The IEE-488 or GPIB bus for controlling instruments by computer has existed now for many decades. It’s often implemented over USB or Ethernet here in 2023, but the familiar connector can still be found on the backs of pricey instruments. In the earlier days of GPIB when a powerhouse Linux laptop was decades away, what computer did the would-be GPIB user reach for? If they were a Rohde and Schwarz customer in the late 1970s the chances are it would have been the R&S PUC process controller, an 8-bit microcomputer that under its smart exterior turns out to be an enhanced Commodore PET. [NatureAndTech] has one for teardown, and you can see it in the video below the break.
Readers with long memories will remember that the PET had an IE-488 bus on a card edge connector, and it’s possible that’s why R&S took it as the basis for their machine. But this isn’t merely a PET in a fancy box, instead it’s a fully new PET-compatible computer, and it has some interesting features. There’s more memory than the original, a set of disk drives, and an expansion bus complete with a high-res graphics card allowing pixel graphics rather than text. Surprisingly though it has a BASIC interpreter it’s a hardware clone of the PET only, the ROM is unique to Rohde & Schwarz.
We think this machine is probably rare enough that we’re unlikely to see one in the flesh, but it’s been a fascinating thing to examine. You can join in with the video below the break, or you can look at the PET’s impact on a more recent scene.
Should be “Rohde & Schwarz” (sorry, language nazi here, and old enough to remember R&S GBIP) :-)
Very rare and interesting fact however, and a curious decision. R&S was never in the habit of taking over other designs…everything had to be theirs. Being a Commodore fan as well, I find this machine fascinating.
Well if an available design that could be easily modified to do what they wanted was out there it probably just made good business sense to use it.
R&S gear, along with HP, is the standard for the EMI test facilities I go to quite frequently. It’s all GPIB controlled, which allows the tech to do hands-off scanning. Even the antenna height, orientation and the turntable position are GPIB controlled.
HP invented GPIB (my dad helped with the design specification) and most of the early HP computers were at least somewhat intended as GPIB controllers for ATE. I don’t know how much the R&S stuff cost in the 70’s, but I know exactly how much the HP stuff cost, and I bet it was at least as much if not more than the R&S stuff. I just bought five R&S scopes, so I’m aware they’re pretty pricey, but some of the HP computers in the 70’s cost more than houses, and this was for a desktop.
When i was at Agilent we occasionally called it HPIB the original name
That looks expensive.
… because it is R&S.
We have an R&S mug in our pool of coffee mugs at the office.
I joke that it cost £2M, but it did come with a free CMW.
Ah! Rohde und Schwarz. Commonly (and affectionately) referred to as “Rost und Schrott” by radio technicians in Germany.
“Rost und Schrott” = “Rust and Junk.”
Ah, yes. Wasn’t it also called “Rohde und Schwarte” unofficially ? ;)
I heard stories about lead plates being installed in pro equipment, to make it “feel” more professional and valuable.
Because, as we know, high weight is a justification for a high price. I believe that Brucker (?) did that at some point. Speaking under correction, of course.
We used to call them Robust und Schwer. I happen to have thrown out one of these boxes 20 years ago. Gave it away to a computer enthusiast.
Not actually a PET then. It’s a 6502 based microcomputer based around off the shelf parts, but in no way compatible with PET code. Don’t forget that the PET came before ULA or PLA chips, so everything was made from off the shelf parts with 74 series logic gluing it together. Superficially they may look similar, but the Devi is in the details.
Those old 8-bit architectures are pretty straightforward. Wouldn’t surprise me at all if the PET and the R&S designs both came from a MOS Technology applications note.
Being that Commodore owned MOS, calling them “off the shelf parts” is a bit strange.
I always wondered why Tramiel never strangled Apple with the 6502. I guess since the 6502 was licensed out to others it would have been hard to strangle Apple. Also, he was happy making money off the chip sales to Apple it seems.
Off the shelf refers to standard, commercially available parts. MOS may have designed the 6502, but they sold it to anybody who asked. There was nothing in the PET that you couldn’t buy direct from the manufacturer on standard commercial terms. Newer computers were made from progressively more custom parts that were specific to a model or manufacturer. Even if you cloned the hardware directly then the ROMs still made the original unique (there was a whole industry around making IBM compatible clone BIOS ROMS without infringing IBMs copyright, and that wasn’t trivial. Without that all you had was a machine that would run some, but not all, PC code).
Oh, and ignoring the fact that the 6502 was licensed to a bunch of other companies, Commodore never really competed with Apple. The PET was the closest they came to the Apple II (which lacked many of the II’s features like expansion slots), but they pivoted towards the home PC end of the market with the VIC20 and C64, which Apple didn’t compete in.
Kind of sloppily written. Twice you refer to IEEE-488, and do so incorrectly both times.
The writer seems to be under the false impression that PETs are “text only”. They do have a serviceable set of keyboard graphics people were using for games since 1977. Apple made you wait almost a decade and buy new computers and new ROMs for such a thing.
And in fact some PETs did have bitmapped gfx. But the ones we had in schools rarely had that, thus the misperception that they did not have bitmap gfx.
Also, does HackaDay not have an editor?
Fascinating that you would be so critical of the HaD editors when your own facts are wildly incorrect. The Apple II line was introduced in 1977, and had bitmapped 280×192 monochrome graphics (or 140×192 with effectively six colors) right from the jump. Where did you get this “Apple made you wait almost a decade and buy new computesr and new ROMs” thing? Did you just pull it straight out of your posterior?
If you recall, bitmapped graphics – even 800×600 on a 4MHz 8-bit processor took a long time to redraw.
Hence the only choices were DMA graphics, or sprite/icon character graphics to get decent update rates.
In my younger engineering days, I worked for an ATE developer building “large” ATE systems driven by the ATLAS programming language for depo-level repair of military avionics. In my mind, that was the heyday of HPIB. Later on, worked on another project on which we had to place IEEE 488 interfaces into a promiscuous mode.
I can confirm that GPIB it is still commonly used in labs today. Better interfaces of course do exist, but IT is for the most part unaware of GPIB, giving it a massive advantage over Ethernet in a corporate setting.
Fun fact: GPIB originated at HP in the early-to-mid ’60s (they called it HP-IB), so it predates microcomputers by about a decade. Since it is relatively simple to implement using discrete logic ICs, a number of 8-bit machines employed it as their system bus using different connectors. The most notable among these was—you guessed it—the Commodore PET.
I worked for a company that made Interactive Voice Response systems (Press 1 for…), their first product was based around the Vic-20, and they actually licensed the design from Commodore so they could make custom hardware
Ah the PET, I remember playing on them in the local library. Damn I’m old. I would love to have one to mess with nowadays. Brings back some memories there.
The Commodore Pet implemented the GPIB port inspired by the Tektronix 4051 (which also inspired other Pet features). HP and Tektronix were the main companies supporting the standard.