanks to reports from various owners (Bo Zimmerman, Dan Benson, Daniel van Heugte, Hans, Arwin Vosselman, Erik DK). Erik has the current highest known serial number at #AA5 000033.
Interestingly, the 232 is not an upgraded descendant of the 116 or 16, as some have believed; the keyboard alone is enough to demonstrate it is in fact a downgraded 264. Oddly, the 232 lacks a user port (as well as an ACIA).
So far, all the 232s I have received reports on were originally purchased in the Netherlands. According to Arwin, the price he paid for his (Hfl. 40, approximately US$20) seems to indicate an inventory dump, probably similar to the dump that ultimately finished the 116. This makes sense geographically because of the nearby Commodore depot in Nieuw-Vennep (Arwin found his in a group of about twenty units being similarly liquidated, without power supply and no production packaging, during the late 1980s). His suggestion is further confirmed by comparing ROMs; the 232 has ROM version number 81, while the 16 and Plus/4 have version 84, with differences in the tape routines, giving strong evidence to the theory that the 232 was in fact a prototype and was never meant to leave the Commodore factory floor. Interestingly, Arwin's model is effectively a 264; his unit has been upgraded to 64K.
Views of Various 264 Prototypes (.jpg, courtesy Marc-Jano Knopp)
Early Prototype 264 (16K) | Late Prototype 264 (20K)
Introduced January 1984
Hardware 7501 @ 1.76 MHz. Not known to have the TMI-Micro 3+1 software, so probably only 32K ROM. 64K RAM. 6551 ACIA.
Graphics and Sound Identical to the 16 and Plus/4.
Eventual Fate The 264 as such was never released. However, an upgraded form, with the added 3+1 software, became the Plus/4 (ACIA in tow). Analysis of Canadian trademark dates seems to imply that the transition from the 264 to the 16 and Plus/4 occurred somewhere around February 1985.
The 3+1 software that the Plus/4 eventually sported was a hasty addition at the last minute that Commodore never originally intended to include with the unit. There were actually supposed to be several flavours of Plus/4 built-in software: a version with a LOGO interpreter (for educational purposes) and a version with Pilot; a business-keyed model with a purportedly better wordprocessor (Superscript) or a spreadsheet (Easycalc); a communications module (COM 264); or a home and business finances module (Financial Advisor). The extra programs not included with the "custom-built" unit could be had as add-on cartridges or possibly through a unique device called the SuperCartridge. According to Dave, however, stores rebelled against stocking multiple units of the same machine, especially one that was not guaranteed to be a big hit because of the lack of 64 compatibility, citing it would make inventory management too complex. Another unmentioned reason was no doubt that it would take stockroom space away from the 64, which was on a rapid rise to superstardom. The option ROM idea was scrapped and the TriMicro 3+1 software added to the base unit instead.
One intriguing option ROM idea, unfortunately never completed, was the Magic Desk II option ROM set. Commodore press releases detailed the option ROM as an updated version of the original 64 cartridge, with "Lisa-like" pulldown menus and icons. It was even supposed to have speech synthesis functions when used with the Magic Voice cartridge, which, since the Magic Voice device was intended for the 64, implies that Commodore was also mulling over a Magic Desk II version for that platform as well. However, the prevailing opinion in Commodore's RandD labs was that the original Magic Desk was little better than a toy application, and an update equally useless. It isn't known if programming even actually commenced on the Magic Desk II project, but Bil Herd in On the Edge comments that Jim Butterfield was demonstrating a 264 and Magic Desk (p336), although this could have been a cross-ported Magic Desk and Jim doesn't really remember.
The 264 had some of the keyboard controversy that the 116 did (but did have a proper keyboard, praise the Lord), just of a different sort. Early 264s, such as the one introduced at Winter CES with the rest of the 264 crowd, had several unlabeled keys. The Commodore's quirky back-arrow key was missing, SHIFT/LOCK was unlabeled, and there was also a perplexingly shaped key where the right SHIFT would normally be. The key shape suggests something along the lines of the old IBM PC's "ENTER" key, but nobody really knows. The final 264 put back the SHIFT/LOCK key and the right SHIFT key, and put an Escape key where the back-arrow would normally go, much the same layout as the Plus/4 eventually sported when it arrived. The cursor arrows on the 264 and Plus/4, the most obvious change to the keyboard which replaced the fused "diamond" deal on the 116 and the CRSR keys of the VIC-20 and 64, are a definite downer -- they have a small contact space, despite their length, so hitting them without being egregiously deliberate is tough; have an odd travel depth, and thus feel funny; and are difficult to touch type with because of their position on the board. Nice aesthetics, though.
