• Please review our updated Terms and Rules here

Turbo Button history?

More likely you're running games that are clocked and not dependant on system timing to function properly. Else how does the system know when to slow down, and when not?
 
Xacalite said:
I heard yet another theory on why turbo button was introduced: some early networking stuff failed to work at higher speeds.

BTW, in certain book I saw recommendation to turn off turbo when running commands like FORMAT, DISKCOPY, DISKCOMP...
And I had at least one machine where it was handled automagically (in BIOS?): even though the button was set to turbo, during diskette operations the system switched itself to de-turbo, it was clearly visible via the turbo LED.
Click to expand...

I don't think so--the turbo was an easy mod to the XT clones and gave a performance boost. Networks weren't widespread at that time. Also consider the 5170's overclocking market interest.
 
KC9UDX said:
More likely you're running games that are clocked and not dependant on system timing to function properly. Else how does the system know when to slow down, and when not?
Click to expand...

I was assuming it has auto-turbo because it has a led but no button for it. Also for some reason the Apogee game Stargunner seems to have massive issues with game speed on my system running at insane speeds even though it is a game made in 1996. Odd.
 
The earliest ad I can find mentioning "Turbo Button" was from Northgate in late 1987.

"Turbo Button" and "Turbo Switch" are mentioned in a 1986 PC Magazine article on XT accelerator options. Interestingly, two of the reviewed products had to go into slow mode in order to format disks and one had to be started in slow mode to boot on a XT.

Edit: I knew it was niggling at me. The Eagle Turbo XL of 1984 had a button on the front panel that changed speed. No fancy name for the button yet. Compaq Deskpro was slightly later with keyboard speed changes. Both Compaq and Eagle provided a list of programs that needed to be run at the lower speed. Anyone got an early Deskpro manual to check it out?

Switches to change clock speed were a fairly common add-on for Kaypro before it showed up in the PC compatible world.
 
Last edited:
There was a good reason for the "turbo defeat" mode back in the day.

Take a look, for example, at the INT 13H diskette code in the 5150 and 5160 BIOS. You'll see at least CPU timing loops there that were written on the basis of an 8088 running at 4.77MHz. Double the speed, and the routines fall apart. Same for many games, particularly the copy-protected ones. Some ERSO BIOSes would save the turbo state on entry to a service routine, reduce the speed to 4.77MHz for the routine and then restore it upon exit.

This didn't work for everything, because not everything used the BIOS for floppy access. So, for instance, the copy-protection check in Harvard Graphics or Lotus 1-2-3 would fail in "turbo mode". The earliest versions of FastBack wouldn't run in turbo mode. I still have an EPROM programmer that fails on anything but a 4.77MHz machine.

Eventually, software publishers caught up to the problem, but it was pretty bad initially. Some way to disable "turbo" was necessary; hence, either the "turbo" button or some sort of 3-key salute was necessary (e.g. Ctrl-Alt-keypad minus).
 
TravisHuckins said:
I've got games from 1985 that work fine on it. Go any earlier than that and nope.
Click to expand...

That would make sense, since in 1984 the AT was introduced, which was the first time that IBM offered a PC-compatible system at a different speed than 8088 4.77 MHz (and they did not include any speed switching).
Perhaps the first turbo-XTs were also released at around the 1984-1985 mark.
Before that, there was only one speed, and it was not clear that the PC platform would become this backward-compatible platform, where you would continue to use the same software for many years, and just get newer/faster hardware every few years.
If you look at eg the VIC-20 and the C64, they were not compatible with eachother. So upgrading to a new computer meant rewriting the software. The PC was a paradigm shift in that sense, in the world of home/personal computing.
 
Last edited:
Let's not forget CP/M though.. with CP/M you *could* upgrade your computer, even get a different brand, and programs would continue to work. If you could get the software moved over (often by a/the vendor though) to the always-different floppy format - at least for 5 1/4" limited systems. And then, for programs that needed it, potentially patching in the necessary terminal support. Of course it could be argued that CP/M was more business than home/personal computing, but then again that was also the original PC target.
As for the introduction of the AT - I'm sure you're right, that would force software designers to stop hardcoding timing loops and the like. I think most of my own "turbo" button experience (where I noticed no problems whatsoever) were from around that time onwards.
 
I didn't have anything with a Turbo button until 1990. Contemporary software at that point, and for the rest of the time that I used actively used PC/XT/AT-clones, for the most part did use hardware timing loops.

It baffled me that in 1994 programmers were assuming that we would be using 80486/33s for decades. Part of the problem I think is that so many people were still using drastically slower machines and no one cared to write timing routines to account for both too-fast and too-slow.

Did CP/M machines span enough different speeds to really warrant a "turbo button" solution? It seems to me with my limited knowledge of CP/M that most things really speed constrained would be addressed in the necessarily different BIOS anyway. PC/XT/AT-clone BIOS' were so generic that you actually could have hardware problems with a new, faster motherboard. Otherwise, on the odd chance that you had a game that ran on grossly different CP/M machines, what was the worst case? Four times as fast? Text-based action games seem to generally be still playable at that speed, and, given how hard video games were back then anyway, the extra challenge wasn't unwarranted, I suspect.
 
There was not a standard speed for Z-80 CP/M-80 systems in general. But programs written for a specific platform, such as the Kaypro II, were very likely to make speed assumptions. As a result, any speed upgrade add-on boards would have to be able to throttle down to the platform's original speed.
 
You must log in or register to reply here.
Back
Top