Information on the 232 seems to allege that there were slight changes in the ROM between the 264 and the Plus/4 (the 232 is at version 81, while the Plus/4 and 16 are at version 84).
Hardware, Graphics and Sound Identical to original unit.
Eventual Fate Commemorative unit.
This was produced as a commemorative tie-in for the Canadian Olympic team in the 1984 Winter Olympics, of which Commodore was a major sponsor. The box is in both English and French.
.jpg Image of the 364 (25K, courtesy Don Judy)
aka V364, CV364, 364V
Introduced January 1984
Hardware Identical to the 264. Keyboard had a numeric keypad, as well (grey keycaps, with the main keyboard in white).
Graphics and Sound Identical to the 16 and Plus/4, but also carried an on-board "Magic Voice" speech synthesizer (the Toshiba chip is visible in the left board view).
Eventual Fate Scrapped prototype. Bil Herd estimates only around two or three were made.
Commodore's final gasp in the TED (d)evolution was that "TED Shall Talk", in the words of engineer Bil Herd, and the 364 was the direct product of that plan. The Magic Voice hardware is identical to the Magic Voice cartridge sold for the 64 and the 364 includes the same BASIC extensions too (but see note at the bottom of the entry).
Dave Haynie was actually the project manager for the 364, and rode the project to its rapid cancellation mere weeks after he took the job.
Unlike the 128's numeric keypad, which was wired into different lines than the normal keyboard matrix, Richard Atkinson on the Plus/4 mailing list says that the Plus/4 ROMs suggest the 364's keypad was nothing more than keys connected to the same lines the regular number keys used in the keyboard matrix, and thus were not read with any special routines. Credence to this theory is added by the fact that the Plus/4's multiple SHIFT and CTRL keys are wired in the same manner (i.e., multiple keys to a single line), and also that legacy features in the old 264 series were never removed from the Plus/4's ROM (there are other goodies around $FD20): if there were separate Kernal routines for reading the keypad, they should have also survived unless they were never there in the first place.
Only one known complete unit is in existence (per Dave, who sadly had to sell it on eBay some time ago; happily, this unit wound up in the possession of Dan Benson, who generously took the photographs above). Several mockups were made with white plastic and painted, but only two moulded cases with real grey production plastic were made, and Dave/Dan's was the only one with a complete functioning keyboard and keypad (most of the others had a real 264 keyboard, but the keypad was non-functional). There seems to be other differences in Dave/Dan's keyboard because Bo Zimmerman reports differing key sizes (the RETURN key on Dave/Dan's seems to be smaller than Bo's 364). However, fortunately, both Dave/Dan's and Bo's work.
In an E-mail interview with Michael Tomczyk, there is a mention of a "MAX Machine" (not the Ultimax) with built-in software applications and speech capabilities in a black case sounding very suspiciously like the 364. In the interview, Tomczyk also mentions a plug-in cartridge patch which might have been related to the SuperCartridge.
Despite having the same electronics, the Magic Voice hardware onboard the 364 has a slightly different vocabulary list (courtesy Bo Zimmerman and Richard Atkinson).
Dave Haynie notes that should you land a V364, but have no power supply, all is not lost: it can use a regular C64 power supply (unlike a Plus/4). Very convenient.
Some of the apparent differences between the 6510 and 7/8501 are actually board-motivated: for example, (intriguingly) there is no way to generate an NMI on the Plus/4, and probably any other 264 series computer, despite the fact the 7501 does accept NMI signals.
TEDs have an annoyingly high failure rate (it's the first thing to check if you get a black or jumbled screen), but unfortunately their only definite source is another working 264, such as a 16 or Plus/4. This means cannibalism in a most displeasing fashion to get your rare 264 working again. (At least there is some hope of turning a more common 6510 into a 7501, but the process is probably out of reach of the average hobbyist.)
TED gets a lot of flak, but the reality is that it remains one of Commodore's better engineering achievements, believe it or not. The graphics alone are extremely good for a early '80's machine, especially given the rich palette, and if TED had sprites and were paired up with a better sound generator (SID comes to mind), the Plus/4 would have given 64 owners pause for thought. As it is, the Plus/4 remains at least as popular as the 64 in Eastern Europe, and there is a surprisingly strong demo scene for it. However, it is also living proof that good engineering can never beat bad management.